年代:1971 |
|
|
Volume 1 issue 1
|
|
1. |
Front cover |
|
Selected Annual Reviews of the Analytical Sciences,
Volume 1,
Issue 1,
1971,
Page 001-002
Preview
|
PDF (39KB)
|
|
ISSN:0300-9963
DOI:10.1039/AS97101FX001
出版商:RSC
年代:1971
数据来源: RSC
|
2. |
Photoluminescence and chemiluminescence in inorganic analysis |
|
Selected Annual Reviews of the Analytical Sciences,
Volume 1,
Issue 1,
1971,
Page 41-131
L. S. Bark,
Preview
|
PDF (10107KB)
|
|
摘要:
P hotolum inescence and C hemiluminescence in Inorganic Analysis L. S. BARK and P. R. WOOD Department of Chemistry and Applied Chemistry The University of Salj%rd Contents Introduction Apparatus and quantum yield standards The determination of metal ions as inorganic complexes Crystal phosphors Alkali metals Beryllium Magnesium Calcium Boron Muminiurn Gallium indium and thallium Silicon germanium tin and lead Arsenic antimony and bismuth Selenium and tellurium Scandium yttrium and the lanthanides Cerium(1V) Titanium zirconium and hafnium Vanadium niobium and tantalum Chromium molybdenum and tungsten Manganese technetium and rhenium Iron Cobalt Nickel Platinum group metals Copper silver and gold Zinc cadmium and mercury Thorium and uranium Halides-Fluoride Miscellaneous ions-Cyanide -Chloride bromide and iodide -S ulphur-con t aining species -Oxygen-containing species -Phosphorus-containing species -Nitrogen-containing species Indicators Conclusion 4 42 BARK AND WOOD Introduction This review covers the period from the middle of 1967 until June 1970 and is restricted to work dealing with the use of photoluminescence and chemi-luminescence in inorganic analysis.There have been several reviews covering periods prior to and overlapping with the period of this review; of these the most detailed is that of Shcherbovl covering the period 1962 to 1966 and including approximately 300 references. Shcherbov has tabulated the methods for the determination of inorganic ions in such a manner that it is possible to make a direct comparison of the sensitivity and the selectivity of the various methods.The greater sensitivity of luminescent analysis compared with that of absorp-tion spectrophotometry has continued to attract considerable and increasing interest. The biennial reviews of White and Wei~sler~-~ contained approximately 500 references in 1966 and over 750 references in 1970. Although these reviews contain few details of any methods and little attempt is made to assess the various methods they are an extremely useful source of information covering many aspects of general fluorimetric analysis including methods for organic analysis, and those involving excitation with X-rays and cathode rays. A review in Esperanto5 includes a description of the fundamental theories of luminescence and examines some of the methods available for the determination of trace amounts of several cations.A somewhat different type of reviewe containing approximately 100 references discusses the advantages and applications of the various types of methods avail-able. The methods include those based on the use of metallofluorescent com-pounds fluorescent ternary complexes fluorescent indicators in titrimetric analy-sis catalysis crystallophosphors and luminescence in frozen solutions. A literature survey was published in 1968,’ but few details are at present available. Two very useful publications have been ‘Volumes I and 11 published in 1967 and 1970 respectively of ‘A Guide to Fluorescence Literature.’a There have been several books and shorter articles published during this period.Some of these deal mainly with luminescence in the solid state or with luminescence applied to biological systems or solely with organic substances. Of these only two ‘Handbook of Fluorescence Spectra of Aromatic Molecules’ by Berlmans and ‘Phosphorimetry ; the Application of Phosphorescence to the Analysis of Organic Compounds’ by ZanderlO are of immediate interest and use. The proceedings of several international conference^^^-^^ dealing with luminescence have been published during this period. These contain both reports of plenary lectures or review papers and research papers; some of the latter are discussed in the appropriate sections of this review. There have been several excellent books14-16 dealing with general aspects of luminescence in solution and with such topics as the modern theories of the origin of luminescence and developments in instrumentation and techniques.One of the most comprehensive and detailed books is that by Parkerf*; a good introductory publication is that of Hercules,l5 which deals with the topic in a simple yet instructive manner and is highl PHOTOLUMINESCENCE AND CREMILUMINESCENCE IN INORGANIC ANALYSIS 43 recommended for the beginner. There are several other books which contain one or more chapters dealing with in~trumentationl~-~~ or with the fluorescence of inorganic substances ,18919 although they are mainly concerned with topics outside the scope of this review. The book by Udenfriend19 has an excellent chapter dealing with the determination of the inorganic constituents of biological materials.There have been several articles reviewing some selected aspects of the field. Phillips20 has briefly discussed the basic theory and some aspects of the design of fluorimeters in an article devoted mainly to the use of fluorescent dyes as tracers. D ~ r r ~ l in a paper dealing mainly with basic theory and instrumentation, discusses the use of polarised fluorescence an aspect of the field that has so far received only scant attention. However only one reference is given for the deter-mination of a metal ion (zinc); the remainder of this review deals with organic compounds. Developments in instrumentation and techniques have been discussed by several workers ; the American Instrument Company (Aminco) include such de-velopments in their series Fluorescence which also includes articles-often of a semi-reviewnature-on the applications of luminescence especially fluorimetric methods to chemical analysis.Another instrument firm the Turner Instrument Company has published a series of articles on fluorimetric analysis; one23 reviews the automated fluorimetric procedures others the determination of selected in-organic i o n ~ ~ - ~ O and the use of dyestuffs as tracers.30 Several reviews of the use of specific compounds as fluorimetric reagents have been published ; Capelin and Ingrarnsl have described the uses of the tetracyano platinate(I1) complex ion as a reagent for the detection of various metal ions by the formation of insoluble fluorescent precipitates ; the metals that can be detected include aluminium, cadmium lanthanum lead mercury(1) mercury(I1) silver thallium(IV) yttrium, zinc and zirconium.The use of flavones as spectrophotometric gravimetric and fluorimetric reagents has been reviewed by Katya1F2 and Korkucs has also dis-cussed the chelating properties of flavanoids in luminescent methods of analysis. The fluorescent properties of the complexes formed between resorcyl aldehyde-acetyl-hydrazones and aluminium gallium scandium and zinc have been reviewed by Urner.= Fairly extensive studies on the analytical properties of some basic dyestuffs have been studied% and it has been shown that the properties can be correlated with the structures and stabilities of the ion association complexes formed; methods of improving the extractability of the complexes molecular orbital calculations of the T electron densities and the polarisation of the various electron shells during complex formation are discussed.The use of fluorescence analysis in air pollution research has been reviewed by Sawicki.36 Although his review is concerned mainly with organic pollutants, references are given to methods for the fluorimetric determination of some of the inorganic pollutants. 44 BARK AND WOOD Apparatus and Quantum Yield Standards One of the present disadvantages of spectrofluorimetry as an analytical technique is that spectra recorded on the single beam instruments commonly used are dependent on the photomultiplier response and the spectral distribution of the excitation source.Thus results obtained by using different instruments are not directly comparable quantitatively and this could account for many of the discrepancies in the sensitivity of various methods that have been reported by different workers. As a consequence considerable interest has been shown in the design and operation of instruments that will record both corrected excitation and corrected emission spectra. The adjustment of energy compensated spectrofluori-meters some of which are now commercially available has been discussed.37 Attachments for the Aminco-Bowman spectrophotofluorimeter which enable cor-rected spectral recording are now available and the design and use of these have been the subjects of two a r t i c l e ~ . ~ * ~ ~ ~ A fully compensated instrument that can be used for the measurement of the luminescence of solutions frozen solutions and solid matrices has been designed by Cravitt and Van Duuren40; the sensitivity of this instrument and examples of its various applications are reported.Similarly, Cundall and Evans41 have reported a fully compensated instrument that can be used as a spectrofluorimeter or phosphorimeter ; results obtained by using this instrument are given. Perkin-Elmer - H i t a ~ h i ~ ~ are marketing three new instru-ments the model MPF-2A is a recording spectrofluorimeter which contains a built-in correction for the spectral distribution of the source; two simpler instru-ments the models 203 and 204 (a recording version of the 203) both use two grating monochromators and can be used with either mercury or xenon arc sources.A similar instrument to the model 203 the Aminco-Bowman SP125 is also available.43 Cher14~ has published a detailed paper on the use of the Aminco-Bowman spectrofluorimeters SPFl and SPF2; in it methods for the calibration of the source and the photomultiplier are compared. This paper should provide a useful reference text for many users of these instruments. Ps~onicki*~ has described a spectrofluorimeter for the automatic recording of fluorescence spectra in the range 220 to 1300 nm; a full-scale deflection of the pen recorder is caused either by a solution of quinine sulphate at a Concentration of 50 ng ml-l or by fluorescein at a concentration of 1 ng ml-l. Two instruments for the measurement of the luminescence of small samples have been reported; Eisingefi6 has described in detail the design and applicatjon of a variable tempera-ture instrument that has been used for recording spectra between 80" and 370 OK.The optical system of a new microspectrofluorimeter has been described.47 An air driven mixer for use in stopped-flow fluorimetry which can be used with a variety of commercially available instruments has been reported48 ; examples of the study of biochemical reactions by using this equipment are given. A stopped-flow apparatus has been described and examples of its use in the study of both fast reactions and chemiluminescence reported.4g The use of flowthroug PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 45 cells with the Aminco-Bowman spectrofluorimeter has been described in an articles0 that contains a design for a micro-flow cell and also descriptions of its use.An instrument in which the emitted radiation is focused directly on to the photomultiplier has been reported.51 This apparatus is reported to be able to record luminescence absorption and the degree of polarisation in the visible and in the ultraviolet bands of a spectrum. The construction of a spectrofluorimeter in which the sensitivity of the detection system is improved by the use of a specially selected photomultiplier tube which is then cooled and a ‘lock-in amplifier’ or ‘time averaging computer’ has been rep0rted.5~ A stabiliser for the xenon arc used is also described. The emission spectra of xenon and krypton arcs have been compared53 by measuring the radiation with a thermopile the total efficiencies were 55 and 35 per cent.respectively. Ness and Hercules54 have devised a spectrograph that uses an image intensifier for the observation of weak light sources; examples of the use of the instrument include the observation of both weak chemi- and electrogenerated luminescences. An ultraviolet lamp mounted in the vicinity of the coils of a short wave generator (64.4 MHz) has been used as an excitation source.55 The high frequency eliminates the background noise and increases the intensity of some of the emission lines; the 254-nm line is intensiiied and 85 per cent. of the over-all energy lies in the 230 to 605-nm band. Several attachments for the conversion of optical instruments to enable fluorescence to be measured have been reported.A fluorimeter attachment (SP860) for the Unicam spectrophotometer is available and some of its applications to biochemistry have been rep0rted.~6 A similar modification for the H.760 Spekker has been described.57 This attachment allows the fluorescence of a 1 ng ml-l of quinine sulphate to be measured. A cell housing for the simultaneous measurement of optical density light scattering and fluorescence has been devised for the Schafer - Phoenix universal spectrophotometer and is claimed to be applic-able to other fl~orimeters.~~ A fluorescence polarisation attachment is now also available for this in~trurnent.~~ Brook and WhiteheadGo have developed a simple reflectance fluorimeter attachment for the Techtron AA3 atomic absorption spectrophotometer and have used it for the determination of uranium in sodium fluoride beads.A Soviet fluorimeter which uses an incandescent lamp rather than a mercury arc is reported to have an increased stability and to cause less photodecomposi-tion? The emission is detected by a photomultiplier tube which is sensitive in the range 300 to 820nm. This instrument is reported to permit the detection of 10-lOgml-l of Rhodamine C and 10-7gml-l of adrenaline. The sensitivity of this instrument has been comparedG2 with that of a Zeiss instrument and three other Soviet models by measuring the apparent fluorescent yield of quinine sulphate, fluorescein Rhodamine S and Rhodamine 62. The results of this investigation are also reported in a review article.6 Winkelman and Grossman63 have used a solid sample attachment for the Aminco-Bowman spectrofluorimeter in the quantitative analysis of opaqu 46 BARK AND WOOD solutions.Methods based on this technique are reported for the determination of tetraphenylporphinesulphonate tetracycline and fluorescein isothiocyanate. An instrument that can be used to study phosphorescence fluorescence and ab-sorbance changes in turbid biological materials has also been described64 and the results obtained by using this instrument in several different modes of operation are reported. The principles and methods of the measurement of the polarisation of fluores-cence excitation spectra have been reviewed65; the construction of apparatus and the procedures are described in detail with Rhodamine B as an example.A unique idea for extending the response of a photomultiplier tube into the ultraviolet region involves placing a cell containing a 2-mm layer of a liquid phosphor in front of the tube; the details for the manufacture of such a cell are given.66 The correction of instrument response time in the measurement of fluorescent lifetimes has been discussed67 and a simple flash unit for the study of transient fluorescent species has been designed.68 The application of this instrument is illustrated by an investigation of the fluorescence of the pyrene eximers. Although a potentially useful technique phosphorimetry has received only little attention in recent years; this lack of attention is probably caused by the inherent practical difficulties in phosphorimetry and the generally poor repro-ducibility obtained when using commercially available instrumentation.Recently, however several instruments and improvements in technique have been reported. The design and construction of a single disc phosphorimeter has been r e p ~ r t e d . ~ ~ * ~ The unique sample cell of this instrument uses a quartz light pipe and its sensitivity is compared with the commercially available Aminco-Bowman phosphorimeter. The resolution with the single disc phosphoroscope is discussed and the deter-mination of the components of several binary mixtures is reported; a study of the effect of variations in the absorption path length on the observed phosphores-cence intensity is included. Three papers of a somewhat more general interest described methods of improving the reproducibility and accuracy of phosphorimetry.Hollifield and Winefordner71 have designed a rotating cell for the Aminco-Bowman. The use of this cell averages out any optical inhomogeneities and minimises positioning errors thus reducing the standard deviation by a factor of up to 10. A modified version of this equipment utilises an n.m.r. cell spinner.72 This latter paper also describes methods of further improving the precision and accuracy of phosphori-metry in quantitative analysis and at the same time reviews many papers pub-lished on the use of phosphorimetry. The errors caused by the background emission from commercial cells have been reported to be minimised by using a simple +clean-up ~r0cedux-e.~~ Two papers have been published on the use of ‘time resolved phosphori-metry’ ; WinefordnerT4 has discussed the principles of and the instrumentation used in this type of analysis and the merits of mechanical phosphoroscopes and pulse excitation systems are further discussed in detail.St. John76 has studied the use of this technique and has derived an expression for the signal-to-nois PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 47 ratio in phosphorimetry. (This work has been further reported in conjunction with McCarthy and Winef~rdner.~~) The theoretical study of signal-to-noise ratio is reported and equations are derived relating both the experimental and the spectral parameters to the photo-detector signal signal-to-noise ratio and an analytically useful monochromator slit width.Calculated detection limits for several organic compounds compare well with the values obtained experimentally. As part of a series of papers on the selection of optimum conditions for spectrochemical methods the sensitivity of absorption fluorescence and phos-phorescence in the condensed phase has been discussed.77 In a similar ~ e r i e s ~ ~ - ~ is discussed the different effects on the sensitivity of fluorimetric analysis of using excitation from a fairly wide spectral band and excitation by using a mono-chromatic source. The automatic correction of the various spectra obtained and the resolution of the various types of emission spectra has occupied several workers; the use of modern electronic equipment has made this correction feasible.The correction of spectra by the use of digital computers has been reported,81 the ALGOL programme given is suggested to be readily adaptable for individual requirements. A method and related equipment that can provide both absolute emission and excition spectra and also the quantum efficiency of luminescence has been described.s2 This system can also be used to measure the half-lives of both phosphorescence and delayed fluorescence it is also possible to resolve overlapping fluorescence delayed fluorescence and phosphorescence spectra. Absolute values of quantum yields have been determined for only a few compounds ; the most convenient method for their determination is a direct comparison with a compound of known quantum yield it is thus essentid that the quantum yields of the accepted standard substances are accurately deter-mined.Testas3 has reviewed the standards available and includes as part of his article those methods that are currently available for the calibration of the various types of spectrofluorimeters. The compounds reported as standards are quinine bisulphate 9,10-&phenylanthracene anthracene D and L tryptophan and 2-aminopyridine. The wavelength regions over which these compounds can be used their quantum yields and the corrected excitation wavelengths are tabulated. Although the most widely used of these compounds is quinine bisul-phate recent publications on its use have led to some considerable doubt as to its suitability as a standard as the quantum yield has been reported to be de-pendent on the wavelength of e x c i t a t i ~ n ~ ~ ~ ~ and at least two values of the quantum efficiency ( 0 ~ 5 5 ~ ~ and 0 ~ 4 6 ~ ~ ) have been reported; these discrepancies are also discussed by other workers.= Gilla8 has reported the advantages and disadvantages of quinine bisulphate as a primary standard but he reports that the quantum yield isindependent of excitation wavelength within &5 per cent.over the range 200 to 400 nm for concentrations of from 10-2 to 10-6 M. In an attempt to explain the discrepancies in the literature Fletcher89 has examined samples of quinine bisulphate from various sources. However with the exception of one sample the relative difference in quantum yield between all of these samples was only 2.2 pe 48 BARK AND WOOD cent. No dependence of the quantum yield on excitation wavelength was shown in the range 240 to 400 nm and it is suggested that some conflicting results may be caused by the practice of using different spectral band widths for absorbance and excitation measurements.The fluorescence quantum yields for eighteen compounds have been reported,90 and these were determined by using a modi-fication of the Weber and Teale method.g1 Fluorescein and anthracene are reported to be the best standards on the basis of agreement with previous work the use of quinine bisulphate as a standard is reported to be complicated by variations in its fluorescence yield with both the sulphuric acid concentration and the excita-tion wavelength. In view of the importance of accurate measurement of quantum yields and the wide use of quinine bisulphate in their determination it is likely that the controversy over its use as a standard will continue for some time.Several new standards have been proposed; 2-aminopyridineg2 is a useful standard for the correction of spectra and for quantum yield determinations in the range 315 to 480nm. Solutions of 2-aminopyridine are stable for several months. This gives it a considerable advantage over quinine bisulphate which is reported to be absorbed on to glass from non-polar solvents and possibly from sulphuric acid.88 Himmel and Mayerg3 have proposed the use of 5-dimethylamjno-naphthalene-1-sulphonic acid PANS acid) as a standard. The use of this com-pound is compared with that of quinine bisulphate and its advantages are dis-cussed.The reported quantum yield of DANS acid (0.36) agrees with the pre-viously reported figure.94 A standard suitable for the longer wavelength range 490 to 800 nm has been preparedg5 by using 4-benzylidene-5-oxo-2-phenyl-oxazo-line in dioxan. The fluorescence maximumis at m n m and indimethylformamide at 625 nm thus by using this compound in these solvents calibration is possible over the above range. A simple visual method for the approximate determination of fluorescence quantum yields has been rep0rted.~6 This method is based on the measurement of the concentration of a test solution that exhibits the same luminescence as a standard solution (quinine bisulphate; fluorescein or Rhodamine B) and deter-mining the absorbance at the excitation wavelength. The luminescence produced on excitation with an ultraviolet lamp is visually compared to that of one of the standards and its quantum yield calculated from the formula-where Q is the quantum yield; E the molar extinction coefficient ; C the concentration ; x an unknown; and s is the standard.The error for the determination of the quantum yield of various compounds does not exceed +25 per cent. An experiment to determine the quantum yield of fluorescein relative to that of quinine bisulphate has been describeds7 as a suitable laboratory exercise PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 49 The extreme sensitivity of luminescence analysis requires that the solvents used are pure. There exists some confusion regarding the required purity of many of the solvents.For example it has been reportedO8 that tap water is adequate and satisfactory for the majority of analyses. While this may be so for organic analyses as tap water generally contains few impurities that fluoresce for in-organic analysis tap water is not sufficiently pure. Many of the trace inorganic impurities especially iron quench luminescence and thus all traces of inorganic impurities should be absent for inorganic work. The best water for use in fluori-metry is probably that produced by triple distillation from glass apparatus; however this procedure is expensive and a combination of an active carbon unit for removal of organic material a resin de-ioniser followed by a 1-pm filter has been suggested as a combination capable of giving water acceptable for most analyses.The use of de-ionisers alone does not produce water of sufficient purity because of organic contamination from the resins used. The purification of organic solvents has been reported by Fisher and CooperJo9 who recommend distillation in the presence of a small amount (0.5 to 1.0 per cent.) of a long chain hydrocarbon or paraffin wax; the fluorescence intensities of some solvents before and after distillation under these conditions are given. According to this paper methanol ethanol and pentanol are best distilled in the presence of cetanol. The solvents which can be used to form optical glass at 77 OK and hence which may be used in phosphorimetry were reviewed in papers published in 1962 and 1963,100-102 and very little more of significance has been published since that time.McCarthy andDunlaplo3have reportedtheuse of solvents that contain chlorine, bromine or iodine. These solvents have an effect on both the phosphorescence and fluorescence intensities of organic ligands dissolved in them. The presence of chlorine or bromine in the solvent generally results in an enhancement in the phosphorescence and causes slightly longer life times than are found with solvent systems containing iodine. Wood and Barklo4 have recently reported the use of a solvent system con-taining up to 10 per cent. v/v of water and indicate that such systems may extend the use of phosphorimetry for the determination of trace amounts of inorganic ions in aqueous solution. There is little doubt that this is a field of work requiring more attention and investigation.The combination of improved solvent systems and instrumenta-tion will do much to increase the potential and use of phosphorimetry. The Determination of Metal Ions as Inorganic Complexes Recently considerable interest has been shown in the luminescence of metal ions with either outer electron shells of the type ndlo (n + l)s2 e.g. T1+ Pb2+, Bi3+ SnZ+ In+ Sb3+ Te(IV) As3+ Se(1V) or of the type ndlo e.g. T13+ Ins+, Sn4+ Sb(V) Ge4+ As(V) Se(VI) Cu+ Au+ Ag+ in frozen hydrohalic acid solutions. The original work on these luminescing halide complex ions wa 50 BARK AND WOOD published in 1960 by Belyi and co-workers since when several papers of con-siderable interest to the analytical chemist have been published by the above worker by Bozhevol'nov and his co-workers and more recently by Kirkbright, Saw and West.Belyi,lo5 in a paper containing thirty references has reviewed and reported the spectral properties of these ions in frozen hydrochloric and hydrobromic acid solutions. The ions characterised by the configuration ndlO(n + l)s2 exhibit luminescence when present in dilute solution to ~ O - * M ) whereas the ions characterised by the ndlO configuration exhibit luminescence in relatively concen-trated solution (0.5 to 1.0 M). The origins of the luminescence from both types of ion are discussed in detail and a process in which the active absorption is that of halide ion is proposed. Although of considerable interest this review does not discuss the possible analytical applications of this luminescence.However a paperlo6 published in the same year describes the possibility of determining ions with the outer con-figuration ndlO(n + l)s2. Some theoeretical aspects of the luminescence emission are reported but the paper is concerned mainly with the determination of metals by luminescence analysis at low temperature and includes reports of methods that use organic reagents as well as those using inorganic complexes. Thallium(I), lead(II) bismuth(II1) and tellurium(1V) have been determined by measuring the luminescence produced when these ions are contained in hydrochloric acid solution (7 M) at 77 OK. The spectral characteristics of these ions and the sensitivity of their determinations are reported and because of their widely separated spectra they can each be determined when all four are present in solution.The authors report that by measuring the luminescence intensity of the niobium complex of 2,2',4'-trihydroxy-5-chloro-(l 1 '-azobenzene)-3- sulphonic acid in solid solution a 100-fold increase in sensitivity is obtained and that the sensitivity of the deter-mination of magnesium with 2-hydroxy-3-sulpho-5-chlorobenzene is increased 20-fold over that of the colorimetric determination by measuring the luminescence at liquid nitrogen temperatures. The possibilities of using phosphorirnetry and crystal phosphors for the analysis of trace amounts of metal ions are also discussed and several examples of the use of these techniques are given. One of the main difficulties of using low-temperature luminescence as an analytical technique is the limitation in the number and types of solvent that form optical glasses on freezing.When using the usual commercially available instruments employing 90" observation of the luminescence the formation of an optical glass is essential because the formation of snows or cracked glasses will obviously result in very variable readings. Several organic solvents or solvent mixtures are known to meet this requirement but few inorganic solvents have been reported to form the required glass. However in recent papers by Kirk-bright Saw and West dealing with the fluoresence of metal ions in hydrochloricl07 and hydrobromic acidslog it is reported that when using thick-walled silica tubing, several inorganic solvents will consistently produce clear glasses.Concentrated hydrochloric hydrobromic sulphuric nitric phosphoric and perchloric acids al PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 51 produce clear glasses at 77 OK whereas concentrated acetic formic boric and oxalic acids invariably form snows that axe white and opaque and hydriodic acid usually forms a yellow extensively cracked glass which does not permit its use. Hydrochloric acid forms optical glasses at concentrations above 5 M and a general studylo' of the luminescence of fifty-five metal ions (2 x M) in hydrochloric acid (6 M) at 77 OK showed that only ten ionic species Sb(III) Sb(V) Bi(III), Ce(III) Cu(I) Pb(II) Te(IV) Tl(I) Sn(1V) and U(V1) fluoresced under these conditions. As the intensities of the solutions of antimony0 and copper(1) were low as observed by Belyi,lo6 and the fluorescence emission of U(V1) in solution and in boric acid beads is well characterised the investigation of the analytical possibilities of this technique was restricted to the remaining seven ions.The uncorrected spectra the effect of hydrochloric acid concentration the sensitivity of the determinations and the effect of standing in various lighting conditions are reported. Continuous irradiation of thallium(1) solutions results in a decrease in intensity which the authors suggest is caused by oxidation to thallium(II1). The optimum hydrochloric acid concentration the wavelengths of maximum emission and excitation and the detection limits are summarised in Table I. No attempt was made by the authors to determine the upper range of any suggested or proposed methods.TABLE I INORGANIC COMPLEXES IN HYDROCHLORIC ACID AT 77°K Hydro- Excitation Emission chloric acid maximum maximum Absolute concen- (uncor- (uncor- detection tration rected) rected) Concentration range of limit, Ion M nm nm calibration curve* pg ml-1 Sb (I1 I) Bi (111) Ce (111) Pb S W ) Te(IV) 7 NR 6 7 7 10 10 7 306 390 330 252 276 326 380 256 272 302 582 580 410 348 390 550 586 380 390 494 strong 514 strong 540 weak 565 weak lo-' - M NR 10-8 - 6 x 10-8 M 10-7 - 10-8 M 10-8 - 10-7 M lo-' - lo-' M lo-' - 8 X lo-' M 10-4 - 10-3 M NR 0.12 NR 0.002 0.014 0.002 0.012 0.2 NR 12 * Concentration range over which linearity is obtained by using a single instrument NR = Not reported.sensitivity setting 52 BARK AND WOOD TABLE I1 INORGANIC COMPLEXES IN HYDROBROMIC ACID AT 77 OK Hydro- Excitation Emission bromic acid maximum maximum Absolute concen- (uncor- (uncor- detection tration rected) rected) Concentration range of limit, Ion M nm nm calibration curve* pg ml-1 S b( I1 I) As (I 11) As (IT) Bi( I I I) Ce( 111) Te(1V) Sn(I1) SbW1 $(I) TV) u (VI) 6 6 6 6 6 6 6 6 6 6 6 6 (in 8 M HBr) 360 360 356 356 378 250 286 304 352 270 314 327 586 586 584 566 350 350 434 414 560 410 550 494 516 strong 540 565 weak 10-8 - 10-7 M 10-7 - 10-6 M 10-5 - 10-4 M 2 x 10-6- 10-4 M 8 x 10-7 - 1 0 - 5 ~ 2 x 10-8 - 4 x 10-7 M 6 X - 3 X lo-' M --2 X lo-' - 10-'M 2 X lo-' - M 6~ 10-7-4 X 10-'M 0.0012 0,012 0.75 1.50 0.012 0.112 0-004 -0.40 0-24 0.14 * Concentration range over which linearity is obtained by using a single instrument sensitivity setting.Table I1 summarises the results obtained from a similar investigation of the luminescent propertiesof fifty-eight ionsin hydrobromic acid at 77 *K.1°8 Of the ions examined only antimony (111) and (V) arsenic(III) bismuth(II1) cerium(II1) , lead(II) thallium(1) tin(I1) and uranium(V1) showed a strong fluorescence. Again, the only ion that shows a significant decrease in intensity on irradiation is thallium(I) and the authors again suggest that this is caused by oxidation to thallium(II1).With the exception of tin(1V) all of the elements that exhibit fluorescence in hydrochloric acid do so in hydrobromic acid; however arsenic (111) and (V) which do not fluoresce in hydrochloric acid do so in hydrobromic acid. In hydrobromic acid tin(I1) can be determined down to 0.12 pg whereas in hydrochloric acid the detection limit for tin [as tin(IV)] is 6pg. Similarly, the methods for the determination of antimony (V) and (111) are more sensitive when hydrobromic acid is used. The use of hydrochloric acid gives a higher sensitivity for the determination of bismuth(III) cerium(II1) lead(II) thallium(1) and tellurium(1V). The authors report that because of the wide separation between the fluorescence emission maxima of several of the elements it is possible to determine some of them simultaneously by a suitable choice of excitation and emission wavelengths.Several authors have reported methods for the determination of individual ions by using this technique. Bozhevol'nov and Solov'evlog have also discussed the possibilities of the determination of trace impurities by fluorescence in frozen solutions of hydrohalic acids and have again reported the possibility of determinin PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 53 thallium(I) bismuth(III) lead (11) and tellurium(1V) in mixtures. Their paper also contains a description of a method used for the determination of antimony in germanium tetrachloride ; the antimony is extracted with hydrochloric acid and after the addition of nitric acid the extract is evaporated to dryness.The residue is dissolved in a small amount of hydrobromic acid and the fluorescence intensity of the resulting solution is measured at 77 OK. The sensitivity is reported to be 0.3 ng m1-1 of antimony; however because only a small volume (0.1 ml) is required for the measurement the sensitivity can be considered to be 0*03ng. The method is highly selective and the interference caused by iron(II1) can be overcome by using the method of standard addition. A method for the determination of arsenic impurities based on the lumines-cence of both arsenic(II1) and arsenicw) in frozen hydrobromic acid solutions has been reported.110 The method used for arsenic(II1) gives the higher sensitivity, 7.5 ng ml-l which is 100 times more sensitive than that reported by Kirkbright et al.(see Table 11) for this ion. When using hydrochloric acid in which according to Kirkbright et aZ.,10* these ions do not form fluorescent species the sensitivity is reported to be lower than in hydrobromic acid because the excitation and emission spectra are broader in hydrochloric acid and also the emissions from these ions lie in the same region as do those of hydrochloric acid itself. These workerslf0 have also described methods for the determination of thaKum(1) and lead(I1)llf and for tin (11) and (IV) indium(III) gallium(III) germanium(II1) and telluriurn(IV).l12 The latter reference includes a discussion of the origins of the luminescence emissions of these metal ions. The fluorescence emission from antimony(II1) in frozen hydrobromic acid (6 M) solution has been used for its determination.l13 The sensitivity is reported as 5 ng and the method is very selective; only iron(II1) and tellurium(1V) interfere when present in amounts exceeding 50 and 20-fold excess respectively.To obtain the maximum sensitivity antimony must be present as antimony(II1) because antimony(V) which exhibits similar spectral properties gives a much lower fluorescence intensity. A method for the determination of antimony(III) bis-muth(II1) and selenium(1V) in mixtures with either hydrochloric or hydrobromic acid has been reported.l14 The dependence of the fluorescence intensity (measured at 77 OK) on the concentration of the ions is linear up to M however no details of sensitivity or interferences are reported in the abstract available.Thallium(1) halides exhibit fluorescence at room temperature as well as in frozen solution and the conditions for the production of the maximum fluorescence at both room temperature and 77 OK have been investigated.ll6 At room tem-perature the maximum fluorescence is observed in a saturated solution of sodium bromide whereas at 77 OK the maximum intensity is observed in hydrobromic acid (7 M); these observations appear to contradict those of Kirkbright et al. (see Tables I and 11) who report the highest sensitivity in hydrochloric acid (10 M) at 77 OK. The luminescence of thallium(1) in hydrochloric acid (2 M) saturated with sodium chloride at room temperature has been used for its determination in alkali-halide single crystals.lf6 An illustrative example is the determination o 54 BARK AND WOOD thallium(1) in sodium iodide; the sample is decomposed by heating with con-centrated nitric acid and then evaporating to dryness with concentrated hydro-chloric acid.Thallium(II1) is reduced to thallium(1) with hydrazine and the excess hydrazine is removed by sublimation (as hydrazine hydrochloride). The residue is then dissolved in hydrochloric acid (2 M) saturated with sodium chloride and the fluorescence intensity measured. The method is reported to be applicable to the determination of down to 0-05pgml-l of thallium in the final solution. A similar procedure with 3-M hydrochloric acid saturated with sodium chloride has been reportedll' as being sensitive to 0.01 pg of thallium and these workers reported that there is some interference from Sb(III) Cu(I) Fe(III) Pb(II) Hg(I), Sn(I1) and sulphites.Kirkbright Saw and West 11* have determined tellurium(1V) by the measure-ment of its fluorescence jn frozen hydrochloric acid (9 M) solution. The excitation and emission spectra vary with the acid concentration and are reported for con-centrations between 6 and 10 M. As the complex formed at high concentrations of the acid is much more fluorescent than that formed at relatively low concen-trations the determination was done in 9 M hydrochloric acid. Of the fifty ions examined at 50-fold weight excess only iron(II1) and iodide which produce yellow solutions and tin(II) which reduces tellurium(1V) to the tellurium (0) interfere.The interference from tin(I1) can be removed by prior oxidation of the ion to tin(1V) and the tellurium can be extracted as its thioglycollate into ethyl acetate or as its diethyldithiocarbamate into carbon tetrachloride thus eliminating the interference from iron(III) which cannot be removed by reduction and is not extracted. The method which is sensitive to 10 ng of tellurium with a standard deviation of 2-2 per cent. is suggested as a possible method for the determination of tellurium in lead and the results of a feasibility study are presented. A method for the determination of lead(I1) by using the violet fluorescence produced by this ion in a solution of hydrochloric acid and potassium chloride at room temperature has been reported.llQ The optimum concentrations of hydro-chloric acid and potassium chloride are 3.3 and 0.8 M respectively; because the intensity of the fluorescence decreases on standing it is essential to measure its intensity within 15 minutes of preparation of the solution.The effect of thirty-one ionic species each in 50 molar excess is reported; bismuth(II1) chromium(VI), copper(II) iron(III) molybdenum(VI) thallium(I) vanadium(V) ascorbate (from ascorbic acid) and metabisulphite are shown to interfere. Procedures for the removal of these interferences are described. The method is reported to be applicable to the determination of lead in the range 10 to 60pg and the results obtained for the analysis of synthetic solutions are reported. The characteristic fluorescence of cerium(II1) in dilute hydrochloric acid has been used for its determination in rare earth mixtures.120 The interference levels for various ions are reported and iron(III) which interferes when present in large amounts can be removed by reduction with hydroxylamine this also reduces any cerium(1V) to cerium(II1).The method is applicable in the range 10-1 to 10-5 per cent. of cerium. Cukor and Weberling121 have proposed the use o PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 55 perchloric acid as a solvent for the determination of cerium(II1). These authors report that when using perchloric acid only iron(II1) and nitrates interfere. The method has been applied to the determination of cerium in yttrium oxide samples, the sample is dissolved in a perchloric acid solution and any cerium(1V) reduced to cerium(II1) by using titanium(II1) sulphate and the fluorescence intensity of the solution then measured.The advantage of using titanium(II1) as a reducing agent lies in its ability to reduce iron(II1) to iron(I1) in perchloric acid solution, thus removing the interference caused by iron(II1). The calibration curve is linear over the range 0.1 to 150 pg of cerium per gram of yttrium oxide and the accuracy is reported as &lo per cent. The results obtained with this method are compared with those obtained from mass spectrometry and spectrophotometry with thenoyl-trifluoracetone as a complexing agent. A further development has been the design of an instrument specifically for the determination of cerium(III).l22 This instrument uses the 265.2-nm line of a mercury lamp as the exciting radiation and the luminescence is measured by using a photomultiplier which is positioned to observe the luminescence from the front surface of the cell.In the same paper a method for the determination of cerium(II1) by using its fluorescence in hydrochloric acid solution is reported. This method is sensitive to 1 x The fluorescence exhibited by uranium(V1) in a frozen solution of tributyl phosphate in synthine (hydrogenated kerosene) has been used for its determina-t i ~ n . ~ ~ ~ The metal is extracted from 7.8 M sodium nitrate solution with a 20 per cent. solution of tributyl phosphate in synthine and the luminescence of the frozen extract is measured by using frontal observation. This technique enables uranium to be determined in the range 1 x These determinations are sensitive and because only a few ions exhibit luminescence under these conditions they are also selective and should find considerable application in the analysis of complex materials.There are however, disadvantages. These are caused mainly by the need to have liquid nitrogen available so as to freeze the solutions and the need for the precise optical arrange-ment of low-temperature accessories which in commercially available instruments , is not particularly reproducible although a recent development the use of a rotating cellJ71J2 should improve the reproducibility of these low-temperature determinations. per cent. of cerium in rare earth samples. to 3 x g ml-l. Crystal Phosphors Relatively few methods have been reported for the determination of para-magnetic metal ions in solution.However their use as activators in crystal phosphors has led to some very sensitive methods for their determination. Holzbecher et a1.W have reviewed various methods for the determination of copper iron and nickel in zinc sulphide or cadmium sulphide and in the raw materials used for their preparation. Copper can be determined by using a zinc sulphide based phosphor in the range '7 x to 1 x per cent. Cobalt, nickel manganese and iron in amounts greater than per cent. interfere 56 BARK AND WOOD Cobalt and nickel (2 x per cent.) can be determined similarly; however iron copper and manganese in amounts greater than 10-6 per cent. cause interference. A similar paper1% reports a systematic investigation of the luminescence of crystal phosphors produced by irradiation with ultraviolet or X radiations.The luminescence resulting from the activation of various carriers by metal ions were investigated; twenty-six carrier substances were investigated and fifty-two metal ions were incorporated in various ways. The possible analytical applications including the detection of gold by using sodium iodide potassium iodide or rubidium iodide as a carrier and the determination of copper nickel or cobalt in a zinc sulphide based phosphor are described. A further reviewl26 reports the luminescence of inorganic materials both with and without activators. Three methods for the determination of manganese based on its use as an activator in crystal phosphors have been published during the period.Bozhevol'nov and Fakeeva127 used a lithium - magnesium tungstate phosphor which is activated by manganese and allows the determination of 1 x 10-8 per cent. of manganese in water or hydrochloric acid. Only those ions that do not interfere are reported. Manganese can also be determined in the range 10 pg to 1 pg by using an antimony tetroxide based phosphor.12* Apparently no foreign ions when present in amounts below 1 ng interfere in the determination of g of manganese. The details of the preparation of the phosphors by using specially prepared antimony tetroxide is described in detail. The most recently published rnethodlz9 also uses an antimony tetroxide based phosphor and is reported to be sensitive to g of manganese.The sample is prepared in a quartz cell containing antimony tetroxide or antimony tetroxide containing 10 mole per cent. of boric oxide the addition of which is reported to increase both the sensitivity and reproducibility of the method. The charge, containing a standard or unknown manganese solution (0-005 ml) is dried under an infrared lamp and then calcined at 1080 "C for 10 minutes; an orange - red luminescence results on irradiation of the phosphor. This luminescence is measured photoelectrically and the error depends on the amount of manganese present; at 10-6 to 10-7 g it is 20 per cent. whereas at 10-lo g if is approximately 90 per cent. Steele and Robert130 have described a method for the determination of uranium in ores and solutions involving the preparation of a sodium fluoride based phosphor.The uranium is pre-extracted as uranyl nitrate in the presence of aluminium nitrate as a salting out agent by using ethyl acetate. A portion of the organic layer is added to a pellet of sodium fluoride in a platinum dish and the mixture is fused. The fluorescence intensity of the resulting bead is measured on a fluorimeter and compared with the intensity of standard samples. For the analysis of solutions the lower limit is 3011gml-l~ whereas for ores the limit is 0.001 per cent. No details of any interferences or the decomposition ores are avail-able. Tarantsova and Nikol'skaya131 have determined uranium in phosphate rocks by using a similar method. However these authors used a co-precipitation tech-nique to concentrate and separate the uranium.The extraction of the uranium from the rock samples and subsequent co-precipitation with zirconium phosphate to 5 PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 57 are reported. The precipitate is ignited at 800 "C for 1 hour and then fused with sodium fluoride. No sensitivity limits are reported but the reported relative error is 9.7 per cent. Various factors affecting the sensitivity of this method have been investigated by Samsoni.la2 The optimum conditions are described the basic bead material is a (30 + 70) mixture of sodium fluoride and sodium potassium car-bonate which to prepare the bead requires heating at 650°C for 10 minutes followed by cooling over a period of 2.5 hours. The authors report that this method is more reproducible (&lo per cent.) than the previously reported methods that use only sodium fluoride beads (525 per cent.).The use of this bead mixture means that a lower temperature is required for the fusion and thus there is less interference from the platinum crucibles1= used in the method but the sensitivity is somewhat reduced. A similar method1% with a bead material consisting of 20 per cent. of sodium fluoride and 80 per cent. of sodium carbonate has been used to determine the uranium contents of thorium nitrate. Thorium quenches the fluorescence of the phosphors and therefore it is essential that the uranium is separated from the major component. The dissolved sample is treated with aluminium nitrate solution and then the uranium and some of the thorium are extracted into ethyl acetate.The thorium is removed from the organic phase by washing with a solution of EDTA and aluminium nitrate. Four separate washings are required to remove all the thorium. The resulting ethyl acetate is evaporated to dryness in the presence of nitric acid and after dissolution is fused with the bead material at 850 "C. The fluorescence intensity which is propor-tional to the uranium content over the range 10-5to 10-*per cent. is then measured. A standard addition method was used to obtain the accuracy of the method and the authors report a standard deviation of 2.67 per cent. A further paper1% on the use of this type of method has been published. The fluorescence of sodium fluoride based beads is used to determine uranium in soil stream sediments and water ; the method used is a direct one requiring no prior separation of the uranium from associated elements.The extraction of the metal ions from these materials and the procedures used are both reported in detail. Kleber136 has reported a fluorimetric spot test for uranium; a neutral solution if the uranyl ion is added to a polyphosphate Na5P3010 or to a (1 + 1) Na,P3010 -CaSO4-2H,0 mixture or to CaPO3F*2H,O and then irradiated with ultraviolet light. The presence of uranium is indicated by a green fluorescence. Silver(1) and thallium(1) interfere with this test. During this period several methods based on crystal phosphor formation have been reported for the determination of the lanthanide elements. These are dis-cussed in the section of this review dealing with the determination of scandium, yttrium and the lanthanides, Alkali metals Since 1967 only two methods have been reported for the determination of an alkali metal.Pitts and Ryanx3' have reported the use of dibenzothiazolyl-methane as a reagent for the fluorimetric determination of lithium. This compoun 58 BARK AND WOOD must be used in dioxan solutions containing 20 per cent. v/v of water because at this concentration of water the fluorescence intensity does not vary with time. At concentrations below 20 per cent. v/v of water the intensity is initially con-siderably higher but decreases rapidly with time to approximately the same intensity as that obtained when using 20 per cent. of water. The authors report that only zinc interferes by giving a fluorescent species but as the pH of the reaction mixture is approximately 11.0 many metals would cause interference by being precipitated as hydroxides and must therefore be removed by precipita-tion and centrifugation or by ion exchange prior to the determination of lithium.Many anions interfere sulphate especially causing a large decrease in the over-all intensity. The method is reported as being able to be used for the determination of 0-25 pg ml-l of lithium in mixed alkali metal salt solutions and to 0.005 pg ml-l in pure lithium salt solutions. Marksman and Strel't~oval~~ have used 8-hydroxyquinoline for the fluori-metric determination of lithium. The solution containing lithium after being neutralised to Congo Red with sodium hydroxide is treated with a solution of 8-hydroxyquinoline.This solution is extracted with chloroform and the fluores-cence intensity of the non-aqueous phase is measured. The method may be used over the range 1 to 20pg of lithium. Although 8-hydroxyquinoline is a non-selective reagent the authors report that only iron(II1) (4 pg) magnesium (2 mg), sodium potassium and calcium (10 to 20mg) interfere. The lower limit of the determination is not reported but the sensitivity seems comparable with that of the method of Pitts and Ryan. Beryllium During the period under review three papers have been published describing new fluorimetric methods for the determination of beryllium. The remaining publications describe applications of the fluorescent beryllium - morin (I) complex to the determination of beryllium in complex materials.OH O (I) Morin Budesinsky and West ,139 prompted by previously reported improvements in the fluorescent and complexing properties of an organic reagent with the intro-duction of an amino-methyl-dicarboxymethyl group and the use of Z-hydroxy-3-methyl naphthoic acid as a reagent for beryllium have prepared and examined the properties of 1 -dicarboxymethylaminomet hyl-2-hydrox y-3-n aphthoic acid PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 59 The enhancement of fluorescence of this compound consequent upon its chelation with beryllium led to the development of a fluorimetric method for the deter-mination of the metal. The factors affecting the fluorescence of the chelate are discussed the optimum pH is 6.8 and is controlled by the addition of hexamine -perchloric acid - perchlorate buffer solutions.The development time is 20 minutes and no change in intensity is observed during a period between 20 and 60 minutes. Calcium-EDTA is added as a masking agent and the interference levels are reported for solutions containing this masking agent. Aluminium magnesium, scandium arsenate citrate fluoride oxalate phosphate and tartrate interfere when present in relatively large amounts. A relative standard deviation of 2 per cent. is reported. Job’s method was used to establish the composition of the chelate which was shown to have a mole ratio of 1 1 two complexes with this mole ratio are formulated. The method is reported as being sensitive to 0.09 pg and usable in the range 0.09 to 1.8 pg of beryllium.The hydrolysis of umbelliferone phosphate which is catalysed by alkaline phosphatase results in the formation of the fluorescent umbelliferone. A kinetic method has been developed140 for the determination of small amounts of beryllium based on the inhibition of this hydrolysis by beryllium. The rates of change of fluorescence of a blank and a sample containing beryllium are determined and the percentage inhibition calculated from a given empirical formula. A calibration curve of the percentage inhibition plotted against the concentration of beryllium is used to obtain the beryllium concentration in the sample. The reported range for the determination is 0.01 to 3eOpgrnl-l. Ions that interfere are those of aluminium cadmium copper dichromate fluoride lead mercury silver and zinc, but only when present in large excess.Bismuth has also been determined by this method and thus will obviously interfere in the determination of beryllium. Bozhevol’nov and S010v’ev~~~ have reported what is probably the first phos-phorimetric determination of a metal. The beryllium complex of dibenzoylmethane is extracted into carbon tetrachloride the extract is frozen in liquid nitrogen and its phosphorescence intensity measured. The method has been applied to the determination of beryllium in waste waters. The sensitivity is reported as 5 pg ml-1 and a list is given of the ions that do not interfere when present in up to 100-fold molar excess. No details of the instrumentation used have been reported in the literature available however because carbon tetrachloride does not form an optical glass on freezing in liquid nitrogenlo2 it would appear that front surface observation of the phosphorescence is used.A previous report142 of this method suggests the use of isopropyl alcohol as the extractant which also does not fomr an optical glass on freezinglo2 and again frontal observation of phosphorescence will be necessary. Several applications of the use of morin for the determination of beryllium in biological and industrial samples have been described. One method for the determination of beryllium in materials of complex composition143 involves a preliminary degradation of the material by fusion with potassium hydrogen fluoride; the beryllium compound is then extracted with butyric acid and the beryllium butyrate so formed is re-extracted with chloroform.and concentrated 60 BARK AND WOOD nitric acid. This extract is evaporated to dryness and after treatment with sulphuric acid the beryllium is reacted with morin at pH 13.0. The method is sensitive to 5 x per cent. beryllium although the amount of sample should not exceed 0-5 g. A procedure for the co-precipitation of beryllium by using tannin and methylene blue followed by recovery and subsequent determination with morin has been reportedlU for the determination of beryllium in urine The method is based on a modification of the Sandell fluorimetric procedure.145 The method is modified by omitting the addition of stannite and using instead triethanolamine to complex heavy metals; the method may be used over the range 0.10 to 1.0 pg of beryllium.In the lower range 0.01 to 0-lOpg the fluorescence intensity of the sample solution decreases fairly rapidly with time. This difficulty is overcome by storage of the solution in an ice-bath. To obtain readings of the intensity, the temperature of a sample in a fluorimeter tube is allowed to rise to 15 -& 1 "C and the intensity measured immediately. By using this method the fluorescence is stable for up to 8 hours. The pH for this determination is 11.6 to 11.9 and the authors report the sensitivity as 0.2 ng ml-l of beryllium in urine with a standard deviation of 10 per cent. The standard deviation for a sample containing 1.0 ng ml-1 is reported to be 9 per cent.Morin has been used146 for the determination of beryllium in ores. These are decomposed by treatment with a mixture of hydrofluoric nitric and sulphuric acids the residue from this treatment is fused with potassium hydrogen fluoride or sodium carbonate and borax and the beryllium from the fused material is dissolved by a solution of EDTA. The pH of this solution is adjusted to 13.0 and a solution of morin is then added. The fluorescence intensity if recorded after 5 minutes. No sensitivity is quoted but procedures for the separation from interfering ions are reported. Titanium phosphate in the presence of EDTA for co-precipitation or solvent extraction of beryllium as its acetylacetone complex (from solutions containing EDTA) with carbon tetrachloride as the extractant are suggested as separation procedures.Mulikovskaya and Sharyhina147 have determined the beryllium content of underground waters. The beryllium is concentrated by co-precipitation with iron(II1) hydroxide followed by absorption on to silica gel from a solution con-taining EDTA and an excess of calcium ions. Following the concentration pro-cedure which is reported in detail the beryllium is determined fluorimetrically by using morin. The preparation of a morin crayon for the determination of beryllium by using the ring-oven technique has been described by West and J~ngreis.1~~ These authors report in detail the procedure for the sepaxation and detection of the beryllium. Beryllium can be deterrnined by visual comparison of the fluorescence of its main complex in the range 0.01 to 0.2 pg.The ions that do not interfere are aluminium thorium gallium indium and scandium which normally form fluorescent compounds with morin but only the effects of the presence of aluminium thorium and gallium have been investigated. These did not interfere in this method PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 61 A review of the methods available for the determination of beryllium up to 1968 has been published by G. K. Turner Associates.% This review includes thirty-three references a brief discussion of the methods both qualitative and quantitative and references to studies of the type of complex formed between beryllium and organic reagents. One reference is of considerable significance in which it is reported14* that there is considerable adsorption of beryllium on to glass and methods proposed to prevent this effect do not appear to have been considered or reported by the authors of recent papers dealing with the deter-mination of beryllium.The use of fluorimetry for the determination of trace amounts of magnesium is apparently well established as a standard method especially in clinical analysis, and the recent literature contains only two descriptions of the use of a new reagent.lsoylsl The remaining papers are devoted to comparisons of previously published fluorimetric methods with other techniques and with each other and the application of these existing methods to the analysis of biological samples. Magnesium The synthesis and use of N,N’-bis ~alicylidene-2~3-diamino benzofuran as a fluorimetric reagent for magnesium is the subject of a paper by Dagnall Smith and West,lso who present a detailed discussion of the development of the method and indicate its application to the determination of the magnesium content of water and plasma samples.The method is carried out at an apparent pH of 10.5 zfi 0-2 which is obtained and controlled by using diethylamine and hydro-chloric acid. (Only apparent pH values can be quoted because the solvent used is 50 per cent. v/v aqueous methanol a mixture in which accurate pH values cannot be measured with conventional equipment.) The masking of ions that interfere either by hydroxide formation or by the formation of coloured or fluorescent chelates is discussed. Some of the results for the magnesium contents of plasma and waters are compared with the values obtained by using flame photo-metry.The method can be used in the range 2 to 100 ng ml-l and in the presence of calcium if EDTA and strontium bromide are used as masking agents. The quantum yield of the magnesium chelate of N,N‘-bis salicylidene-2,S-diamino benzofuran is compared with those of the magnesium chelates of N,N’-bis sali-cylidene o-phenylene-diamine ; N,N’-bis salicylidene ethylene diamine and calcein (11) (fluorescein-2,7-bis methyl iminodiacetic acid). As part of a paper on the quantum efficiencies of some analytically useful chelates,lS1 the ‘sensitivity factor,’ S calculated from the fonnula-where E is the molar extinction coefiicient at the wavelength of excitation; 0 is the quantum efficiency; and H is the half band width of the fluorescence spectrum in nanometers 62 BARK AND WOOD (11) Calcein ( Fluorexone) is reported for the various chelates.Parker and R e e ~ l ~ ~ have previously reported the use of a similar factor with H in wavenumbers units. From the results obtained it has been proposedl53 that N,N’-bis salicylidene ethylene diamine which has previously been regarded as the most sensitive reagent for magnesium may well be equalled in sensitivity by other reagents and that they may show different properties with regard to selectivity thus making them more suitable for a given application. With the exception of calcein the spectral, structural and chelating properties of all the above reagents have been previously compared with one another.The dissociation constants for the magnesium chefates are reported in the same paper. Dale Turnbull and Radle~15~ have also described the preparation and use of N,N’-bis ~alicylidene-2~3-diamino benzofuran and report this reagent as being the most sensitive for the determination of magnesium. They propose a method for the determination of up to 2 x per cent. of magnesium in nickel. The development of separation techniques for the removal of microgram amounts of magnesium from gram amounts of nickel and molybdenum is also reported. Other reagents 8-hydroxyquinoline bis salicylidene ethylene diamine and lumo-magneson (111) have been examined for the determination of magnesium in Lumomagneson substances of complex composition.155 The colour of the fluorescence of the com-plexes in various extractants is tabulated and the conditions for the use of each of these reagents are described.The sensitivities reported are 0.5 pg ml-l, 0.025 ng ml-1 and 20 ng ml-l respectively. Although it is well known that many ions interfere with the fluorescence of the magnesium chelate of 8-hydroxyquinoline PHOTOLUMINESCENCE AND CHEMfLUMfNIlSCENCE fN I&oRGANjfC ANALYSIS 63 a method is proposed for the determination of magnesium in iron ores by using this reagent.155 The authors use hexamethylene tetramine (urotropine) for the precipitation of iron and report that the method is applicable to iron ores con-taining 0-05 per cent. of magnesium. Patrovsky has reported the use of 8-hydroxyquinoline-5-sulphonic acid for the fluorimetric determination of magnesium.156 ~ ~ 5 7 The preparation of the reagent and the experimental conditions for its use are described.The method is done at pH 9.5 and the removal of many interfering ions is reported. Cadmium, cobalt(I1) copper(I1) manganese(I1) molybdenum(V1) nickel(I1) tungs-ten(v1) and zinc may be masked with triethanolamine and hydroxylamine ; calcium which interferes by forming a fluorescent species with the reagent is masked by the addition of a slight excess of 1,2-bis-(2-aminoethoxy) ethane N,N,N’,N’-tetra-acetic acid (EGTA) followed by the removal of the excess EGTA by adding barium chloride solution. The method is applicable over the range 0.1 to 2-Opgml-l and in the presence of 20mg of calcium the coefficient of variation is reported to be 6 per cent.The determination of the magnesium content of biological fluids by rapid and accurate methods is an ever important problem. Two papers have recently been published comparing methods for the determination of magnesium in bio-logical fluids. E n d ~ l ~ ~ has compared the use of atomic absorption spectroscopy and fluorimetry by using 8-hydroxyquinoline-5-sulphonate. Good correlation between these methods is reported for the range 0.19 to 2.7 pg ml-l of magnesium in serum. The average recovery for the fluorimetric method was 101 per cent. The presence of bilirubin which did not affect the results from atomic absorption, lowered the recovery to 73 per cent. Four methods for magnesium in serum have been c0mpared.l5~ The reported methods are (i) colorimetric with Titan yellow; (ii) flame emission spectrophotometric ; (iiz] atomic absorption spectrophoto-metric and (iv) fluorimetric by using the method of Schachter.16O The results obtained for thirty-one normal subjects whose sera were analysed by all four methods are reported.The coefficients of variation obtained from multiple deter-minations on one sample by using the above four methods are 4.0 3-0 1.8 and 2.3. The fluorimetric method was adopted by the authors as being the most suitable for routine analysis because of the simple procedure with few attendant sources of errors. This method gave the narrowest range of values for the normal subjects examined. The reagent lumomagneson has been reported by GusevlG1 for the deter-mination of magnesium in both urine and serum samples.He recommends that standards be determined simultaneously with each determination to eliminate errors caused by temperature variations which considerably influence the fluores-cence intensity of the chelate. The use of calcein for the determination of mag-nesium in sera has been reported.lGa The reagent reacts to form fluorescent complexes with aluminium calcium magnesium and zinc ; interferences are pre-vented by masking with EGTA and BAL (2,3-dimercaptopropanol). The stability constants for calcein - magnesium and calcein - calcium complexes are K = 7. 64 BARK AND WOOD and K = 6.6 respectively the values for the EGTA complexes are 5.4 (Mg) and 10.7 (Ca). Interference by phosphate is also reported but no indication is given of the methods necessary to prevent this interference.who have determined magnesium in biological samples. Calcium and phosphate interfere when present in large amounts the calcium is removed by precipitation with oxalate and the removal of phosphate by an ion-exchange method is proposed. The magnesium contents of samples of sera urine bone and faeces by using the modified procedure are reported. Relevant techniques for the prior treatment of these samples are given in detail and the results obtained with this method compare favourably with those obtained by atomic absorption spectrophotometry. Automated fhorimetric procedures for the analysis of biological samples have been described with 2,2’-dihydroxyazobenzene1@ and 8-hydroxyquinoline-6-sulphonic acid165 as the fluorimetric reagents.Breen and Marshall1@ used 2,2’-dihydroxyazobenzene for their automated method in which a Technicon AutoAnalyzer is used in conjunction with a Turner fluorimeter (model 111); the arrangement of the instrument the flow system and the preparation of solutions are described in detail. The authors report that no preliminary treatment of serum was necessary but that urine samples required acidification and dilution before analysis. The advantages of this method over automated methods previously describedlGs 916‘ are indicated. With this pro-cedure there is no need to remove calcium by precipitation as the oxalate or serum proteins by dialysis hence the need for a dialysis module is eliminated. Only one buffer solution is used instead of the two previously required and the method can be used for smaller samples (0.1 to 0.3 ml).The equipment is capable of analysing forty samples per hour and results obtained from the determination of magnesium levels in the serum and urine of various subjects are reported. The standard deviation is 1 per cent. on triplicate analyses. The influence of gluconate and glucogalacto-gluconate on the fluorimetric determination of magnesium with 2,2’-dihydroxyazobenzene has been investigated.ls8 The presence of these com-pounds at various concentrations up to 500 mg 100 ml-I did not significantly alter the magnesium values obtained. Klein and Oklanderl65 have reported and compared the results of two auto-mated methods for serum magnesium by using 8-hydroxyquinoline-5-sulphonic acid as the reagent.The interference caused by calcium is obviated by the addition of EGTA in carefully calculated concentrations such that no interference is caused by chelation of any excess with magnesium. A direct method in which protein is not removed is compared with a method involving prior deproteinisation by dialysis. The samples showed a lower blank fluorescence after dialysis. However the direct method can be used if suitable compensation is made for the higher blank reading. Results for both methods are given. Forty samples per hour can be analysed satisfactorily with the instru-mentation described; later experiments indicated that if a sampling rate of sixty Schachter’s methodlS0 is also used by Clark an PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 65 samples per hour was used the fluorescent response was only 95 to 97 per cent.of that obtained at a rate of forty samples per hour. These automated procedures may possibly be improved in selectivity and sensitivity by using N,N’-salicylidene-2,3-diamino benzofuran which has already been shown to be applicable to the determination of magnesium in sera.lS0 Magnesium has also been determined in soil samples by using 2,2’-dihydroxy-azobenzene. The reagent dissolved in a mixture of ethanol ethylene diamine and hydrochloric acid is reacted with neutral (ammonium acetate) extracts of the soil ~amp1es.l~~ The method is reported to be more sensitive than the colori-metric method with Titan yellow and gives similar results to those obtained with atomic absorption techniques.Chloride nitrate phosphate sulphate aluminium, copper manganese and sodium do not interfere in the levels normally encountered in these mineral samples but the presence of calcium gives high results. Theuse of a reagent solution containing calcium prevents this interference. Schachter’s method160is the basis of the automated determination of the principal exchangeable ions in ~ 0 i l . l ~ ~ The values found are claimed to be superior to those obtained by other methods. Calcium The use of fluorimetry for the determination of trace amounts of calcium in biological samples has been reported extensively and as with magnesium fluori-metry appears to be a well established technique in laboratories dealing with this type of sample.Although there are many examples of fluorimetry being used for the deter-mination of calcium only two new reagents for its fluorimetric determination have been reported during the period of this review. One is 8-hydroxyquinalde-hyde-8-quinolyl hydrazonel71 and a method for the determination of calcium with this reagent has been ~atented.17~ The preparation and properties of this new reagent are reported171; the dissociation constants of the reagent and the approxi-mate instability constant of the calcium chelate are also given. The conditions for the determination are reported in detail; the optimum pH is in the range 11.0 to 13.0 (0.1 M solution of potassium hydroxide is used) the development time is 10 to 15 minutes and the intensity of the solutions remains constant for up to 1 hour.The main advantages of this reagent are its selectivity and stability, both of which are considerably better than those of the most widely used reagent, calcein. By using the specified conditions up to 10-fold excess amounts of stron-tium and 100-fold amounts of barium and magnesium do not interfere; inter-ference levels for several other ions are reported. The apparatus used in this work is constructed of quartz although no reasons are given for this choice. Methods for the determination of calcium in the range 1 x to 1 x per cent. in potassium chloride and methyltrichlorosilane are also reported in the paper. The method of standard additions is used to confirrn the results obtained. A separate paper by the same workers reports the determination of 5 x per cent.of calcium in rnethyltrichloro~ilane.~~~ The determination of calcium i 66 BARK AND WOOD magnesium oxide has been described174 by using 8-hydroxyquinaldehyde-8-quinol-ylhydrazone. When present in amounts greater than 4 pg ml-l magnesium inter-fered but could be removed by extraction with [2- [(2-hydroxy-l-naphthyl) azol-phenyl azoxyl-4-methyl phenol in a solution of tributyl phosphate in carbon tetrachloride. At lower concentrations of magnesium extraction was not considered to be necessary. The work of Budesinsky and West on the introduction of dicarboxymethyl amino methyl groups into fluorescent reagents has resulted in the publication of a paper on the use of l,5-bis (dicarboxymethyl amino methyl)-2,6- dihydroxy naphthalene for the determination of ca1~ium.l~~ The preparation of the reagent is described and a method for the determination of calcium in the range 10 to 500 ng is proposed.The optimum pH is 11.7 at which the reagent is unstable. This lack of stability requires that readings must be made at an exact time (5 minutes) after the addition of the reagent. The decomposition of the reagent is reported to be increased by irradiation with ultraviolet radiation and the authors suggest that total destruction of the naphthalene skeleton is probable. Despite this decomposition the pK values of the reagent have been determined by using a spectrophotometric method at wavelengths of 220 and 230nm. Although many ions may interfere if cyanide is used to mask certain ions and others are extracted as 8-hydroxyquinolates only magnesium strontium and barium ions give significant effects.This method suffers from the disadvantage that the reagent is very unstable in alkaline solution. (A solution of the reagent is stable for only 5 hours. This is far less stable than the widely used reagent, calcein .) A comparison of atomic absorption spectrophotometry and fluorimetry for the determination of serum calcium has been published.l7* The fluorimetric method used was that of Kepner and Her~u1es.l'~ The recovery figures are reported and in the range 2 to 4 meq. 1-1 Values obtained by using the fluorimetric tech-nique were considerably lower than those obtained by atomic absorption spectro-photometry. These higher results from atomic absorption are a definite trend even at higher levels.Interference with the fluorimetric method was observed when haemolysed or icteric serum was used. Calcein has been used by Uemura17* for the determination of calcium in biological materials. The author recommends the use of perchloric acid for the deproteinisation of muscle prior to the deter-mination of the calcium. The useable range is 0.8 to 3.6 pg of calcium over which range the calibration plot is linear. The fluorescence intensity is compared to that of a standard containing 4pg of calcium. The use of this reagent has also been reported by Lewin et aE.,179 who report that the calibration curve is 'S'-shaped because of the formation of both mono- and di- complexes and the range over which this plot is linear is dependent on the calcein concentration.The effect of variations in the calcein concentration is reported in detail and a concentration of 8 mg 1-1 of calcein gives a linear calibration between 0.40 and 1-40 mg of calcium. However one great disadvantage with the use of calcein is that the optimum re-agent concentration must be established for each new batch of the commercia PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 67 reagent used. Conversely this concentration effect may be turned to advantage, as raising or lowering the calcein concentration enables calcium to be determined at either higher or lower concentrations than before. No interference is observed with magnesium phosphate or citrate even at concentrations 100 times those generally encountered at physiological levels.Both bilirubin (at concentrations greater than 1 mg ml-1) and haemoglobin (at 50 mg ml-l) interfere by reducing the fluorescence intensity. The analysis of plasma of various subjects has been carried out and in some cases the results have been compared with those obtained by using the method of Clark and Collip.lSo The recovery was checked by adding calcium to both standard and plasma solutions; the results varied between 98.6 and 104.3 per cent. Although no reasons are given this method is not recom-mended by Lewin et ~E.17~ for the determination of urine calcium. Moser and Gerarde 181 have also modified the method of Kepner and Hercules. Their modification involves the use of the ‘Unopette system.J1*2 All of the reagents for a single determination are pre-measured and pre-packed ; to prevent errors caused by the instability of calcein solutions the calcein is pre-packed in a dry form.The method is rapid and reported to be applicable to the determination of calcium in urine serum plasma and cerebrospinal fluid. There is a welcome trend to automate many routine clinical analyses and several automated methods have been proposed for the determination of calcium in sera all of which use calcein as the reagent. The use of an automated method for the determination of calcium in a variety of serum samples has been des-cribed.lS3 The method involves the use of dialysis to remove the interference of bile pigments and lipids; details of the instrumentation used are given and the results obtained are shown to be comparable with those obtained by the manual method of Clark and Collip.Magnesium (5 mg 100 ml-l) phosphate sulphate, copper(I1) and iron(II1) ions did not interfere when present in two to four times the normal serum level. The reported recovery is 98 to 101 per cent. with standard deviation of A0912 mg 100 ml-l. The samples are treated with hydro-chloric acid prior to dialysis to ensure complete recovery of the calcium. At p H values greater than 4-5 some calcium is bound to protein the use of a solvent system with a pH of approximately 1.6 ensures complete dissociation of the calcium - protein complexes and aids the recovery. Classen Marquardt and SpathfS4 have investigated the effect of dilution and dialysis on the determination of calcium in whole blood and serum by using a similar automated procedure.The method has been used for determinations i.n vivo. In a paper describing the simultaneous determination of calcium and phos-phorus calcein has been used1% for the determination of calcium. The sample is diluted with hydrochloric acid and dialysed against a tin(I1) chloride - hydrazine reagent which gives a solution suitable for the spectrophotometric determination of the phosphate. For the calcium determination the undialysed effluent is further diluted and mixed with an alkaline solution of calcein; the calibration curve obtained for calcium is linear over the range 5 to 14.5 mg 100 ml-l. Lipemic serum which gives high calcium values is pre-treated by using ether extractio 68 BARK AND WOOD to give a serum which when analysed fluorimetrically gives comparable results to those obtained from atomic absorption spect ropho tometry.Quantinl'O has compared automated procedures for a colorimetric method (with Murexide) a flame photometric method and a fluorimetric method (with calcein) for the determination of calcium in soils. Details of the instrumentation and procedures used are reported. The fluorimetric method is apparently the best as it has the advantage of being applicable over a wider range of concentrations (0.1 to 7.5 meq.) than is possible with the other methods. The only stringent precaution required is to ensure that the ambient temperature does not change abruptly during determinations. Boron Acetyl salicylic acid has been reported as a reagent for the fluorimetric determination of boron.lS6 The boron-containing sample is pre-treated by dissolv-ing it in potassium hydroxide solution and then evaporating this solution to dryness in a quartz crucible.The residue is reacted with salicylic acetic and sulphuric acids and the fluorescence of the resultant solution measured after 5 to 10 minutes. The complex is reported to be stable for several hours. The method is sensitive to 0.01 pg ml-l but no details of interferences are available. This seems to be a fairly rapid and sensitive method. The fluorescence of the ternary complex of boron with salicylic acid and Rhodamine 6G is the basis of a method for the determination of boron.lS7 A sample treated with Rhodamine 6G (IV) and salicylic acid is evaporated to dryness (IV) Rhodamine 6G (6ZH) and the excess salicylic acid removed by complexing with iron(II1).The ternary complex is then extracted into benzene and its fluorescence measured. During an investigation of the complex formation between boric acid and a series of hydroxyanthraquinones in concentrated sulphuric acid Holmels8 found that chinizarin (1,4-dihydroxyanthraquinone) was the only compound producing a fluorescent complex with boron. This is the basis of his proposed method. The sulphuric acid concentration used is 91 to 96 per cent. and intensity readings are taken after allowing the mixture to stand for 2 to 3.5 hours. Variation of tempera-ture in the range 10" to 40 "C had no effect on the difference in intensity between the sample and blank. The method is sensitive to 10 ng ml-l.Solutions that wer PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 69 stored in volumetric flasks were examined for their boron content by the use of the spectrophotometric method by using quinalizarin which is capable of detecting 0402 pg of boron,l89 and is thus more sensitive than the reported fluorimetric method. The determination of boron as the tetrafluoro borate ion by using butyl-rhodamine is the subject of a Russian patent.lgO Prior to its determination, the boron present in solutions containing a high concentration of sodium chloride or present in very small amounts is precipitated with sodium salicylate and butyl-rhodamine. No other details are given in the abstract. Benzoin has been used for the determination of boron in salts and also in samples of high purity tin.Pod-chainova and Skornyakovalgl have described amethod for the determination of trace amounts of boron in soil by using benzoin the extraction and the determination of the boron are described in detail. The analysis takes 30 to 40 minutes and the method appears to be fairly sensitive. Although no determination or detection limits are reported in the abstract a coefficient of variation of 5 to 10 per cent. for boron in the range 0.1 to 1.0pgml-l is reported suggesting that this is the usable range of the method. The determination of boron in high purity tinlg2 is similar to the above method. The extraction procedure involves the removal of tin and other heavy metals by precipitation as sulphides after the dissolution of the sample with a mixture of hydrochloric acid and (1 + 1) hydrogen peroxide.By using the method of standard addition boron contents as low as 5 x per cent. can be determined with an error of less than 20 per cent. The determination of boron in high purity silicon tetrachloridelg3 with Thoron I Carseno-2-benzene (l-azo-l')-2-oxynaphthalene-3,6-sulphonic acid] is reported to be sensitive to 3.5 x per cent. of boron (with a 20-ml sample) with a relative error no greater than 40 per cent. A method of standard addi-tions is used to check and verify the results obtained. As commercial samples of such azo dyestuffs as Thoron I are often mixtures the reagent is purified by three re-crystallisations from ethanol - acetone mixtures and dried at a tempera-ture of less than 60 "C in a nitrogen atmosphere.Concentration and separation of the boron is achieved by the addition of triphenylchloromethane in diethyl-aniline; this results in the formation of a non-volatile boron complex and the silicon tetrachloride is then removed by evaporation. Using a method of this sensitivity means that great care must be taken to ensure complete or standard cleanliness of all of the apparatus used and to achieve the necessary high purity of all the reagents. These stringent requirements result in the most serious dis-advantage of this method that is the need to distil in platinum apparatus an already highly pure sulphuric acid. The cost of the apparatus would possibly be prohibitive unless many determinations were to be carried out.Another point arising from this distillation is the possibility of boron contamination from platinum. Such contamination has been reported by Couchl33 who studied the leaching of boron from platinum crucibles during the fusion of rock and soil samples with sodium carbonate. The possible leaching of boron from borosilicate glasses may be a limiting factor in any method proposed for boron 70 BARK AND WOOD A review of the published fluorimetric methods for the determination and detection of boron has been produced by G. K. Turner Associates.26 This review contains twenty-six references and briefly describes the various methods making it an excellent article for those workers interested in the use of fluorimetry for the determination of inorganic ions. Aluminium Several papers have been published during this period dealing with the use and sensitivity of some of the reagents previously proposed for the determination of aluminium.The relative sensitivities of the aluminium complexes of 8-hydroxyquinoline, salicylidene-o-aminophenol morin querce tin (V) lumogallion (VI) and Eriochrome black T (VII) have been inve~tigated~~*J~~ and calibration curves for the deter-mination of aluminium by using these reagents presented. The authors conclude (W Quercetin S03Na (VII) Eriochrome black PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 71 that the slopes of these calibration curves depend on the wavelength used for excitation; the more radiation that is absorbed by excess reagent then the smaller is the slope of the calibration curve.The effects of solvents ethanol, methanol acetone and dimethyl formamide on the fluorescence quantum yields of the aluminium chelates of lumogallion morin quercetin and salicylidene-o-aminophenol have been investigated.lQ6 For all quantum yield determinations, a 10-fold excess of aluminium was present to ensure there was no interference from uncomplexed reagent. The authors conclude that the increases in absorption (approximately two times) and in luminescence intensity (three to eight times) on the addition of a non-aqueous solvent are primarily the result of the formation of a solvate-type compound between the molecules of the complex and solvent molecules and not the result of a shift in the equilibrium of complex formation; the dielectric constants of the solvents do not directly affect the luminescence intensity.Calibration curves whose slopes agree with the sensitivity factor calcu-lated as in references 194 and 195 are presented the increase in sensitivity of the reagents as a consequence of the non-aqueous solvents is attributed mainly to an increase in the quantum yield. The Turner Instrument Company has issued a reviewlg7 of the published methods for the fluorimetric determination of alu-minium which appeared in the literature prior to 1967. This review consists primarily of tables showing the sensitivity interferences and details of filter and wavelength of maximum emission etc. and a list of thirty references. During this period two new reagents for the determination of aluminium have been reported.Ho~man~~~examined three dyestuffs Eriochrome red B (VIII), the sodium salt of alizarin sulphate and alizarin red SW (IX) as possible reagents. PH Eriochrome red B (1x1 Alizarin red S 72 BARK AND WOOD However the fluorescence of the aluminium - alizarin dye chelates show con-siderable dependence on both the pH conditions and the dye-to-aluminium ratio, and mainly because of these factors the author rejected these compounds and developed a method using Eriochrome red B. The pH of the system is maintained at the optimum value of pH 3.6 by using an acetate buffer and the calibration plot is linear over the range 0.6 to 6 pg of aluminium. The method is apparently simple in operation but the fluorescence intensity continues to show a slight increase even after standing for 5 days.Although the changes are small it is necessary to measure the intensity of simultaneously prepared sample and standard solution at a definite time interval after the addition of the reagent. A suitable period is suggested to be 24 hours although this time effectively removes the method from consideration by most analysts and the sensitivity is not sufficiently great to compensate for the time factor. The use of N-salicylidene-2-amino-3-hydroxyfluorone has been proposed2M as a reagent for the determination of aluminium and its sensitivity compared with that of the alternative reagents salicylidene-o-aminophenol ; 2-hydroxy-3-naph-thoic acid ; 2,2’,4-trihydroxyazobenzene-5-sulphonic acid ; Acid Alizarin Garnet R (X); Pontachrome Violet SW (XI); Pontachrome Blue Black R (XII) and morin.The proposed reagent is insoluble in water and therefore the determination must be carried out in aqueous solutions containing 10 per cent. v/v of ethanol. Increasing the alcohol concentration increases the fluorescent intensity of both the blank and the chelate but this is not a sufficiently large increase to justify its use. The pH of the system is controlled by the addition of acetate buffers and the chelate complex shows a maximum fluorescence intensity in the pH range (XI Acid Alizarin Garnet R (XI 1 Pontachrome Violet S PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 73 (XI11 Pontachrome BBR 5-1 to 5.4 the authors use and recommend a pH of 5.2. The development time required is 30 minutes and after this time the fluorescence intensity remains constant for up to 80 minutes and is not temperature dependent in the range 0" to 30 "C.A 4-fold excess of the reagent at the maximum limit of the method is recommended. The reported lower limit is 2 ng ml-l. Interference is caused by the presence of equimolar amounts of iron(III) which the authors suggest should be removed by a mercury cathode cell or by solvent extraction. Ions such as fluoride and phosphate which form stable complexes with aluminium also interfere. Gallium can be tolerated only in less than 10-fold excesses but anions such as nitrate sulphate and chlorate do not interfere. The comparison of N-salicylidene-2-amino-3-hydroxyfluorone with the other fluorimetric reagents listed show that when using a filter fluorimeter it is the most sensitive.The advantages and disadvantages of the various reagents are discussed and although some further comparisons were apparently made by using spectro-fluorimetry no critical observations are reported. The determination of aluminium and gallium by using lumogallion has been described.200 The method for aluminium may be used in the range 0.1 to 2 pg in a final volume of 25 ml and requires that solutions of the reaction mixtures at the optimum pH of 5.0 are heated for 20 minutes at 80" C to ensure complete reaction. When aluminium is present in the range 0-01 to 0.2 pg then the complex may be extracted into pentanol. Despite the negative interference caused by the presence of chromium cobalt copper iron nickel scandium tin(IV) titanium or vanadium(V) the method has been applied to the determination of both suspended and dissolved aluminium in sea water.201 The use of the reagent in conjunction with 1,lO-phenanthroline to mask the iron interference enables trace amounts of aluminium in natural waters to be determined.202 The effect of the anions of buffer solution on the reaction between salicylidene-o-aminophenol and aluminium has been investigated.203 Acetate and biphthalate ions compete with the reagent for the aluminium and hence the authors recommend that hexamethylene tetramine should be used to prepare buffer solutions.These findings suggest that better sensitivity may also be obtained from other methods by using hexamethylene tetramine instead of acetate buffers which seem to be used in most methods.Bognar and Pataky204 have utilised the red fluorescence produced by th 74 BARK AND WOOD aluminium - Pontachrome Blue Black R complex to determine aluminium how-ever because the fluorescence intensity is time dependent the metal is determined by comparison of the fluorescence intensity of a sample solution with that of a series of standard solutions after the simultaneous addition of the reagent. At pH 5.7 0.02 to 2 pg ml-l if aluminium and at pH 4-9 0.2 to 2 pg ml-1 of alu-minium can be determined with an error of less than 10 per cent. Pontachrome Violet SW which gives an orange-red fluorescence with aluminium can be substituted for Pontachrome Blue Black R. The exchangeable aluminium content of soils and the total aluminium content of rocks and minerals have been determined205 by using the fluorescence of the aluminium chelate of 8-hydroxyquinoline in chloroform.The method used is a modification of that of Goon et aL206 An aliquot of a potassium chloride extract of soil containing less than 20pg of aluminium is treated with the reagent and a pH 9.0 buffer solution and chloroform is added. After shaking for exactly 2 minutes the chloroform layer is separated. The extraction is repeated and the organic extracts are combined. The fluorescence intensity of the combined extracts is measured. Iron(II1) interferes unless present in two to three-fold excess of the aluminium and as the exchangeable aluminium in soils is generally in excess of the exchangeable iron this interference may usually be ignored.No other interferences are reported. The aluminium content of rocks and minerals is determined after extraction of the aluminium with sodium hydroxide solution, by the same method used for soil samples. The author reports that the method is considerably better with regard to stability sensitivity reproducibility and interferences than the titration technique which uses a metallochromic indicator. Results for standard samples and a comparison of the results obtained for various soil samples by using both fluorimetric and titrimetric techniques are reported. Gallium indium and thallium In recent years the determination of metal ions in this group especially the determination of trace amounts of gallium has received considerable attention.Reagents containing the 2,2’-dihydroxyazo grouping have been used for the fluorimetric determination of gallium. The reactions of o,o’-dihydroxyazo com-p o u n d ~ ~ ~ ~ and o,o’,@’-trihydroxyazo compounds208 with gallium have been studied by using spectrophotometric techniques. In both types of compound the reagent reacts as its tautomeric quinone hydrazone form and produces complexes of the type GaL (positively charged and fluorescent) at a pH of less than 5 and GaL, (neutral and non-fluorescent) at a pH greater than 5. The complex-forming ion is Ga(OH)2+ for both the charged and neutral complexes formed with the o,o’-di-hydroxyazo compounds (Eriochrome Blue Black R (XIII) and chlorophenol-azo-naphthol) and for the neutral species of the trihydroxy compounds (o,o’,p-tri-hydroxyazobenzene and sulphonaphtholazoresorcinol) .The complex-forming ion in the case of the charged complex of the trihydroxy compounds is Ga3+ PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 76 (XIII) Eriochrome BBR The reaction of gallium with quercetin has also been studied209 and the type of complex formed is dependent upon the pH at which the reaction is carried out. In the pH range 1.0 to 4.5 Ga(OH)2+ forms a 1 1 positively charged fluorescent species with the reagent in this complex bonding to the hydroxyl group ortho to the carbonyl is presumed. At pH values greater than 5.0 a 1 2 neutral non-fluorescent complex is formed with the ion Ga(OH)2+ and in this case the para-hydroxyl group is assumed to be involved in the complex formation.Two papers have been published in which the sensitivities of various reagents that have been proposed for the determination of gallium are compared. Babko et aL2l0 compared the reagents morin quercetin salicylidene-o-aminophenol, sulphonaphtholazoresorcinol lumogallion and Rhodamine B (XIV) in a variety ow Rhodamine B of solvents. They recommend Rhodamine B in a 3 1 mixture of benzene - diethyl ether sulphonaphtholazoresorcinol in 95 per cent. of methanol and lumogallion either in 95 per cent. of methanol or using extraction with isoamyl alcohol as being the most sensitive. Shcherbov and Matveets2fl compared the use of 8-hydroxyquinoline Rhodamine S (XV) salicylidene-o-aminophenol morin and lumogallion. The calibration curves for the determination of gallium and the spectral properties of its complexes with each of these reagents are reported.The authors report that whereas Rhodamine S is the most sensitive spectrophoto-metric reagent lumogallion is the most sensitive fluorimetric reagent. 76 BARK AND WOOD ow Rhodamine S Both 2-(2'-hydroxyphenyl) benzoxazole and 2-(2'-hydroxyphenyl) benzothia-zole have been used for the fluorimetric determination of gallium.212 The gallium complex of either of the reagents is extracted from a saturated aqueous solution of sodium perchlorate solution with iso-amyl alcohol and the fluorescence intensity of this extract is then measured. A higher sensitivity (0.6 pg) is reported for the method when using 2-(2'-hydroxyphenyl) benzoxazole as the reagent.Barium, calcium chromium magnesium manganese mercury and silver in amounts up to 100-fold excess and cadmium cobalt lead zinc and zirconium up to 20-fold excess do not interfere. 4-(5-Chloro-2-hydroxyphenylazo) resorcinol has been used as a fluorimetric reagent for the determination of gallium in sea water.213 The metal is concen-trated prior to its determination by extraction of the gallium reagent complex into hexanol followed by re-extraction into hydrochloric acid solution and evaporation to dryness. The residue is ignited in a muffle furnace at 600" to 700 "C for 10 to 15 minutes after which it is dissolved in 6 M hydrochloric acid and the gallium halide complex produced is then extracted into diethyl ether. The etkrer extracts are evaporated to dryness; the residue dissolved in 0.01 M hydrochloric acid and then treated with reagent acetic acid and ethanol.The gallium complex of 4-(5-chloro-2-hydroxyphenylazo) resorcinol is extracted into hexanof and its fluorescence intensity is measured and compared to standards containing up to 5 pg ml-1 of gallium in the aqueous solution. In a later paper214 a comparison of seven phenolic azo compounds as possible reagents for gallium is reported, and the above reagent is reported as the most sensitive (0-5 ng ml-l of gallium). A detailed description of the application of this reagent to the determination of gallium in silicates is given. After dissolution of the silicate samples in hydro-fluoric and sulphuric acids the procedure is similar to that used for the deter-mination of gallium in sea water.21s The authors report variations of this method that enable gallium to be determined in sulphide ores silica or silicon.Lumogallion [5-chloro-3-(2,4-dihydroxyphenylazo)-2-hydroxybenzene sulph-onic acid] has been used for the determination of aluminium and gallium.200 The optimum pH for the determination of gallium is pH 3.0 and ammonium acetate buffers are used to control the pH. The method is reported to be of use for the range 0.5 to 5 pg of gallium in the final volume of solution (25 ml). How-ever no details are given of the volumes of either the reagent solution or of th PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 77 buffer solutions used in making the final volume up to 25 ml and therefore the actual concentration of the gallium solution submitted to analysis is not known.The authors report that the complex can be extracted into iso-amyl alcohol when gallium is present in the range 0.03 to 0-2pg. The main disadvantages of this method are the need to heat the solutions at 80°C for 20 minutes to ensure complete chelate formation and the interference caused by the ions cobalt copper, chromium(VI) rhenium nickel scandium tin(IV) titanium and vanadium(V). No details are given of methods for the removal of interferences. Matveets and Shcherbov213 have compared spectrophotometry and fluorimetry for the deter-mination of gallium by using lumogallion. Procedures for the determination both before and after extraction of the 1 1 gallium reagent complex by using iso-amyl alcohol are reported.For both procedures the spectrophotometric method is sensitive to 0-05 to 2 pg ml-l of gallium whereas the direct fluorimetric method is sensitive to 0-002 pg ml-l and the fluorimetric method in which the complex is extracted is sensitive to 0.001 pg ml-l. No interferences are reported. The decrease in fluorescence intensity of the gallium - lumogallion chelate observed in the presence of excess lumogallion has been investigated.216 This effect is attributed to competition between the excess reagent and the fluorescing gallium complex for the exciting radiation. The authors also report that a similar effect is observed for the gallium - morin complex but that in this case it cannot be attributed to the competition for the exciting radiation. 8-Hydroxyquinoline has been used to determine the gallium content of silicate rock and fly ash The sample is treated with hydrofluoric acid and evaporated to dryness this procedure is repeated twice first with a mixture of hydrofluoric and hydrochloric acids and then with hydrochloric acid.The resulting residue is dissolved in 1 1 hydrochloric acid solution and any insoluble material remaining is fused with sodium carbonate and the product added to the hydro-chloric acid solution. The interfering ions copper(II) iron(II1) and vanadium(V), are then removed by the addition of hydroxylammonium chloride and evaporation of the mixture to near dryness. After the addition of 1 10 hydrochloric acid, the solution is buffered with potassium hydrogen phthalate and the pH is adjusted to 2.8 to 2.9; oxine solution is then added and the metal complex is extracted into chloroform.The non-aqueous phase is successively washed with potassium cyanide solution (pH 9-5) and hydroxylammonium chloride (pH 2.8) to remove the interfering ions; molybdenum(VI) vanadium(V) copper(I1) and nickel(II) and any residual iron(III) produced by the rapid oxidation of iron(I1) on shaking with chloroform. The authors report that sulphate is the only serious interference in the determination which is completed by measuring the fluorescence intensity of the chloroform extract and comparing this with the intensities of blank and standard solutions that have been subjected to the same procedure. The results obtained by using this method which has a coefficient of variation of 4.2 per cent.at a level of 20 p.p.m. of gallium compare well with the results obtained by activation analysis. The value obtained for the U.S. Geological survey standard silicate rock sample (Granite G1) agrees with that reported in the literature 78 BARK AND WOOD Bark and Ftixon218 have described the use of 2-(2’-pyridyl) benzimidazole as a reagent for the fluorimetric determination of gallium and indium. In the range 70 to 700 ng ml-l of gallium the proposed method has a standard deviation of 4.8 per cent.; for indium in the range 110 to 1 000 ng ml-1 the standard deviation is 2.3 per cent For gallium the maximum intensity is observed at pH 4.09 in the presence of a minimum of a 30-fold excess of reagent; for indium the optimum pH is 5.2 and a minimum of 10-fold excess of reagent is required.Mole-ratio plots indicate the formation of a 1 1 gallium - reagent complex whereas with indium a 2 1 complex is formed. The recommended development time is 30 minutes after which period no significant changes in intensity are observed for up to 10 days. Although several metal ions interfere the method is made more selective by extracting gallium and indium as their benzoates with ethyl acetate followed by evaporation of the extract to dryness. The residue is then dissolved in hydrochloric acid and treated with the reagent and buffer solution. The glassware used with the exception of the sample cells was treated with tetrabutyl titanate in cyclohexane which on slow hydrolysis in air gives a protective polymeric layer of butoxy - titanium groups and prevents absorption of the metal ions on to the glass.The formation and extraction of the ion association complexes formed between Rhodamine 6G and the halogen complexes of gallium and indium have been studied by using fl~orimetry.~~9 The extraction procedure is the same in each case; the complexes are extracted by shaking a mixture of equal volumes of the aqueous phase and benzene for 1 minute the fluorescence intensity of the organic phase is measured 15 minutes after the separation. The maximum extraction of gallium at gallium-to-indium ratios of less than 1 2500 is obtained by using 6 M sulphuric acid and M hydrochloric acid however at a gallium-to-indium ratio of 1 10000 the maximum extraction is obtained by using 6 M hydrochloric acid only.For indium the maximum extraction is obtained from a solution that is 0.3 M in hydrobromic acid and 5 M in sulphuric acid. Complete extraction of gallium as its bromide complex was not achieved even in 12 M sulphuric acid. The authors also report the conditions for the extraction of the iodide complexes of these metals. The effectiveness of a number of fluorimetric reagents for the determination of indium have been compared.220 The sensitivities are compared by calculating the product of the molar absorption at the wavelength of the exciting radiation and the quantum yield. The reagents examined were 8-hydroxyquinoline lumo-gallion morin quercetin Rhodamine B Rhodamine 6Y and salicylidene-o-aminophenol. Morin in 50 per cent. v/v methanol at pH 3.6 (ammonium acetate buffers) and Rhodamine B in benzene are reported to be the most sensitive.Rhodarnine 6Zh (IV) which does not react with tellurium has been des-cribed as a reagent for the determination of indium in indium tellurides.221 Samples are dissolved in 1 1 hydrochloric acid the resulting solution is evaporated to dryness and the residue is then dissolved in 2 M hydrochloric acid. An aliquot of this solution containing approximately 1 pg of indium is treated with th PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 79 reagent and 7.5 M sulphuric acid and the chloride complex formed with the reagent is extracted into benzene. The fluorescence intensity of the benzene extract is measured. Details of sensitivity or interferences are not available.Rhodamine S has been proposed as a reagent for determining the indium content of underground waters.222 The method is based on the fluorescence of the indium(II1) bromide - Rhodamine S complex in benzene solution. Details of the procedure for the concentration of indium by co-precipitation with iron (111) hydroxide and the removal of interferences by solvent extraction are given. The method is applicable to water samples containing 0.05 to 0.1 pg 1-1 of indium. The reaction between pyrrhonine [3,6-bis(dimethylamino) xanthylium chloride] and indium in strongly acid solution results in the formation of a fluorescent brick-red complex that can be extracted into benzene. The optimum conditions for this reaction have been established2=; 0.2 pg ml-l of indium can be determined spectrophotometrically and it is proposed that the fluorescence of the aqueous layer may also be used to determine the indium concentration.However the intensity decreases with time. Although calibration curves are linear for the ranges 5 to 20 and 10 to 40 pg ml-l the method is far less sensitive than other fluorimetric methods for indium and hence has little to recommend it. During the period of this review no organic reagents have been reported for the determination of thallium. However several workers have described methods for its determination based on the fluoresence of thallium(1) chlorides both at room temperature and in frozen solution. These proposed methods are described in the section of this review dealing with the determination of metals as halide complexes.Silicon germanium tin and lead During this period the only method that has been reported for the deter-mination of silicon is an indirect one2% that is based fluorimetrically on the determination of molybdenum by using carminic acid (XVI) . Silicomolybdic acid, ow Carminic acid which is produced at pH 1.5 in aqueous solution is extracted into isoamyl alcohol and traces of excess molybdenum are removed by washing the organic phase with dilute sulphuric acid. The complex is then re-extracted from the organic laye 80 BARK AND WOOD with dilute ammonium hydroxide solution and the molybdenum present in the extract is determined fluorimetrically. The limit of detection is equivalent to 3 ng ml-l of silicon and the method has been used to determine the silicon content of hydroftuoric acid ammonium hydroxide sodium bicarbonate and ammonium molybdate.A new reagent rezarson (2,2',4'-trihydroxy-3-arsono-5-chloro azobenzene) has been proposed for the determination of germanium,226 both fluorimetrically and spectrophotometrically. The reagent is non-fluorescent but the germanium chelate is and this fluorescence enables 4ngml-l of germanium to be determined the upper limit of linearity being 4pgml-l. The method is reported to be very selective only a 1000-fold molar excess of aluminium caesium indium or zinc interfere. The method has been used to determine the germanium content of coal samples. The reagent rezarson contains three chelating sites with which ger-manium could possibly react and to establish the functional group Lukin, Efremenko and Petrova226 studied spectrophotometrically the reagent and its germanium complex.The functional group was shown to be the 2,2'-dihydroxy-azo grouping by comparing the reactions of germanium with rezarson and with two other compounds of similar structure but differing in that in one the 2'-hydroxy group is absent and in the other the arsono group is absent. Both tin and lead have been determined by the measurement of their fluorescence in hydrohalic acid solution and these methods are described in the appropriate section of this review. Pal and Ryan227 have reported the determination of both tin(I1) and tin(1V) by using 8-hydroxyquinoline-5-sulphonic acid as the reagent both these oxidation states giving the same fluorescence spectrum with the reagent.A detailed des-cription of the procedures necessary to determine tin is given; the maximum fluorescence intensity is obtained in the pH range 4.0 to 5.2 with a 125 to 250 molar excess of the reagent. The addition of ethanol methanol dioxan or dimethyl formamide increases the intensity of fluorescence but to obtain consistent results, carefully controlled volumes of these solvents must be added. However the authors report that 1 ng ml-1 of tin can be determined by using ethanol methanol or dioxan as solvents. As the reported method does not involve the use of these solvents it is only applicable for the determination of tin in the range 5 to 250 ng ml-1. The interference levels for various ions are reported; copper(II) mercury(I1) and iron(II1) interfere but can be masked by the addition of thioglycollic acid [iron(III)] sodium thiosulphate [copper(II)] and chloride ions [mercury(II)] or by the addition of hydroxylammonium chloride.The other most serious inter-ferences are caused by fluoride and EDTA even at concentrations of 5 and 2.5 pg ml-1 respectively. The complexes formed between the reagent and both tin(I1) and (IV) are of the form ML and because of the similarity between these com-plexes the authors suggest the formation of octahedral complexes of the type [SnQ2,H,0]2- and [SnQ2.(0H) 2]2- where HQ represents the reagent. No methods for the fluorimetric determination of lead by using organic reagents have been reported during the period of review PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 81 Arsenic antimony and bismuth During this period no methods have been reported for the determination of arsenic by using organic reagents.However methods based on the luminescence of arsenic halides have been described as have similar methods for the deter-mination of bismuth and antimony. These methods are reported in the section of this review dealing with the determinations of metal ions as halide complexes. Guilbault et a1.l4O have described the use of the inhibition of the catalysed hydrolysis of umbelliferone phosphate by both bismuth ions and beryllium ions for the determination of either metal ion. This method can be used to determine bismuth in the range 1 to 70 pg ml-1 with an average relative error of k1.3 per cent.The procedure used is identical for the determination of either of these metals. (Details of the conditions for the determination of beryllium are described in the appropriate section of this review.) The possibility of determining antimony fluorimetrically by using Rhodamine 6Zh or Rhodamine S has been investigated22* and compared with the spectro-photometric methods. The fluorescence of the ion-association complexes formed between the Rhodamine dyes and the chloride or bromide complexes of anti-mony(II1)or (V) is the basis of the method. However despite a reported sensitivity of 0-1 pgml-l when extracting the antimony chloride complex with Rhodamine 6 Zh the reproducibility is so poor (standard deviation 30 to 35 per cent.) that the authors conclude that spectrophotornetry cannot be replaced by fluorimetry for the determination of these ions.A cherniluminescent method has been described229 for the determination of antimony(V) the luminescence resulting from the reaction between luminol (XVII) 0 (XVI I) Luminol and antimony(V) at pH 11 to 12. The total amount of the luminescence (measured photographically) is proportional to the concentration of antimony0 and inversely proportional to the concentration of luminol. Only when antimony is present as [SbClJ- is a chemiluminescent reaction observed so that suitable methods of oxidising antimony(II1) to anthony(V) are reported. Ceric sulphate, potassium bromate potassium permanganate and sodium nitrite are reported as possible oxidising agents and of these sodium nitrite is the most convenient because in acid solution any excess can be easily removed by the addition of urea or hydroxylammoniwn chloride.The addition of a solution of [SbCl,]- ion 82 BARK AND WOOD to an alkaline solution of luminol gives the maximum chemiluminescence and hence results in the highest sensitivity. This the authors report as 0.05 pg ml-1, no upper limit for the method is reported. Zinc (up to 50-fold excess) cadmium and tin(1V) (up to 200-fold excess) do not interfere although titanium(II1) as chloride interferes by producing a chemiluminescent reaction. Iron and copper ions quench the fluorescence but this interference may be prevented by previously sequestering these ions with EDTA. Selenium and tellurium Although no new reagents have been reported for the determination of selenium or tellurium several papers have been published on the use of 3,3’-di-aminobenzidine and 2,3-diaminonaphthalene as reagents for the fluorimetric determination of selenium in biological and industrial samples.Methods for the determination of both selenium and tellurium based on their fluorescence in frozen solutions of halogen acids have also been reported and are discussed in the appropriate section of this review. W a t k i n ~ o n ~ ~ ~ has reviewed the methods for determining selenium in biological materials. The section of his review dealing with fluorimetry consists mainly of a comparison of the use of 3,3’-diaminobenzidine and 2,3-&aminonaphthalene. The use of 2,34aminonaphthalene which has previously been reported to be more sensitive than 3,3’-diarninoben~idine~~~ offers some advantages.When using 3,3’-diaminobenzidine under the normal analytical condition of excess reagent, only the monopiazselenol is formed. This means that because two free amino groups are present in piazselenol it can only be extracted from basic solution, a condition under which several ions will precipitate and consequently cause interference with the determination. However the piazselenol resulting from the reaction of selenium(1V) with 2,3-diaminonaphthalene can be readily extracted into organic solvents under mildly acidic conditions thus eliminating the need to remove or mask ions that form insoluble hydroxides. The only solvents that extract the complexes and in which the piazselenol of 3,3’-diaminobenzidine fluoresces are reported to be toluene xylene and mesitylene of which toluene and xylene are generally used.Similarly the solvents that may be used for the extraction of the piazselenol of 2,3-diaminonaphthalene are cyclohexane toluene, decalin and petroleum spirit; decalin and xylene are the most frequently used. As the reaction of selenium with o-diamines to form piazselenols is specifically with selenium(IV) care must be taken to ensure that the selenium is present in this form and the methods of degradation that result in solutions containing selenium(1V) are discussed. The methods generally used involve oxygen-flask combustion or degradation by using nitric and perchloric acid mixtures. Both the reagents are subject to photodecomposition to yield fluorescent products thus commercial batches require careful purification before use and reactions should take place in yellow light or darkness.The fluorescence of the piazselenols is quenched by increased temperature and by the presence of water but despite these quite serious disadvantages both reagents have been widely used. No othe PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 83 fluorimetric methods are mentioned in Watkinson's review and because the two reagents discussed are so well established recent publications are concerned primarily with the methods of degradation. Costa232 has examined the influence of the type of biological material on the determination of selenium by using 3,3'-diaminobenzidine by comparing the results obtained from a simple standard curve and the method of standard additions.The results and conclusions resulting from this study are reported. Despite the possible use of both oxygen-flask combustion and wet oxidation for the decomposition of samples prior to the determination of selenium only one method2= has been reported in which an oxygen-flask combustion is used. The details of the decomposition by wet oxidation differ from method to method. In some papers methods for the separation of selenium from interfering ions prior to its determination with 3,3'-diaminobenzidine or 2,3-&aminonaphthalene are reported. Methods involving 3,3'-diaminobenzidine. Selenium has been deter-mined in bone and hard dental tissues.234 These materials are first decomposed by wet oxidation with a mixture of hydrogen peroxide nitric acid and perchloric acid the selenium is then isolated by distillation as SeBr and determined fluori-metrically by using the benzidine reagent.A method that involves the use of a modified wet oxidation mixture has been described.= The mixture recommended is sodium molybdate sulphuric acid and nitric acid and because this mixture boils in the range 150" to 170 "C the loss of selenium that occurs at temperatures greater than 180 "C is thereby minimised although some loss does still occur. The authors recommend that the selenium in the wet ashing mixture should be extracted as its dithiol complex because distillation as SeBr causes losses and also that co-precipitation with elemental arsenic is time consuming.In addition, the use of zinc dithiol ensures separation of selenium from a number of elements that interfere with the reaction between selenium(1V) and the reagent. The selenium dithiol complex is extracted into chloroform and after evaporation of the solvent is decomposed by heating with perchloric acid. The authors report that per cent. of selenium can be determined by this method with an error of up to 25 per cent. The conditions for the fluorescent determination of selenium by using diaminobenzidine are re-examined and the authors report that the optimum pH for the formation of the piazselenol is pH 0.9 to 1.1 compared with the pH of 1.5 & 0.3 reported by other workers.236~237 However the use of this lower pH increases the reaction time required to 60 to 70 minutes compared with the 30 to 40 minutes required at the higher pH.The solvents into which the piazselenol may be extracted have also been examined. In the optimum pH range 8.0 to 8.5 toluene is reported to be the best extractant. This method has also been used by Karelina and S a l ~ n a n e ~ ~ ~ for the detemination of selenium in plants, although these workers also propose the co-precipitation of selenium with elemental arsenic for the prior separation of selenium when animal tissue is to be analysed. Co-precipitation of selenium by this method has been used239 for the concentratio 84 BARK AND WOOD and separation of selenium from the solution resulting from wet oxidation prior to its determination with 3,3’-diaminobenzidine. The results for the analysis of fungi are reported.The material is fust decomposed by wet oxidation with nitric and perchloric acids care being taken to avoid evaporating it to dryness, which would inevitably result in a loss of selenium. Traces of nitric acid are then removed by the addition of water and evaporation to the appearance of perchloric acid fumes. The selenium is concentrated by co-precipitation and is determined fluorimetrically. The determination of selenium in natural waters has been reported.%* After concentration by evaporation and co-precipitation with elemental arsenic the selenium is determined fluorirnetrically by using the benzidine reagent. For samples containing 0.6 to 20 mg 1-1 the reported relative error is &SO per cent. 3,3’-Diaminobenzidine and o-phenylene diamine have been used for the deter-mination of selenium in non-ferrous metal samples.241 The use of the former reagent enables 1 x 10-4 per cent.of selenium to be determined in high-purity tellurium. The piazselenol from 3,3’-diaminobenzidine is formed in the pH range 0 to 2.5 and is extracted by using an organic solvent at pH 10 to 12. Bismuth, copper and iron(II1) are the only ions that are reported to interfere in this extrac-tion procedure. The fluorimetric determination of selenium in mineral rawmaterial by using this reagent has been reportedM2 in a brief review of reagents available for the fluorimetric and spectrophotometric determination of selenium. Methods involving 2,3-diaminonaphthalene. Five methods for the de-termination of selenium in biological samples involving 2,3-diaminonaphthalene have been reported during this period.Hoffman et ~1.24~ recommend decomposition of the material to be analysed by heating it with a mixture of perchloric and sulphuric acids with the addition of small amounts of hydrogen peroxide. Potential interferences are masked by the addition of EDTA and the rate of piazselenol formation is increased by boiling the solution for exactly 2 minutes after which the piazselenol is extracted into cyclohexane and its fluorescence intensity measured. Ewan Baumann and Popew4 used an oxidising mixture of sodium molybdate perchloric acid and sulphuric acid for the decomposition of biological materials. The decomposition is followed by co-precipitation of the selenium with elemental arsenic and the solution obtained on dissolution of the precipitate is then reacted with the 2,3-diaminonaphthalene by heating for 45 minutes at 50 “C.The piazselenol stabilised by the addition of ethylenedinitrilo-tetra-acetic acid is extracted into decalin and its fluorescence intensity measured. Recovery values for the individual stages of the analysis are reported and the over-all recovery is 94.6 & 1 per cent. Results are reported for the various types of bio-logical sample analysed by this method which is stated to require 6 hours for the analysis of ten samples. Watkinson’s meth0d~~6 has been slightly modified by Oison,246 and the results of analyses and of a collaborative study bv using this method are reported. The procedure is described in detail and the method is reported to be more sensitive precise and selective than those previously reporte PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 85 by Robinson et aLa47 and Klei11,2*~ which are the official A.O.A.C.methods for determining selenium in food and plants respectively. On this basis it is recom-mended by Olson that the method be adopted as the official method for selenium in plants. The degradation methods for the analysis of selenium in plants have been investigated by Hall and G ~ p t a 2 ~ ~ who used 2,3-diaminonaphthalene as the fluorimetric reagent. Detailed results are presented showing that uncontrolled boiling of the oxidation mixture of perchloric and nitric acids with small amounts of hydrogen peroxide gives low results whereas controlled heating of the mixture in a silicone fluid (F190) bath at 195" to 200 "C minimises the losses.The authors report that oxidants such as vanadate and molybdate were unpredictable in behaviour and quite frequently resulted in violent reactions with a corresponding ejection of the digestion mixture from the flasks. The optimum pH for the forma-tion of the piazselenol when using pure selenium(1V) solutions and acetate -hydrochloric acid buffers is pH 1.1. However the optimum pH when analysing plant samples is pH 2-3 to 2.5 and is controlled by the use of formic acid buffers. Separation of the selenium is reported to be unnecessary when the piazselenol is extracted with decalin. have determined the selenium content of various samples and although no detailed method is reported the one used was probably that of Olson.246 During this period Lamand and A ~ t i e r ~ ~ ~ are the only workers to use an oxygen-flask combustion as a means of decomposing biological products prior to the determination of selenium with 2,3-diaminonaphthalene.Details of the procedure used are given and the results obtained are tabulated in their paper. 2,S-Diaminonaphthalene has also been used to determine the selenium content of cast ir0n.~~1 The metal is dissolved in a mixture of perchloric and nitric acids and the oxidation is completed by the addition of an excess of sodium perman-ganate the excess then being removed by the drop-wise addition of sodium nitrite. As the piazselenol is extracted into cyclohexane from acid solution only iron(II1) interferes and this interference is easily removed by the reduction of iron(II1) to iron(I1) by hydroxylammonium chloride.The method is applicable in the range 0.01 to 0.1 per cent. of selenium and at the higher limit the standard deviation is reported to be 0.003 per cent. Prior to the extraction and measurement of the fluorescence intensity of the piazselenol complete reaction is ensured by heating the reaction mixture at 50 "C for 30 minutes. The results obtained from the analysis of various cast-iron samples are reported. Several other reagents o-phenylenediamine 1,4-diphenylthiosemicarbazide, dithizone diantipyrinylmethane and 3,3'-diaminobenzidine have been examined as reagents for the fluorimetric and spectrophotometric determination of selen-ium.242 Dithizone is reported as being the most useful forming a complex with selenium in both 6 M hydrochloric acid and 5.5 M sulphuric acid.The complex can be extracted into carbon tetrachloride from these acid solutions. Results are presented on the simultaneous fluorimetric determination of selenium by reaction with dithizone and tellurium by reactions with Butylrhodamine S after the separation of both elements by co-precipitation with arsenic. Patrias an 86 BARK AND WOOD Despite the fact that both 2,3-&aminonaphthalene and 3,3'-diaminobenzidine have been used as fluorimetric reagents for the determination of selenium for some years there is much conflicting evidence as to the best methods of digesting the materials and separating the selenium when necessary.The work of Hall and G ~ p t a ~ ~ ~ examines the possible sources of error and of the papers reviewed is probably the most detailed investigation of the optimum condition for the deter-mination of selenium. For biological samples the best method of decomposition is apparently very dependent on the type of material thus a detailed examination of the literature should be made to establish the best method for the type of sample to be analysed. Scandium yttrium and the lanthanides Many papers have been published on the luminescent determination of the metals of this group. The lanthanides have been determined in crystalophosphors by using a variety of excitation techniques including cathode rays X-rays and electron probes. However these methods are not included in this review and for further information readers are referred to the reviews of White and Wei~sler.~-~ The polyhydroxyflavones morin and quercetin react with scandium under certain conditions to form fluorescent complexes and because these reagents have been used to determine scandium the of the composition of the complexes is of considerable interest.The complexes were studied spectrophoto-metrically by using the isomolar series molar ratio and isobestic point methods and in all cases the complex is shown to be of the form ML and the complexed reagent molecule is the singly charged polyhydrox yflavone anion. Two complexes are formed with morin a fluorescent and mono-positively charged complex at a pH between 1.2 and 2.8 and a non-fluorescent neutral complex at a pH between 6.0 and 7.6.With quercetin two fluorescent complexes are formed. One is positively charged (pH 5-5 to 6.0) and the other neutral (pH 6.0 to 6+) and a further complex which is formed between pH 7.6 and 8.8 is neutral and non-fluorescent. The co-ordinating scandium ion in the positively charged species is ScOH2+ whereas the co-ordinating ion in the neutral species is Sc(0H)i. These authors have also reported253 the formation of a morin - antipyrine complex that can be extracted into chloroform from a solution containing perchlorate ions at a pH of 3.3 to 3-4. The ions aluminium gallium indium titanium thorium, zirconium and lutetium interfere by forming extractable complexes under the conditions used for the determination of scandium whereas yttrium and most lanthanides which interfere with the direct method involving the use of morin, do not interfere except when present in fairly large excess.The sensitivity of the method developed on the above basis is 0.01 pgml-l of scandium and the ratio of scandium morin antipyrine perchlorate is reported to be 1 1 3 1. A method for the determination of scandium by using salicylaldehyde semi-carbazone2M has been applied for the determination of scandium in materials of fairly complex composition.255 The procedures necessary for the separation o PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 87 scandium from interfering ions are reported and by them 0-03 per cent. of scan-dium may be determined with a relative standard deviation of 10 per cent.A comparison256 of the scandium complexes of the 5,7-dihdo derivatives of 8-hydroxy-quinoline showed that the dichloro-derivative gave the most intense fluorescence, and thus a method was developed for the determination of scandium by using this derivative as the reagent. The complex formed in aqueous solution at pH 9.5 is extracted into chloroform and its fluorescence intensity measured. The method is reported to be more sensitive than that using 8-hydroxyquinoline and to have an error of &3 per cent. Yttrium lanthanum and lutetium interfere but the scandium content of rocks can be determined after the removal of iron(II1) by extraction with 5,7-dichloro-8-hydroxyquinoline at pH 3.0 and concentration of the scandium by co-precipitation with calcium oxalate.The determination of yttrium by using 5,7-dibromo-8-hydroxyquinoline has been reported.257 The complex that is formed between yttrium and this reagent at pH 6.5 in aqueous solution is extracted with benzene and the fluorescence intensity of the organic phase is measured. However because aluminium also forms a fluorescent chelate with the reagent the intensity is dependent on the total concentration of both metals. The addition of EDTA results in the decom-position of the yttrium - quinolate complex and hence after the addition of EDTA the fluorescence is that of the aluminium complex alone and the yttrium concentration can then be determined by difference. The method is reported to be sensitive to 0.002 per cent. of yttrium oxide in lanthanum oxide. 4-Dicarboxymethylaminomethyl-2-hydroxy-3-naphthoic acid has been pro-posed as a reagent for the determination of beryllium lanthanum and lutetium.1sB The development of the method which is applicable to the determination of between 7 and 28 pg of lanthanum or lutetium is reported in detail.The procedures for the determination are simple requiring only the adjustment of the pH of the test solution to a pH between 4.0 and 7.0 followed by the addition of the reagent and adjustment of the pH to 10.0 by a hexamine - perchlorate - perchloric acid solution and sodium hydroxide. Many foreign ions in amounts greater than 0.5,umole interfere seriously and the authors suggest that these could probably be removed by ion exchange or extraction with tri-iso-octylamine in xylene. Lanthanum and lutetium can be determined however in the presence of a three-fold molar excess of the other lanthanides.This is the only published method during this period that describes the determination of these particular lanthanides. Cerium(II1) has been determined by using the fluorescence of its halide complexes in acid solution and methods using this fluorescence are reported in the section of this review dealing with the determination of metal ions as inorganic complexes. 4- [ (2,4-Dihydroxyphenyl)-azo]-3-hydroxy-l-naphthalene sulphonic acid has been used as a reagent for the determination of cerium(III).258 The reagent itself does not fluoresce whereas its cerium complex does. The effects of buffer solution and solvents on the fluorescence intensity of the complex were examined and the results obtained show that the maximum fluorescence occurs in the presence o 88 BARK AND WOOD acetate or hexamethylenetetramine buffers at pH 4-5 in a 75 per cent.v/v acetone -water mixture. The composition of the complex was shown to be 1 l l of cerium(II1) - reagent - acetate by Job’s method. The only reported interference is that caused by the presence of thorium(IV) which also forms a fluorescent complex and an ion-exchange procedure that is described in detail is required to separate these ions from each other. This method is sensitive to 0.25 pg ml-1 of cerium(II1). Cerium(1V) The rate of the chemiluminescent reaction between luminol hydrogen peroxide and copper(I1) ions is decreased in the presence of cerium(1V) ions.This effect is the basis of a method proposed for the determination of c e r i ~ m ( I V ) ~ ~ ~ at con-centrations greater than 0.02 pg ml-l (with a relative error of &15 per cent.). The authors suggest that the reduction in rate is the result of the formation of a competing triple complex formed between luminol hydrogen peroxide and cerium(1V) ions. This complex allows the formation of the copper (11) complex to proceed to give a chemiluminescent reaction but at a much slower rate and hence over a measurably longer period of time. An investigation260 of the reaction of cerium(1V) with siloxene (polymeric siloxane) has led to the development of a method for its determination. Measure-ment of the chemiluminescence resulting from this reaction in sulphuric acid allows 7 pg ml-l of cerium(1V) to be determined.The lanthanide ions P9+ Sm3+ Eu3+ Gd3+ Tb3+ Dy3+ Ho3+’ E9+ and Tm3+ have been determined261 by their fluorescence in yttrium oxide in a powder form. A mixture of the lanthanon and yttrium oxides is dissolved in hot nitric acid, and the metal oxalates are then precipitated by the addition of oxalic acid. The precipitate is heated at 1200°C for 1 hour giving a powder with an average particle size of 5 pm. The excitation and emission wavelengths the transitions occurring during luminescence and the detection limits for the determination of these ions are tabulated and discussed. The lower limit of detection is of the order 10-8 to moles of lanthanon oxide per mole of yttrium oxide and the upper limit of linearity of the calibration curve is approximately moles.The lower limits of gadolinium and samarium are somewhat higher than that for the other elements (10-5 moles) because of the poor emissivity of the powders produced. During the period under review the majority of the papers published on the determination of the lanthanides were concerned with the determination of samarium europium gadolinium and terbium in mixtures of the lanthanide oxides. A procedure for the chromatographic separation of the lanthanides and their subsequent determination by a spectrographic or luminescent method has been reported.262 The separation which is achieved by chromatography using a silica gel with an impregnant of bis-(2-ethyl-hexyl) phosphoric acid as a reversed stationary phase followed by elution with various concentrations of hydrochloric acid is reported in detail whereas the subsequent luminescent analysis based on the formation of a vanadium trioxide - lanthanide crystalophosphor is not.Th PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 89 authors do however report the sensitivities as 2 x per cent. for gadolinium and terbium 5 x per cent. for dysprosium and 4 x per cent. for europium. A method for the determination of samarium europium dysprosium hol-mium erbium and thulium in yttrium oxide has been reported.263 The maximum luminescence of the yttrium orthovanadate based phosphors prepared by calcina-tion of yttrium oxide and ammonium vanadate at 900" to 1100 "C occurs when the yttrium-to-vanadium ratio is 1-25 to 2.0 and not at the stoicheiometric ratio of 1 l .The details for the analytical procedure by the method of standard additions are given. The sensitivity is reported to be (1.6 to 8.2) x per cent. for these ions in yttrium oxide. The samarium content of cerium(1V) oxide has been determined2a by measuring the fluorescence of a lead sulphate - lithium fluoride phosphor which is prepared by mixing the cerium(1V) oxide with lead sulphate and a solution of lithium fluoride. This mixture after drying is heated at 860 "C for 30 minutes. The concentration of samarium is determined by measuring the luminescence of the phosphors prepared both with and without the addition of samarium the difference in signals corresponds to the original amount of samarium present. Details of the sensitivity of the method are not available.Several methods have been reported for europium and samarium in which the fluorescence of a lanthanide - ,8 diketone - organic base ternary complex is measured. These complexes exhibit well separated line-like spectra thus allowing the determination of these ions in mixtures. The fluorescent complex formed by europium 1 ,lo-phenanthroline (phen) and thenoyl-trifluoroacetone (TTFA) [Eu (TTFA) phen] has been used for the determination of europium in a variety of salts rocks and waters.265 The complex is formed in aqueous ethanol solution and is then extracted into benzene and its fluorescence intensity measured. The method can apparently be used to determine between lo- and lo-' per cent. of europium in aqueous solution and can also be used to determine samarium.This method has been reported in a paper266 that also describes the determination of these metals by using their fluorescent ternary complexes formed with 1,lO-phenan-throline and 2-phenylcinchoninic acid. Few details are given in the abstract available; however it is reported that when using the phenanthroline - thenoyl trifluoroacetone method the benzene extraction procedure is less sensitive than the procedure involving the direct measurement of the fluorescent intensity of a suspension of the complex in aqueous alcohol. The direct procedure is however, subject to more interference from the presence of foreign ions including several lanthanides. The possibility of measuring the fluorescence of the ternary com-plexes [Eu TTFA-phen] in the presence of interfering ions without solvent extraction has been investigated by Kononenko et 41.267 They report that acetone and dioxan are suitable solvents for the precipitate whereas it does not dissolve on the addition of ethanol or methanol.The effect of using various bases is also reported but the method is most sensitive when using TTFA and 1,lO-phenan-throline. The optimum solvent mixture i s 1 3 of water - acetone and by usin 90 BARK AND WOOD the method of standard additions 0.17 ng ml-l of europium and 17 ng ml-1 of samarium can be determined. The results obtained for the analysis of various rare earth oxides by using this modified method are comparable with those obtained by using the extraction procedure. A method has been described2*8 for the determination of samarium and euro-pium on the basis of the fluorescence of their complexes with TTFA collidine and diphenylguanidine.The complexes are of the form [M (TTFA),HB] where M represents the metal and B the base. The optimum pH for the determination is 6.5 to 7.5 and the fluorescence intensity of either a suspension of the complex in aqueous ethanol or of a benzene extract can be measured. The direct method is sensitive to 1 to 5 x per cent. of Eu203 and from 1 to 10-3 per cent. of Sm,O in the oxides of lanthanum gadolinium terbium yttrium ytterbium and lutetium ; other lanthanides interfere. The extraction method is somewhat less sensitive as only 5 to 10 x 10-4 per cent. of Eu203 and 2 to 5 x 10-2 per cent. of Sm20 can be determined however it is more selective for the determination of the elements than is the direct procedure.In both procedures the methodof standard additions is used to obtain accurate results. A method using the tri-octylphosphine oxide adduct of the europium complex of benzoyltrifluoroacetone has been reported269 for the determination of europium. The complex formed in aqueous solution at pH 4.5 is extracted into n-hexane by shaking it for 30 minutes and the fluorescence is measured against Rhodamine B, which is used as an internal standard. Samarium and iron(II1) interfere with the method which is applicable to the determination of concentrations of europium less than 5 x M. The complexes formed between 2-naphthoyltrifluoroacetone (NTFA) and samarium or europium can be quantitatively extracted from an aqueous solution at pH 6.0 with benzene by using tri-octylphosphine a synergistic agent.The NTFA complexes of these ions are more fluorescent than the corre-sponding benzoyl- and thenoyl-trifluoroacetone complexes and this has been utilised for the determination of samarium and europium.270 The effects of pH, reagent concentration temperature and irradiation time are reported. Few foreign ions interfere iron(II1) causes a negative error whereas large amounts of samarium cause a positive error in the determination of europium. Because of the high fluorescence intensity of the europium complex compared with that of the samarium complex europium interferes with the determination of samarium even when present in small amounts.However by measuring the fluorescence intensity of both complexes and then the intensity of the europium complex separately both metals can be determined in mixtures. The authors report that the sensitivity is 0.1 ng ml-1 for europium and 0.1 pg ml-l for samarium and include in this paper results of the analysis of samarium and europium and the analysis of europium in artificial mixtures of the lanthanides. Belcher Perry and Stephen2'l have investigated the formation of complexes of europium with /I-diketones in dimethyl formamide solution. The detection limits, excitation and emission wavelengths and the range of linearity of the calibration curves are reported for the various /3-diketones examined for the determinatio PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 91 of europium.Of the compounds examined TTFA gives the highest sensitivity (0.75 ng ml-l) and a detailed procedure for the determination of europium by using TTFA in dimethyl formamide solution is reported. A solution containing europium is evaporated to dryness and the residue is dissolved in dimethyl formamide. After adjusting the pH of this solution to 7.5 with ammonium hydroxide solution TTFA is added and the intensity of the fluorescence produced is then measured. The authors also report that europium and terbium can be determined simultaneously in the range M by measuring the fluores-cence of their complexes formed with dimethyl formamide alone. Results obtained by the use of these methods are reported. The luminescence of europium(II1) in borate glasses formed by heating a borax - boric acid mixture at 1100 "C for 15 minutes has been investigated.272 The origins of this luminescence are discussed and the authors conclude that the best procedure for the determination of europium(II1) in these glasses is by the measurement of the emission at 617 nm with excitation at 395 nm (conditions under which the calibration is linear over the range 0.001 to 2 per cent.).A similar paper273 reports the determination of gadolinium (10 p.p.m. to 20 000 p.p.m.) in borate glasses. Terbium has been determined by its fluorescence in a yttrium oxide - sodium sulphate phosphor prepared by heating yttrium sulphate and sodium sulphate at 1OOO" to 1050 0C.274 The intensity of the fluorescence is directly proportional to the terbium concentration up to 0-6 per cent.At higher concentrations a plot of log (intensity) against log (concentration) is linear. The presence of equal amounts of other lanthanides does not cause interference in the determination of 2 x per cent. equal amounts cause serious interference by reducing the intensity. A number of yttrium oxide samples were analysed by using this method and the results are reported with the corresponding mean deviation which varies from 0 to 18 per cent. Dysprosium has been dete1mined~7~ by using the pyrazalone derivatives, 3-methyl-1-phenylpyrazol-5-one and 3-methyl-1-tolylpyraol-&one. The rare earth oxide (250 pg) is mixed with a solution of hexamine and the reagent and after allowing the mixture to stand for 40 minutes the fluorescence of the super-natant solution is measured.This method is sufficiently sensitive to allow the determination of 0.005 per cent. of Dy203 in Gd203; 0.4 per cent. in Sm203 and 0.08 to 0-1 per cent. in the oxides of cerium neodymium or thulium. Pyrazolinone derivatives (4-sulphophenyl- 3-methyl- phenyl-3-methyl- tosyl-3-methyl- 5-pyrazolinone) and an antipyrine-sodium salicylate mixture have been proposed276 as reagents for the determination of terbium and dysprosium in the tributyl-phosphate solutions that can result from the separation of rare earths from admixture. The preparation and the use of bis- [ l-(2-pyridyl)-3-methyl-5-pyrazolonyl]-4,4'-methane for the determination of terbium and dysprosium have been des-~ribed.~77 The spectral properties of the terbium dysprosium and samarium to per cent.of terbium whereas for the determination of 2 x 92 BARK AND WOOD complexes and the dependence of the fluorescence intensity on pH metal-to-ligand ratio and the presence of other lanthanides are reported in this very detailed paper. Results obtained for the analysis of rare earth oxides are presented. The effect of foreign lanthanide ions on methods for the determination of samarium euro-pium terbium and dysprosium has been inve~tigated.~'~ The methods examined were those using Atophan (2-phen ylquinoline-4-carboxylic acid) or TTFA with 1 :lo-phenanthroline phenyl- and tolyl-3-methyl-5-pyrazolinones 4-sulphophenyl-3-methyl-5-pyrazolinone and antipyrine salicylates. It is noted that methods in which the technique involves the measurement of the fluorescence of the complex present as a suspension or as polydentate complexes containing two atoms of the lanthanide element per mole are subject to interference by the other rare earths, whereas methods in which the fluorescence of solutions of the relatively simple complexes is measured are not susceptible to such interferences.Theoretical reasons for these interferences are reported. Kononenko et aZ.279 have established the conditions for the determination of terbium in mixtures of lanthanide chlorides with phenyl salicylate as the fluoro-phore. The optimum pH for the formation of the fluorescent complex is 9.3 (glycine buffer) and the method which is sensitive to 0-006 per cent. of terbium with respect to the composition of the final mixture requires that solutions are allowed to stand for 20 minutes before the fluorescence intensity is measured.The method of standard additions is used and the mean error is k4.5 per cent. The results obtained for the analysis of various materials containing terbium are reported. A similar methodZ80 for the determination of terbium in the range 0.0064 to 3-2pgml-1 is based on the fluorescence of the sulphosalicylic acid-EDTA complex of the metal. Under the conditions described at pH 11.9 (diethylamine -hydrochloric acid buffer) no interference is caused by the presence of the other lanthanides or by the thirty-three metals or fourteen anions tested. Interference is however caused by the presence of 1000-fold excesses of lead rnercury(1) and uranium(V1).Kirillov Vitkun and PoluektovB1 have described a method for the deter-mination of thulium by using a calcium fluoride phosphor. The phosphor is pre-pared via the precipitation of calcium fluoride from a solution of calcium nitrate by a solution of thulium in hydrofluoric acid. The precipitate is then annealed at 1050" to 1100 "C for 5 to 20 minutes depending on the yttrium content of the material being analysed. The fluorescence intensity varies linearly with concentration of thulium for up to 200pg in 200mg of calcium fluoride. The elements of the cerium sub group with the exception of lanthanum and neo-dymium interfere. The sensitivity of the method is reported to be between 0.001 and 0.20 per cent. depending on the composition of the material to be analysed.Three methods involving spark phosphorimetry for excitation of crystal phosphors containing lanthanides have been reported during this period. Gadolinium, samarium and europium have been determined in metallic uranium282 by using luminophores based on yttrium oxide for gadolinium and yttrium vanadate for europium and samarium. Details are given of the prior concentration of thes PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 93 elements by ion-exchange chromatography and of the preparation of the phosphors. The methods are sensitive to 2 x 10-6 per cent. with a standard deviation of 30 per cent. Gadolinium has also been determined2s3 by the luminescence of its silica based phosphor. The effect of various salts on this luminescence was investigated and of these sodium sulphate is reported to cause the largest increase.Thus the phosphor is based on silica - sodium sulphate mixtures that are fused at 1050 "C in the presence of gadolinium. The only other lanthanide to exhibit luminescence under these conditions is terbium although uranium(V1) exhibits a green lumines-cence. Serious quenching is caused by the presence of small amounts of cerium, chromium cobalt copper iron and manganese. This method which is sensitive to 5 x per cent. of gadolinium requires that the metal is concentrated before its determination. Melamed et aL2= have proposed a method for the simultaneous determination of gadolinium terbium dysprosium and europium in yttrium oxide phosphors which are excited by a condensed spark between the tungsten elec-trodes of a spark phosphoroscope.The phosphors emit at widely differing wave-lengths depending on the lanthanide vix. gadolinium 311 nm; terbium 543 nm ; dysprosium 571 nm; and europium 602 nm. This enables their determination in mixtures. The method of preparation of the phosphors is described and involves precipitation of the yttrium and lanthanide as oxalates followed by calcination at 1300°C. Titanium zirconium and hafnium Few publications have appeared on the fluorimetric determination of the metal ions in this group. T i t k o ~ ~ ~ ~ has patented a luminescent method for the deter-mination of titanium. The metal is extracted as its ternary complex with hydrogen peroxide and 8-hydroxyquinoline using chloroform as the extractant.The organic phase is washed with sulphuric acid and then re-extracted with a solution of sodium tetraphenyl borate in dilute sulphuric acid. The fluorescence of the resulting diphenylborylhydroxyquinolate is proportional to the amount of titanium originally present. Two methods have been described for the determination of zirconium. Babko et aZ.286 measured photographically the inhibition by zirconium ions of the chemiluminescence of the copper - hydrogen peroxide - luminol system and used this inhibition for the determination of the zirconium. The possible reasons for the inhibition are discussed. At the optimum pH which is in the range pH 9 to 11.5 the formation of a sparingly soluble 1 1 zirconium - hydrogen peroxide complex is suggested. Thus hydrogen peroxide is withdrawn from the reaction mixture and the amount of luminescence from the system is therefore reduced.The effects of variations of the concentrations of the individual components on the inhibiting effect of zirconium are reported in detail. Calibration curves at both pH 10 and 10.5 were obtained and within a given range a linear relationship between the zirconium concentration and the total luminescence is obtained. The sensitivity is reported a s 0.01 pg ml-1 with a relative error of &25 per cent. A 94 BARK AND WOOD pH 11.0 the same sensitivity is obtained. One of the main disadvantages of this method a disadvantage that is common to all chemiluminescent methods involv-ing photographic measurements is the need for a dark room. Although inter-ferences are not reported it is known that vanadium and titanium also show an inhibition of this system and thus would probably interfere in the determination of zirconium.The zirconium content of industrial samples has been determined by using m0rin~~7 as the reagent. The samples are pre-treated by decomposition with hydrofluoric and sulphuric acids. After evaporation to dryness the residue is fused with sodium bicarbonate and extracted with water. The resulting residue is dissolved in nitric acid and the solution is then extracted with tributyl phosphate. On shaking the tributyl phosphate with an alkaline solution of EDTA the metal is re-extracted into the aqueous phase. Hydrochloric acid (sufficient to give a final concentration of approximately 7 M) ascorbic acid and morin are added.The fiuorescence intensity which is stable for 1 to 2 days is then measured. This method can be used for the determination of zirconium in the range 0.002 to 0.004 per cent. with a deviation of 0.001 per cent. The method seems simple in operation. Quercitin has been used for the fluorimetric determination of zirconium288 in 8 per cent. of ethanol and 2.4 M hydrochloric acid. Increasing the acidity and reducing the ethanol concentration causes a decrease in fluorescence of the zirconium - quercitin complex. Brookes and Town~hend~~~ report that in 9 M perchloric acid containing 2.5 per cent. of ethanol zirconium does not form a fluorescent complex with quercetin whereas hafnium gives an intensely fluorescent species. The chemical similarity between zirconium and hafnium makes the determination of one in the presence of the other difficult thus this method, which is sensitive to 1 to 2 pg of hafnium is extremely useful.At concentrations of above a 5-fold molar excess zirconium interferes by producing a weak fluores-cence. The authors report that it is a simple matter to compensate for this slight increase in intensity but do not give details of the procedure to be followed. Vanadium niobium and tantalum No methods for the determination of either niobium or tantalum have appeared in the literature during this period and only one method for the determination of vanadium has been reported.290 This novel method depends on the shortening of the inhibition period of the reaction between bromate bromide and ascorbic acid by the catalytic action of vanadium.The bromine liberated by the reaction quenches the fluorescence of Rhodamine B cresyl violet or trypaflavin (XVIII) at pH 5.0 and of acridine at pH 5.0 and 2.0 (150 compounds were examined as indicators for this system). The use of the method of simultaneous comparison with a series of standard solutions allows the determination of 0.02 to 20 pg ml-l of vanadium. The inhibition period (t) measured by observing the quenching of fluorescence of the added dye is inversely proportional to the vanadium concen-tration so that a linear calibration plot is obtained by plotting the concentratio PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 95 (XVI II)* Try paflavin of vanadium against l/t.The paper includes detailed descriptions of the method and of the interference caused by foreign ions. Chromium molybdenum and tungsten A detailed investigation of the properties of six C-substituted bis-triazinyl amino stilbene derivatives has led to a method for the quenchimetric determination of chromium( 111) .2g1 The glycine substituted derivative (triazinyl stilbexone) (XIX) 9 p \ R' R (XIX) R = -NHCH2COONa Triazinyl stilbexone was shown to be the most sensitive reagent for chromium(II1) over the pH range 2.5 to 3.5. The reaction cannot be explained by a stoicheiometric reaction between stilbexone and chromium(II1) and the authors suggest that the reaction may be catalytic ; during the reaction the chromium(II1) is presumably converted to an unreactive state.From an examination of the absorption spectrum of the reagent in the presence of chromium(II1) and the apparent lack of reaction with oxidising and reducing agents it is suggested that redox reactions are not involved The dependence of the intensity of the fluorescence on the chromium concentration is not linear. In the presence of EDTA (1 x lo-* M) 0-1 pg ml-l of chromium can be determined in the presence of a 200-fold excess of aluminium germanium, cobalt nickel tungsten(IV) platinum(IV) zinc manganese and chromium(V1) and in 20-fold amounts of silver calcium iron(III) molybdenum(V1) palladium(II), selenium(V1) and vanadiumo. Amounts greater than 2 x moles of alkali metal salts and of the following anions I- C1- F- NO,- POaS- SO,2- S2-, tartrate nitrilotriacetate peroxidisulphate bromate and thiosulphate interfere.In the same paper procedures are reported for the determination of chromium in hydrochloric acid and in germanium tetrachloride. The measurements are made 10 minutes after mixing the reagent and sample solutions. Tables showin 96 BARK AND WOOD the accuracy and reproducibility of the method are given. The method is very sensitive (4 ng ml-l) and is simple in operation. However interference from fairly low concentrations of both alkali metal salts and the very large number of anions would seem to place severe restrictions on the use of the method for any samples of complex composition. Another disadvantage is the apparent necessity to produce a calibration plot for each series of determinations an analytical procedure not likely to be favoured in many laboratories.The only method described for molybdenum is a titrimetric one involving a fluorescent indicator,292 and is reported in the section of this review dealing with fluorescent indicators. A previously reported method293 for the determination of tungsten in steels by using the fluorescence of the tungstate - flavonol (XX) complex suffers from 0 ow Flavonol the very severe practical restriction that nearly all of the metals commonly present in steels cause interference with the tungsten determination in one way or another. Bottei and T r ~ s k ~ ~ ~ have designed an ion-exchange separation procedure for tung-sten which enables it to be determined in steels by the above meth0d.~93 After dissolution of the steel in hydrochloric and nitric acids the tungsten is separated by an ion-exchange procedure that is reported in detail.The sodium form of Dowex 5W-X4 50 to 100 mesh is used for the preparation of the ion-exchange column. As the cations in the solution of a steel are removed by the ion-exchange resin only the effect of anions present in the effluent was investi-gated. The most serious interference is caused by chromate although treatment of the solution with formaldehyde before the removal of cations reduces chromium(V1) to chromium(II1). This is then removed by the ion exchanger. Vanadates and molybdates also interfere and despite the possibility of masking the ions with potassium cyanide it is recommended that this method be limited to steels in which the vanadium-to-tungsten ratio does not exceed 2 and the molybdenum-to-tungsten ratio does not exceed about 4.Within these limitations the method gives accurate results that are comparable with the reported tungsten contents of the various steels investigated. The same authors have extended the use of flavonol to the determination of tungsten in almost pure tungsten and cobalt -nickel - chromium refractory Detailed procedures for the dissolution o PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 97 these materials and the subsequent determinations with flavonol are reported. As with the previously reported method chromium(V1) is reduced to chromium(II1) by the action of formaldehyde. The interfering ions are then removed by precipi-tation as hydroxides followed by filtration or centrifugation.The main difficulty encountered in the analysis of the refractory alloys was their dissolution for which a mixture of perchloric and sulphuric acids must be used. The results obtained from the analysis of these materials are given and agree with the reported tungsten contents. The refractory alloys examined contain approximately 4 per cent. of tungsten and the authors suggest that samples containing larger or smaller quantities could be analysed by this method. Manganese technetium and rhenium During this period the only metal of this group for which fluorimetric methods of determination have been reported is manganese. Methods for the determination of manganese by its use as an activator in inorganic phosphors have been reported and are included in the section of this review dealing with the use of phosphors in luminescence analysis.Pal and R ~ a n ~ ~ 6 have reported in detail a method for the determination of manganese based on the oxidation of 8-hydroxyquinoline-5-sulphonic acid by the permanganate ion to produce a highly fluorescent water soluble species that could not be isolated or identified. The dependence of this method on the presence of permanganate means that prior oxidation of the manganese is essential. This is achieved by a catalytic oxidation using silver(1) and ammonium peroxidi-sulphate in phosphoric acid solution. At the low concentrations of silver(1) (lo-* to M) used it is necessary to boil the solution for 1 to 2 minutes to ensure complete oxidation of the manganese.However the boiling time is not critical and there is no difference in the fluorescence intensity for heating times between 1 and 10 minutes. Following the oxidation a 5 to 20 molar excess of the reagent is added and the fluorescence intensity measured after allowing the mixture to stand at room temperature for between 10 minutes and 1 hour a period over which no change in the intensity is observed. The intensity is constant over a wide range of acidities (M phosphoric acid is normally used) however above pH 5.0 a bathochromic shift of the fluorescence maximum is observed. The authors report that the method is applicable to the determination of manganese in the range 2.5 ng ml-l to 2.5 pg ml-l with a standard deviation of 1.9 per cent. The only serious interference is caused by a 50-fold excess of cerium(1V) ; although both thorium and cerium are precipitated as phosphates the thorium interference can be removed by centrifuging.The interference levels of other ions are reported and the results obtained from the analysis of four steel samples are given and these agree with the reported manganese contents. A chemiluminescent method based on the catalytic effect of manganese on the reaction between luminol and hydrogen peroxide has been reported.297 The optimum pH is 9-6 and is maintained by using an ammonium hydroxide - ammonium chloride buffer that is purified before use by the extraction of any manganese present with a chloroform solutio 98 BARK AND WOOD of 8-hydroxyquinoline. The selectivity of the method is increased by the addition of 1 ,lo-phenanthroline and sodium citrate although vanadium (as ammonium metavanadate) nickel cobalt chromium(II1) and iron(II1) ions interfere.The iron(II1) interference can be removed prior to the addition of the reagent by irradiating the solution with a 250-W mercury lamp at a distance of 10cm for 5 minutes. The luminescence is measured photoelectrically over a period of 5 minutes. The method is applicable to the range 0.05 to 5ng of manganese. This method is the subject of a patent.298 The manganese content of germanium tetrachloride and trichlorosilane has been determined.299 The method used is based on the catalysis by manganese of the oxidation of the magnesium chelate of lumomagneson by hydrogen peroxide. No details of the measurement of the increased rate of oxidation which results in a reduction in the fluorescence intensity by decomposition of the chelate are reported in the abstract available.The optimum conditions are pH 11.0 at a hydrogen peroxide concentration of 0.01 M, and give a sensitivity of 0*06ngml-.-l. The ions that interfere with the deter-mination of 0.01 lug of manganese are reported. Iron Temkina et aLm have prepared and examined as analytical reagents a series of twelve new fluorescent complexans each containing the iminodiacetic acid grouping. They propose that several of these compounds may be used as reagents for the quenchimetric determination of iron(I1) (and copper) with which they form complexes of the type ML,. Another method for iron is based on the quenching of the fluorescence of 2,2’,2”-terpyridine by reaction with iron(I1) .301 The optimum pH for the determination is 3.6 and in such conditions the reagent will be present as the monoprotonated form.Even though the iron(II1) - terpyridyl complex will be self reducing (in a manner analogous to the auto-reduction of the iron(II1) - 1,lO-phenanthroline complex) the authors consider it necessary to have only one oxidation state of iron present and accordingly propose the prior reduction of any iron(II1) to iron(I1) by using hydroxylamine as the reductant. The method is reported to be applicable in the range 50 to 500 ng ml-l (it is not clear from the literature available whether thisis the final concentration or the concentration of the original sample solution).The standard deviation reported is &5 per cent. Although cobalt(II) copper(I1) and nickel(I1) interfere the interference can be removed by preliminary extraction of the iron(II1) from 6~ hydrochloric acid solution by using a suitable organic extractant and then re-extraction of the iron(II1) into a more dilute acid solution. (The review authors recommend the use of /3/3’-dichloro ðyl ether and M hydrochloric acid as the dilute acid for re-extraction.) Babko and Kalinichenkom2 have determined iron impurities in sodium chloride by a chemiluminescent method. The oxidation of luminol by hydrogen peroxide in the presence of tri-ethylene-tetramine in the pH range 7-5 to 11.0 is catalysed by the presence of iron(III) and the luminescence resulting from this reaction is dependent upon the concentration of luminol hydrogen peroxide an PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 99 iron(II1).It is however independent of the concentration of the amine within the concentration range of 3 x M. It is reported that cobalt interferes and that the error is between 30 and 50 per cent. The compound stilbexone (4,4‘-diamino-N,N,N’,N’-tetrakis carboxymethyl stilbene-2,2’-disulphonic acid) originally p r o p o s e d ~ 3 ~ ~ as a reagent for the kinetic determination of iron(II1) has been used to determine iron in germanium tetra-chloride,299 trichlori~ilane2~9 and calcium ~ulphide.~Os The oxidation of stilbexone by hydrogen peroxide to give a non-fluorescent product is accelerated by iron(III), thus by measuring the rate of the oxidation reaction the amount of iron(II1) can be calculated.The rate is unaffected by small amounts (0.04 to 0.06 g) of calcium sulphide so that prior separation of the iron(II1) is not proposed. The sensitivity is reported to be 1 ng ml-l of iron(II1). to Cobalt A method for the determination of cobalt based on its quenching of the fluorescence of the aluminium-Pontachrome Blue Black R complex has been rep0rted.~06 The optimum conditions for the operation of the method which is sensitive to 1 ng ml-l of cobalt(II) are described. The sensitivity can be improved by extracting the aluminium chelate into iso-amyl alcohol before measuring its intensity. The effects of more than sixty metal ions on the fluorescence intensity are reported and the procedures necessary for the determination of cobalt(I1) in the presence of interfering ions are given.During an investigation of the quenching of the fluorescence of the aluminium 1-(2-pyridylazo)-2-naphthol (PAN) chelate by nickel Schenk et aL307 observed that the cobalt - PAN chelate fluoresced even though the paramagnetic (d7) cobalt ion would be expected to cause a quenching of the fluorescence. The authors suggest that this anomaly may be caused by aerial oxidation of the cobalt(I1) to cobalt(II1) in the presence of nitrogen donor ligands and that cobalt(II1) would be diamagnetic in the presence of a strong ligand field such as PAN. These sug-gestions were based on the observed colour change of the cobalt - PAN chelate over a period of 30 minutes and the reaction of cobalt(II1) with PAN to give the same absorption and fluorescent spectra as obtained from the reaction of PAN with cobalt(I1).This reaction to form a fluorescent chelate can be used to determine cobalt in the concentration range of 1 x to 1 x l o - 4 ~ with a relative standard deviation of 5.6 per cent. However because a large excess of PAN is required the determination must be done in almost absolute ethanol, and unless the cobalt to be determined is present in ethanolic solution it is un-likely that the reported sensitivities could be achieved. The authors do however, suggest that this method is of mainly theoretical interest because of the formation of a fluorescent cobalt complex. The colorimetric procedure using PAN is only slightly less sensitive but the fluorimetric method is reported to offer some advantage in that it can be used in the presence of aluminium cadmium iron(II1) and nickel(I1) without prior separation or the use of masking agents.Th 100 BARK AND WOOD interferences caused by foreign ions are reported and it is suggested that these are caused. by competition with cobalt for the PAN molecules. The cobalt content of germanium tetrachloride and trichlorosilane has been determix~ed~~g by using a previously reported kinetic method308 that is based on the catalysis by cobalt of the oxidation of salicylfluorone with hydrogen peroxide. Nickel In a review of the use of luminescent solids in analysis Holzbecher and Novak309 have described a method for the determination of nickel and this is reported in the appropriate section of this review.The only other method that has been reported is that of Schenk et aL307 who used the quenching of the fluores-cence of the aluminium chelate of PAN by nickel for its determination. The paper gives the details of the method which is reported to be applicable in the range to ~O-’M with a standard deviation of 7.7 per cent. A stock solution of aluminium- PAN prepared by allowing a mixture of ethanolic solution of aluminium and PAN to stand for 24 hours is reacted with a nickel solution by heating at 40 “C for 40 minutes or by allowing the mixture to stand at room temperature for 4 hours and the fluorescence intensity is measured. Preparation of fresh stock solutions of the reagent requires that new calibration curves are obtained.The nickel solution used by the authors is prepared in 95 per cent. of ethanol and no description or suggestions for the determination of nickel in aqueous solutions are made. An investigation of the effect of some metal ions on the determination is reported and of the ions examined only chromium inter-feres seriously. No investigation of the effect of anions is reported. This method, although very sensitive is unlikely to be used for the analysis of complex samples because of the apparent need to obtain solutions of nickel in ethanol so as to obtain the reported sensitivity . Platinum group metals During the period under review only one method has been reported for any of these metals. Iridium(II1) has been determined310 by using 2,2’,2”-terpyridine in the range 2 to 20 p.p.m.with an accuracy of -&5 per cent. The optimum conditions for the method are reported. The chelate is formed by boiling a solution of iridium, terpyridyl and ethanol for 2 to 2-5 hours any iridium(1V) present being easily reduced to iridium(II1) by the addition of hydroxylammonium chloride. After heating the solution is diluted with a buffer solution (ammonium hydroxide-ammonium chloride at pH 8-0) and the excess of the reagent is removed by extraction with chloroform. As the luminescence of the chelate is subject to oxygen quenching the sample solution is flushed with nitrogen for 10 to 15 minutes before measuring the luminescence which the authors suggest is room-temperature phosphorescence. This they deduce from measurements of the lifetime of the luminescence.The spectral properties and the composition of the chelate ar PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 101 discussed in detail and the interferences caused by relatively small amounts (4 x per cent.) of the other group VIII metal ions are reported. Of these ions iron(II) rhodium(II1) and ruthenium(II1) caused the most serious inter-ference. No other ions were examined as possible interferences. The authors report that this is the only luminescent method for the determination of iridium and that only a few spectrophotometric methods are known. Copper silver and gold It can be seen from the previous sections of this review that relatively few methods have been published for the determination of transition metal ions.The copper group of metals is the exception and several methods have been reported for the determination of these metals. Temkina et aZ.300 have prepared and examined the properties of several new fluorescent complexones containing the iminodiacetic acid group and have pro-posed that some of these compounds could be used for the quenchimetric deter-mination of copper(I1). Similarly the preparation and properties of some methyleneiminodiacetic acid derivatives of 7-hydroxycoumarin have been reported311 and the use of the complexones of 7-hydroxycoumarin and 3-carboxy-7-hydroxycoumarin for the determination of copper(I1) has been proposed. No experimental details of the investigation of the conditions or of any interferences are given but it is reported that by using these reagents copper(I1) can be determined in the range 0.6 to 6 pg with the 7-hydroxycoumarin complexone and in the range 0.06 to 1.2 pg with the 3-carboxy-7-hydroxycoumarin complexone.Sufficient information to enable these determinations to be carried out is given. Harris and Ritchie,312 in a report of their work on the biochemistry of 1,1,3-tricyano-2-amino-1-propene noted a reaction that occurred between this com-pound and copper(I1) to yield a compound that is highly fluorescent in acid solution. These authors subsequently proposed this reaction as the basis of a method for the fluorimetric determination of copper(11) .313 The reaction proceeds rapidly in neutral or slightly alkaline solution (pH 7.5) and is complete after warming at 37" to 40 "C for 15 minutes.Subsequent to the reaction the solution is acidified and the fluorescence intensity of the acidic solution measured. To maintain a reasonable reaction rate a 38-fold molar excess of 1,1,3-tricyano-2-amino-1-propene is required compared with the expected stoicheiometry of 2 moles of reagent to 1 of copper(I1). No sensitivity Limits are reported but the authors claim that the method is sensitive and specific for the determination of copper in the nanogram range and suggest that the use of dimethylformamide as the solvent or the addition of ethyl acetate to the pre-formed copper - reagent complex in aqueous solution will increase the sensitivity. The addition of ethyl acetate to an aqueous solution would appear to be a doubtful procedure but it may be possible that this is intended as an extractant or that a misprint has occurred and a water miscible solvent should be used.A number of ions were examined as potential interferences and when present in a 20-fold excess the individual ions Al(III) Ba(II) Ca(II) Cu(I) Co(II) Fe(II) Fe(III) Au(III) Mg(II) Mo(V1 102 BARK AND WOOD Ni(II) Ag(I) Na and Zn did not interfere. However when a mixture of these ions is present a reduction in the rate of the reaction is observed and to obtain reproducible results suitable for quantitative work under these conditions the reaction mixture should be allowed to stand overnight. The results of the analysis of biological materials which are decomposed by wet oxidation are reported and recovery values which range from 95.5 to 102.5 per cent.are obtained. The authors also suggest that the copper content of other materials such as steels or ores may be determined by this method after prior extraction of the copper as its dithizone complex. A method for the detection of copper(1) by using thiamine has been reported.314 Two procedures are described one is a spot test that is sensitive to Oe3pg of copper per drop and the other is an extraction method in which the fluorescent species is extracted into iso-amyl alcohol to give a sensitivity of 6 pg ml-1. This extraction procedure has been further developed315 as a quantitative method. Copper(I1) is reduced to copper(1) by using hydroxylamine and after adjusting the pH to 7.0 to 7.2 with sodium hydroxide solution the solution is cooled in an ice-bath for 5 minutes and an ice-cold alkaline (pH 11.0 to 11.5) solution of thiamine is added.The mixture is allowed to stand and the fluorescent species after the addition of sodium sulphate is extracted by using iso-amyl alcohol and its fluores-cence intensity measured. Interference is caused by the presence of several ions. Equimolar quantities of CN- PO$- and 10-fold excess of Co(II) Fe(III), Hg(I1) and Ag(1) ; 100-fold excess of Pb(II) Sn(II) Zn F- Mn042- and a 104-fold excess of C1- are reported to interfere. Two papers316y317 describe the use of N-(p-hydroxy propylanabasine) as a reagent for the fluorimetric determination of copper(1) ions. The fluorescence intensity of the complex formed between copper(1) and the reagent at pH 9.5 to 11.0 (borate buffer) in the presence of a 5 to 10-fold excess of the reagent is measured and is proportional to the concentration of copper(1).The abstracts of these papers differ only in the reported sensitivity. Reference 316 reports that linear calibration plots are obtained for the range 0.03 to 0-5 and 1 to 14 pg ml-l of copper whereas reference 317 reports ranges of 0.1 to 1.0 and 2 to 12 pg ml-l. Although it is reported that palladium and uranyl ions quench the fluorescence, this method appears to be fairly selective. Of these fluorimetric methods the most sensitive and selective and with which there is no obvious disadvantage in the technique required is that using TRIAP.313 Lumocupferron (XXI) and its two analogues a-(9-methoxybenzylidene) hip-puric acid and a- (P-diethyl aminobenzylidene) hippuric acid have been compared as reagents for the kinetic determination of ~opper(II)~l* and of these a-(p-diethyl-aminobenzylidene) hippuric acid gives the highest fluorescence in the presence of copper(I1).The optimum pH range is 8.0 to 10.0 and a sensitivity of 0.2 to 0.4 ng ml-1 can be achieved if the copper impurity in the ammoniacal acetate buffers used is masked with diethyldithiocarbamate. The rate of formation of the fluorescent compound (it is suggested319 that this is a dimer of the reagent) is dependent on the copper concentration so that by measuring the rate o PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 103 COOH Lumocupferron development of fluorescence the copper concentration can be determined.This method which is selective for copper(II) was originally reported in 1963319 and has since been used for the determination of copper in germanium tetrachloride and trichloro~ilane.~~~ Several fundamental studies of the luminescence of metal porphyrin compounds have been made as under suitable conditions the compounds exhibit very narrow phosphorescence spectra at liquid nitrogen temperature. Solov’ev et have used this emission for the determination of copper in the presence of several other ions that also phosphoresce when chelated with porphyrins. A mixture of the solution containing copper and etioporphyrin(I1) in acetic acid is evaporated to dryness and the residue dissolved in butyl iodide. The use of butyl iodide results in the formation of an adduct in which the heavy atom effect on the porphyrin molecule causes it to emit in the infrared region.However the position of the emission of the copper complex is barely affected and thus the intensity of the blank is reduced. The method can be used to selectively determine 1 ng of copper. No details of the errors or the instrumentation used are reported. Silver has been determined by using 8-hydroxyquinoline-5-sulphonic acid,321 and the optimum conditions for the method which can be used for the range 1.25 x to 5 x lo-* per cent. of silver with a standard deviation of 2.1 per cent. are described. The reaction between silver(1) and persulphate ions results in the formation of silver(I1) which the authors suggest then oxidises the reagent to a highly fluorescent compound and therefore by measuring the fluorescence intensity produced the concentration of the silver present can be determined.At the optimum pH of between 1.5 and 3.5 only a few metal ions interfere, although more than 1 x per cent. of copper(II) mercury(I1) and palla-dium(I1) cause quenching while zirconium(1V) and hafnium(1V) cause a large increase in intensity only 1 x per cent. of chloride can be tolerated. Three methods have been published for the determination of silver all of which make use of the fluorescent ion-association complexes formed between rhodamine dyes and silver halide complexes. Perminova and Shcherbov322 used Butyl Rhodamine S for the determination of silver in minerals. The sample is decomposed by treatment with nitric sulphuric and hydrofluoric acids and the silver extracted from a 0.25~ sulphuric acid solution with dithizone in benzene.After washing the organic layer the metal is extracted into an aqueou 104 BARK AND WOOD solution of potassium bromide and sulphuric acid the sulphuric acid concentration is adjusted to 4~ and the resulting solution is shaken with a solution of the reagent in benzene. The fluorescence intensity of the non-aqueous layer is measured and is directly proportional to the concentration of silver. The relative standard deviation is reported to be 5 per cent. and the sensitivity to be 1 x per cent. Other authors have described323 a modification of this method which involves co-precipitation of the silver with dithizone. A method324 very similar to one previously reported322 uses Rhodamine 6 Zh as the reagent.No limits for this method are reported in the abstract. The chemiluminescent reaction between lucigenin (XXII) and hydrogen to 2 x (XXI I) Lucigenin peroxide in alkaline media is catalysed by silver and the optimum conditions for the use of this reaction for the determination of silver have been reported.326 The authors recommend that to obtain the maximum sensitivity of 0-1 pg ml-l the total luminescence should be measured photographically. Under the optimum conditions for the reaction viz. pH 13.5 (obtained by using 0.3 M potassium hydroxide solution) M lucigenin and 0.02 M hydrogen peroxide the ions of chromium cobalt copper lead manganese nickel and osmium also catalyse the reaction so that the method is not selective.The solubility of silver chloride was determined by this method and the authors report that the result obtained (1-05 x Two methods for the quenchimetric determination of silver have been re-ported. 2,3-Naphthotriazole has been used as a reagent for the gravimetric, spectrophotometric and fluorimetric determinations of silver.326 The optimum pH for the reaction is 10.5 and 0.025 to 0.1 pg ml-l of silver can be determined by measuring the reduction in the fluorescence intensity of the reagent in the presence of silver. The gravimetric and spectrophotometric methods are made highly selective by the use of suitable masking agents although these masking agents complex enough silver to interfere with the fluorimetric method and their use is g 1-l) agrees satisfactorily with previously published values PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 105 not permitted.Thus the fluorimetric method is more subject to interference by foreign ions than are the other methods. The authors also report that this reagent can be used as a titrant for the fluorimetric determination of silver in the range 0.1 to 2.0 pg ml-l. The quenching of the fluorescence of tetrabromofluorescein (TBF) by silver ions in the presence of 1,lO-phenanthroline (caused by the forma-tion of a non-fluorescent ternary complex) has been used by El-Ghamry et for the determination of silver. The conditions for the maximum quenching of the fluorescence are reported. The reaction is not very dependent on pH or time, maximum quenching of the fluorescence occurring between pH 3.0 and 8.0.For simplicity an ammonium acetate solution (pH 7.0) is used as the buffer and the intensity of the solutions shows no change for up to 2 weeks. The nature of the complex was investigated by Job’s method and on the basis of the results obtained the authors propose that the complex is [Ag(Phen),TBF2-]. The method is applicable in the range 4 to 40 ng ml-l of silver in the final solution which means that the highest possible sensitivity is of the order of 5 ng ml-l of silver in the solution to be analysed. The interference levels of a number of anions and cations are reported. The authors despite the interference caused by equimolar amounts of iridium(IV) platinum(1V) and Br- and 10-fold molar excesses of osmium(IV), rhodiurn(III) I- and S2- report that only palladium(I1) and CN- interfere seriously with the determination.Only two methods have been reported for the determination of gold. Pod-berezskaya et aZ.328 have investigated the use of various rhodamine dyes as reagent for the determination of gold(II1). Of these dyes Rhodamine 6J Rhodamine S, Ethylrhodamine and Butyl Rhodamine S the latter is the most sensitive. Under the optimum conditions for the formation and extraction of the ion association complex formed between the [AuClI- and the dyestuff cation the method is applicable to the determination of 1 x 10-5 to 1 x per cent. of gold in ores. A detailed description of the procedure for the decomposition of sulphide ores and the subsequent determination of gold is given.Another dye Rhodamine B has been used for the determination of gold in a variety of ores.329 The decomposition of the ores the extraction by co-precipitation of gold (by using tellurium) and the subsequent determination are reported in detail. The procedure is long and involved but this disadvantage is amply compensated for by the high sensitivity (in the p.p.b. range). The results obtained for the analysis of various samples are reported and compared with the results obtained by using other methods. Zinc cadmium and mercury Reviews of various methods for the fluorimetric determination and detection of both zincz9 and cadmium27 were published in 1967 by the Turner Instrument Company. Bark and Rixon218 have described a new reagent 2-(2’-pyridyl) benzimi-dazole for the spectrofluorimetric determination of zinc gallium and indium.By using this reagent zinc can be determined in the range 15 to 800 ng ml-1 with a standard deviation of 1.1 per cent. The development of the method is describe 106 BARK AND WOOD in detail. The optimum pH is 5 to 8 and a minimum of a 10-fold molar excess of the reagent is required to obtain the maximum difference in intensity between a solution of the reagent and a solution of the reagent containing zinc. After an initial development time of 30 minutes the intensity of solutions containing zinc is stable for up to 10 days. The removal of most of the interferences is achieved by the addition of sodium hypophosphite but cadmium still interferes by forming a fluorescent complex with the reagent.The addition of a 1 to 2-fold molar excess of sulphide (calculated on the cadmium concentration) followed by reading the intensity after 10 minutes gives only the fluorescence of the zinc chelate the cadmium complex being decomposed. If a large excess of sulphide is present, the intensity of the zinc chelate decreases after 20 to 30 minutes. Mole ratio plots showed the formation of a 1 1 metal-to-ligand complex that could not be isolated. The method is sensitive reproducible and easy to operate. A somewhat more sensitive method involving dibenzothiazolylmethane has been reported.=O This reagent has previously been used for the fluorimetric determination of zincs1 and 1ithi~m.l~' Reference 330 describes a spectrofluorimetric investigation of the reagent and its use in the determination of zinc.Purification of the reagent is achieved by extracting a chloroform solution of the reagent with 0.1 M EDTA at pH 10.0 the background fluorescent intensity of the reagent after its recovery by evaporation of the chloroform being one half that of the reagent re-crystallised from ethanol. Trace metals are removed from the potassium hydroxide used in the method. This is accomplished by extraction with a solution of dithizone in carbon tetrachloride. The residual coloration in the potassium hydroxide is removed by washing with chloroform followed by ion exchange. Purification of the reagents allows 2 ng ml-f of zinc to be determined. In neutral solution the complex formed is that in which the zinc is chelated by the nitrogen atoms of the thiazole rings.The fluorescence of this complex is susceptible to quenching by water which places a practical restriction on the method. However the complex formed in alkaline solution involves deprotonation of the methylene bridge resulting in a complex that is more resistant to the quenching action of water. Only zinc and lithium (ionic radii 0.072 and 0.071 nm respectively) form strongly fluorescent chelates with this reagent. The natures of the zinc chelates formed in neutral and basic solutions are confirmed by infrared and nuclear magnetic resonance spectroscopy. The increased sensitivity of the method when using the purified reagent demonstrates the importance of the purification of reagents for this type of determination. N-8-Quinolyl-~-toluene sulphonamide is described332 as a reagent for the qualitative and quantitative determination of zinc and cadmium.The zinc complex is formed in aqueous solution in the pH range 8.0 to 8.3 the aqueous solution is then extracted with chloroform and the fluorescence intensity of the non-aqueous extract is measured. Mole ratio plots show that the zinc complex is of the form ZnL,. The analytical procedure described enables zinc to be deter-mined in the range 0.5 to 6.4 pg ml-l and although no interferences are reported, cadmium will almost certainly interfere because the spectral characteristics o PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 107 the cadmium complex are similar to those of the zinc complex. Of the few ions examined as possible interferences in the qualitative method only aluminium, copper chromate and cyanide are reported to interfere and these ions will, presumably also interfere with the quantitative method.Studies of the photo-decomposition of the complex are reported but no reasons for this decomposition are put forward. However the results obtained indicate that the fluorescence intensity should be measured immediately following the extraction of the chelate with chloroform. The conditions reported for the determination of zinc also apply to the determination of cadmium but for cadmiurn mole ratio plots show that both 1 1 and 2 1 ligand - metal complexes are present in solution. The method is reported as being applicable to the determination of cadmium in the range 5 to 50 pg ml-l.N-8-Quinolyl-&toluene sulphonamide has been used to determine zinc in titanium dioxide= in the range 5 x 10-4 to 0.16 per cent. with a maximum relative error of 16 per cent. Details for the dissolution of the sample and the subsequent extraction with iso-amyl alcohol of the zinc as its pyridine - thio-cyanate complex are reported. The zinc is removed from the organic phase by extraction with a solution of ammonium hydroxide and ammonium chloride the aqueous phase is buffered with glycine N-8-quinolyl-$-toluene sulphonamide is added and the fluorescence intensity measured after 10 to 15 minutes. Suitable analytical standards are prepared by using zinc free titanium dioxide and the results are compared with those obtained when using a method of standard additions.Nahanand and Houck= have described the use of 8-hydroxyquinoline for the determination of zinc in plasma and urine. The optimum pH for the formation of the fluorescent zinc chelate of 8-hydroxyquinoline is 8.0 a pH at which magnesium also forms a fluorescent chelate. However the authors report that using a universal buffer (sodium acetate and sodium barbiturate mixtures) in the presence of gum arabic prevents the formation of the fluorescent magnesium chelate. Hence under the conditions described for the method the reagent is fairly selective for zinc as neither calcium (up to 20 mg 100 ml-l) nor magnesium (up to 6 mg 100 ml-l) interferes. No other ions are reported as interferences, although the authors have investigated the effects of cadmium cobalt(II) iron(II), iron(II1) and lead (1 mg ml-l) on the fluorescence and report that only cadmium causes any significant increase in the fluorescence.The results obtained by the proposed method are reported and the authors state that these results agree with previously published values obtained by using atomic absorption spectrophoto-metry and the dithizone method. Recovery values obtained by the method of standard additions are between 100 and 109 per cent. for plasma and 99 and 120 per cent. for urine. The amount of zinc in the skin of patients with erythematiosis has been determined335 by fluorimetry with 8-benzenesulphonamidoquinoline as the reagent. The skin is decomposed by heating in a mixture of concentrated sulphuric per-chloric and nitric acids until a clear solution is obtained.This solution is then evaporated to dryness the residue dissolved in water and a buffer solution 108 BARK AND WOOD (pH 8.0 by using glycine) and a solution of the reagent are added. The fluorescence intensity is measured after 10 to 15 minutes. The results obtained for the zinc content of the skin of various subjects are given. 3,5’- Bis - (dicarboxymethylaminomethyl) - 4,4‘- dihydroxy - trans stilbene has been useds6 for the spectrofluorimetric determination of cadmium. This work is part of the work of these authors on the effect of introducing a dicarboxymethyl-aminomethyl group into parent molecules which are fluorimetric reagents. The preparation and reactions of this reagent with metal ions are described the pH values at which the metal chelates exhibit maximum fluorescence are pH 10.9 for barium calcium magnesium and strontium pH 7-9 for cadmium gadolinium, lanthanum lutetium yttrium and zinc and for aluminium and beryllium pH 5.2 and 6.4 respectively.Two procedures are described for the determination of cadmium. In mixtures with other ions and in the presence of zinc in amounts greater than 6 pg cadmium is extracted from a solution containing tartrate and cyanide at pH 11.0 with a solution of sodium diethyl dithiocarbamate in carbon tetrachloride. The metal is then re-extracted into dilute hydrochloric acid solution, the pH adjusted to 7.0 and the solution treated with hexamine and 3,5’-bis-(dicarboxymethylamino-methyl)-4,4’-dihydroxy trans stilbene. The fluorescent intensity of the resulting solution is then measured within 1 hour of the addition of the reagent as on keeping solutions for periods of longer than 1 hour a slight decrease in the fluorescence intensity is observed.If zinc is present in amounts less than 6 pg then a mixture of zinc and cadmium can be analysed without prior extraction of the cadmium. When using the extraction procedure 2 pg of lead, 10 pg of thallium 400 pg of EDTA and 500 pg of DPTA interfere. The method is sensitive to 0-5 to 25 pg of cadmium with a precision of k3.2 per cent.; over this range a linear calibration plot is obtained. This method is more sensitive than that involving iV-8-quinolyl-~-toluene sulphonamide.332 The dissociation constants of the stilbene derivative and the over-all stability constants of its cal-cium cadmium and lanthanum chelates are given.The authors suggest that the fluorescent cadmium species is of the form CdL2- however they also report mole ratio plots that indicate the formation complex with a 2 1 cadmium-to-reagent ratio i.e. Cd,L. The only method reported for the determination of mercuryB7 uses Rhodamine S and can be used in the range 0.002 to 0.01 per cent. with a standard deviation of approximately 20 per cent. The sample is decomposed with nitric and hydro-chloric acids and the resulting mixture is filtered. After diluting and adjusting the over-all acid concentration to 2 M the mercury is extracted by using dithizone. A solution of 4.5 M sulphuric acid containing 0.4 M bromide ions is then used to re-extract the mercury into an aqueous medium.After adding Rhodamine S and adjusting the bromide ion concentration to 0 . 2 ~ by dilution the complex is extracted with benzene and the fluorescence of the organic extract measured. Thorium and uranium Thorium has been determined in monazite by using Aavonol in hydrochlori PHOTOLUMINESCENCE AND CHEMILUMINESCENCE I N INORGANIC ANALYSIS 109 acid medium.338 After dissolution of the sample in hydrochloric acid and digestion with sulphuric acid to convert any fluorides to sulphates the solution is diluted and passed through an anion-exchange column (in the nitrate form) to remove interfering ions. The thorium is then eluted from the column by using dilute hydrochloric acid. The subsequent procedures necessary for the determination of thorium are reported in detail.The method is applicable to the determination of 1 to lOOpg of thorium with a standard deviation of 3.6 per cent. Although interference limits for foreign ions are repoIited they are not discussed and from the table of interfering ions it is apparent that several ions when present at fairly low concentrations relative to the thorium concentration (0.4 pg ml-I) will interfere. The condition for the formation of a luminescent complex between thorium and quercetin have been inve~tigated.~~ The use of acetate buffers reduces the absorbance of the complex. However although hexamethylenetetramine at pH 4.0 does not affect the absorbance it completely quenches the luminescence. On this basis the authors propose that the luminescent 1:1 complex formed contains the acetate ion because in the absence of acetate a 2 1 ligand - metal complex is formed that is non-fluorescent.The ion-exchange behaviour of both complexes show them to be cationic and the authors report that 0.5 pg mi-1 of thorium can be determined by using this reagent. In a further papee0 these authors describe the effect of the solvent on the detemination of thorium with morin. The luminescence of the complex formed at pH 2.0 in aqueous methanol, aqueous ethanol and aqueous acetone was measured as a function of thorium concentration and on the basis of the results of this investigation the authors recommend that for maximum sensitivity 50 to 75 per cent. v/v of methanol-water should be used in place of the 25 per cent. v/v of ethanol - water previously used.341 The methods for the determination of uranium are reported in the section of this review dealing with the determination of metal ions as halo complexes.The use of uranium activated sodium fluoride based phosphors seems to be estab-lished as the standard method for the determination of uranium in a variety of materials despite some discrepancies as to the most suitable bead material. The Turner Instrument Company has published a review28 of the methods available for the determination of uranium most of which are based on the fluoride bead formation. However methods involving morin Rhodamine B Rhodamine 6G and zinc phosphate suspensions are also reported. The methods that have been used to determine uranium in various materials are tabulated and methods for its fluorimetric detection are also reported.This excellent review contains forty-one references. A book that reviews the luminescent method for the determination of uranium has been published,342 but the abstract available gives no details of the methods described. Halides Fluoride. The fluoride ion interferes with the fluorimetric determination of many metal ions as the metals form more stable complexes with the fluoride tha 110 BARK AND WOOL, they do with the metallo-fluorescent molecule. Several workers have utilised such interferences and have reported methods for the determination of fluoride by using the quenching of the fluorescence of a metal chelate. A review24 produced by the Turner Instrument Company describes the methods published for the determina-tion of fluoride (prior to 1968) dl of which are based on the quenching of the fluorescence of a metal chelate.Hocman and his co-workers34s have adapted a method of Powell and Saylo? for the determination of fluoride in dental material. The method is based on the quenching of the fluorescence of the Eriochrome Red B - aluminium complex by the fluoride ion. The reagent is prepared such that a slight excess of Eriochrome Red B is present thus ensuring that all of the aluminium in solution is in the complexed form. This solution is stored for 3 to 5 days before use as storage for this period ensures that consistant readings are obtained. Dental samples were subjected to a lengthy ashing procedure requiring 53 hours the resulting ash was then heated in concentrated sulphuric acid and the hydrofluoric acid removed by steam distillation.The detection limit quoted is 0.025 pg ml-l of fluoride and the limit of determination is given as 0.05 pg ml-l. No interfering ions are reported. Fluoride and phosphate are the only common anions that interfere with the fluorescence of the zirconium - flavonol chelate. This inherent selectivity prompted an investigation for the quenchimetric determination of fluoridew; an investigation of the parameters necessary to establish the method is reported in detail. The optimum pH is 1-78 but providing that the standards and test solutions are at the same pH strict control is not essential. Sufficient zirconium solution is added to a mixture of flavonol and the test solution so that flavonol is present in a slight excess (zirconium and flavonol form a 1 l complex).Many foreign ions affect the determination the most serious interferences being those caused by the presence of aluminium citrate molybdate oxalate and tartrate. These ions cannot be tolerated even in only a 10-fold excess. Interference levels for some of the other thirty-seven ions tested are given and many of these ions can be tolerated in several thousand fold excess. This method is quoted as being applicable to amounts of fluoride between 2 and 100ngml-l. As part of his work on the determination of fluoride in sera and urine, Tavesa6 describes the use of the morin - thorium complex as a reagent for fluoride. The effects of sulphate phosphate nitrite and fluoride on the fluorescence of the complex are reported.All of them cause a reduction in the intensity of the fluorescence although nitrite does not show an immediate effect but causes a marked reduction after 18 hours. The calibration curve which is linear down to a fluorimeter reading of 50 per cent. (setting the blank at 100 per cent.) indicates that fluoride can be determined in the range 2 to 65 ng ml-l. Larger amounts can be determined by the use of larger amounts of reagent. No lower limits are given. The development of the method is not reported in detail but the method seems comparable in sensitivity and ease of operation with that of Guyon et aL"5 The values for serum fluoride measured by this method after diffusion at roo PHOTOLUMINESCENCE AND CBaMfLUMINESCENCE IN INORGANIC ANALYSIS 111 temperature agree with those obtained by using a fluoride electrode and also with those predicted by the renal clearance of radioactive fluoride.The relative standard deviation when measuring 10-6 M fluoride in 2 ml of serum is 10 per cent. A method involving the quenching of the fluorescence of the aluminium -PAN chelate is reported to be sensitive to 2 x 10-SEA fluoridex7; however as the method involves the use of almost absolute ethanolic solutions of all of the reagents and of the fluoride the sensitivity for the analysis of fluoride in aqueous solutions will be considerably lower. The aluminium - PAN reagent is prepared so that the mole ratio of PAN to aluminium is 4 1 ensuring complete chelation of the aluminium. Procedures that are applicable in the range 38 pg ml-l to 6 ng ml-l of fluoride and over the wider range of final concentrations of 19 pg ml-l to 19ngml-l have been described.In both procedures it is necessary to allow equilibration of the solutions for 5 hours at room temperature. Interferences at equimolar levels were examined and only phosphate and iodide interfered. No mention is made of the interference of cations some of which will undoubtedly interfere for example nickel which these authors report3O7 can also be determined by quenching of the aluminium chelate of PAN must interfere. The interferences in the methods described can presumably be overcome by the distillation of the fluoride as hydrofluoric acid using the method originally described by Huckbay et aLx8 and modified slightly by H ~ c m a n .~ ~ A further development has been the design of an automatic fluorimetric fluoride analyser by Thompson Zielenski and Ivie.s9 This is a simplified version of an instrument previously described by Ivie et aLS0 The modified instrument is more easily used because a single photomultiplier is employed the previous instrument used two matched photomultipliers and considerable time was involved in the selection of suitable photomultipliers and the necessary continual adjust-ment of these to the same sensitivity. The measurement is based on the reaction of gaseous hydrofluoric acid with the magnesium salt of 8-hydroxyquinoline. The complex which is highly fluorescent is distributed by impregnation on strips of filter paper. After reaction the intensity of the fluorescence of the magnesium oxinate is reduced and the amount of quenching is proportional to the amount of the hydrofluoric acid present.The sampling is through two warmed glass tubes, in one of which the hydrofluoric acid is absorbed on to a thin coating of sodium bicarbonate which serves as a blank. The other tube is empty. The two air streams are then passed on to the tape which is irradiated by ultraviolet radiation. The reported sensitivity is 0-14 pg m-3 of hydrogen fluoride. The intensities of the fluorescence of the blank and the sample areas are compared. Chloride bromide and iodide. Karyakin and Babi~heva~~l have described a method for the luminescent determination of chloride. The method depends on the adsorption of fluorescent dyes on to silver chloride.A number of dyes (Eosin Fluorescein (XKIII) Fluorexone (11). Rhodamine B Rhodamine 6G and Diiodofluorescein) were examined. Fluorescein gave the best sensitivity of between 2 and 1Opg1-1 (reproducibility 10 per cent.) or 10 and 5Opg1-112 BARK AND WOOD (XXI Ill Fluorescein (reproducibility 5 per cent.). For fluorescein to be adsorbed the silver chloride particles must be positively charged by adsorption of an excess of silver ions. By breaking down the solvent sheaths of the reacting silver ions and chloride ions with ultraviolet radiation it is possible to increase the amount of adsorption of the dyestuff with a consequent increase in the over-all sensitivity of the method. Chloride ions in water are determined by the addition of fluorescein and silver nitrate followed by irradiation with ultraviolet radiation (254nm) for 3 to 10 minutes.The fluorescence intensities of solutions before and after irradiation are measured on a fluorimeter. Although this is a rapid and sensitive method it has the disadvantage that it cannot be applied to water purified by the use of ion-exchange resins unless the amount of organic material washed off the resins is extremely low and is constant. A modification to the titration of chloride with silver ions by using dichloro-fluorescein as an adsorption indicator has been proposed by Friend.352 The modi-fication is the addition of excess ethanol which causes a sharpening of the end-point when interfering ions are present. Reasons for this effect are suggested. A fluorimetric method has been developedm for the determination of chloride in conductivity water.The basis of the method which is sensitive to 0.05 pg ml-l, is the 10-fold increase in fluorescence obtained from the reaction product of the luminol - hydrogen peroxide reaction in the presence of the hypochlorite ion. The intensity is proportional to the concentration of hypochlorite. Only the abstract of this method is currently available so that working details of the procedure cannot be reported. No methods for the fluorimetric determination of bromide have been described during the period covered by this review. The reaction between luminol and iodine in alkaline solution has been shown3% to result in the emission of an intense blue luminescence and has been used for the determination of iodine.This is the subject of a patent.= The factors in-fluencing the reaction for both the production of maximum intensity of the chemi-luminescence and the total amount of light emitted are reported in detail. The optimum conditions are pH 13.0 (0.2 M sodium hydroxide) at a temperature of 20" C with a luminol concentration of 2 x 1 0 - 4 ~ . The intensity of the chemi-luminescence which is measured photographically attenuated ten times by th PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 113 use of a suitably treated photographic plate. The authors report the sensitivity of the method to be 1 pg ml-l with a standard deviation of 0.1 pg for iodine con-centrations of 1 to 3pgml-l. The molar luminescence coefficient which is a function of the relative quantum yields of chemiluminescence reactions is suggested as a basis for comparing the sensitivity of chemiluminescent methods.A report on an investigation of the reaction of 2‘,7’-bis(acetoxymercuri)-fluorescein (XXIV) with anions that are known to form stable mercury com-(XXIV) 2’ 7’-Bls (acetoxy mercuri) fluorescein plexesS6 includes a description of a method for the determination of iodide. The effects of time and pH are reported the optimum pH is 7.8 and intensity readings must be taken 10 & 1 minutes after mixing the solutions. There is an initial decrease in intensity of a solution of the fluoresceinate complex on the addition of iodide but on standing for longer than 15 minutes there is a gradual increase in intensity.The authors attribute this effect to the fission of the carbon - mercury bond. No determination limits are quoted but 0.87 pg of iodide is the lowest reported figure and the maximum sample volume that can be used is 14 ml. The relative deviation is 16 per cent. at this level. A 1 2 stoicheiometry of the mercury complex to iodide which would be expected if both mercury atoms were complexed, is reported. Although most common anions do not interfere chlorides and thio-cyanates at concentrations above 0.005 M and bromide at all concentrations do. A much simpler procedure based on the quenching by iodide of the fluorescence of the uranyl ion has been rep0rted.~57 The effects of time sodium hydroxide concentration and uranyl acetate concentration are reported in detail.The fluorescence intensities of solutions containing iodide are stable for between 30 minutes and 8 hours although if the intensity is measured after a given time then acceptable readings can be obtained within 30 minutes. The sensitivity is approximately 20 ng ml-l. The interference levels for thirty-two ions are reported. Miscellaneous ions Cyanide. Cyanide despite its ability to form stable complexes with mercury, does not affect the fluorescence of 2’,7‘-bis(acetoxymercuri) fluorescein. However, it does inhibit the reaction of iodide with the reagent to give a fluorescent product. Thus by measuring the reduction intensity of a cyanide - mercury fluoresceinat 114 BARK AND WOOD solution to which iodide is added the amount of cyanide can be determined.m The optimum pH for the determination is pH 8.3 and although most common anions do not interfere chloride and thiocyanate interfere at concentrations greater than 0.005 M and bromide interferes at all concentrations.No sensitivity is reported but the minimum amount of cyanide used was 3.6pg. Guilbault and KrameP8 have patented a method for the fluorimetric detection of cyanide based on the formation of fluorescent products with various quinones [p-benzoquinone N-chloro-9-benzoquinoneimine 2,5-dichlorobenzoquinone and o-($-nitrobenzene sulphonyl) quinone monoxime] in solvents such as dimethyl formamide and dimethyl sulphoxide. The compounds formed at pH 6.5 to 7.5 in these solvents were not identified but their fluorescence enables the detection of 5.0 pgml-l of cyanide.Sulphur-containing species. A reference359 to the determination of trace amounts of thiocyanate as bromine cyanide by using dithiofluorescein has appeared in the recent literature. Grunert Ballschimter and TolgSSO have reported a method for the deter-mination of sulphide in the range 1 to 10 ng ml-l based on the quenching of the fluorescence of 2’,7’-bis(acetoxymercuri)fluorescein complex in alkaline solution and have used this method for the determination of sulphlde in organic compounds. The reactions of various anions with this mercury complex and the effect of some cations on its fluorescence are also reported. This method has been used by Axelrod et ~1.36~ for the determination of hydro-gen sulphide in the atmosphere. The calibration curves are prepared by using sodium sulphide solutions and the preparation of the reagent and of synthetic air samples are reported in detail.Sampling is achieved by drawing air through a bubbler containing a solution of sodium hydroxide (0.1 or 1.0 M) at a rate of 2 1 min-l for up to 1 hour. The resulting solution is treated with the reagent and its fluorescence intensity measured. However when sampling with 1.0 M sodium hydroxide solution is used sufficient sulphuric acid is added to reduce the sodium hydroxide concentration to 0-1 M. The efficiency of the sampling, for which a detailed procedure is given is reported to be at least 80 per cent. and to increase with increasing sulphide concentration. Of the ions investigated as possible sources of interference only bromide sulphite nitrite and iodide of the common anions interfere organic sulphur compounds would be expected to interfere and cysteine and cystine do cause a reduction in the fluorescence intensity.The authors also report that no interference is caused by the presence of 1OOO-fold molar excesses of cadmium cobalt(II) copper(II) iron(II) iron(III) lead, manganese(II) nickel(II) potassium and sodium. This is contrary to the effects of cobalt(II) copper(II) iron(II1) and nickel(I1) reported by Grunert et aZ.360 Sulphite and nitrogen dioxide interfere although sulphite can be removed by sampling through potassium bicarbonate filters whereas nitrogen dioxide cannot be removed by absorption or by the addition of reducing agents and thus remains a major interference. Another source of error is caused by the absorption o PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 115 fluorescent materials from the atmosphere but this error is easily eliminated by measuring the fluorescence intensity of the solution both before and after the addition of the reagent.The method is sensitive (2 x lo-* per cent. is the lowest figure reported) and is fairly selective only nitrogen dioxide at concentrations above 1 x per cent. interferes. The results obtained for the analysis of various samples are reported. Sulphur dioxide has been determined by using the quenching of the fluores-cence of 5-amino-fluorescein by the addition compound formed between formalde-hyde and b i ~ u l p h i t e . ~ ~ ~ Simulated samples are prepared from sodium metabisulphite solutions containing sufficient mercuric chloride (0.1 M) and sodium chloride (0.2 M) to give a 0.1 M solution of sodium tetrachloromercurate(I1).The analysis is performed by mixing the sulphur dioxide - sodium tetrachloromercurate(I1) solution with formaldehyde allowing the mixture to stand for 5 minutes and then adding the reagent. The resulting solution is allowed to stand for a further 20 minutes and its fluorescence intensity is then measured. The optimum con-ditions for the reaction are reported in detail. For M reagent the quenching is independent of acid concentration in the range 0.02 to 0.05 M and of formalde-hyde concentration in the range 0.2 to 0-7 per cent. but for between and M solutions of 5-arnino-fluoresceinJ the intensity is constant over the range 0.05 to 0.1 M acid and 0.6 to 1.5 per cent.of formaldehyde. The use of low concentrations of the reagent should increase the sensitivity but at a reagent concentration of lo-' M (a concentration at which its fluorescence is easily measured) no reaction occurs under the conditions used. The sensitivity is reported as 0.02 pg ml-l of sulphur dioxide. The effects of various ions on the determination are reported and only iron(II1) interferes at low concentration (10-4 M) some interference is caused by nitrogen dioxide. Although this method is intended to be used for the determination of sulphur dioxide in the atmosphere by using the sodium tetrachloromercurate(I1) as a trapping solution no details or results of the determination of sulphur dioxide in gaseous samples are reported.The chemiluminescent reaction between luminol hydrogen peroxide and iodine in alkaline solution at pH 13.0 is inhibited by the presence of sulphide and this effect has been to determine sulphide. The measurement of the lumines-cence can be either photographically as the total luminescence or photoelectrically as the maximum luminescence. The photographic procedure is described in detail and this permits the determination of 5 ng of sulphide with a standard deviation of 2 ng. Reasons for the reduction in intensity are proposed and the method can be used for the determination of sulphur-containing organic compounds. Care must be taken in the analysis of these compounds to follow the recommended order of addition because although this is not critical for the determination of sulphide in sodium sulphide it is apparently critical for organic compounds.Oxygen-containing species. Although oxygen quenching is well known in fluorescence and phosphorescence analysis little use has been made of this effect for the determination of oxygen. Orban et a1.W have however designed a 116 BARK AND WOOD instrument for the detection and determination of oxygen based on the quenching of the green phosphorescence of solid trypaflavin. A Plexiglas plate coated with trypaflavin is placed in a vacuum chamber which is then de-gassed. The light source and photomultiplier are placed on the same side of a light chopper with a 90" open segment such that when the rotating disc cuts off the incident light the photomultiplier is exposed to the phosphor.The quenching of this phosphorescence enables the detection of oxygen at partial pressures of less than 4 x 10-6mm of mercury. At 2 x mm of mercury the intensity of the phosphor is reduced to half its original value. Hydrogen peroxide has been determined3G5 at concentrations down to 8 x 10-8 molar by using its chemiluminescent oxidation of luminol which for this purpose, is catalysed by haemin in the presence of the triethanolamine copper(I1) complex salt. The luminescence is measured by using a photomultiplier in an instrument described by the authors. No other details are reported in the abstract. Atmospheric ozone has been determined by Watanabe and N a k a d ~ i . ~ ~ ~ The basis of the method is the oxidation in acidified ethanolic solution of 9,lO-dihydro-acridine to acridine a highly fluorescent compound.Nitrogen dioxide interferes with this method but the results obtained compare well with the often used phenolphthalein method which is less sensitive. A specific automatic ozone analyser has been de~igned,~67 the basis of its operation being the chemiluminescent reaction occurring between ozone and luminol in the presence of haematin. The chemiluminescent material is impregnated into filter-paper and a disc of this material is mounted in the instrument. The air is then drawn through the pad and the luminescence caused by the ozone is detected by a photomultiplier and is recorded continuously. At a flow-rate of greater than 0.2 1 min-l the readings are directly proportional to the ozone concentration.For continuous operation one piece of the impregnated filter-paper can be used for periods of up to 1-5 to 2 hours, after which it must be replaced. When used intermittently the paper must be replaced after 5 to 6 hours. The instrument is sensitive to 0.3 pg 1-1 of ozone. Phosphorus-containing species. Two similar procedures for the deter-mination of phosphate in water samples have been reported.368 For amounts of phosphate less than 1 pg ml-l the recommended method is the quenching of the fluorescence of the aluminium - rnorin complex whereas for amounts greater than 10 pg ml-l quenching of the tin - ffavonol chelate is recommended. Although no details of the investigation of the conditions for these determinations are reported, the procedures described use the conditions for maximum sensitivity.Of the common anions only fluoride interferes seriously and its effect may be considerably reduced by boiling an acidified sample for a few minutes. There are cationic interferences and a few of these can be removed by a single hydroxide precipitation. The authors report that several water samples were analysed for phosphate by these techniques and that satisfactory results were obtained. Nitrogen-containing species. Rubin and Knott3@ have reported a method for the determination of ammonia in biological samples. The basis of the metho PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 11 7 is the oxidation of the fluorescent compound the nicotinamide adenine diphospho-pyridine nucleotide (NADH) to the non-fluorescent NAD+ during the reaction of ammonia with a-ketoglutamic acid in the presence of glutamic dehydrogenase to form glutamic acid.The optimum pH is 7.6 and is controlled by the use of a TRIS [tri-(hydroxymethyl)aminomethane] buffer which is mixed with sodium hydroxide glutamic dehydrogenase and sodium chloride to give a stock substrate solution. The determination requires that readings are taken before and after the additions of a-ketoglutamic acid (5-minutes reaction time is allowed). These are recorded for a blank a sample solution and for a solution containing a standard addition of ammonia. In each case the change in intensity is recorded and the ammonia concentration calculated by using an empirical formula. The investi-gation of the optimum conditions for this determination is reported in detail and the recovery values ranged from 92 to 104 per cent.for the analysis of plasma. The main advantages of this method over those previously used are the selectivity and the sensitivity that allows smaller samples to be used and also permits a more rapid completion of the reaction. Ammonia in blood or plasma has also been determined370 on the basis of its reaction with acetyl acetone and formaldehyde to produce the fluorescent compound 3,5-diacetyl-1 ,tl-dihydrolutidine (DDL) . The ammonia is separated by a diffusion process requiring 30 minutes and then deter-mined by reacting it with acetyl acetone and formaldehyde for 10 minutes in a boiling water-bath. No sensitivity is reported but the calibration curve is linear up to 0.25 pg ml-l of ammonia and the lowest point on the curve is 0.05 pg ml-l.Recovery values that range from 88.1 to 95.9 per cent. for plasma and 93-1 to 98.9 per cent. for blood are reported. The authors report that this technique is an improvement over existing methods from the standpoint of sensitivity and time, but no reference is made to the previously reported method of Rubin and K n ~ t t . ~ Indicators During the period of this review several papers have been published dealing with the use of fluorescent pH indicators and metallo fluorescent indicators. Although many of the chelating agents used in fluorimetric analysis could possibly be used as indicators or titrants especially in view of the high selectivity of some of the methods their application to this type of analysis does not appear to have been exploited to any great extent.In 1969 Bermejo et aL371 reviewed the use of fluorescent indicators in chelometric analysis. The review contains references to thirty-eight publications. Nishikawa372 has also reviewed the use of fluorescent indicators and reports that reliable fluorimetric methods for the determination of twenty-one elements are known. Kat0h~7~ has investigated the behaviour of the fluorescent acid - base indica-tors uranine and dichlorofluorescein in various solvent mixtures-water - ethanol, water - methanol and water - acetone. The organic solvent concentrations are 40 60 and 80 per cent. v/v and the fluorescent colour changes during variation of pH conditions for each of these concentrations are reported.Initially the intensities increase with increasing pH however at high pH an additiona 118 BARK AND WOOD adsorption band appears which overlaps with the emission band causing a decrease in intensity. The pH dependence of several substituted quinolines (6-methoxy- .Q-cyano- 6-ethoxy 6-ethoxy-2-methyl- and 8-methoxy-) has been reported.374 The colour of the fluorescence and the pH ranges for fluorescence are given. The 4-cyano derivative exhibits fluorescence both in acid and alkaline solution whereas the remainder fluoresce only in acid solution. Fluorescent pH indicators have the advantage over the use of coloured indicators in that in cloudy or coloured solutions the fluorescence change can often be observed. This advan-tage is illustrated by the use of the chemiluminescent acid - base indicators pyrogallol luminol and lucigenin for the titration of turbid soil extracts with hydrochloric The quenching observed with these solutions is negligible.4-Methylumbelliferone is a useful fluorescent indicator for acid - base titri-metry and Chen376 has made a detailed study of the dependence of its fluorescence on pH. The fluorescence is blue in alkaline and weak blue in acidic solution with a transition interval between pH 7-0 and 8.0. The corrected fluorescence spectra in both 0.01 M sodium hydroxide solution and 0.01 M hydrochloric acid are reported and the quantum yields of the anionic form and the non-ionised form are calculated by comparison with quinine sulphate and are reported to be 69 and 70 per cent.respectively so that the apparent decrease in intensity in acid solution is caused by the blue shift of the adsorption band. The author notes that ‘most of the published data on fluorescent indicators is of a qualitative nature from which it is impossible to tell how fluorescent the compounds are and what the properties are based on.’ Thus he appeals for ‘better characterisation of the spectral properties of such indicators.’ The ionisation constants of a series of twenty-five indigo dyes which are described as potential indicators in luminescent analysis have been determined377 and the effects of substituents are discussed. Wiersma and Lott378 have modified a Farand Fluorimeter Model 2-A for the instrumental measurement of end-points in the chelometric titration of metal ions with fluorescent indicators.Examples of the determination of aluminium(II1) , lead(I1) mercury(II) iron(II1) and samarium(II1) by the back titration of excess EDTA with gallium in the presence of morin as the indicator and of the analysis of mixtures of gallium and indium by a double titration with EDTA and TTHA, again using morin as the indicator are given. The end-points of the titrations of calcium strontium and barium with EDTA have been detected by using the chemiluminescent indicator lumin01.~~~ The reaction between luminol and hydrogen peroxide in the presence of either the cobalt(I1) or copper(I1) complexes of EDTA is catalysed by the presence of the alkali earths and a luminescence is observed. However at the end-point of their titration with EDTA this effect ceases and thus no luminescence is exhibited.A dark room is required for this method which allows the determination of 2.7, 5-8 or 11.0 pg ml-l of calcium strontium or barium respectively with an error of h0.3 per cent PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 119 Calcium has been determined in urine and serapBo by visual observation of the end-point during its titration with EDTA using calcein as the indicator. A simple apparatus that enables the end-point to be seen more easily is described. The calcium content of serum is determined by titration with EDTA to the disappearance of the green fluorescence of the calcein - calcium complex at a pH above 10-5. Urine calcium is determined similarly after prior evaporation ashing at 600 “C and dissolution of the residual calcium in hydrochloric or sulphuric acid.A similar method which is an adaptation of the method of Kepner and Hercules,l?? has been reported.381 However in this method the end-point is measured instru-mentally. The titrant (EDTA) is added automatically in 1-p1 portions and the fluorescence is recorded by an instrument equipped with an event marker to record each pulse of the delivery pump. The method is sensitive to 0.04pg of calcium with a coefficient of variance of less than 2 per cent. and it is also highly selective as proteins magnesium phosphate and other substances commonly present in biological materials do not interfere except when present in large excess. The results for the analysis of various biological samples and the corresponding recovery values are reported.Four new metallofluorescent indicators analogous to calcein but prepared from glycine instead of iminodiacetic acid have been reported.3s2 The products obtained frqm the Mannich condensation of glycine with fluorescein and its derivatives were examined chromatographically and in each case two or three new products were obtained. Examination of these compounds as fluorescent indicators for the compleximetric titration of copper(I1) with EDTA is reported and the best results are obtained with bis-2’,7’-N,N-glycinemethylene-4’,5’-di-chlorofluorescein. However as the starting material 4’,5’-dichlorofluorescein is no longer commercially available bis-4’,5’-N,N-glycinemethylene-2’,7’-dichloro-fluorescein is recommended for this determination.Two similar papers have been published. A comparison of three new bis-N,N-glycinemethylenedichlorofluores-ceins has been made and these are proposed383 as metallo-fluorescent indicators for the determination of copper(II) with which they form 1 1 complexes. The 2’,7’-bis- (bis- (carboxymethyl) aminomethyl) derivatives of 4’3’- 2’,7 ’- and 3,6-di-chlorofluorescein have been compared38* as fluorescent indicators and again these especially the 3’,6’-dichloro derivative are proposed as indicators for the complexometric determination of copper(I1) with EDTA. Temkina et aL3% have prepared the N-carboxyalkyl derivatives of some aminonaphthalene sulphonic acid compounds and of these complexons two of t hem-1 -amino- (N,N-dicarboxymet h yl) -2-napht hol-4-sulphonic acid and ethylene-diamine-N-(4-~ulphonaphthalene) -N,N‘,N’-triacetic acid-are suggested as fluores-cent indicators.Copper(I1) and nickel(I1) completely quench the fluorescence of these reagents and hence their direct titration with EDTA to a blue fluorescence is possible and by back titration with copper(I1) or nickel(II) bismuth(II), calcium cobalt (11) lead(I1) manganese(II) thorium(1V) and zirconium(1V) can be determined. The titration of bismuth(II1) and mercury(II) which do not completely quench the fluorescence of the indicators is done in the presence o 120 BARK AND WOOD Rhodamine B which screens the residual fluorescence so that the end-point corresponds to a change in the fluorescence colour from pale rose to blue.Several dyes obtained from the coupling reaction between the diazonium salt of 4-amino-3-hydroxynaphthalene-1-sulphonic acid with a number of fluorescent coumarin derivatives have been proposed3ss as fluorescent indicators of aluminium. The compounds isolated from the reaction mixture are impure and require careful purification by column chromatography. The 4-methylumbelliferone derivative, which was the most widely studied of these compounds gives a pink fluorescence with 0.2 to 10 p.p.m. of aluminium at pH 4.0 to 5.0 a condition under which the reaction is apparently fairly selective. IE~carilla~~~ has described the use of Calcein Blue (XXV) ,CHpCOOH CH2-N 0 ‘CH~COOH (XXV) Calcein Blue determination of silver by titration with potassium iodide.as an indicator for the A detailed procedure for the determination of silver and the results of some titrations at different pH values and concentrations are reported. A nitrogenous base fraction of Uzbekistan petroleums has also been used as a fluorescent indicator in the titration of silver ions with iodide or bromide.M8 The blue fluorescence disappears at the end-point. The results obtained for the analysis of a silver - copper alloy are reported and compared with those obtained by a potentiometric method. The authors report that this indicator can be used in coloured solutions and that it gives more reproducible results than either eosin or the potentiometric method. The complexones formed by a Mannich condensation of several coumarins with iminodiacetic acid and formaldehyde have been investigated as fluorimetric reagents and indicat01-s.~~~ Their properties as indicators were examined by the complexometric titrations of copper(I1) and calcium.Of the fifteen complexones examined the 3-carboxy-7-hydroxycoumarin complexone is recommended as an indicator for the titration of copper(I1) with EDTA and for the back titration with copper(I1) in the determination of other metal ions. Lukovskaya and Markova3S9 have described the use of luminol as a chemi-luminescent indicator for the determination of sulphite by iodimetric titration. The stoicheiometric reaction between sulphite and iodine prevents the chemi-luminescent oxidation of luminol by iodine in alkaline solution so that at the first excess of iodine chemiluminescence is observed and its onset indicates the end-point .M~lybdenum~~~as molybdate has been determined titrimetrically by using PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 121 lead(I1) salt as the titrant and primulene as the fluorescent indicator. The solution is irradiated with ultraviolet radiation from a mercury source. The end-point is indicated by a fluorescent colour change from blue to violet and is reported to be sharp. Although ions that react with lead or molybdate interfere it has been used for the determination of the molybdenum contents of some steels. The indicator-lead molybdate reaction is reported to be reversible and therefore the method might be of use for lead. Talipov et a1.390 have investigated the determination of lead by a similar method.Excess molybdate or tungstate ions are added to the lead solution and the excess molybdate or tungstate is then back-titrated with a standard lead solution using an indicator obtained from sulphur-containing petroleum (the fraction of nitrogenous bases with a boiling point of 140" to 200 "C at 4 mm of mercury pressure). Conclusion From a consideration of the various papers reviewed it is apparent that there is a great need in this field for corrected spectra ie. for spectra that have been compensated for variations in lamp intensity and in photomultiplier response efficiency. This places a load on the designers and manufacturers of equipment. In the field of phosphorimetry outside that section of phosphors there is a need for the development of suitable cells so that highly reproducible optical geometry can be achieved and hence eliminate the error often inherent in the equipment at present in use.Closely allied to this problem is the need for research on solvent systems capable of having a high proportion of water and yet able to form transparent glasses at the low temperatures generally used for phosphori-metry. While the use of frontal illumination and viewing of snows has met with some success the need to modify existing optical systems to enable measurements to be made tends to make the methods somewhat esoteric. All of these improvements are necessary and will probably be time consuming. One other improvement that is necessary and need not take any time to bring about is an agreement on the method of reporting the sensitivity of methods.Too often one sees a method whose claimed sensitivity is the concentration of the metal ion in the final solution after large and necessary dilution has occurred. This is especially so in phosphorimetry. It would be better if all authors would quote as the limit of sensitivity the minimum concentration of the ion that must be present in the sample before treatment so that a satisfactory measurement and determination can be made. Allied to the above improvement is that which would be obtained if all authors would state all of the ions that have been investigated as possible interferences. In this way one would know not only those ions that interfere in the method, but also those whose presence can be tolerated.While in a very few cases this might indicate the paucity of evidence on which claims are based and would clearly outline the limits of the method it would greatly enhance the value of most papers to most research workers 122 BARK AND WOOD 8 9 10 11 12 13 14 16 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 References Shcherbov D. P. Zav. Lab. 1968 34 641. English Translation 763. White C. E. and Weissler A. Analyt. Chem. 1966’38 155R. - and - Ibid. 1968,40 116R. - and - Ibid. 1970,42 57R. Bozhevol’nov E. A. Kern. Int. 1967 3 127. - Zh. Analit. Khim. 1967 22 1692. English Translation 1418. Kayser E. G. and Hall T. N. US. Clearing House Federal Science Technical Information A.683673 1968 pp.121 (Available from US. Govt. Res. dev. Depts., 1969 69 86). Passwater R. A. Editor ‘A Guide to Fluorescence Literature,’ Vol. I 1967; Vol. 11, 1970. Plenum Press Ltd. Berlman I. B. ‘Handbook of Fluorescence Spectra of Aromatic Molecules,’ Academic Press 1965. Zander M. ‘Phosphorimetry ; the Application of Phosphorescence to the Analysis of Organic Compounds’ (Translated by Goodwin T. A.) Academic Press 1968. ‘12th Conference on Luminescence Lvov U.S.S.R. 1964,’ Columbia Technical Translations Bull. Acad. Sci. U.S.S.R. 1965 29 Nos. 1 and 3 Physical Series. Szigeti G. Editor ‘International Conference on Luminescence Budapest Hungary, 1966,’ Vols. I and 11 Kultura 1968. Lim E. C. Editor ‘International Conference on Molecular Luminescence Loyola University Chicago U.S.A.1968,’ Benjamin Publications 1969. Parker C. A. ‘Photoluminescence of Solutions with Applications to Photochemistry and Analytical Chemistry,’ Elsevier 1968. Hercules D. M. Editor ‘Fluorescence and Phosphorescence Analysis ’ Interscience Publications Inc. 1966. Guilbault G. G. Editor ‘Fluorescence ; Theory Instrumentation and Practice,’ Edward Arnold and Marcel Dekker 1967. Ewing G. W. Editor ‘Instrumental Method of Analysis,’ Third Edition McGraw-Hill 1969. Bowen E. J. Editor ‘Luminescence in Chemistry,’ Van Nostrand Company Limited, 1968. Udenfriend S. ‘Fluorescence Assay in Biology and Medicine,’ Academic Press 1969. Phillips R. E. Amer. Lab. 1969 8. Dorr F. 2. Analyt. Chem. 1963 197 241. Passwater R. A. Editor Fluorescence News Biochemical Instrumentation Division, American Instrument Company.‘Automated Fluorimetric Procedures,’ Acc. No. 10033A G. K. Turner Associates, California 1968. ‘Fluoride,’ Acc No. 10014 G. K. Turner Associates California 1968. ’Beryllium,’ Acc. No. 9945 G. K. Turner Associates California 1968. ‘Boron,’ Acc. No. 10032 G. K. Turner Associates California 1968. ’Cadmium,’ Acc. No. 8981 G. K. Turner Associates California 1967. ‘Uranium,’ Acc. No. 9944 G. K. Turner Associates California 1968. ‘Zinc,’ Acc. No. 8979 G. K. Turner Associates California 1967. ‘Fluorescent Tracers,’ Acc. No. 9941A2 G. I(. Turner Associates California 1970. Capelin B. C. and Ingram G. Talanta 1970 17 187. Katyal M. Ibid. 1968 15 95. Korkuc A. Wiad. Chews. 1969 23 345; Chem.Abs. 1969 71 56224j. Urner Z. Colln Czech. Chews. Commun. 1968 33 1078. Pilipenko A. T. Savranskii L. I. and Nguen M. S. Zh. Analit. Khim. 1969,24 460. Sawicki E. Talanta 1969 16 1231. Becsay J. G. and Scheller K. Rev. Scient. Instrum. 1967 38 1793. Howerton H. K. in Guilbault G. G. Editoor ‘Fluorescence,’ Marcel Dekker Inc., Fluorescence News 1966 1 5. Cravitt S. and Van Duuren B. L. Chem. Inst. 1968 1 71. Cundall R. B. and Evans G. B. J. Scient. Instrwrn. 1968 1 (2) 305. English Translation 337. 1967 PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 123 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 76 76 77 78 79 80 81 82 83 84 86 86 Manufacturers Bulletin Perkin Elmer 1969.Fluorescence News 1968 3 1. Chen R. F. Analyt. Biochem. 1967 20 339. Pszonicki L. Chewtia Analit. 1967 12 375; Analyt. Abs. 1968 15 4477. Eisinger J. Photochetn. Photobiol. 1969 9 247. Karnaukov V. N. Kulakov V. I. Mel’nikova E. V. and Yashin V. A. Tsitologiya, 1968 10 654; Chem. Abs. 1968 69 33416t. Chen R. F. Schechter A. N. and Berger R. L. Analyt. Biochem. 1969,29 68. Hiromi K. Ono S. Itoh S. and Nagamura T. J . Biochem. (Tokyo) 1968,64 897; Chem. Abs. 1969 70 59186~. Teller D. N. and Denber H. C. B. Fluorescence News 1969 4 4. Barenboim G. M. Domanskii A. N. Rozanov Yu. M. and Turoverov K. K., Prom Khim Reaktivov Osobo Chist Veshchestv. 1967 No. 8 263; Chem. Abs. 1968, 69 111834~.Langelaar J. De Vries G. A. and Bebelaar D. J . Scient. Instrum. 1969 2 (2) 149. Newell P. B. and O’Brien J. D. J . Quantum Electron. 1968 4 291; Chem. Abs., Ness S. and Hercules D. M. Analyt. Chem. 1969 41 1467. Shcherbov D. P. and Voinov S. A. Prom Khim Reaktivov Osobo Chist. Veshchestv, Fell G. S. and Tilstone W. Spectrovision 1969 (21) 4. Pszonicki L. Chemia Analit. 1967 12 431; Analyt. Abs. 1968 15 4476. Phoenix Precision Inst. Co. Manufacturers Bulletin Instrument Data Sheet SA-6 69, Phoenix Precision Inst. Co. Manufacturers Bulletin Instrument Data Sheet 2BP-864, Brook R. R. and Whitehead N. E. J . Scient. Instrum. 1968 1 (2); (8) 879. Shcherbov D. P. Plotnikova R. N. and Kaptil’nya M. A. Prom Khim Reaktivov Bozhevol’nov E. A. Kreingold S.U. and Plotnikova I. M. Zav. Lab. 1968 34, Winkelman J. and Grossman J. Analyt. Chem. 1967 39 1007. Nishimura M. Legallais V. and Mayer D. Rev. Scient. Instrum. 1969 40 (2) 271. Suzuki S. Oyo Denki Kenkyusho Hokoku 1967,19 20; Chem. Abs. 1969,70 33029q. Rast H. E. and Caspers H. H. Appl. Optics 1967 6 1577, Munro I. H. and Ramsey I. A. J . Scient. Instrum. 1968 1 (2) 147. Selinger B. and Speed R. Chem. Inst. 1969 2 (I) 91. Hollifield H. C. Ph.D. Thesis 1968 126 pp.; Diss. Abs. B 1969 30 (l) 89. Hollifield H. C. and Winefordner J. D. Chem. Inst. 1969 1 (4)’ 341. - and - Analyt. Chem. 1968 40 1759. Zweidinger R. and Winefordner J. D. Ibid. 1970 42 639. Saunders L. B. Winefordner J. D. and Zweidinger R. Analytica Chim. A d a 1967, Winefordner J. D. Accounts Chem.Res. 1969 2 361. St. John P. A. Ph.D. Thesis 1967 Diss. Abs. B 1968 29 (l) 82; Chem. Abs. 1968, St. John P. A. McCarthy W. J. and Winefordner J. D. Analyt. Chem. 1966 38, Cetorelli J. J. McCarthy W. J. and Winefordner J. D. J . Chem. Educ. 1968, Shcherbov D. P. and Plotnikova R. N. Zh Analit. Khim. 1967 22 1146. English - and - Ibid. 1968 23 (lo) 1443. English Translation 1270. - and - Ibid. 1968 23 (ll) 1597. English Translation 1411. Byrom P. and Hundson J. B. Talanta 1968 15 714. Costa L. Grum F. and Paine D. Appl. Optics 1969 8 1149. Testa A. C. Fluorescence News 1969 4 (a) 1. Fletcher A. N. J . Phys. Chem. 1968 72 2743. Chen R. F. Analyt. Biochem. 1967 19 (2) 374. Melhuish W. H. J . Phys. Chem. 1961 65 229. 1968 69 101400k. 1967 (8) 249; Chetn. Abs.1969 70 53620d. 1969. 1969. Osobo Chist Veshchestv 1967 (8) 243; Chem. Abs. 1969 70 63754q. 618. English Translation 739. 47 558. 69 113359s. 1828. 45 98. Translation 9 66. BARK AND WOOD 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 Rusakowicz R and Testa A C Ibzd 1968 72 793 Gdl J. E Photochem Photobzol 1969 9 313 Fletcher A N Ibzd 1969 9 439 Dawson W R and Windsor M W J Phys Chem 1968 72 3251 Weber G and Teale F W. J Trans Faraday Soc 1957,53 646 Rusakomcz R and Testa A C J Phys Chem 1968 72 2680 Himmel C M and Mayer R T Analyt Chem 1970 42 130 Chen R F Nature 1966 209 69 Eisenbrand J and Hauprich H E Pharmazze 1967 22 652 Analyt Abs 1969, Babko A K Baranov S P and Kalabina L V Zh Analzt Khzm 1969 24 485 Russo S F J Chem Educ 1969 46 (6).375 Passwater R A and Hewitt J W Fluorescence News 1969 4 (4) 15 Fisher N and Cooper R M Chem G. Ind 1968 619 Scott D R and Allison J B J Phys Chem 1962 66 561 Smith F J Smith J K and McGlynn S P I Rev Sczent Instrum 1962 (33) 1367 Winefordner J D and St John P A Analyt Chem 1963 35 2211 McCarthy W J and Dunlap K L Talanta 1970 17 305 Wood P R and Bark L S Proc Soc Analyt Chem 1969 7 149 Belyi M U zn Szigeti G Edztor ‘International Conference on Luminescence, Bozhevol’nov Solov’ev E A and Fakeeva 0 A zn Szigeti G op cat Vol 11, Kirkbright G F Saw C G and West T S TaEanta 1969 16 (1).65 Bozhevol’nov E A Solov’ev E A Trudy vses nauchno zssled Inst Khzm Belyi M U and Kushnirenko I Ya Z h Przkl Spectrosk 1969 10 84 Chem -and - Ibzd 1968 9 272 Chem A b s 1969,70 438241 - and - Ibzd 1969 10 810 Chem Abs 1969,71 45419a Kirkbright G F Saw C G Thompson J W and West T S Talanta 1969, 16 1081 Belyi M U and Kushnirenko I Ya Zh Przkl Spectrosk 1968,9 442 Chem Abs , 1969 70 63824n Solov’ev E A Golovina A P Bozhevol’nov E A and Plotnikova I M Vestn Mosk Unzv Ser 11 1966 21 (5) 89 Chem Abs 1967 66 32868t Schmidt K and Staude H Z analyt Chem 1968,234 241 Shcherbov D P and Ivankova A I Prom Khzm Reaktzvov Osobo Chzst Veshchestv Kirkbright G F Saw C G and West T S Analyst 1969 94 457 Kirkbright G F and Saw C G Talanta 1968 15 570 Furukawa M Sasaki S Nakashima R and Shibata S Nagoya Kogyu Gzjutsu Shzkensho Hokoku 1968 17 251 Chem Abs 1969 70 120780~ Cukor P and Weberlmg R P Analytzca Cham Acta 1968 41 404 Poluektov N S Kirillov A I Tishchenko M A and Zelyukova Yu V Zh Analzt Khzm 1967 22 707 English Translation 604 Dobrolyubskaya T S and Anikina L I Ibzd 1967,22 1841 English Translation, 1541 Holzbecher 2 Divis L Novak J Felcmanova-Rauchova D and Reznicek J , Sb Vys Sk Chem-Techno1 Praze Anal Chem 1968 3 65 Chem Abs 1970 72, 96320q Holzbecher 2 and Novak J Colln Czech Chem Commun 1967 32 2956 Gunther G Kem Tzdskr 1969 81 (6-7) 16 Bozhevol’nov.E A and Fakeeva 0 A Prom Khzm Reaktzvov Osobo Chzst Allsalu M L and Kil’k I R Z h Analzt Khzm 1967 22 167 English Trans-16 1092 English Translation 361 Budapest Hungary 1966,’ Kultura 1968 Vol I p 807 p 2068 and - Analyst 1969 94 538 -Reakt 1967 30 202 Analyt Abs 1968 15 3264 Abs 1969 70 111327b 1967 No 8 191 Chem Abs 1968 69 64394p Veshchestv 1967 No 8 218 Chem Abs 1968 69 64356c lation 14 PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 125 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 Allsalu M.L. Kil’k I. R. and Kerikmae M. Mitt. Forstw. Abt. Univ. Tartu, Steele T. W. and Robert R. V. D. U.S. Atomic Energy Commission Report NIM-163, Tarantsova M. I. and Nikol’ksaya Yu P. Izv. Sib.Otd. Akad. Nauk. S.S.S.R. Samsoni Z. Mikrochim. Acta 1967 (1) 88. Couch E. L. Clay and Clay Minerals 1969 17 38. Desai S. R. and Kum K. Sudhalatha Analyst 1969 94 699. Smith A. Y. and Lynch J. J. Can. Dep. Energy Mines Res. Geol. Surv. Can. Pap., 1969 69-40; Chem. Abs. 1969 71 19371~. Kleber H. 2. Analyt. Chem. 1968 234 115. Pitts A. E. and Ryan D. E. Analytica Chim. Acta 1967 37 (4) 460. Marksman A. L. and Strel’tsova S. A. Trudy Tashkent Politekh Inst. 1962 (42) 50; Budesinsky B. and West T. S. Analytica Chim. Acta 1968,42 455. Guilbault G. G. Sadar M. H. and Zimmer M. Ibid. 1969 44 361. Bozhevol’nov E. A. and Solov’ev E. A. Prikl. Spektrosk. Muter Soveshch. 1965, Bozhevol’nov E. A. bst. Chem. Z. 1965 66 74. Galkina L. L. Byull. nauchno-tekh. I n . . Min.Geol. S.S.S.R. Ser. Izuch. Veshchestv Sostava mineral’n Syry’a. tekhnol Obogasch. Rud. 1967 (3) 19; Analyt. Abs. 1968, 15 5873; Chem. Abs. 1968,69 32665m. Desai S. R. and Kum K. Sudhalatha Talanta 1967 14 1346. Sandell E. B. ‘Colorimetric Determination of Trace Metals,’ Third Edition Inter-science 1959. Plotnikova R. N. Ashaeva R. P. and Shcherbov D. P. Issled. Razrab Fotornetrich. Metod Opred Mikrokolichestv Elem. Miner. Syr’e 1967 56; Chem. Abs. 1969 71, 18495r. Mulikovskaya E. P. and Sharyhina I. N. Novye Metody Analiza Khim. Sostava Podzemn Vod 1967 65; Chem. Abs. 1968 69 5122x. West P. W. and Jungreis E. Analytica Chim. Acta 1969 45 188. Sill C. Willis C. P. and Flygare J. K. jun. Analyt. Chem. 1961 33 1671. Dagnall R. M. Smith R. and West T. S. Analyst 1967 92 20.Dagnall R. M. Pratt S. J. Smith R. and West T. S. Analyst 1968 93 638. Parker C. A. and Rees W. T. Ibid. 1960 85 587. Dagnall R. M. Smith R. and West T. S. J . Chem. SOL ( A ) 1966 1595. Dale A. R. Turnbull G. B. and Radley J. A. U.S. Clearing House Fed. Sci. Tech. Inf. A.D. 1967. Available CFSTl from U.S. Govt. Res. Dev. Rep. (1968) 68, (24) 60; Chem. Abs. 1969 70 83980~. Shcherbov D. P. Plotnikova R. N. and Skvortsova T. N. Prom. Khim. Reaktivov Osobo Chist Veshchestv. 1967 No. 8 166; Chem. Abs. 1968 69 643524.. Patrovsky V. Colln Czech. Chem. Commun. 1967 32 2656. - 2. Analyt. Chem. 1967 230 355. Endo R. Kyosac Iho. 1969 81 124; Chem. Abs. 1969 71 98867b. Ryan M. P. and Hingerty D. J . Clin. Path, 1968 21 220. Schachter D. J . Lab. Clin. Med.1961 58 (3) 495; Analyt. Abs. 1960 7 3388; Gusev G. P. Lab. Delo. 1968 (3)’ 157; Chem. Abs. 1968 69 647r. Zepf S. Clin. Chim. Acta 1968 20 473. Clark I. and Hou G. Analyt. Biochem. 1967 19 14. Breen M. and Marshall R. T. J . Lab. Clin. Med. 1966 68 701. Klein B. and Oklander M. Clin. Chem. 1967 13 26. Hill J. B. Ann. N.Y. Acad. Sci. 1962 102 108. Whitmore D. N. and Evans D. L. K. J . Clin Path. 1964 17 644. Oreopoulous D. G. Soyannwo M. and McGeown M. G. Clin. Chim. Acta 1968, Swanson R. A, Hovland D. and Fine L. O. Soil Sci. 1966 102 244; Analyt. Abs., Quantin A. Ann. Sci. Univ. Besancon Bot. 1966 3 (3) 11. 1968 No. 219 168; Chem. Abs. 1969 71 66317s. 1967. Nucl. Sci. Abs. 1968,22 (la) 27541; Chem. Abs. 1968 69 102817~. Ser. Khim. Nauk. 1968 (6) 48; Chem.Abs. 1969 70 73992e. Analyt. Abs. 1969 17 577. 2 (16) 166. Analyt. Abs. 1962 9 2408. 20 349. 1968 15 5070 BARK AND WOOD Bozhevol’nov E. A Fedorova L F Krasavin I A and Dziomko V M Z h -- and - Russian Patent 210,458 February 6th 1968 Chem Abs , Bozhevol’nov E A and Fedorova L F Metody Analzt Khzm Reaktzvov Prep, - and - Ibzd 1968 No 15 48 Chem Abs 1969 70 8611r Budesmsky B and West T S Talanta 1969 16 399 Rodgerson D 0 and Moran I K Clan Chem 1968 14 1206 Kepner B L and Hercules D M Analyt Chem 1963 35 1238 Uemura T SGZ Rep Tohoku Unzv Fourth Ser 1968,34 (l) 31 Chem Abs 1968, Lewm M R Wills M R and Baron D N J Clzn P a t h 1969 22 222 Clark E P and Collip J B J Bzol Chem 1925 63 461 Moser G B and Gerarde H W Clan. Chem 1969 15 376 Gerarde H W Mzcrochem J 1965 9 340 Fingerhut B Poock A and Miller N Clzn Chem 1969 15 870 Classen G H Marquardt P I and Spath M Arznezmzttel Forsch 1968 18 211 Klein B Kaufman J H and Isaacs J Clzn Chem 1967 13 1071 Podchainova V N Skornyakov L V and Dvinyaninov B L Isv Yyssh Ucheb Khzm Khzm Tekhnol 1968 11 241 Chem Abs 1968,69 48939q Babko A K and Vasilevskaya A E Ukrazn Khzm Zhur 1967 33,314 Analyt Abs 1968 15 3208 Holme A Acta Chem Scand 1967 21 1679 Smith G S Analyst 1935 60 735 Chem A b s 1936 30 45 Babko A K Chalaya 2 I and Mikitchenko V F Russian Patent 210,460, February 6th’ 1968 Chem Abs 1968 69 15996~ Podchainova V N and Skornyakova L V Trudy ural’ polztekh Inst 1967,163 60, Analyt A b s 1969 16 2204 Lel’chuk L Yu and Ivanshina V A Isv Tomsk Polztekh I n s t 1967 148 152, Chem Abs 1969 70 92875k Rigin V I and Mel’nichenko N N Zav Lab 1967,33 3-4 English Translation 1.Babko A K Volkova A I Get’man T E and Baranov S P T r Kom Anal Babko A K Volkova A I and Get’man T E Zh Analzt Khzm 1967,22 1004 and - Ibzd 1968 23 39 English Translation 28 ‘Aluminium,’ Acc No 8980 G K Turner Associates California 1967 Hocman G Acta Fac Rerum Natur Unzv Comenzane Chzm 1968 No 13 75 White C E McFarlane H C E Fogt J and Fuchs R Analyt Chem 1967, Nishikawa Y Hiraki K Morishige K and Shigematsu T Japan Analyst 1967, Shigematsu T Nishikawa Y Hiraki K Nagano N Ibzd 1970 19 55 Babko A K and Lisichenok S L UKrazn Khzm Zhur 1969,35 98 Chem Abs , 1969 70 839252 Bognar J and Pataky S M Mzkrochzm Acta 1969 (l) 221 Cook M G Sod Scz Soc Amer Proc 1968 32 292 Goon E Petley J E McMullen W H and Wiberley S E Analyt Chem 1953, Nazarenko V A Thu L N and Dranitskaya R M Zh Analzt Khzm 1967, - and - Ibzd 1967 22 346 English Translation 302 Nazirenko V A Biryuk E A Antonovich V P and Ravitskaya R V Ukrazn Babko A I( Volkova A I and Get’man T E Ibzd 1969 35 69 Chem A b s , Shcherbov D P and Matveets M A Prom Khzm Reaktzvov Osobo Chzst Veshchestv , Analzt Khzm 1969 24 531.English Translation 399 1668 69 15997d 1968 No 15 44 Chem Abs 1969 70 8612s 69 74406h Khzm Akad Nauk S S S R 1969 17 73 Chem Abs 1970 72 96253v English Translation 842 - 1 39 367 16 692 Analyt A b s 1969 16 74 and - Ibzd 1968 17 1092 Chem A b s 1969 70 405659 --25 608 22 518 English Translation 457 Khzm Zhur 1968 34 504 Chem Abs 1968 69 48817y 1969 70 83954h 1967 No 8 157 Chem A b s 1968 69 92637p 126 17 1 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 21 PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 127 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 23 1 232 233 234 235 236 237 238 239 240 24 1 242 243 244 245 246 247 248 Holzbecher Z.Sb. Vys. Sk. Chem. Technol. Praze Anal. Chem. 1967 1 63; Chem. Abs. 1969 70 73878~. Thu L. N. Zh. Analit. Khim. 1967 22 636. English Translation 552.Thu L. N. Dranitskaya R. M. and Nazarenko V. A, Ukrain. Khim. Zhur. 1968, 34 186; Analyt. Abs. 1969 16 2916. Matveets M. A. and Shcherbov D. P. Issled Razrab Fotometrich Metod Oped Mikrolkolichester Elem. Miner. Syr’e 1967 122; Chem. Abs. 1969 71 27100k. Babko A. K. Get’man T. E. and Volkova A. I. Ukrain. Khim. Zhur. 1969 35, 190; Chem. Abs. 1969,71 9293w. Desai S. R. and Sudhalatha K. K. Ind. J . Appl. Chem. 1967 30 (3-4) 116. Bark L. S. and Rixon A. Analytica Chim. Acta 1969 45 425. Alimarin I. P. Zorov N. B. Golovina A. P. and Tsintsevich E. P. Izv. Akad. Nauk. S.S.S.R. Ser. Khim. 1968 12 2678; Chem. Abs. 1969 70 7387931. Volkova A. I. Get’man T. E. and Kukibaev T. I. Ukrain. Khim. Zhur. 1969, 35 844; Chem. Abs. 1970 72 28094~. Vesene T. B. Zav. Lab.1969 35 32. English Translation 38. Mulikovskaya E. P. and Sharyhina I. N. Novye Metody Analiza Khim. Sostava Bordea A. Bull. Inst. Politech. Iasi 1969,13 (3-4) 209; Chem. Abs. 1968,69 92646r. Kasaiura K. Chem. Analyt. Warsaw 1969 14 1325. Lukin A. M. Serebryakova G. V. Bozhevol’nov E. A. and Zavarikhina G. B., Tr. Vses. Nauch-Issled Inst. Khim. Reaktivov. Osobo Chist. Khim. Veshchestv. 1967, No. 30 161; Chem. Abs. 1968 69 8127s. Lukin A. M. Efremenko 0. A. and Petrova G. S. Zh. Analit. Khim. 1967,22 1234. English Translation 1040. Pal B. K. and Ryan D. E. Analytica Chim. Acta 1969 48 (2) 227. Ivankova A. I. and Shcherbov D. P. Issled Razrab. Fotometrich. Metod. Ofired. Mikrokolichestv. Elem. Miner. Syr’e 1967 138; Chem. Abs. 1969 71 18492n. Komlev 0. I.and Zinchuk V. K. Visn l’vivsk’k Univ. Ser. Khim. 1967 (9) 50; Analyt. Abs. 1968 15 6604. Watkinson J. H. in Muth 0. M. Editor ‘Selenium in Biomedicine,’ Avis Westport, 1966. Parker C. A, and Harvey L. G. Analyst 1962 87 558. Costa M. M. Revta. Port. Quim. 1966 8 (3) 136. Lamand M. and Astier C. Ann. Fals. Expert. Chim. 1969 62 (684) 4. Mamedova F. M. Nikolaeva K. and Bozhevol’nov E. A. Stomatology Moscow, 1968 47 ( 5 ) 81; Chem. Abs. 1969 70 34950~. Koval’skii V. V. and Eermakov V. V. Zh. Analit. Khim. 1966 21 447. English Translation 399. Rossum J. and Villaruz P. J . Amer. Wat. W k s Ass. 1962 58 746. Stanton R. E. and McDonald A. J. Analyst 1965 90 497. Karelina L. and Salmane R. Opred Mikrochem. Biol. Obe’klakh 1968 151 ; Chem. Abs. 1969 71 109741j.Ryabchikov D. I. Nazarenko I. I. and Anikina L. I. Zh. Analit. Khim. 1968, 23 1242. English Translation 1095. Lushnikov V. V. and Kondrateva E. N. Novye Metody Analiza. Khim. Sostava Podzemn Vod. 1967 84; Chem. Abs. 1968 69 5123b. Shkrobot E. P. and Shebarhina N. I. Sb. Nauch. Tr. Gos. Nauch-Issled Inst. Tsvet Metal 1968 No. 28 18; Chem. Abs. 1969 70 53705k. Shcherbov D. P. Ivankova A. I. and Gladysheva G. P. Issled Razrab Fotometrich Metody. Opred. Mikrokolichestv. Elem. Miner. Syr’e 1967 10; Chem. Abs. 1969, 71 18569t. Hoffmann I. Westerby R. J. and Hidiroglou J . Ass. 08. Analyt. Chem. 1968, 51 1039. Ewan R. C. Baumann C. A. and Pope A. L. J . Agr. Fd Chem. 1968 16 212. Watkinson J. H. Analyt. Chem. 1966 38 92. Olson 0. E. J . Ass. Ofl. Analyt. Chem. 1969 52 627.Robinson W. O. Dudley H. C. Williams K. T. and Byers H. G. Ind. Engng Klein A. K. J . Ass. Ofl. Analyt. Chem. 1943 26 346. Podzemn Vod. 1967 78; Chem. Abs. 1968 69 5117c. Chem. Analyt. Edn 1934 6 274 128 BARK AND WOOD Hall R. J. and Gupta R. L. Analyst 1969 94 292. Patrias G. and Olson 0. E. Feedstuffs 1969 41 32. Clarke W. E. Analyst 1970 95 66. Nazarenko V. A. and Antonovich V. P. Zh. Analit. Khim. 1967,22 1812. English Translation 15 17. - and - Ibid. 1969 24 358. English Translation 254. Kirkbright G. F. West T. S. and Woodward C. in Shallis P. Editor ‘Proceedings of the SAC Conference Nottingham 1965,’ W. Heffer and Sons Ltd. 1966 p. 474. Shcherbov D. P. and Nikolaeva V. P. Prom. Khim. Reaktivov. Osobo Chist. Vesh-chestv. 1867 No. 8 186; Chem.Abs. 1968 69 64380f. Nishikawa Y. Hiraki K. and Shigematsu T. Nippon Kagaku Zasshi 1969 90 (5), 483; Chem. Abs. 1969 71 357102. Krillov A. I. Lauer R. S. and Poluektov N. S. Zh. Analit. Khim. 1967,22 (9) 1333. English Translation 1123. Chan Ti Huu Volkova A. I. and Get’man T. E. Ibid. 1969 24 (5) 688. English Translation 536. Dubovenko L. I. and Chan Ti Huu Ukrain. Khim. Zhur 1969,35 (6)’ 637; Chem. Abs. 1969 71 54723~. Grigorenko F. F. and Dubovenko L. I. Ibid. 1968 34 (12) 1294; Chem. Abs., 1969,70 102739k. Ozawa L. and Toryu T. Analyt. Chem. 1968 40 187. Shmanenkova G. I. Zemskova M. G. Melamed Sh. G. Pleshkova G. P. and Sukhov G. V. Zav. Lab. 1969 35 (S) 897. English Translation 1073. Poluektov N. S. Vitkun R. A. and Gava S. A. Zh. Analit. Khim.1969,24 (5) 693. English Translation 640. Bozhevol’nov E. A. and Fakeeva 0. A. Trudy Khom. Anal. Khim. Akad. Nauk. S.S.S.R. Inst. Geokhim. Anal. Khim. 1968 16 67; Chem. Abs. 1968,69 2427%. Kononenko L. I. Melent’eva E. V. and Poluektov N. S. Khim. Transuranovykh. Oskolochnykh. Elem. Akad. Nauk. S.S.S.R. Otd. Obslich. Tekh. Khim. 1967 156; Chem. Abs. 1968,69 15773~. Kononenko L. I. Melent’eva E. V. Vitkun R. A. and Poluektov N. S. Prom. Khim. Reaktivov. Osobo. Chist. Veshchestv. 1967 No. 8 223; Chem. Abs. 1968, 69 643818. Kononenko L. I. Tishchenko M. A. Vitkun R. A. and Melent’eva E. V. Zav. Lab. 1968,34 (12) 1432. English Translation 1727. Melent’eva E. V. Poluektov N. S. and Kononenko L. I. Zh. Analit. Khim. 1967, Shigernatsu T. Matsui M. and Sumida T. Bull.Inst. Chem. Res. Kyoto Univ., Shigematsu T. Matsui M. and Wake R. Analytica Chim. Acta 1969 46 101. Belcher R. Perry R. and Stephen W. I. Analyst 1969.94 26. Reisfeld R. and Greenberg E. Analytica Chim. Acta 1969 47 155. Reisfeld R. and Biron E. Talanta 1970 17 105. Gava S. A. and Poluektov N. S. Zav. Lab. 1969,35 (1). 20. English Translation 24. Tishchenko M. A, Kononenko L. I. Vitkun R. A. and Poluektov N. S. Ukrain. Tishchenko M. A. Kononenko L. I. and Poluektov N. S. Prom. Khim. Reactivov. Butter E. Kolovos U. and Holzapfel H. Talanta 1968 15 (9) 901. Poluektov N. S. Kononenko L. I. Vitkun R. A. and Tishchenko M. A. Prom. Khim. Reaktivov. Osobo. Chist. Veshchestv. 1967 No. 8 140; Chem. Abs. 1968, 69 64376h. Kononenko L. I. Mishchenko S. A. and Poluektov N.S. Zh. Analit. Khim. 1966, 21 (ll) 1392. English Translation 1237. Dagnall R. M. Smith R. and West T. S. Analyst 1967 92 358. Kirillov A. I. Vitkun R. A. and Poluektov N. S. Sovrem. Metody. Khim. Spectral. Anal. Mat. 1967 211; Chem. Abs. 1968 68 26676d. Anikina L. I. Bagreev V. V. Dobrolyubskaya T. S. Zolotov Yu. A. Karyakin, A. V. Miklishanskii A. Z. Nikitina N. G. Palei P. N. and Yakovlev Yu. V., Zh. Analit. Khim. 1969 24 (7) 1014. English Translation 810. 22 (2) 187. English Translation 158. 1968 46 (6) 249; Chem. Abs. 1969 71 56289j. Khim. Zhur. 1966 32 (5) 608; Analyt. Abs. 1967 14 6327. Osobo. Chist. Veshchestv. 1967 No. 8 231; Chem. Abs. 1968 69 73714~. 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 28 1 28 PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 129 283 284 285 286 287 288 289 290 29 1 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 3 15 316 317 318 319 Anikina L.I. Dobrolyubskaya T. S. and Karyakin A. V. Le Viet Binh 26 PrikZ Spektrosk 1969 10 (3) 518; Chem. Abs. 1969 71 93682. Melamed Sh. G. Antonov A. V. and Kulevskii L. V. Zav. Lab. 1967,33 (6) 712; Chem. Abs. 1968 68 18352e. Titkov Yu. B. Russian Patent 256,332 November 4th 1969; G e m . Abs. 1970, 72 96425~. Babko A. K. Dubovenko L. I. and Mikhailova L. S. Zh. Analit.Khim. 1966, 21 (5) 648. English Translation 491. Plotnikova R. N. Perminova D. and Shcherbov. D. P. Prom. Khim. Reaktivov Osobo Chist Veshchestv. 1967 No. 8 197; Chem. Abs. 1968,69 64412t. Hercules D. M. Talanta 1961 8 485. Brookes A. and Townshend A, Chem. Commun. 1968,24 1660. Bognar J. and Jellink O. Mikrochim. A d a 1968 (5) 1013. Temkina V. Ya. Dyatlova N. M. Kreingold S. U. Yaroshenko G. F. Antonov, V. N. Lasyovskii R. P. and Bozhevol’nov E. A. Zh. Analit. Khim. 1967,22 (12), 1830. English Translation 1532. Andrushko G. S. Maskinycheva Z. and Talipov Sh. T. Uzb. Khim. Zh. 1969, 13 (2) 24; Chem. Abs. 1969 71 77019m. Bottei R. S. and Trusk B. A, AnaZyt. Chem. 1963 35 1910. - and - Analytica Chim. Acta 1967 37 409. - and - Ibid. 1968 41 374. Pal B. K.and Ryan D. E. Ibid. 1969 47 35. Kalinichenko E. Ukrain. Khim. Zhur. 1969,35,755; Chem. Abs. 1969,71 11932811. - Russian Patent 252,712 September 22nd 1969; Chem. Abs. 1970 72 62527~. Kreingold S. U. and Bozhevol’nov E. A. Tr. Anal. Khim. Akad. Nauk SSSR Int. Geokhim. Anal. Khim. 1968 16 194; Chem. Abs. 1968 69 56745~. Temkina V. Ya. Sidorenko V. V. Yaroshenko G. F. and Lastovskii P. P. Prom. Khim. Reaktivov Osobo. Chist. Veshchestv 1967 No. 8 101; Chem. Abs. 1968, 69 86967e. Fink D. W. Pivnichny J. W. and Ohnesorge W. E. Analyt. Chem. 1969 41 833. Babko A. K. and Kalinichenko I . E. Metody. Anal. Khim. Reaktivov Prep. 1966, No. 13 82; Chem. Abs. 1968 68 9057k. Kreingold S. U. Bozhevol’nov. E. A, Lastovskii P. P. and Sidorenko V. V., Zh. Analit. Khim. 1963 18 ( l l ) 1356; Chem.Abs. 1964 60 6204f. Kreingold S. U. Bozhevol’nov E. A. Lastovskii P. P. and Sidorenko V. V. Dokl. Akad. Nuuk. S.S.S.R. 1963 153 (l) 97; Chem. Abs. 1964 60 74273. Laanmaa M. Allsalu M. L. and Kokk H. Tartu Riikliku. Ulikooli Toim 1968, No. 219,199; Chem. Abs. 1969 71 77012d. Possidoni de Albinati J. F. An. Asoc. Quim. Argent. 1967 55 (1-2) 61. Schenk G. H. Dilloway K. P. and Coulter J. S. Analyt. Chem. 1969 41 (3) 510. Kreingold S. U. and Bozhevol’nov E. A. Metody Analiza Khim. Reaktivov i Prepara-Holzbecher Z. and Novak J. Colln Czech. Chem. Commun. 1968 32 (S) 2956. Fink D. W. and Ohnesorge W. E. Analyt. Chem. 1969 41 (1)’ 39. Salam-Khan M. A. and Stephen W. I. Analytica Chim. Acta 1970 49 255. Harris J. and Ritchie K. Ann. N . Y . Acad.Sci. 1969 153 706. - and - Analyt. Chem. 1969 41 ( l ) 163. Yamane Y. Yamada Y. and Kunihiro S. Bunseki Kagaku 1969 17 (9) 973; Chem. Abs. 1969 70 43723a. Yamane Y. Muyazaki M. and Ohtawa M. Ibid. 1969 18 (6) 750; Chem. Abs., 1970 72 18150~. Talipov Sh. T. Makismycheva 2. T. and Zeltser L. E. DokZ. Akad. Nauk. Uzbek. S S R 1968 25 (lo) 25; Chem. Abs. 1909 70 83949k. Bozhevol’nov E. A. Kreingold S. U. and Sosenkova L. I. Trudy vses. naucho-issled. Inst. Khim. Reaktivov Osobo Chist. Khim. Veshchestv. 1967 (30) 176; Analyt. Abs. 1968 15 3806; Chem. Abs. 1968 69 73657d. Bozhevol’nov E. A. and Kreingold S. U. Zh Analit. Khim. 1963 18 (81 942. English Translation 818. tov 1965 No. 11 49; Chem. Abs. 1966,65 9724g. - f and - Uzb. Khim. Zh. 1968 12 (6). 16; Chem.Abs. 1969,70 73867t 130 320 321 322 323 324 325 326 327 328 329 330 33 1 332 333 334 335 336 337 338 339 340 34 1 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 BARK AND WOOD Solov’ev E. A. Bozhevol’nov E. A. Lebedeva N. A. Mironov A. F. and Evstig-Ryan D. E. and Pal B. K. Analytica Chim. Acta 1969 44 (2) 385. Perminova D. N. and Shcherbov D. P. Prom. Khim. Reaktivov Osobo Chist Vesch-Shcherbakov D. P. Perminova D. N. Issled Razrab Fotometrich Metod Opred Taskarin B. T. and Shcherbov D. P. Ibid. 1967 85; Chem. Abs. 1969’71 18572~. Babko A. K. Terletskaya A. V. and Dubovenko L. I. Zh. Analit. Khim. 1968, Wheeler G. L. Andrejack J. Wiersma J. H. and Lott P.F. Analytica Chim. El. Ghamry M. T. Frei R. W. and Higgs G. W.. Ibid. 1969 47 (l) 41. Podberezskaya N. K. Sushkova V. A, and Shilenko E. A. Zav. Lab. 1967,33 (2), Marinenko J. and May I. Analyt. Chem. 1968 40 (7) 1137. Ryan D. E. and Afghan B. K. Analytica Chim. Acta 1969 44 115. Trentholm R. and Ryan D. E. Ibid. 1965 32 317. Haworth D. T. and Boeckeler R. H. Microchem. J. 1968 13 158. Zholin A. V. and Serebryakova G. V. Trudy Vses. Nauch. Issled Inst. Khim. Reak-tivov Osobo Chist. Khim. Veshchestv. 1967 No. 30 242; Chem. Abs. 1968,68 56328j. Mahanand D. and Houck J. C. Clin. Chem. 1968 14 6. Konstantinov A. V. Korobochkin L. M. and Anastasina G. V. Trudy Nov. Ap$arature Metodikam pew. Mosk med. Inst. 1967 (5) 167; Chem. Abs. 1968, 68 57314~. Budesinsky B.and West T. S. Analyst 1969 94 182. Ivankova A. I. Perminova D. N. and Shcherbov D. P. Prom. Khim. Reaktivov Osobo Chist. Veshchestv. 1967 No. 8 174; Chem. Abs. 1968 69 92661s. Bottei R. S. and D’Alessio A. S. Analytica Chim. Acta 1967 37 (3) 405. Babko A. K. Chan Ti Huu Volkova A. I. and Get’man T. E. Ukrain. Khim. Zhur., - and - Ibid. 1969 35 (6) 642; Chem. Abs. 1969 71 56353a. MilkLy R. G. and Fletcher M. H. J . Amer. Chem. SO~. 1957 79 5425. Dobrolyubskaya T. S. Nauka Moscow 1968 99 pp.; Chem. Abs. 1969 70 8656j. Hocman G. Lacko G. and Hegedus L. Acta Fac. Rerum Univ. Comeniane Chim., Powell W. A. and Saylor J. H. Analyt. Chem. 1953 25 960. Guyon J. C. Jones B. E. and Britton D. A. Mikrochim. Acta 1968 (6) 1180. Taves D. R. Talanta 1968 15 1015. Schenk G.H. and Dilloway K. P. Analyt. Lett. 1969 2 379. Huckbay W. B. Welch E. T. and Metler A. W. Analyt. Chem. 1947 19 154. Thompson C. R. Zielenski L. F. and Ivie J. O. Atmos. Envir. 1967 1 (3) 253. Ivie J. O. Zielenski L. F. Thomas M. D. and Thompson C. R. J . Air. Pollut. Karyakin A. V. and Babicheva G. G. Zh. Analit. Khim. 1968,23 ( 5 ) 789. English Friend J. P. Talanta 1969 16 617. Babko A. K. Terletskaya A. V. and Dubovenko L. I. Ukrain. Khim. Zhur. 1966, Babko A. R. Markova L. V. and Lukovskaya N. M. Zh. Analit. Khim. 1968,23 (31, - and - Russian Patent 217,026 April 26th 1968; Chem. Abs. 1968, GoIovos G. Haro M. and Frieser H. Talanta 1970 17 273. Britton D. A, and Guyon J. C. Microchem. J. 1969 14 1. Guilbault G. G. and Kramer D. N. US. Patent 3,432,269 March llth 1969.Wronski M. Chew. Anal. Warsaw 1969 14 1183. Grunert A. Ballschimter K. and Tolg C. Talanta 1968 15 451. Axelrod H. D. Cary J. H. Bonelli J. E. and Lodge J. P. jun. Analyt. Chem., neeva R. P. Ibid. 1969 24 (2) 231. English Translation 125. chestv. 1967 No. 8 181; Chem. Abs. 1968 69 92683a. Mikrokolichestv Elem. Miner. Syre 1967 149; Chem. Abs. 1969 71 18574r. 23 (6) 932. English Translation 809. Acta 1969 46 (2) 239. 152. English Translation 174. 1969 35 (3) 292; Chem. Abs. 1969 71 9404h. 1968 (13) 71. Control Ass. 1965 15 195. Translation 684. 32 (7) 728; Analyt. Abs. 1967 14 6807. 401. English Translation 330. 69 64439g. 1969 41 (13) 1856 PHOTOLUMINESCENCE AND CHEMILUMINESCENCE IN INORGANIC ANALYSIS 131 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 Axelrod H. D. Bonelli J. E. and Lodge J. P. jun. Ibid. 1970 42 (a) 512. Lukovskaya N. M. and Markova L. V. Zh. Analit. Khim. 1969 24 1862. Orban G. Szentirmay Z. and Patko J. in Szigeti G. Editor ‘International Con-Kubal J.,Chem. Listy 1968 62 (12) 1478; Chem. Abs. 1969 70 740212. Watanabe H. and Nakadoi T. J . A i r Pollut. Control Ass. 1966 16 (ll) 614; Dmitriev M. T. and Kitrosski N. A. J . Phys. Chem. U.S.S.R. 1968.42 (12) 1672. Guyon J. C. and Shults W. D. J . Amer. Wat. W k s Ass. 1969 61 (S) 403. Rubin M. and Knott L. Clin. Chim. Acta 1967 18 409. Sardesai V. M. and Provido H. S. Microchem. J. 1969 14 550. Bermejo Martinez F. Monserrat Gras Gonzalez de Lopidana and Antonia Berrera Nishikawa Y. Bunseki Kakagu 1968 17 (7) 888; Chem. Abs. 1968 69 113115j. Katoh K. Ibid. 1968 17 ( l l ) 1377; Chem. Abs. 1969 70 53627m. Nikolic K. Velasevic K. and Buric I. D. Glasn. Hem. DruSt. Beogr. 1968 31 (9-lo) 393; Chem. Abs. 1969 70 73781k. Stawinski J. Rocz Glebozn 1967 18 191; Chem. Abs. 1968 68 11906511. Chen R. F. Analyt. Lett. 1968 1 (7) 423. Lysenko G. M. Kislyak G. M. and Ponochovnyi V. I. Zh. Prikl Spectrosk 1968, 9 (3) 492; Chem. Abs. 1969 70 38874~. Weirsma J. H. and Lott P. F. Analyt. Lett. 1968 1 (lo) 603. Szarvas P. Korondan I. and Raisz I. Mag. Chem. Fdly. 1966 72 (lo) 441. Rudolph G. G. Holler J. J. jun. and Ford W. J. Clin. Chim. Acta 1967 18 187. Borle A. B. and Briggs F. N. Analyt. Chem. 1968 40 (2) 339. Bermejo Martinez F. and Monserrat Gras Gonzalez de Lopidana Analytica Chitn. - and - Inform Quim. Anal. Madrid 1969 23 151. Monserrat Gras Gonzalez de Lopidana Acta Cient Compostelana 1966 3 (a) 173; Temkina V. Ya. Dyatlova N. M. Yaroshenko G. F. Lavrova 0. Yu. and Lastov-Aguila J. F. Talanta 1967 14 (lo) 1195. Escarrilla A. M. Analytica Chim. Acta 1968 43 (2) 353. Talipov Sh. T. Gorina S. N. Maksimycheva 2. T. and Kanumikova V. V., Nauch. Tr. Tashkent. Gos. Univ. 1968 No. 323 107; Chem. Abs. 1970,72 96376n. Lukovskaya N. M. and Markova L. V. Zh. Analit. Khim. 1969 24 1893. English Translation 1539. Talipov Sh. T. Maksimycheva 2. T. and Andrushko G. S. Dokl. Akad. Nauk. Uzbek. SSR 1968 25 (S) 24; Chem. Abs. 1969 70 83975r. ference on Luminescence,’ Vol. 1 Kultura 1968 p. 611. Analyt. Abs. 1968 15 1095. Ramallo Inform Quim. Anal. 1969 23 (4) 109. Acta 1969 47 139. Chem. Abs. 1969 69 8009e. skii. P. P. Zh. Analit. Khim. 1969 24 (2) 240. English Translation 135
ISSN:0300-9963
DOI:10.1039/AS9710100041
出版商:RSC
年代:1971
数据来源: RSC
|
3. |
Recent developments in activation analysis |
|
Selected Annual Reviews of the Analytical Sciences,
Volume 1,
Issue 1,
1971,
Page 133-175
T. B. Pierce,
Preview
|
PDF (5360KB)
|
|
摘要:
Recent Devetopments in Activation Analysis T. B. PIERCE Analytical Sciences Diuisiow A .E.R,E. HavwdL Berks. Contents General considerations Introduction Techniques of activation analysis Chemical separation techniques Radiation measurement Reactor activation analysis-Delayed techniques -Prompt techniques Accelerator neutron sources-Delayed techniques -Prompt radiation Radioisotope neutron sources Delayed techniques Prompt techniques-Elastic scattering Methods of isolation of activity Neutron-activation analysis Charged-particle activation analysis -Prompt gamma radiation -Particle group measurement -Nuclear positive ion microprobe Gamma photon activation analysis 13 134 PIERCE General Consideration Introduction Activation analysis has been recognised as an analytical technique for over 30 years and for the last 20 years of that time has been intensively investigated at specialist establishments that have had available the experimental facilities necessary for supporting the work.The cost of these facilities particularly of suitable radiation sources has often been high limiting both the number of centres at which the full range of activation methods can be investigated and the avail-ability of shared irradiation facilities. As a result activation analysis has not been able to enjoy the support of many of the organisations dedicated to the develop-ment of analytical techniques for specific applications particularly for industrial use which have played so important a part in bringing other analytical techniques from the research phase through the development stages needed to mould the method into an approved tool that can be entrusted with routine widespread application.In spite of the fact that only a limited number of centres have been sufficiently well equipped to pursue the development of a range of activation analy-sis techniques potential advantages of the method have been sufficient to justify the investment of substantial effort a t some large establishments with more specialised effort at smaller laboratories with restricted access to irradiation and other facilities so that the technique has been under continuous development. This interest has been reflected in the number of papers published; a recent bibliography contains more than 4,000 references and shows that the growth of publications on the subject of activation analysis is almost exponential with a doubling time of little over 3 years.l Initially the technique of activation analysis was fairly clearly defined.A sample would be irradiated in a field of particles or electromagnetic radiation usually neutrons and most frequently those from a nuclear reactor. After the completion of irradiation the elemental concentration of one or more components of the sample was determined by measuring the induced radioactivity usually after isolation by radiochemical separation of the activity to be counted from all others produced in the sample. However the last 5 years have seen increasing interest on the part of the activation analyst in the application of techniques that are no longer based solely on the measurement of induced radio-activity but which still involve an irradiation step as an integral part of the analytical procedure and are consequently often grouped together under the general heading of activation analysis.Thus activation analysis has become a very much more diffuse subject than it was some years ago as analytical scientists have harnessed more of the properties of the nucleus to provide relevant information, and instrumentation and techniques have been developed that have resulted in the less convenient methods being put on a more competitive analytical basis. To the non-specialist activation analysis can now appear to be a somewhat arbitrary accumulation of separate techniques a view that is not helped by the fact that there is no comprehensive up-to-date text-book available which treats the subject in a systematic manner.While a detailed theoretical discussion and comparison of th RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 135 nuclear interactions exploited by activation analysis as a means of obtaining analytical information is out of place in a review article such as this a short sub-section gn techniques of activation analysis has been included in the first part of this review in an attempt to give the reasons for the structure chosen for the article and also in the hope that it will give an idea to the non-specialist albeit in a very superficial way of the interconnection between the techniques discussed. The majority of the references used for this review have appeared in print within 3 years of June 1970 but reference is made to earlier papers when these are of special interest and illustrate some particular facet of activation analysis not well covered elsewhere.It is pertinent to conclude this short introduction with a few comments on the general trends that have become apparent in activation analysis over recent years. Many developments in techniques have followed directly from improvements in instrumentation and it is encouraging to see that instrument manufacturers and machine operators are becoming more prepared to provide equipment tailored to meet the exacting demands of the analyst rather than the activation analyst having to make do with equipment primarily designed for an entirely different purpose.Improvements to methods of y-ray spectroscopy common to so many activation techniques have extended the scope of intact methods of analysis. These have consequently taken over some of those determinations that would previously have required chemical manipulation of the sample ; y-ray spectroscopy of partially separated mixtures of radionuclides is now also a more attractive proposition. However the strength and permanence of chemical separations in activation analysis are clearly illustrated by work in laboratories that are well versed in sophisticated instrumental techniques as well as radiochemical methods, and which have still found the need to develop new methods of chemical separation in spite of being well equipped with advanced instrumentation for y-ray spectro-scopy.Clearly activation analysis has as yet made little impact on industrial applications where analyses of large numbers of samples are required and for which in some ways activation analysis is well suited. Partly because of the cost (although some activation systems can be produced that are competitive in price with larger conventional analytical techniques such as X-ray fluorescence or mass spectroscopy) but also because of lack of effort at existing activation laboratories, the possibility of using activation analysis for specific determinations of interest to industry has not always been investigated in sufficient detail to warrant the expenditure by a potential user necessary to equip a laboratory with suitable equipment. It is therefore of considerable interest that a very large British industrial organisation with previous experience of reactor neutron-activation analysis has seen fit to invest in a nuclear reactor of its own to be used in part for analytical work.Techniques of activation analysis As activation analysis now embraces a number of different techniques, apparently chosen in a somewhat haphazard way (elastic scattering based o 136 PIERCE Coulomb repulsion is generally included in works on activation analysis whereas Mossbauer spectroscopy is normally considered to be an entirely separate subject) to avoid confusion a review such as this one must identify the different fields of interest by sub-division. To explain the method by which sub-divisions used in this review have been chosen and to acquaint the non-specialist with the inter-relationship of the various methods a very short discussion of the techniques of activation analysis and the relevant processes occurring during a nuclear reaction are outlined below.In the author’s opinion the usual procedure normally followed in works on activation analysis of considering solely the formation and decay of radioactive species is no longer adequate for describing activation procedures as an increasing number of publications are appearing that are concerned with the detec-tion of radiations other than those derived from the decay of radioactive nuclei. The irradiation stage of an activation procedure is an intrinsic part of the analytical method and distinguishes activation analysis from isotope techniques such as radioactive isotope dilution radio release and isotopic exchange in which production of the radioactive species is carried out quite independently of the analysis.One way of illustrating the different techniques of activation analysis is by considering a nuclear reaction occurring according to Bohr’s theory of compound nucleus fomation. The nuclear reaction occurs by two independent steps the first being the absorption of the incident particle to form a compound nucleus and the second the subsequent decay of the compound nucleus. The first step assumes that the incident particle is absorbed by the target nucleus to form a compound nucleus in a highly excited state with an energy level E in the nucleus +’ *$ as shown in Fig. 1. A+a A t a - I C z + z z t z - I Fig.1. Schematic representation of a nuclear reaction A + a - l B z + RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 137 The excitation energy derived from the binding energy of the incident nucleus and the kinetic energy of the incident particle can be rapidly re-distributed amongst the nucleons in the compound nucleus and if at any stage sufficient energy is concentrated on a single nucleon or a group of nucleons in the nucleus these may be ejected. The second stage of the nuclear reaction therefore is the decay of the compound nucleus feeding levels in the product nucleus shown in Fig. 1 as emission of the proton groups Po PI and P, although of course other particles, for example neutrons or a-particles may be emitted during nuclear decay.If a particle is not ejected the excited levels of the compound nucleus may lose their excitation energy in the form of y-radiation (yl y2 ys and y4 in Fig. l) which is characteristic of the nuclear transitions occurring. All this takes place very rapidly, nearly always in less than 1 ps and thus radiations resulting from compound nucleus decay cannot usually be measured after physical transfer of the sample from the place of irradiation to the radiation counter as is usual in traditional methods of activation analysis They are ‘prompt’ radiations which must be measured while the irradiation is in progress. Particle emission may leave the product nucleus in either the ground state or in excited states and if an excited state is produced the excitation energy may again be lost in the form of y-radiation as represented by transitions y6 and y7.The product nucleus ;$ ZiC having been formed artificially may be unstable in its ground state decaying by the emission of radiation familiar in the decay of radioisotopes. In Fig. 1 the product nucleus is assumed to decay by /3-emission (Po and /?,) to form the final stable nucleus 2 $ -’B although emission of /?+ electron capture etc. might be alternative modes of decay. If /3-emission feeds excited states of this residual nucleus then again the excitation energy may be lost by emission of y-radiation occurring very rapidly after emission of the /3-particle (yJ. As the half-lives associated with /3-decay are generally relatively long emission of the /3-particles and those radia-tions following /3-decay (y8) can often be detected after removal of the sample from the place of irradiation to a counter.Prompt radiations are therefore usually nuclear reaction products ; delayed radiations are most frequently derived from the decay of radionuclides. The decay scheme shown in Fig. 1 is by no means universal as not all reactions occur through compound nucleus formation and decay may take place by more complex modes than those shown in Fig. 1. Nevertheless the scheme does illustrate many of the types of decay that are now used as a basis for activation analysis. The general types of over-all reaction are summarised in Table I below. TABLE I TYPES OF ACTIVATION PROCEDURE ;A + [;$:B] -$A + :a . . . . (1) __+ $A+ + fta * . * (2) -+ZA$rtB + y .. . (3) -++$:ItD+td .. .. (4 138 PIERCE In equations (1) and (2) the same particle is emitted as is absorbed and the reactions are called scattering reactions. In equation (1) kinetic energy is conserved and therefore the reaction appears to the observer to be a typical ‘billiard-ball’ interaction but in equation (2) kinetic energy is lost and is converted to excitation energy of the product nucleus. Equations (3) and (4) are transmutation reactions in that the product nucleus is different from the target nucleus but in equation (3) the product nucleus is formed by particle-capture with the emission of y-radiation, whereas in equation (4) another particle is emitted resulting in a different product nucleus. Neutrons charged particles and y-photons are all regularly used to irradiate samples and to initiate nuclear reactions for analytical purposes so that three general categories of activation analysis based on the type of irradiating radiation, can be distinguished.The character of the different radiations and the methods of their production profoundly influence the nature of the analytical information produced and the experimental techniques that are acceptable. The analytical procedure chosen will of course be dependent on the type of interaction upon which measurement is based. However the options open must clearly differ for prompt and delayed radiation as for the measurement of prompt radiation counting conditions will be influenced by the proximity of the radiation sources and there is no chance of using chemistry to isolate the radiation to be measured from all other emitted from the sample between irradiation and counting.The following chart summarises the different categories of radiation available for sample irradiation, but neutrons have been so extensively used for activation analysis that neutron techniques have been sub-divided into three categories based on the available Particles available for irradiation I I y-Phbtons Neutrons Chaiged particles I I I I I Reactor neutrons Neutrons from accelerators Radioisotope sources sources of neutrons reactor accelerator and radioisotope. The second chart shows the options available for activation procedures that are based on methods available for isolation of the radiation to be determined. This review therefore is divided into three major sections one for each of the three types of particles used to irradiate samples in activation procedures as shown in the charts but the section dealing with neutron-activation analysis is sub-divided into three further section RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS Radiation measurement procedures I 139 Methods based on measurement of delayed radiation I I Methods based on measurement of prompt radiation I I Instrumental Chemical methods separation of activity covering analytical methods based upon reactor accelerator and radioisotope neutron sources.Each section consists of two parts wherever appropriate the first being concerned with methods based upon the measurement of delayed radiation which make up the greater part of activation procedures as they have been more extensively investigated and applied while the second considers those methods that are dependent upon the measurement of prompt radiation for their success.Where prompt and delayed techniques axe combined as in certain neutron generator applications these are included in the section covering prompt methods. In addition a separate section is included that is concerned with methods of isolating the radiation to be measured from other emission from the sample as these are common to most activation procedures. Methods of Isolation of Activity During irradiation it is extremely fortuitous if the only activity induced in the sample is that upon which the analytical measurement is to be based; conse-quently some means must be found for isolating the required activity from all other radiations emitted from the sample.Initially chemical separation was used almost exclusively to perform this function but following extensive development, instrumental methods of radiation measurement have assumed an increasingly important role in activation analysis. Over the last few years significant advances have been made in the manufacture of new y-ray detectors in instruments to handle detector output and in mathematical and computer methods used for data processing. It must remain very reassuring however to the trained chemist that, in spite of sophisticated instrumentation now available to the activation analyst, chemical separation is often the only satisfactory method of achieving the high decontamination factors necessary for isolating very small amounts of active material and its activity from one delayed source from very large amounts of activity from others.Intact methods of analysis are most frequently based on y-ray spectroscopy particularly for measurement of the induced radiation emitted by radioisotopes as contributions from several sources can often be distinguished 140 PIERCE from the same sample and the convenience of the technique is a great attraction where the alternative chemical manipulation is complex or when large numbers of samples must be processed consequently making heavy demands upon laboratory effort. Therefore when induced radioactivity is measured and y-ray spectroscopy cannot adequately distinguish between the induced radioactivities of a complex mixture there has been a noticeable trend to combine y-ray spectroscopy with chemical separation by carrying out group separations in which y-emitters are not isolated individually but are separated into a few convenient groups and the contribution of individual emitters to the total activity of the groups is then distinguished by y-ray spectroscopy.By this means a compromise is achieved whereby chemical separation is used to remove interferences that restrict the application of y-ray spectroscopy but the heavy investment in laboratory effort that would be needed to carry out complete chemical separation of the components to bring them to a high state of radiochemical purity is reduced by using y-ray spectroscopy to distinguish the several y-ray emitters in the simplified mixtures.Chemical separation techniques Chemical separation techniques are primarily applied to the separation of a radionuclide from an irradiated sample between irradiation and counting and are particularly valuable when small amounts of active species are to be isolated from a much more active matrix. Inactive carriers are usually added to permit calcula-tion of chemical yield and many procedures are still designed to provide the product in a state of high radiochemical purity so that counting can be carried out with relatively simple equipment and the heavy investment in instrumentation required for more complex techniques is avoided. Separations must clearly be completed within the time-scale permitted by the half-life of the active species so that the preferred chemical and experimental methods may have to be modified to cope with the rate of decay of the radionuclide measured.Pre-separation of specific components of a sample is occasionally carried out when separation between irradiation and counting is not feasible but the method is susceptible to errors caused by reagent blanks which are avoided by post-irradiation separations to provide one of the major advantages of activation analysis for trace-element determinations. Much information is available in the literature about chemical separation procedures and the series of monographs issued by the U.S. National Academy of Sciences2 remains a useful source of information on the behaviour of individual elements that can be used as a basis for devising separation procedures.Also included are a number of works on specific separation techniques. Precipitation for so long used as the principal separation tool of the radio-chemist continues to appear with regularity in published radiochemical separation procedures although the general lack of selectivity of the method frequently limits its use to primary fractionation of the components of a mixture for example by scavenging or for the preparation of sources for chemical yield determination. Salutsky’s review3 provides a useful introduction to the precipitation method and Bock Werthmann4 has summarised the use of some organic reagents for precipita RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 141 tion thus employing a technique for so long applied to non-active analytical work.5 Solvent extraction figures prominently in separation procedures by virtue of its versatility convenience and the simple equipment needed.Good selectivity can often be achieved by careful choice of extractant and conditions. The book by Morrison and Freiser,G although first published some years ago provides a valuable introduction to the use of solvent extraction for analytical separations and contains much useful data. A more recent book by Stary' considers the use of chelates in solvent extraction techniques. In addition an immense amount of data is available in open literature describing extraction mechanisms and experimental behaviour of solvent extraction systems. Separation on organic ion exchangers continues to be the most widely used technique based on differential migration methods and will be as familiar to those analysts concerned with the separation of inactive materials as to those requiring the separation of radioactive species.The monograph by Samuelson8 continues to be a useful source of information for the practicing analyst and a newer book by Inczedy9 is also biased towards the analytical use of ion exchangers. Reversed-phase partition chromatography which aims to combine the choice of extractants available for solvent extraction with the multi-plate effect experienced in differen-tial migration by immobilising an extractant on a suitable carrier has been used for radiochemical separations for more than 10 years and is being increasingly applied to activation analysis.Comprehensive reviews on the subject are difficult to find but the article by CerrailO demonstrates the capability of the technique. Thin-layer chromatography so widely applied to organic chemistry can be adapted to provide inorganic separationsll and its use extended for radiochemical separations by reversed-phase techniques,12 but like paper chr~matographyl~ it suffers from the limitation of having only a low capacity if spot distortion is to be avoided. Gas chromatography has been extensively developed for organic applica-tion and is an attractive method for fast separations provided that the active species can be converted to a suitably volatile form. The method is applicable to many element combinations as a very large number of all the elements in the periodic table have been separated by gas chromatography.14 Radiochemical separation on inorganic precipitates is a field of separation technology that has been receiving increasing attention over recent years as reaction with ions in solution can occur by mechanisms such as re-crystallisation precipitation isotopic exchange and redox reactions to alter the elemental concentrations and the results of about 2,000 absorption experiments designed to assess the retention of a large number of ions on several precipitates have been summarised by Girardi et ~ 1 .~ 5 Distillation, which has long been used for the separation of certain radioactive species in activa-tion procedures for example for the separation of fluorine-18 produced by charged-particle activation analysis can offer quick and rapid separations of suitable elements,la and other less widely used methods such as isotopic exchange,17 electrochemical methods1* and separation on micro-crystalline cellulose columns1g have all been used to augment the repertoire of the activation analyst 142 PIERCE Of the two further aspects of radiochemical processing sample solution and chemical yield measurement rapid and complete solution of the analytical specimen often proves to be the most troublesome stage of sample treatment.The method chosen must ensure complete solution of major and minor constituents and Girardi2* has summarised a number of techniques in general use for biological materials. Chemical yield determination by weighing a precipitate is still widely used but other established methods of quantitative analysis axe increasingly used to determine the amount of element associated with the radioactivity measured.Re-act ivat ion and radioactive isotope dilution provide alternative radio analytical methods of mass measurement but considerable interest has been shown in recent years in so-called substoicheiometric finishes to activation procedures. These methods require the measurement of activity associated with a predetermined amount of active species sampled by adding to the system a known amount of reactant which is insufficient to combine with all the active species present in the system,21 and has the advantage that small masses of material can be examined by this means. Procedures have been developed for very many elements and these have been summarised in a book by Ruzicka and Stary.22 From the data available on separation techniques chemical procedures can be devised that permit one or a number of elements to be separated from complex matrices and determined at very low levels.For example ten trace elements have been determined in sixty-seven different meteorites by using radiochemical separation following neutron-activation analysis in some cases at levels of as low as 0*0005 ~ . p . m . ~ ~ Wile complex radiochemical separations are still necessary for many trace element determinations there has been a noticeable increase over the last few years in the use of group separations to avoid overlap of interfering y-peaks and to prevent the dominating activities induced in the sample from masking the radiations of low intensity.Examples of group separation are to be found in many fields of application of activation analysis. For example six ele-ments arsenic chromium cobalt copper molybdenum and tungsten have been determined in steels by using first a-benzoinoxime in chloroform followed by an ether extraction of chloride complexes and then counting the induced activity with either a germanium counter or a sodium iodide scintillator when high resolu-tion or high efficiency were required.24 Rock standards have been fractionated into six groups by volatilisation extraction with hydrated antimony pentoxide, ion exchange and solvent extraction with tributyl phosphate and then forty-five elements determined by y-ray spectroscopy on the separated groups.25 Group separation has also been used for the automatic processing of irradiated biological samples the separation being obtained by absorption on columns of specific materials.Column materials included Kiesehlguhr impregnated with di- (2-ethylhexyl) orthophosphate anion exchangers chelating resins and zirconium ortho-phosphate providing fully automatic separation in 40 minutes.26 In view of the substantial effort that is being devoted in non-active laboratories to automated chemical procedures in an attempt to reduce the investment in labora-tory effort required for each analysis automation of radiochemical separation RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 143 is an expected development where large numbers of samples are to be handled.Relatively simple systems have been devised to operate in a fixed mode and are clearly adequate for many routine operations where similar chemical operations can be guaranteed for a substantial period of time. More complex systems based on computer programmes27 have also proved to be of value when greater flexibility is required; digital computers have been used for treatment of experimental data.% It would seem logical to assume that with the rapid decrease over recent years in the cost of very small digital computers these machines will find their way into radiochemical laboratories for controlling larger chemical separation units and for carrying out some reduction in the handling of experimental data where this is feasible. Radiation measurement The primary purpose of radiation measurement is to isolate or determine the radiation emission derived from the sample upon which analytical measurement can be based.Consequently the counting system must be chosen not only to be sensitive to the type of radiation emitted from the sample (y-photons charged particles or neutrons) but also to provide data that can be identified with the component of the total radiation spectrum which is to be measured. A simple counting system is likely to be adequate for counting a radiochemically pure source produced by chemical separation but systems capable of discriminating between radiation from several sources present in the same sample are clearly liable to demand more complex instrumentation. Many radioactive isotopes and excited species decay by y-ray emission and the high potential information content of y-ray spectra has resulted in the widespread use of y-ray spectroscopy for the examination of multi-component mixtures while the convenience of the technique has ensured its use for more simple counting applications.Consequently y-ray detection is now the major measurement tool in the radiochemical laboratory and all other counting techniques must to some extent be regarded as special-purpose applications. Undoubtedly the most important single development that has occurred in y-ray spectroscopy over recent years has been the appearance of the lithium-drifted germanium diode as a viable y-ray detector to supplement the long established thallium-activated sodium iodide scintillator and Heath29 has reviewed the various aspects of germanium counter usage in detail.The high resolution of the germanium counter is the characteristic that is of major interest to the analyst, and full width at half maximum peak height of about a few kilovolts or better can now be achieved which is substantially better than that attainable with sodium iodide. This enables individual y-sources emitting lines of closely similar energies, to be distinguished in mixtures in a way that was not previously possible with non-dispersive y-ray detectors. The disadvantages that accrue from the use of germa-nium counters quite apart from the need to keep the detectors permanently cooled include low efficiency and high cost a 40-cm3 germanium counter together with cryostat and pre-amplifier costing about the same as a sophisticated atomic absorption spectrophotometer.The basic advantage of the germanium counter i 144 PIERCE derived from the low average energy of 2.98 eV required for the production of an electron-hole pair in germanium. However to exploit the inherent energy resolu-tion limited only by statistical considerations not only must the detector itself be as free from defects as possible but the other components of the counting system must have suitable properties. A major problem is the deterioration in performance of the over-all counting system at high count rates and new signal processing systems have been devisedSo that enable high counting rates to be accepted with relatively little deterioration in system resolution.A consequence of the improved resolution is the need for multi-channel pulse-height analysers with larger numbers of channels and a compromise must usually be made between the number of channels theoretically desirable and the financial investment called for. However, 4000 and 8000 channel systems are encountered relatively frequently in larger installations. Fixed-wire multi-channel pulse-height analysers are generally used for accumulation of data but small digital computers provide an alternative method of direct data acquisition and can offer very flexible systems as operation is under programme control and can often be varied relatively easily. The fall in price of small computers over recent years now makes them an altogether more attractive proposition for pulse-height analysis although the ‘word-length’ of the smallest machines may be inadequate to permit a single ‘word’ being allocated to each channel.81 Increase in computer size permits a more flexible system to be developed and the additional functions required for activation analysis such as sample transfer can be more easily put under computer control.32 The small computer has the additional attraction of being capable of being applied to other problems such as data processing and presentation when not in use as a multi-channel pulse-height analyser.= Very small computers have also been used to permit rapid transfer of information from a number of hard-wired analysers on to a magnetic tape store,w and direct connection between multi-channel analysers and a large computer capable of carrying out complex calculations has also been estab-lished.% To exploit the full potential of the high-resolution germanium counter for isotope identification precise y-line energies of nuclear transitions are required and recently amongst the tabulations that have been produced is one containing 2000 y-line energies from about 250 isotopes produced by neutron activatiomS6 The relative merits of using sodium iodide scintillators and germanium detec-tors for radiation counting in activation-analysis procedures will of course be entirely dependent on the composition of the radiation emitted but the two detectors have been compared for the analysis of a number of elements in lanthanum - tungsten bronzes and the examination of mixtures of rare earths.37 Comparison of the use of the two detectors with each other and with group separation procedures for the analysis of glass suggested that automated group separation probably provides the best cornpromi~e.~~ Multi-detector systems have been developed to suppress Compton event~393~0 and for y,y-coincidence spectro-metry,41 but although these techniques have proved invaluable for certain applica-tions the competitive position of the activation technique with respect to other analytical methods becomes weaker as greater capital investment is required fo RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 145 satisfactory instrumentation.Correct processing of accumulated data be it a simple y-ray spectrum or a more complex mixture of many sources is an essential part of any activation procedure and Yule42 has traced the development of methods used for the processing of y-ray spectra and given examples of the methods of mathematical manipulation of data in general use.Manipulation of data accumulated in y-ray spectrometers with the aid of large computers is now in widespread use and programmes are used for peak location, the measurement of peak energies the determination of y-line intensities and the calculation of elemental content of the sample. Many variations on the method of least squares fitting have been developed and are now extensively used where libraries of suitable standard spectra are available for mathematically constructing the necessary equations.43 These techniques undoubtedly increase the power of activation analysis and lighten the burden of onerous manual calculations but perhaps comments on y-ray spectroscopy should finish with a note of encourage-ment to those prospective activation analysts who do not have at their disposal the exceedingly expensive instrumentation required to pursue highly sophisticated techniques of y-ray spectroscopy.Not only can complex calculations be success-fully carried out on laboratory c a l c ~ l a t o r s ~ ~ but many valuable activation deter-minations can still be achieved with more mundane but in very many cases highly effective peak-area calculation techniques such as that of C~veIl*~ or the familiar base-line method. Although y-ray counting dominates the techniques of radiation measurement used in activation analysis certain other types of counter have found limited application for specialised purposes.Demands for charged-particle spectrometry have increased with the greater interest in the measurement of the prompt products resulting from nuclear inter-actions. Magnetic techniques** offer high resolution which permits detailed examination of complex spectra but semiconductor detectors are more convenient:' are non-dispersive and provide a resolution that is adequate for many applications. Neutron time-of-flight spectroscopy a technique widely applied in nuclear physics laboratories for the determination of neutron energies has been used to identify neutron groups emitted after sample irradiation with pulsed charged-particle beams.48 The measurement of tracks in polymeric materials such as cellulose acetate can provide a valuable positional indication of the presence of impurities when highly ionising radiations are and the work involved in track counting can sometimes be reduced by the use of instrumentation primarily intended for metallurigcal investigation,50 which enables tracks in a given field to be automatically totalled.Neutronlactivation Analysis Sample irradiation with neutrons has been extensively used in activation procedures and consequently is considered first in this review. Neutrons are most frequently produced in a nuclear reactor but accelerator and radioisotope neutron sources as shown on pp. 138 and 139 provide alternatives that although providing 146 PIERCE substantially lower neutron output can be used for certain applications and indeed may have some advantages where a high flux is not required.Neutrons themselves are a penetrating radiation and therefore interaction can be obtained with rela-tively large volumes of sample. However reactor irradiations are more frequently used to examine small samples perhaps of the order of a few hundred milligrams as the sensitivity of the technique is often adequate and the production of large amounts of unwanted activity is thereby avoided. Reactor activation analysis Delayed techniques. The use of a nuclear reactor as a neutron source is attractive to the analyst carrying out measurements at trace levels as not only are high sensitivities attainable as shown in Fig. 2 (taken from Ref. 51) but the high -8 z V Q) v) > > .-CI .- .- CI UJ .-% s $2 r r .-L.UJ +-’ 0) - G Q, y. 0 13 t z 5 I0 I 10-1 10-2 I O - ~ I O - ~ Sensitivity pg per 10Xn cmm2 s-’ Fig. 2. Sensitivities available for thermal neutron-activation procedures sensitivity of the method can often be exploited in realistic systems by using chemical separation as reagent blanks can be avoided and corrections can be made for chemical yield. It is advisable for the analyst to regard the reactor as something more than a ‘black box,’ particularly taking the trouble to find out the relevant characteristics of the irradiation station in so far as they influence his measure-ments. Guinn52 has summarised many of the characteristics of water reactors of the TRIGA type which are likely to be of interest to the activation analyst.Although most analytical applications are based on interaction of thermal neutrons high energy neutrons are available in a reactor and in some cases these may prove to be preferable for a specific determination. Assessment of the value of a particular reaction can often be made from a knowledge of cross sections if these are available (a) ( b ) Fig. 3. prepared for microcircuit production Photomicrograph (a) and the corresponding autoradiograph ( b ) of a silicon slice [To face Page 14 RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 147 both for the reaction to be used as the basis for measurement and also for those other reactions occurring with constituents in the sample that could produce activities likely to interfere with the determination.Tables of cross sections have been compiled which facilitate this type of calculation.63 Reactor fast neutrons, besides interfering with thermal neutron procedures offer the opportunity of determining a number of elements for which normal capture reactions provide inadequate sensitivity or which do not yield activation products suitable for measurement. A detailed study of the behaviour of thirty-five elements has shown that six, oxygen silicon phosphorus iron yttrium and lead have detection limits for reactor fast neutron-activation analysis that are lower than for thermal neutron irradiation while a further fifteen elements have similar sensitivities for fast and thermal activation.= However the limitation to the use of reactor fast neutrons is often not the inherent sensitivity of the method but the accompanying flux of low energy neutrons which may cause interfering nuclear reactions.Resonance neutrons also have uses for sample irradiation and their applica-tions have been reviewed.55 The possibility of nuclear interference must always be considered in activation procedures and a compilation has appeared of information on second-order reaction^.^^ Reactor pulsing was proposed some years agos7 so as to improve the sensitivity of methods employing short-lived nuclides and a number of recent applications have been described that emphasise the value of the technique.58 Considerable interest has also been devoted to devising methods for measuring short-lived nuclides to extend the scope of activation procedures and sensitivities and cadmium ratios for the production of a number of nuclides with half-lives varying from 2.5 s to 21 min have been determined practically and the results reported.59 Activation procedures based on the measurement of short-lived nuclides introduce special problems of counting timing and sample handling but equipment used for the determination of lead by the measurement of 800 ms lead-207 m has been described,BO and add - subtract sequence being used to distinguish between long-lived and short-lived components of the y-ray spectrum.The implication of counting short-lived nuclides has previously been considered in some detaiP and a more recent report has been published that is concerned with problems of decay curve analysis.62 Positional location of impurities by counting tracks formed in polymeric materials such as cellulose acetate,s3 or the long established technique of auto-radiography has found application when the radiation from the component to be investigated can be distinguished from other sources of radiation induced in the sample.When the over-all composition of the sample is suitable useful informa-tion can be obtained from such position sensitive techniques as shown in Fig. 3, (Pierce T. B. and Peck P. F. unpublished work) in which the photomicrograph of a silicon slice prepared for microcircuit production [Fig. 3(a)] is compared with an autoradiograph of the induced activity [Fig. 3(b)] taken after neutron irradiation, when the silicon activity has decayed away.The major component of the induce 148 PIERCE activity was caused by gold and the distribution of the element can be seen to be correlated with structural features of the slice. The distribution of copper and other elements in silicon has been examined by autoradiography to assist in the production of pure materials and the half-lives of components found by photo-metric examination of autoradiographs taken over an extended period of time.s4 Autoradiographic techniques have also been used to examine the inhomogeneous distribution of tantalum and tungsten in zone-refined niobiumJ65 while auto-radiography of oil paintings following neutron irradiation has revealed structural detail enabling pigments to be identified.66 In the latter case an absorbed dose of 50 rads was needed to provide adequate activity for autoradiography but even after a dose two orders of magnitude greater there was no sign of any changes in colour hardness flexibility or solubility in the paintings which were examined up to 3 years after irradiation.Secondary particles produced as a result of the interaction of neutrons with some component in the sample can provide a worthwhile alternative to neutron activation in certain cases. The best known application that of the determination of oxygen by tritons produceds7 as a result of the reaction 6Li (n,cc)T has found further application for the determination of oxygen in microcrystalline carbon68 and organic compo~nds.~~ Following the earlier work of Born a secondary reaction technique has also been used for the measurement of lithium,'O and deuterium has been determined by counting fluorine-17 produced by the reaction 160(d,n)17F which was induced by recoiling deuterium nuclei from the (n - d) collision.In this case the reaction 160(n,p)16N was used for internal standard-isation. Application-s. Reactor neutron-activation analysis has found very wide application which makes a critical survey impossible in the space available in a review such as this. However as the general theory and experimental techniques involved in many programmes of work utilising activation analysis are well understood interest now often lies in the results and their interpretation rather than in the method whereby the results are obtained always assuming that this has been reliably carried out.Applications must therefore figure in any discussion of activation analysis and the purpose of this section is to outline some of the sectors of application that have been found for neutron-activation analysis so that the reader can subsequently pursue the subject in more detail if he so desires. The high sensitivity attainable by neutron-activation analysis together with the lack of any reagent blanks and the ability to calculate chemical yields has made neutron-activation analysis a favoured method for the determination of many elements at low levels in a variety of complex matrices. One of the major fields of application of reactor neutron-activation analysis has therefore been the examination of a very wide range of materials of geological cosmological and bio-logical interest and examples of these applications are given below.In addition, the convenience of reactor activation analysis particularly where examination of the intact sample is possible or when only limited chemical separation is necessary, has established its use as the preferred technique for other purposes RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 149 Rocks meteorites minerals etc. Neutron-activation analysis has always been attractive for the analysis of geological and allied materials and there has been no dearth of examples of this particular type of usage in recent years. Undoubtedly one of the most interesting applications of neutron-activation analysis that has been reported recently has been the analysis of the lunar samples brought back by the Apollo 11 mission and reported in a copy of Scieme devoted to the proceedings of the Apollo 11 Lunar Science Conference held at Houston from the 5th to 8th January 1970.A number of analytical papers on the composition found for lunar material are included and neutron-activation analysis figures largely in the analytical methods chosen. Some laboratories have applied several analytical methods of which neutron activation is only ~ n e ~ ~ ~ ~ ~ but many data have been accumulated on the composition of rocks soil and core ~amples.~~-~7 Spallogenic manganese-53 was determined after reactor irradiation by counting 300-day manganese-54. In this case chemical yield was measured with carrier-tree manganese-52 as tracer.7s The extensive information obtained by activation analysis for a very large number of meteorites has been summarised demonstrating the scope of the technique for the analysis of both stony and iron meteorite^.^^ Neutron-activation analysis has been used to obtain information about the elemental abundances of various elements in the Allenda meteorite that fell on February 8th 1969 and was of the relatively rare carbonaceous chondrite type.791so An interesting develop-ment following an application of neutron-activation analysis established by Smales in 1958 was the determination of nickel cobalt manganese and iridium in cosmic spherules between 100 and 350 pm in diameter which were separated from 1.5 tons of Pacific Ocean sediment.For the measurement of short-lived nuclides the spherules were mounted on mica sheets for irradiation and y-ray spectrometry was used to isolate contributions from the individual elements.sl Among the wealth of papers that have been devoted to the analysis of rocks and ~ltramafic:~ metamorphics4 and igneous rockss5 have been amongst those examined.The value of germanium counters in the examination of geological materials has been clearly demonstrated by Gordon,86 who has deter-mined twenty-three elements in igneous rocks by high resolution y-ray spectro-metry following neutron irradiation and has considered the possibility of measur-ing several others. Identification of the source of y-lines in the complex spectra may prove difficult and principal y-rays are listed together with likely interferences to aid identification.Chemical separations have been used after instrumental examination of the sample to extend the range of elements determined,s7 while y - y coincidence spectrometry has also been found to provide extra selectivity in certain cases.s* The measurement of uranium in geological material has again been made by measurement of a fission product of uranium-235,89 in this instance tellurium-132, to complement the more usual technique of delayed neutron measurement and the ratio of protactinium-231 to uranium-235 in rocks has been measured by neutron activation and a-particle spectr~metry.~~ Activation analysis has been used t 150 PIERCE demonstrate that gold is not homogeneously distributed through standard rocks G-1 and VV-1 by division of a large sample and subsequent measurement of the smaller fractions .91 Soils have also been examined in some detail by neutron-activation techniques, and an attempt has been made to determine whether the y-spectra of soils can be used to trace their geographical Those nuclides that could be detected after sodium-24 activity had decayed were examined by y-ray spectrometry but soils of dissimilar origins sometimes gave rise to similar y-spectra thus throwing doubts upon the possibility of identifying the source of soil specimens from the induced y-ray spectra.Neutron-activation analysis and atomic-absorption analysis have been recommended as a particularly suitable combination of analytical techniques for soil analysis because of the range of elements that can be deter-mined with good precision and accuracy.93 Epithermal neutrons can sometimes prove more satisfactory than thermal neutrons for sample irradiation and rubi-dium caesium strontium antimony tantalum and uranium are elements that are favourable for this type of a c t i ~ a t i o n .~ ~ Where they can be applied the convenience of rapid methods of intact analysis permits the high throughput of samples and consequently enables the areal distribution of elements to be examined.g5 Similar techniques have been used to examine certain elements present in volcanic ash from the Taal volcano.98 Analysis of certain components in sediments from the South Pacific have enabled the behaviour of trace elements during sedimentation processes to be under~tood.~~ Application to life sciences. Life sciences provide a field where activation techniques have found profitable application and over the last few years a large number of reviews have a~peared.~8-lO~ These cover the use of reactors both for conventional activation procedures and for more general applications and include the use of secondary reactions for the determination of some elements that are not conveniently activated by reactor neutrons directly.Fields of application include metabolic and kinetic studies the determination of stable added tracers elemental determination for research and diagnostic purposes and even activation analysis in vivo. The possibility of obtaining an early diagnosis of cystic fibrosis by the analysis of clippings from toe and fingernails for sodium potassium and chlorine has been inve~tigated,l03~~04 and the ability to determine the sodium-to-potassium ratio by neutron-activation analysis improves the prospect of performing a success-ful sweat test.Of the many measurements carried out on the elemental content of various types of tissue a study of differences in trace element contents between normal and atherosclerotic aorta~,~05 investigations into the relationship between the occurrence of amyotrophic lateral sclerosis and manganese metabolism,lo6 and analysis of lung tissue from workers from certain mineral mining and processing industries107 have been reported. Neutron-activation analysis has also been used to monitor the levels of sodium magnesium manganese and sulphur in commercially available DNAse I (bovine pancreatic)lo8 and for dose control of suppositories of atropine methobromide contained in sealed polythene tubes ; measurement was based on the determination of the induced bromine-82 activity.log The value o RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 151 reactor neutrons for in vivo activation analysis is limited by the relatively low energy of the neutrons available but iodine has been locally activated in the thyroid by using an internal standard to allow for neutron absorption in tissue,l1° while the mass ratio of calcium to chlorine in human tibia has been measured by a neutron capture y-ray method.lll Substantial use has also been made of activation analysis in dental investiga-tions,l12 and the antimony arsenic copper gold iron mercury silver sodium and tin contents of dental pulp from diseased teeth have been examined so as to assess the biological importance of these elements.l13 Marine life and waters.Neutron activation has been used to support pro-grammes of investigation into fish and marine organisms. Elemental concentra-tions in tissue of species of Pacific salmon have been investigated114 and variations found between species and between sexes of the same species. All salmon were found to concentrate caesium relative to potassium. Levels of various elements in midline muscle dorsal muscle brain spinal cord liver heart spleen kidney and blood of rainbow trout raised under controlled conditions have been found and compared with that of the food fed to the fish.ll6 Variations in the trace element concentrations of corals have shown that it may be possible to distinguish reef from near-reef specimens by the higher uranium content of the latter.l16 The calcium content of two groups of zooplankton from four lakes Lago Maggiore Lago di Varese Lago di Comabbio and Lago di Monate have been compared,l17 and the effect of salinity on the deposition of calcium magnesium manganese strontium and sodium has been investigated.lls In spite of the difficulties associated with the relatively high induced radio-activity from sodium and chlorine the sensitivity and convenience of activation analysis has proved attractive for the examination of trace elements in a variety of natural waters.For example vanadium has been determined at concentrations ranging from 0.2 to 49.2 pg 1-1,119 and several elements have been measured directly in sea water after a 4 to 5-week cooling period has allowed the sodium-24 activity to decay.120 Activation analyses have supported studies with sea water,121 spring water,122 river rain water124 and mine water.125 Lanthanide distribu-tions in water from the Gulf of Mexico have also been determined by using neutron activation and a seaward decrease in elemental concentration was observed which has been correlated with river run-off.(A maximum in the depth profile was observed similar to that obtained with other elements in the vicinity of 1000 m.12s) High purity materials. The high sensitivity of neutron-activation analysis has been well exploited for the examination of the trace element contents of a variety of high purity materials including semi-conductors,l27 to aid impurity studies and to assist in improving manufacturing processes.The influence of surface impurities on the reproducibility and accuracy of the determination of the copper gold and silver contents of high purity germanium and silicon have been studied and correct surface treatment before activation has been found to be essential so as to obtain reliable results.128 Frequently concentration profiles or elemental contents of thin layers are required and an anodisation-peeling technique has been used in conjunc 162 PIERCE tion with neutron-activation analysis to obtain the concentration profile of energetic ions implanted in silicon crystals.12@ Neutron-activation techniques have also been used to follow the presence of impurities during the growing of doped laser crystals and to assess the extent of crystal contamination.130 Among the many other pure materials analysed are s e l e n i ~ m ~ ~ ~ ? ~ ~ ~ aluminiumls~l~ metallic uranium1% and plates of crystalline quartz.l= High induced activities of gallium-72 and arsenic-76 formed from the matrix when gallium or gallium arsenide is irradiated present special problems of handling as the sample activity may be 1 to 10 Ci and the separation of induced activity particularly of short-lived radio-nuclides that must be measured before the matrix activity has had time to die away can present special problems.These have been considered in some detail, and sensitivity under optimum irradiation conditions has been investigated for various impurities present at very low 1e~els.l~' Separation and determination of contaminants in the processing agents used in the telecommunication industry, such as silicochloroform distilled water nitric acid and carbon tetrachloride have been made by using a paper-chromatographic separation followed by y-spectro-scopy after neutron irradiation.l58 Oil pesticides and tobacco.The use of activation analysis in the mineral oil industry can be used to provide information concerning the composition of crude oil finished products wear elements engine deposits and residues. A knowledge of the source of an oil causing pollution of waterways is clearly important in sub-sequently deciding legal responsibility and in an attempt to provide a means of characterising oil sources sixteen different marine fuel oils have been analysed by neutron-activation analysis for vanadium manganese sodium cobalt antimony, arsenic copper and zinc.As part of this project prolonged exposure of the oils to sea water has been examined to assess the effect it may have on their elemental composition.139 The use of neutron-activation analysis to assess the content of pesticides present in feeding stuff has continued and the determination of bromine in wheat and bread after thermal neutron-activation analysis has been done as a means of determining traces of bromine-containing germicides.140 An extensive programme of work into the arsenic content of fifty-two samples of thirty-three brands of American cigarettes has been facilitated by the fact that no chemical separation has been needed in the neutron-activation determinati~n,~~~ and the content of a number of elements in a standard cigarette have been deter-mined by neutron activation to support work on tobacco and health research.142 Freedom from complication in sample preparation has proved attractive for the examination of the distribution of arsenic residues in tomato plants.Discrimina-tion within the plant indicated that the soil concentration of arsenic would have to be at a level that caused reduction in the size or yield of the fruit before the con-centration in the fruit exceeded tolerance limits.143 Further examples of the application of neutron-activation analysis to pesticide residue measurementla demonstrate the relatively rapid measurement of bromine and arsenic in a wide variety of plant material such as fruit tobacco or wheat flour.An extensive com-parison between neutron-activation analysis and a variety of other analytica RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 163 techniques for the determination of a large number of elements in a standard biological material has been made by a number of collaborating laboratories.14s Trace elements in the atmosphere. An increasing amount of work has been concerned with the determination of trace elements present in the atmosphere so as to obtain information about sources of pollution. The seasonal variations of manganese and aluminium have been followed from deposits collected in the Chiba area of Japan.146 Instrumental neutron-activation analysis involving germanium counters has permitted measurement of the concentration of more than twenty elements in atmospheric aerosols of urban areas and the technique was shown to be more sensitive and accurate than the other analytical techniques in current use.147 Sodium iodide scintillators have proved satisfactory for the determination of a number of elements present in airborne particulate matter collected on cellulose-base filter-papers.l@ The distributions of a number of elements at altitudes varying from lo00 to 50,000 f t have been examined by instrumental activation analysis in an attempt to establish the origin of aerosols in the atmosphere and the concentra-tion of iron was found to decrease by a factor of eight between the altitudes 10,000 and 50,000 ft.149 Anomalously high concentrations of both bromine and chlorine have been detected above the surface of Lake Michigan150 near urban localities and the distributions of trace elements in marine aerosols have also been examined.151 Forensic ap$&cations.The application of neutron-activation techniques to the analysis of samples of forensic interest has long been discussed in neutron-activa-tion circles and has been used to provide legally acceptable evidence in a number of cases. As more detailed information has been accumulated there has been an opportunity to assess the value of the technique in greater detail. Trace element content of hair for long suggested as a possibility of identifying an individual has been considered in depth with samples from a large number of volunteers and the analytical results have been treated statistically to assess their ~alue.l5~ Incorpora-tion of thallium into hair and nails has been correlated with the time a patient was known to have taken the element,l= and the results from neutron activation have been compared with those obtained by analysis using spark source mass spectro-scopy for the determination of a number of elements in human hair.154 The elemental content of samples of glass from a suspect's shoe has been compared with that from the broken window in a house that was burgled and results showed that glass from the shoe and from the window probably came from the same batch of manufactured glass155 ; the origin of glass fragments found duringan investigation of an accident between two cars was also used to ascribe responsibility for the acci-dent.lW Other investigations have included the detection of gun-shot residues by the presence of barium and antimony the determination of antimony arsenic, aluminium and silver in bullet lead for bullet characterisation discrimination between paint pigments and the determination of mercury in oats alleged to have poisoned animals.15' Art and archaeology.Use has been made of neutron-activation analysis to examine the elemental contents of small samples of materials removed from paintings to ascertain if trace element contents of pigments could be sufficientl 164 PIERCE characterised for dating to be possible and thence to reveal forgeries. The trace element content of white lead has been examined in some detail showing that impurity concentrations can give some indication of age.15*v15Q Contents of certain elements in pottery sherds from Hajar bin Humeid and related areas have been examined to study correlations in composition,160 and those from Knossos and Mycenae have also been investigated; in the latter case some samples from each group examined appeared not to belong to the supposed compositional category.lsl The geologic origins of prehistoric obsidian artifacts have been found by matching the sodium manganese lanthanum samarium scandium iron and rubidium contents of the artifacts with those of obsidian from geologic sources.162 An attempt has also been made to establish the origin of amber samples found during archaeological investigation with the assistance of neutron-activation analysis,16s and the silver content of coins dating from the Roman period has been traced by the analysis of 700 ~amp1es.l~~ Non-active tracers.The use of inactive materials as tracers that can sub-sequently be determined by activation analysis is attractive for many applications as the hazards associated with the handling of radioactive tracers is avoided and although the technique is more liable to contamination from the same element present in the medium to be investigated careful choice of a tracer can limit this occurrence. Selection of the element is also governed by the desired nuclear properties as detection at levels of adequate sensitivity ought to be relatively easy, and chemical behaviour may dictate the response of the tracer to some of the processes examined.Stable tracers in conjunction with neutron-activation analy-sis have been used for large-scale environmental studies and for the examination of the behaviour of very much smaller systems The movement of sand on the Mondello Beach Palermo Sicily has been studied with silver and cobalt stable tracers mixed with sand and the distribution of tracer was subsequently found by sampling and subsequent measuring by neutron-activation analysis.16s Indium, injected at a cloud base as finely divided particles from pyrotechnic flares was subsequently determined in rain collected in an array of fourteen samplers spaced over 11 km. The sensitivity of the method was limited by the natural background of indium which was found to be 6 & 3 ng of indium per litre.166 Stable tracers have been applied to the tracing of estuary waters,l67 and the technique has also been considered for examining the surface distribution of agricultural insecticides sprayed by helicopter.16* The addition of 0.1 per cent.w/w of a rare earth oxide to gunpowder permits detection of the firing of a weapon with a high degree of probability and a cost analysis suggests that the labelling of gunpowder with europium will increase the cost of ammunition by less than 2 per cent.169 Refrac-tory materials containing tracer elements have been used to identify non-metallic inclusions in steels which originate from such refractory components as ladle-linings or the hearths of electric furnaces.170 Isotopic analysis. As a radionuclide is produced by the interaction of a neutron with a specific isotope neutron-activation analysis is essentially a method of isotopic analysis and offers the possibility of determining isotope ratios in specifi RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 156 samples.The ratio of uranium-235 to uranium-238 has been measured from the change in intensity of y-peaks of neptunium-239 and fission produ~ts,l~l-l~~ while manganese43 has been determined in meteoritic iron by a neutron-activation t e ~ h n i q u e . ~ ~ ~ Miscellarceoiws applications. Neutron-activation analysis has been used to support other programmes of work that are not conveniently covered by the head-ings given previously and a number of these studies have been chosen to demon-strate the wide field of application of activation analysis.The rate of evaporation of some solids has been measured by condensing the evaporating substance on a water-cooled lead plate and then irradiating the plate and condensate with neutrons. Subsequent measurement of the induced radioactivities permitted the measurement of evaporation rates in the range 10-7 to 10-9gcm-2 ~ e c - l . l ~ ~ Neutron-activation analysis has also provided a method for re-determining the vapour pressure of liquid cerium176 and has been used to measure individual corrosion rates of iron and chromium in sodium systems.177 The behaviour of selenium mixed with lithium fluoride and added to highly purified graphite powder as a spectrographic buffer has been followed under arc discharge by neutron techniques and rates of vaporisation of the materials have been determined under a number of different conditions.l78 Prompt techniques.An alternative to the measurement of induced radioactivity produced as a result of sample irradiation with reactor neutrons is the measurement of prompt events from nuclear interaction which occur very rapidly after the initial interaction of the reacting nucleus with the incident neutron. The capture of thermal neutrons results in the emission of y-radiation that is characteristic of the nuclear transitions occurring and can consequently provide the basis of analytical determination. The initial excited states resulting from neutron capture may have energies of 7 MeV or more so that spectra are frequently complex and the y-ray emission from multi-component samples is often difficult to analyse.However ‘capture spectra’ for a very large number of elements have been compiled for sodium iodide scintillator~l7~ and germanium countersls0 so that suitable reference data are available if use of the technique is being considered. Although the inherent sensitivity of prompt techniques is high,lsl and for many elements may exceed that available for conventional activation analysis only relatively low neutron fluxes are normally available for capture work as neutron beams are usually extracted from the side of a reactor through the biological shield, and the decrease in neutron flux over that available in the reactor may be 6 or 7 orders of magnitude hence good sensitivity can only be obtained for those elements that possess high neutron capture cross-sections.Boron which cannot be con-veniently determined by more usual neutron-activation techniques undergoes the reaction l0B (n,a)7Li with thermal neutrons producing lithium-7 in its first excited state; this isotope then decays to ground-state by the emission of the 478 keV y-line. The y-ray spectrum resulting from the interaction of boron is therefore un-complicated and the high cross-section for the production of lithium-7 provides 156 FIERCE the basis for a sensitive determination. Moreover the importance of the presence of boron in certain materials makes simple determinations of the element a matter of real interest. This method has been successfully applied to the determination of boron in steels and sensitivities of the order of a few tens of parts per million have been a c h i e ~ e d .~ ~ ~ J ~ ~ The technique has also proved useful for the determination of a number of other elements including gold,lS4 samarium and gadolinium,lS5 and a number of elements in biological materials including hydrogen calcium phos-phorus and chlorine.lS6 Several methods for the analysis of fissile materials based upon the measurement of prompt radiation measuring neutrons or y-rays have been developed.187 Accelerator neutron sources While nuclear reactors have received the major attention as neutron sources for activation analysis neutrons may also be produced by the interaction of accelerated charged particles with target nuclei of suitable materials. Much information is available in nuclear physics literature on the nuclear parameters of neutron-producing reactions and monographs have been producedlS8 giving neutron energy as a function of reaction particle energy and angle for a number of the more widely-used nuclear reactions.While a careful choice of neutron energy can provide selectivity of value in reducing the effect of nuclear interferences in practice large accelerators are generally needed to achieve this variation in neutron energy and it is therefore not surprising that much analytical work carried out with accelerator neutron sources has been based on the use of so-called neutron generators which are usually low-voltage accelerators producing neutrons by means of the reaction A major attraction of this particular reaction is that substantial outputs of 14-MeV neutrons can be produced with accelerating voltages of the order of 120 KV.Thus the use of accelerator neutron sources can fulfil two functions firstly to provide irradiation facilities that are smaller and cheaper than nuclear reactors but which are still capable of producing a useful neutron flux and secondly to provide a neutron output with an energy distribution that is substantially different from that obtained in a nuclear reactor but which can be valuable for the determination of certain elements. Early neutron generators were almost all large drift-tube machines requiring substantial shielded space for safe operation. This sometimes made them uncompetitive when their requirements and performance were com-pared with those of other analytical techniques.However from the point of view of the analyst requiring an instrument that is relatively easy to install the newer, sealed-tube neutron generators appear to have a number of advantages.ls9 One is that they can be installed in the now well known hole-in-the-floor shielding. Although such shielding presents difficulties if frequent access to the generator target is required it makes relatively little demand on laboratory space. Delayed techniques. The primary use of neutron generators is for analytical methods based on the measurement of radioisotopes formed as a result of the *H + SH -+*He + In + 17.6 Me RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 157 mteraction of the high-energy or ‘fast’ neutrons with elemental constituents of the sample and the reactions most likely to occur are (n,a) (n,p) (n,2n) (n,n’).In addition there may also be a component in the induced activity produced by capture of low-energy neutrons particularly if substantial amounts of moderator are present near the neutron emitting target. Fast neutron-activation analysis is therefore capable of providing radioisotopes from irradiated elements which differ from those produced in largely thermal irradiation facilities and in certain cases permit the determination of those elements that cannot be conveniently determined as a result of thermal neutron irradiation. The variety of nuclear reactions that may occur under fast neutron irradiation, while giving some scope for the choice of radionuclide upon which quantitative measurement may be based also leads to the most serious limitation of the applica-tion of fast neutron-activation analysis to practical problems.The limitation is created by the production of the same radionuclide from more than one element (nuclear interference). For example chromium can be determined after irradiation with fast neutrons by measurement of the vanadium-52 produced by the reaction 52Cr(n,p)52V. Vanadium-52 is a 7-emitter decaying with a half-life of 3.76 min and with E = 1-43 MeV. However should any manganese be present in the sampIe the vanadium-52 will also be produced from the manganese by the reaction 55Mn(n,a)52V and if an appreciable flux of thermal neutrons is also present at the irradiation position vanadium-52 could be produced from vanadium-51 by the capture reaction 51V(n,y)52V.Thus the total yield of vanadium-52 will be a function of the amounts of all three elements present in the sample. If an accelera-tor is available which can substantially vary the energy of the neutrons produced, then a judicious choice of neutron energy can sometimes reduce the contribution made by an interfering reaction or decrease the yield of a major component to the y-ray spectrum.1g0 However neutron generators do not normally have such flexibility and therefore the scope for such choice of neutron energy is very limited. The ratio of thermal to fast neutrons can be varied by the use of a moderator placed near to the neutron-emitting target sometimes allowing the contribution from capture reactions to be but the penalty of using a moderator may well be one of lowering the induced radioactivity in the sample and thus limiting the scope of the analytical determination.The use of moderators alters the energy spectrum of neutrons irradiating a sample and can in some cases lead to changes in the relative induced activities formed by fast reactions,192 but in general these effects are relatively small and are not sufficiently large to improve substantially the scope for many analytical determinations where interference is serious. Another problem that critically affects the design of neutron generator systems for activation analysis stems from the rapid variation in flux with distance from the neutron-emitting target varying along the central axis as l / d near the target and l / d 2 at a greater distance where d is the distance between the target and the point at which the flux is being measured.lg3 Flux distributions for typical generator systems have been calculated194 and the results emphasise that not only will the gradient through a thick sample vary substantially even if there are no self 168 PIERCE shielding corrections to be made but also that if a sample is irradiated very near to a neutron-emitting target careful positioning is necessary to ensure that successive samples experience the same neutron flux for a given generator output.However, by careful design high precisions can be obtained by using a single-transfer system,ls5 but alternatives are to rotate the sample,lgg to use a dual rotatorl97 in which the sample and some form of standard are irradiated simultaneously or to irradiate the sample and standard together without rotation.lS8 Because of the limited target life of many high-output neutron generators and the decrease in a flux with distance from the neutron-emitting target that limits the number of samples that can be realistically irradiated at any one time neutron generator applications are normally limited to relatively short irradiations and measurement of the induced activity of short-lived radionuclides.The major interest in fast neutron-activation analysis initially stemmed from the possibility of determining oxygen by the reaction 160(n,p)16N,19s as 7.3 s nitrogen-16 is a convenient radio-nuclide for production by generator irradiations and the unusually high y-line energy of 6-1 MeV emitted as a result of the decay of the nitrogen-16 can be measured relatively easily.As a result fast neutron-activation analysis has been used for the determination of oxygen in a very wide variety of matrices. Some of the corrections involved in the application of fast neutron-activation analysis have been considered200 and must be applied when the highest accuracy is required. A significant development in the use of neutron generators has been a careful evaluation of the technique for the determination of oxygen in steel,201 and sub-sequent use of the technique in industrial operation.202 Special standards have been produced for the determination of oxygen in steels under industrial conditions.20s Another light element which because of its importance has received con-siderable attention is nitrogen particularly when present in foods and feeding stuffs as very large numbers of nitrogen determinations are at present carried out by the feeding stuffs industry to assess the protein content of the material.A detailed investigation has been made involving several hundred determinations on sixteen different types of material and good linear correlation was obtained between fast neutron-activation results and those obtained by the more traditional Kjeldahl techniques204 A similar method has been applied to the determination of nitrogen in maize seeds205 and in flour,206 and the same nuclear reaction [14N(n,2n)13N] has been used to measure the amount of nitrogen in petroleum products. In this case the effect of recoil protons was considered in some detail as their interaction with carbon present in the sample can produce nitrogen-13 by secondary reactions, including 12C(p,y)13N and 13C(p,n)13N but the yield of nitrogen-13 was found to be a function only of the weight fraction of carbon or hydrocarbons e~amined.~O' Neutrons (14MeV) have also been used for the determination of oxygen, silicon and aluminium in lunar rocks and soils and oxygen abundances were found to be lower than those in most common terestial rocks comparable to the levels present in certain types of stony meteorites.208 Among the many other applications of fast neutron activation reported is the analysis of biological materials for silicon, chlorine potassium phosphorus calcium and aluminium209 ; halogens have bee RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 159 determined in photographic emulsions2lO and copper silver and gold in ancient coins.211 Fissile material remaining after the dissolution of irradiated fuels has been determined by neutron-generator irradiation and delayed-neutron counting techniques and the sensitivity of the methodhas been assessed at 20 mg of uranium-23tL212 A mixture of thermal and fast neutrons has proved to be satisfactory for the determination of aluminium and chlorine in intact solid composite propellents.With a precision of better than &l per cent. the technique appears to be adequate for quality control purposes.213 To extend the scope of fast neutron-activation analysis to the measurement of radionuclides having a half-life of less than 1 s, fifteen nuclear reactions have been studied to obtain information about the forma-tion of a number of radioisotopes with half-lives varying from 0.5 to 870 ms.214 Unfortunately only a very limited number of radioisotopes can be produced with half-lives of this order.This somewhat restricts the application of the technique, but the measurement of short-lived nuclides may provide a useful alternative in limited cases when the counting of other radioisotopes presents some difficulty, for example the measurement of the sodium-24 nuclear isomer with a half-life of 20 ms has been proposed as an alternative to the use of 15-h sodium-24 as a basis for magnesium analysis.216 The penetration of fast neutrons has stimulated an interest in the possibility of in vivo measurements by activation analysis; an interesting example is the determination of total nitrogen and protein in mice.21e While most generators operate on the basis of the 3H(d,n)4He reaction an alternative is the reaction 2H + 2H+3He + ln + 3-28 MeV.This can be produced in small positive ion accelerators of the neutron generator type although with rather lower cross-section. A survey of sensitivities for 3-MeV neutron irradiations has been made for a large number of elements217 and an assessment of the possi-bility of using both 3-MeV and 14-MeV neutrons for precious metal exploration218 has been made. In this case a mobile 150-kV positive ion accelerator was used as the neutron source and the measurements were based upon the determination of the induced radioactivity.14MeV neutron-activation analysis has been used for compositional mapping of sea floor sediments by analysing collected samples but an in situ instrument could be designed to operate on the same principles.2fQ The relatively small size of generator neutron sources has prompted some workers to consider the possibility of using fast neutron systems in conjunction with 'y-ray spectroscopy for on-stream analysis. The high degree of penetration of both neutrons and y-rays enables relatively large volumes of material to be sampled. This is clearly desirable when the sample stream is inhomogeneous and rugged containment of both generator and detector without the need for thin windows is possible thus permitting the equipment to withstand an industrial environment.Extensive assessments of the technique have been made and some experience has been obtained with an experimental conveyor A number of cells have been designed to permit both liquids and solids to be examined including one with provision for irradiation of both sample and standard streams.=l Application of the method to the determination of a number of materials of industrial interest ha l60 PIERCE been assessed; the materials included water salt and sulphur in crude oil= and copper in copper ore .223 Although from considerations of cost neutron generators are an attractive form of accelerator neutron source for analytical purposes when they are available and the experimentation can stand the higher operating costs such large accelera-tors as electrostatic generators or even cyclotrons provide a higher neutron output and greater flexibility.Interest in small cyclotrons for charged-particle activation analysis has also been extended to the use of these machines for the production of neutrons and assessments have been made of their characteristics for neutron-activation analysis.224s226 Photoneutrons produced by high energy Bremsstrahlung, generated in an electron linear accelerator by interaction of the electron beam with a suitable converter can also be used for neutron-activation analysis.226 The ability to vary neutron energy which is usually feasible with these larger machines by choosing suitable target materials and accelerating energy can be used to improve conditions for analysis. Neutrons generated by cyclotrons have been used for whole-body assays of sodium and calcium by irradiation of cadavers and of human volunteers analytical measurement being based on the activity of calcium-49 and sodium-24.227 Irradiation doses received by three volunteers were approximately 1300 mRem.Prompt radiation. Prompt events induced by accelerator neutrons are clearly dependent upon the energy of the neutrons being used to irradiate the sample. Although reactions such as (n,p) and (n,a) produce particles that can be measured directly and are likely to yield residual nuclei in excited states that may subsequently de-excite by the emission of characteristic y-rays the primary reaction that has been of interest to analysts is inelastic scattering as y-rays emitted during de-excitation can provide the basis of analytical determination.Interest in inelastic scattering has been centred around the measurement of light elements such as carbon which are difficult to determine by more conventional activation techniques or in methods operating under specialised analytical con-ditions such as those required for on-line sensors where the immediacy of the y-line yield may prove an advantage. Substantial information about inelastic scattering is available in nuclear physics literature and y-rays have been produced from a variety of materials.228 A major limitation to the analytical use of inelastic scattering methods is imposed by the high radiation background in the presence of which the necessary measurement must be made ; anticoincidence systems have been found to improve the accumulated y-spectra but these add substantially to the cost of the detector systems.Because of the relatively high level of background events detected by the counting system up to the present time inelastic scattering has been primarily concerned with the determination of major element levels and the possibility of determining carbon in and aluminium iron and magnesium in silica based materialsm0 has been examined. Moderation of fast neutrons will often occur in large samples so that capture spectra will be observed in addition to the contribution from inelasticneutro RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 161 scattering. The two types of radiation can be distinguished by the time of radiation emission following a neutron pulse and further analytical information about the analytical composition of the sample can be derived from a measurement of thermal and epithermal neutron die-away.Neutron die-away inelastic neutron scattering, neutron capture and activation have been combined in a single experiment that has been devised for the analysis of rocks and bore holes.=l Radioisotope neutron sources Conventional radioisotope neutron sources which are manufactured by mixing two components a radioactive isotope and a material that emits neutrons upon interaction with the radiation emitted by the radioisotope have the advantages of stability of output small size reliability and zero power requirements. Unfortu-nately the neutron outputs of these sources are relatively low typically varying from 106 to 10s neutrons s-1 total for those in general use although larger installa-tions have been devised.Isotopic neutron sources can be produced with a number of neutron energy di~tributions,2~~ although in some cases the neutron outputs available are very low. Analytical applications have been devised that are based on the measurement of induced radioactivity or of prompt events but the former have been more numerous. The use of fast and thermalised neutrons from a polonium -beryllium neutron source has been exploited to provide a determination of alumina and silica in bauxite samples and the method has been found to be less time consuming and more accurate than the wet chemical methods previously In some cases the lower neutron energy of radioisotope neutron sources can prove an advantage over the 14-MeV neutrons obtained from neutron generators as inter-fering reactions may be avoided.Thus fluorine can be determined by the measure-ment of nitrogen-16 produced by the reaction l9F(n,a)l6N but in this case there is no interference from nitrogen-16 produced from oxygen-16 by the (n,p) reaction as would be experienced with generator activation as the neutron energy is below the threshold value.= High output neutron sources can be obtained by irradiating americium - beryllium mixtures to produce (241Am-2Q2Cm-Be) sources and the y-spectra and sensitivities produced from elements irradiated with this type of source have been Fluorine has been determined with such a source by counting nitrogen-16 activity and a sensitivity of 0.4 mg of fluorine in a 10-g sample was found for a source emitting 4.8 x lo9 neutrons s-l total; the available flux was 1.4 x lo8 fast neutrons cm-2s-1.B6 The reliability of the radioisotope neutron sources and their minimum maintenance requirements make them attractive for on-stream applications for industrial purposes.The low sensitivity achieved with small sources that are relatively easy to incorporate into industrial systems remains a limitation and in an effort to improve the over-all sensitivity sample re-circulation has been proposed237 and the benefits of the technique demonstratedJB8 although the response of such a system must inevitably be relatively slow. Very much larger systems have been considered including one based on an antimony-124 - beryllium source capable of handling up to 7,200 Ci of antimony-124.This system can pro 162 PIERCE vide a thermal neutron flux of approximately 2 x lo8 neutrons cm-2s-1 on the inside surface of a cylindrical irradiation but clearly very special facilities are required for handling such large amounts of radioactivity. A radio-isotope system has also been devised which is based upon the use of a much smaller neutron source (5 Ci of americium - beryllium) for the off-line analysis of discrete samples for process control; silicon aluminium and chromium have been deter-mined simultaneously and the advantages of the technique were found to be the simplicity of sample preparation and the insensitivity to particle size and matrix effects.240 The fluorine reaction mentioned above has provided a means of determin-ing fluorite in ore by a probe-type technique.241 Prompt radiation measurements based on the use of radioisotope neutron sources have been carried out by the measurement of both inelastic scattering and capture y-rays.The determination of carbon in iron ore in sinter mix and in fly-ash containing 2 to 16 per cent. of carbon has been based upon the 4-43-MeV y-ray produced by the (1) -+ (0) transition in carbon-12 resulting from the inelastic scattering of isotope source neutrons; the precision found was 0.2 per cent. of carbon.242 Prompt radiation from source neutron irradiation has also been ex-amined in some detail particularly in Russia with the aim of using the method for mineral exploration and bore hole logging.A very large number of papers have appeared considering different aspects of the subject and a combination of neutron sources can help to improve the quality of the analytical data derived from the measurements. A pulsed radioisotope neutron source has been constructed for neutron die-away systems by separating several alpha sources from beryllium targets with a rotatable shutterN3 operating at 12,000 r.p.m. thus giving 500-ps neutron bursts. Field equipment based on an antimony - beryllium neutron system a germanium counter and a small computer has been installed in a mobile laboratory for ore assay rock identification and exploration.244 Undoubtedly one of the most interesting developments that has taken place recently in the development of radioisotope neutron sources has been the avail-ability to certain workers of the spontaneous fission emitter californium-252.The neutron output of this isotope is 2-34 x 10l2 neutrons s-l g-l from a volume of less than 1 cm3 and the half-life of decay is 2.6 yearsa5; the neutron energy spectrum is somewhat similar to that of the thermal-fission neutron spectrum of uranium-235. It thus provides a neutron energy distribution that differs from many of the radioisotope sources generally used for neutron-activation analysis. Investigations have been carried out to assess the value of californium-252 for activation analysis, and a 0-37-mg source with a neutron emission intensity of 8.6 x lo* neutrons per second placed at the centre of a 1-ft moderator cube was found to give thermal, epithermal and fast fluxes of 9.7 x lo6 9.4 x 104 and lo8 cm-2 s-l mg-l respec-tively with a cadmium ratio of 7*6.248 Interference-free sensitivities and detection limits have also been found by using this source.A substantial programme of work has been carried out by a number of laboratories to assess possible outlets for californium-252 and these have included bore-hole logging applications247 and in sit% analysis of the ocean floor.= In the latter case neutron capture gamm RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 163 radiation was measured to identify manganese in nodules and gold in ore in a simulated marine environment. However while the potential of this new neutron source for a number of activation applications has clearly been demonstrated its real value to activation analysis will only become apparent when supplies of the isotope become more generally available.Charged -particle Activation Analysis Charged particles lose their energy rapidly during passage through matter and so differ radically from neutrons in their behaviour. Low penetration endows charged-particle activation analysis with characteristics that may be either dis-advantageous or advantageous depending upon the type of analysis being carried out but many of the recent applications of the technique have been designed to exploit low penetration and to provide specific analytical information from a very limited thickness of material. However radiation yields can be obtained from sample thicknesses that exceed 100 ~ m 2 * ~ so that the mass of sample analysed can often compare favourably with that examined by other instrumental analytical techniques particularly if the ion beam is spread to cover a large surface area.The yield will however vary with depth because of the change in cross-section with energy. Delayed techniques The majority of the applications of charged-particle activation analysis have been based upon the measurement of induced radioactivity and have been designed to exploit the high sensitivity of the technique for the determination of light elements such as carbon oxygen and nitrogen which are of considerable importance as impurities in many fields of material science but for which conventional neutron-activation techniques often do not provide adequate sensitivity. Variations of both the type of particles (for example protons deuterons helium-3 and helium-4) and their energy are frequently possible with positive ion accelerators so that the options open to the activation analyst can be numerous.The most suitable reaction is normally chosen only after consideration of both the element to be determined and the sample matrix as the interfering nuclear reactions may well limit apphca-tion of the technique. The convenience of the method must also be considered as a suitable particle accelerator must be available. Samples must of course be suitable for irradiation under the normal conditions that apply when using particle accele-rators and they must be capable of dissipating the heat resulting from interaction of the sample with the incident ion beam.One of the problems that has accom-panied the use of charged-particle activation analysis has been standardisation in view of the different stopping powers of various matrices and the use of the concept of equivalent thickness250 or average ~ross-section~~1 has simplified the application of the technique. Computer transformation methods carried out in conjunction with accurate activation curves for specific nuclear reactions hav 164 PIERCE simplified the analysis of different matrices by establishing a curve for one matrix and converting it for use in another with the aid of differential range-energy table~.2~~ Helium-3 was proposed for sample irradiation by Markowitz and Mahony2= some time ago as the particle undergoes many exoergic reactions with a large cross-section at low kinetic energy and has now been shown to be of con-siderable interest to workers at a practical level.Not surprisingly semi-conductor materials in which light element impurities play an important role in modifying behaviour have been examined by helium-3 activation and measurements of the oxygen content of gallium phosphide silicon and germanium by this method have yielded results that are in good agreement with theoretical estimates2= Deter-mination of oxygen and carbon by helium-3 irradiation has been supplemented by proton activation analysis of nitrogen for a combined determination of the three elements at concentrations as low as 10 parts per lo9 in semiconductor sili~on,2~~ and interferences in the use of helium-3 and helium-4 activation for the determination of carbon and oxygen in high purity iron nickel and chromium have been studied with helium-3 ions of up to 30 MeV and helium-4 of up to 54 MeV.256 Comparison standards for charged-particle activation analysis particularly for helium-3 irradiations have been produced by the anodic oxidation of tantalum foil and the isotopic composition of the oxide has been varied by using isotopically enriched water in the ele~trolyte.~~' Oxide films were thus produced with thickness varying from 0.5 mg cm-2 to over 35 mg cm-2 and were capable of reproducing beam degradation conditions from negligible energy loss to complete stopping of a 10-MeV helium-3 beam.Total reaction cross-sections of the reactions 1°B(d,n)llC, 14N(d,n)150 and 160(d,n)l7F have been examined to assess their value for activation analysis and at 3-MeV deuteron energy theoretical detection limits of 3 x 10-lo g, 5 x 10-11 g and 4 x 10-ll g have been calculated for boron nitrogen and oxygen, respectively.258 Deuteron irradiation has also been used to support studies into the diffusion of oxygen in gallium ar~enide,2~~ while protons with energies higher than is usual for activation analysis (185 MeV) have been used to determine the nitrogen content of seeds by the reaction 14N(p,d)13N.260 The excitation function of the reaction 160(3H,n)18F has been examined so as to assess the possibility of using accelerated tritons for oxygen determination261 and provides an alternative to the use of secondary tritons produced by reactor neutron irradiation of lithium for oxygen measurement (considered in a previous section).Oxygen has been deter-mined in pure zirconium and aluminium at surface oxygen levels of from 100 to 10-3 pg cm-2 by this method. The low depth of penetration of helium-3 ions has been exploited to enable the depth profile of oxygen to be determined in silicon.262 Recoil ranges of fluorine-18 predicted from considerations of total momentum transfer were found to agree with experimental values. Charged-particle irradia-tion has been combined with autoradiography to examine the positional location of light elements.263 After irradiation the sample was placed in contact with a suitable photographic film to record the distribution of the induced activities and the different activities identified by calculating half-lives from a number of exposures of the same sample taken at different times after the completion of irradiation RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 165 Prompt techniques A major change in emphasis that has been apparent in the application of charged-particle techniques over the last few years has been the increasing interest in analytical methods based upon the measurement of the prompt nuclear reaction products emitted during irradiation.Prompt events occurring during charged-particle irradiation can often be detected fairly readily and background from the incident radiation is less of a problem than for example when fast neutrons are used to irradiate the sample. The growth in interest in prompt methods has stemmed not only from an improvement in instrumentation which has enabled the activation analyst to apply the techniques to samples of real interest rather than to demonstration materials but also from a growing awareness that the methods more often associated with nuclear physics investigations can sometimes be adapted to provide specific analytical information.Elastically scattered particles prompt gamma radiation and particle groups emitted as a result of nuclear interaction have all been used as a basis of analytical methods and are considered briefly below. In general particle energies used for the production of prompt events are lower than those used for conventional charged particle activation although there are exceptions. Elastic scattering. The principle of elastic charged particle scattering as applied to analysis has been described by R ~ b i n ~ ~ in some detail and examples are given that have been based on the use of a magnetic spectrograph to examine scattered ions.A more recent review264 considers not only the general principles of elastic particle scattering but also use of the technique to provide structural information in single crystals by making use of the phenomenon of channelling. The energy loss of a particle scattered elastically is a function of mass of the scattering and of the scattered nuclei and a plot of scattered particle yield against particle energy will show a peak for scattering from a thin film or a Rutherford Plateau with a high-energy cut-off characteristic of the scatterer when the targets are thick.Thus elastic scattering is best applied to the determination of a thin film of a heavy element present on a lighter substrate as the contribution from the heavy film is clear of the substrate plateau and is therefore easier to measure quantita-tively; in some cases depth distributions have been calculated from the shapes of particle peaks. Elastic scattering has been used to examine the surfaces of a variety of materials and 1-MeV helium ions have been shown to be capable of achieving a limit of detectability of atomic fraction with a depth resolution of 300 A.2s6 The technique has been used to examine surface contamination on silicon wafers occurring during washing with a contaminated solution of hydrofluoric acid in de-ionised and less than one half of a monolayer of gold and copper con-tarninant could be resolved and identified.Peak shape analysis has been used to study the diffusion of gold into copper at temperatures in the range 360" to 500" C267 and enables elemental variations with depth to be followed with greater resolution than is possible by using mechanical sectioning techniques. An interesting use of elastic particle scattering has been the development of a relatively compac 166 PIERCE instrument based on a curium-242 a-particle source and semi-conductor detectors for instrumented lunar missions,268 which subsequently operated successfully on the moon’s surf ace.269 Prompt gamma radiation. Prompt gamma radiation can be excited from most of the light elements under suitable irradiating condition~~~o and application of the traditional methods of y-ray spectroscopy can enable several elements to be determined simultaneously in a sample.Several workers have determined fluorine by the reaction 19F(p,a y)l60 counting the 6.1-MeV y-line and measurements have been carried out in gases solids and liquids down to levels of the order of 20 ~ . p . m . ~ ~ l The use of sharp resonances in this reaction has permitted depth profiles to be obtained by gradually increasing the accelerating particle energy so that the major reaction yield occurs at different distances below the sample ~urface.~~2 Sensitivities of 2 x and 1 x per cent. have been obtained for the determination of boron in silicon by measuring y-rays with energies greater than or equal to 11 and 15 MeV produced by the reaction l l B ( ~ y ) l ~ C .~ ~ ~ Particle group measurement. Measurement of particle groups emitted as a result of nuclear reactions provides yet another basis for analytical measurement. By a suitable choice of experimental conditions particle groups can often occur in a region of the spectrum where there is a very low background and hence the sensi-tivity of the method can be very high. The technique is well applied to the examination of thin films of light elements such as carbon and oxygen and particle group measurement can be used in conjunction with resonance techniques to gain information about sub-surface distribution.274 Surf ace densities of the order of 10-7 g cm-2 of sulphur can be easily identified on coins by the reaction 32S(d,p)33S?75 and depth distributions have also been obtained by the measurement of neutrons resulting from the (d,n) reaction as the energy of the emitted neutron depends on the energy of the deuteron at the position of reaction.48 Carbon nitrogen and oxygen have been determined on steel surfaces in this way and the estimated sensitivity was 0.1 pg cm-2 with a depth resolution of 4500 A.Nuclear positive ion microprobe. Use of the methods described above has not been generally concerned with the examination of small areas of sample but clearly if the size of the incident positive ion beam can be reduced a positive ion microprobe can be devised that can operate on the basis of a measurement of nuclear reactions.276 Initially relatively large beams of about 25 to 100 pm were 0btained,277,~78 but by using a quadrupole lens system in the form of the Russian quadruplet a 3-MeV proton beam of less than 4pm in diameter has been pro-d~ced.27~ The principal interest in the technique stems from the possibility of deter-mining the spatial distributions of light elements which are just those elements that are difficult to determine by more established methods of elemental distribu-tion analysis such as the electron microprobe.Applications reported include the examination of the distribution of oxygen in welds277 and particle scans of minerals and thin f i l r n ~ 2 ~ 8 while the determination of the ratios of carbon-13 to carbon-12 in biomedical samples2*0 has been proposed RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 167 Gamma Photon Activation Analysis If a y-photon with an energy that is greater than the separation energy of a particle is absorbed by a nucleus particle emission may follow and photodisinte-gration of the nucleus occurs.Interest in y-photon activation was intially stimu-lated by the sensitivity of the method for the determination of light elements such as carbon and oxygen particularly as the penetration of y-photons permits the analysis of bulk samples. Gamma photons of high energy are required to obtain reactions with cross-sections adequate for high sensitivity determinations and are normally produced as ‘Bremsstrahlung’ by allowing accelerated electrons to strike a radiator of a heavy element. Betatrons or linear accelerators are used for electron acceleration but in general the latter have found widest application as high beam currents can be produced with these machines.Like accelerator neutron sources accelerator y-photon sources produce radiation beams that show very rapid flux variations. In spite of this however special irradiation facilities including pneumatic transfer systems have been installed in a number of laboratories to satisfy the requirements of the activation analysts. Interferences in y-photon activation may arise as in the use of other high energy incident radiation by nuclear interference as well as by secondary neutrons and investigations have been carried out by a number of workers to define the products formed by irradiating different elements. These have included a comprehensive investigation of products resulting from irradiation of a large number of elements in the periodic table281 with 30-MeV photons while interaction of 35-MeV photons with eighteen elements had also been examined.282 Control of the energy of the photons enables interferences to be avoided and optimisation techniques have been developed to exploit this.283 Interest in the determination of light elements by y-photon activation analysis has been maintained and further papers have appeared describing applications.These have included the determination of carbon in vanadium at the 10 to 150 p.p.m. level284 and in high purity iron chromium nickel and molybdenum.2S6 The deter-mination of oxygen in sodium has been possible by using y-photon activation analysis at levels down to a few parts per million,286 while another method for the determination of oxygen in sodium employs a radiochemical separation involving the exchange of the active oxygen with inactive oxygen present in water and subsequent distillation of a portion of the water.%’ The technique has also been found to be satisfactory for the determination of carbon and oxygen in 0-1 cm3 of a lead - bismuth eutectic mixture; in this case the sensitivity for carbon was found to be 5 p.p.m.while the minimum level at which oxygen was determined was 3 & 2 p.p.m.288 Carbon and oxygen present as major constituents have been determined in humic acid by irradiation with Bremsstrahlung of maximum energy 24.5 MeV,S9 and the use of laboratory betatrons has also been proposed for the determination of oxygen in coals.2g* Gamma photon activation analysis has been used for the determination of other elements either where neutron-activation analysis has not been easily applied or where the existence of intense y-photon sources has made the technique convenient.Thus yttrium has been determined in rare earths which because o 168 PIERCE their high thermal neutron absorption cross-sections would provide self-shielding problems if thermal neutron activation were used.291 Fluorine chlorine bromine and iodine have been determined in a single sample by means of (y,n) reactions292 and fluorine determined in urine with a precision of 10 per cent. at the 1 to 2 p.p.m. level with a sensitivity of 0.01 pg after concentration on an anion-exchange resin.293 A number of geological applications have been devised for y-photon activation, including the measurement of zirconium and titanium complex titano - zirconium iron nickel and cobalt in iron meteorites295 and the zirconium content of zirconium ores and concentrates.296 Photo nuclear activation has been compared with fast neutron activation for the determination of copper in ores and flotation products so as to develop a method less sensitive to particle size and matrix effects than X-ray fluorescence and photo-nuclear activation has been preferred by virtue of the fewer interferences produced.223 (y,y‘) reactions can sometimes be produced with y-photons of a few MeV that is to say of very much lower energies than are normally used for y-photon activation analysis.297 While cross-sections are often very low it is possible to use a smaller and consequently a cheaper accelerator.Among the applications that have been reported have been the determination of selenium yttrium silver barium hafnium and gold based on the production of metastable states of stable isotopes with an accelerator energy of 4.2 MeV and a current of 5 p a m p ~ . ~ ~ ~ An alternative method of producing the (7,~‘) reaction is by using an intense radioactive y-source. An 80 kCi cobalt-60 source has been used for such measurements and methods have been developed for determining selenium bromine strontium silver cadmium and indium in certain mixtures and lower levels of from 1 to 70-mg amounts of individual elements could be determined.29B The low neutron binding energies of deuterium and beryllium have enabled (y,n) reactions to be used for the deter-mination of these elements and an on-line method has been developed for monitor-ing the deuterium oxide concentration of process streams at deuterium oxide levels of 0.015 to 0.13 per cent.by measuring neutrons emitted from the 2D(y,n)1H react ion .300 References 1 2 3 4 5 6 7 8 9 10 Lutz G. J. Boreni R. J. Maddock R. S. and Meinke W. W. Editors ‘Activation Analysis A Bibliography,’ National Bureau of Standards Technical Note 467 1968. Nuclear Sciences Series NAS-NS U.S. National Academy of Sciepces. Salutsky M. L. in Kolthoff I . M. and Elving P. J. Editors Treatise on Analytical Chemistry,’ Interscience Publishers Vol. 1 part 1 1959 p. 733. Bock Werthmann W. ‘Nuclear Activation Techniques in the Life Sciences,’ International Atomic Energy Agency Vienna 1967 p.173. Welcher F. J. ‘Organic Analytical Reagents,’ D. Van Nostrand and Co. 1948. Morrison G H. and Freiser H. ‘Solvent Extraction in Analytical Chemistry,’ John Stary J. ‘The Solvent Extraction of Metal Chelates,’ Pergamon Press 1964. Samuelson O. ‘Ion Exchange Separations in Analytical Chemistry,’ John Wiley and Sons, Inczedy J. ‘Analytical Applications of Ion Exchangers.’ Pergamon Press 1966. Cerrai E. Chromatog. Revs 1964 6. Wiley and Sons Inc. 1962. 1963 RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 169 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Stahl E.‘Thin-layer Chromatography,’ George Allen and Unwin Ltd. 1969. Pierce T. B. and Flint R. F. Analytica Chim. Acta 1964 31 595. Schneer A. E. Kem. Kozlem 1969,31 23. Juvet R. S. and Cram S. P. Analyt. Chem. 1970 42 14R. Girardi F. Pietra R. and Sabbioni E. J . Radioanal Chem. 1970 5 141. De Voe J . R. ‘Application of Distillation Techniques to Radiochemical Separations,’ Qureshi J. H. and Nagi F. I. J . Inorg. Nucl. Chem. 1967 29 2879. De Wet W. J. J . S . Afr. Chern. Inst. 1969 22 168. Fritz J. S. “;“d Peters M. A. Talanta 1969 16 575. Girardi F. Nuclear Activation in the Life Sciences,’ International Atomic Energy Pierce T. B. and Peck P. F. Analyst 1963 88 603. Ruzicka J. and Stary J. ‘Substoichiometry in Radiochemical Analysis,’ Pergamon Smales A.A. Mapper D. and Fouch6 K. F. Geochemica et Cosmochinzica Acta 1967,31, Thompson B. A. and LaFleur P. D. Artalyt. Chem. 1969 41 853. Morrison G. H. Gerard J. T. Travesi A. Currie R. L. Peterson S. F. and Potter, Samsahl K. AnaZyst 1968 93 101. Goode G. C. Baker C. W. and Brooke N. M. Ibid. 1969 94 728. Girardi F. in De Voe J. R. Editor ‘Modern Trends in Activation Analysis,’ National Bureau of Standards Special Publication 312 1 577. Heath R. L. in De Voe J. R. op. cit. Special Publication 312 2 1959. Kandiah K. Sterling A. Trotman D. L. and White G. ‘Proceedings of the Inter-national Symposium on Nuclear Electronics,’ September 1968 La Socidtd Frangaise des Electroniciens et des RadioBlectriciens. Lewis A. U.K. Atomic Energy Authority Report A.E.R.E.-R.5844 1968. Thompson C. J. Nuclear Applications 1969,6 559. Pierce T. B. Webster R. K. Hallett R. and Mapper D. in De Voe J. R. op. cit., Christensen P. and Christiansen E. M. Euratom report 4289 1969 317. Cola G. D. Girardi F. and Guzzi G. Euratom report 4289 1969 313. Adams F. and Dams R. Radiochem. Radioanal. Lett. 1969 1 163. Voigt A. F. Becknell D. E. and Menapace L. in De Voe J. R. op. cit. Special Publica-tion 312 2 1035. Coleman R. F. in De Voe J. R. op. cit. Special Publication 312 2 1262. Parker J. L. Holm D. M. and Barnes B. K. in De Voe J. R. op. cit. Special Publica-tion 312 2 1075. Cooper J. A. Rancitelli L. A. Perkins R. W. Haller W. A. and Jackson A. L. in De Voe J. R. op. cit. Special Publication 312 2 1054. Michelsen 0.B. and Steinnes E. Talanta 1969 16 1436. Yule H. P. in De Voe J. R. op. cit. Special Pulication 312 2 1155. Salmon L. Nucl. Instrum. Meth. 1961 14 193. Op de Beeck J. and Hoste J. Analytica Chim. Acta 1966 35 427. Covell D. F. Analyt. Chem. 1959 31 1785. Rubin S. in Kolthoff I. M. and Elving P. J. Editors ‘Treatise on Analytical Chemistry,’ Peisach M. and Poole D. O. ‘Modern Trends in Activation Analysis,’ Proceedings of the Moller E. Nilsson L. and Starfelt N. Nucl. Instrum. Methods 1967 50 270. Fleischer R. L. and Price P. B. Science 1963 140 1221. Loveridge B. A. and McInnes C. A. J. The Microscope 1968 16 105. Smales A. A. in Meinke W. W. and Scribner B. F. Editors ‘Trace Characterization Guinn V. P. in Lenihan J. M. A, and Thomson S. J. Editors ‘Advances in Activation Aliev A.I. Drynkin V. I. Leipunskaya D. F. and Kasatkin V. A. ‘Nuclear-Physical Nuclear Sciences Series NAS-NS 3108 US. National Academy of Sciences 1962. Agency Vienna 1967 p. 117. Press 1968. 673. N. M. Ibid. 1969,41 1633. Special Publication 312 2 1116. Vol. IV Part 1 Interscience Publishers 1963 p. 2075. 1965 International Conference. Texas A and M University 1965 p. 206. Chemical and Physical,’ Monograph 100 1967 National Bureau of Standards. Analysis,’ Vol. I Academic Press (London) Ltd. 1969 p. 37. Constants for Neutron Activation Analysis Atomizdat Moscow 1969 170 PIERCE 54 Yule H P Lukens M R and Guinn V P Nucl Instrum Methods 1965 33 277 55 Proza 2 and Rakovic M Chem Lzsty 1969,63 257 56 Op de Beeck J P J Radzoanal Chem 1970,4 137 57 Lukens H R and Yule H P Nucl Instrum Methods 1965,33 273 58 Naughton W F and Jester W A zn De Voe J R op czt Special Publication 312, 1 490 59 Leushkina G V Lobanov E M Dutov A G and Matveeva N P A t Energ.(U.S S R ) 1969,26 381 60 Wiernik W and Amiel S J Radzoanal Chem 1969 3 393 61 Junod E Report CEA-R 2980 Parts A and B 1966 62 - Report CEA-R 2980(2) Part C 1970 63 Fleischer R L Price P B and Walker R M Sczence 1965 149 383 64 Martin J zn De Voe J R op czt Special Publication 312 1 517 65 Reed R E 0 R N L Report 4334 1969 p 3 66 Sayre E V Lechtman H N and Heather N Stud Conserv 1968 13 161 67 Osmond R G and Smales A A Analytzca Chzm Acta 1954 10 117 68 Mattson J S Crittenden J C and Mark H B Nucl App Technol 1969 7 383 69 Rafaeloff R Razdochem Radzoanal Lett 1969 1 199 70 Nikolov K and Todoruvsky D Isotopenpraxzs 1969 5 408 71 Smales A A Mapper D Webb M S W Webster R K and Wilson J D Sczence, 1970 167 509 72 Morrison G H Gerard J T Kashuba A T a Gangadharam E V Rothenberg A M , Potter N M and Miller G B Ibzd 1970 167 505 73 Reed G W Jovanovic S and Fuchs L H Ibzd 1970,167 501 74 Goles G G Osawa M Randle K Beyer R L JCrome D Y Lindstrom D J , 75 Waenke H Begemann F Vilcsek E Rieder R Teschke F Born W Quijano-76 Schmitt R A Wakita H and Rey P Ibzd 1970 167 512 77 Turekian K K and Kharkar D P Ibzd 1970 167 507 78 Herr W Herpers U Hess B Skerra B and Woelfle R Ibzd 1970 167 747 79 Emery J F Strain J E O’Kelley G D and Lyon W S Radzochem Radzounul Lett 1969 1 137 80 Morgan J W Rebagay T V Showalter 14 L Nadkarm R A Gillum D E , McKown D M and Ehmann W D Nature 1969,224 789 81 Millard H T zn De Voe J R op czt Special Publication 312 1 395 82 Landstrom 0 Samsami K and Wenner C -C Report A E 296 1967 83 Gangadharam E V and Reddy G R Radzochem A d a 1969 11 90 84 Green T H Brunfelt A 0 and Heier K S Eavth Planet Scz Lett 1969 7 93 85 Babaev A Babakhodzhaev S M Khaidarov A and Khudoiberganov U Dokl Akad Nauk Tadzh SSR 1969 12 24 86 Gordon G E Randle K Goles G G Corliss J B Beeson M H and Oxley S S , Geochem Cosmochzm Acta 1968 32 369 87 Fdby R H and Haller W A zn De Voe J R op czt Special Publication 312 1 339 88 Ehmann W D and McKown D M Analyt Lett 1969,2 49 89 Suttle A D O’Brien B C and Mueller D W Analyt Chem 1969 41 1265 90 Rosholt J N and Szabo B J zn De Voe J R op czt Special Publication 312,1 327 91 Fritze K and Robertson R zn De Voe J R op czt Special Publication 312 2 1279 92 Kline J R and Brar S S Isotop Radzat Technol 1970 7 319 93 Brunelle R L Hoffman C M Snow K B and Pro M J J Ass Analyt Chem 1969, 52,911 94 Brunfelt A 0 and Steinnes E Analytzca Chzm Acta 1969 48 13 95 Wainerdi R E Uken B A Santos.C G and Yule H P ‘NucIear Techniques and Mineral Resources,’ International Atomic Energy Agency Vienna 1969 p 507 96 Santos G I and Wainerdi R E zn De Voe J R op czt Special Publication 312 1 379 97 Robertson D E BNWL-1051 (Part 2) 1967 p 59 98 Lux F 2 analyt Chem 1968,243 107 99 Kasperek K Atomkernenergze 1969 14 143. Martin M R McKay S M and Steinborn T L Ibzd 1970 167 497 RICO M Voshage H I and Wlotzka F Ibzd 1970 167 523 RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 171 100 Frigerio N.A. CNM-R-2 (Vol. 2). 1967 p. 797. 101 Rakovic M. Chem. Listy 1969 63 679. 102 Comar D. in Lenihan J. M. A. and Thomson S. J. Editovs ‘Advances in Activation 103 McAndrew R. G. Smathers J. B. Wainerdi R. E. Harrison G. M. and Doggett R., 104 Woodruff G. L. Babb A. L. Wilson W. E. Yamamoto Y. and Stamrn S. J. Nucl. 105 Strain W. H. Rob C. G. Pories W. J. Childers R. C. Thompson M. F. Hennessen, 106 Yoshimasu F. Wakayama Igaku 1969,20 31. 107 Rancitelli L. A. Perkins R. W. and Renzetti A. D. BNWL-1051 (Part Z) 1969 p. 6. 108 Akaboshi M. Maeda T. Noda M. and A. Waki Zool. Mag. Tokyo 1967 76 235. 109 Holm V. Steinnes E. and Waaler T. Meddr Nwsk. Farm Selsk 1968 30 17.110 Lenihan J. M. A. Comar D. Riviere R. and Kellershohn C. Nature 1967 214 1221. 111 Comar D. Crouzel C. Chasteland M. Riviere R. and Kellershohn C. Nucl. Appl., 1969 6 344. 112 Nixon G. S. Smith H. and Livingston H. D. ‘Nuclear Activation in the Life Sciences,’ International Atomic Energy Agency Vienna 1967 p. 455. 113 Higashi T. and Kawai J. Kokobyogakkai Zusshi 1969 18 357. 114 Rancitelli L. A. BNWL-1051 (Part 2) p. 146. 115 Rancitelli L. A. BNWL-1051 (Part 2) p. 142. 116 Livingston H. D. and Thompson G. NYO-2174-96 1969. 117 Merlini M. Ravera O. and Bigliocca C. in De Voe J. R. op. cit. Special Publication 118 Gordon C. M. and Larson R. E. in De Voe J. R. op. cit. Special Publication 312,1 142. 119 Linstedt K. D. and Kruger P. Analyt. Chem. 1970 42 113.120 Piper D. Z. and Goles G. G. Analytica Chim. Acta 1969 47 560. 121 Robertson D. E. and Prospero J. M. BNWL-1051 (Part 2) 1969 p. 53. 122 Oda T. Onken Kiyo 1968,20 205. 123 Rancitelli L. A. and Tanner T. M. BNWL-1051 (Part 2)’ 1969 p. 137. 124 Bhatki K. S. and Dingle A. N. Radiochem. Rudioanal. Lett. 1970 3 71. 125 Abdullaev A. A. Zakhidov A. S. and Nishanov P. K. Im. Akud. Nauk. Uzb. SSR, 126 Hayes D. W. Ph.D. Thesis Texas A. and M. University 1969 Nucl Sci. Abs. 13779. 127 Cali J. P. ‘Trace Analysis of Semiconductor Materials,’ Pergamon Press 1964. 128 Nagy L. G. and Giber J. Period Polytech. Chem. Eng. 1968 12 37. 129 Przyborski W. Roed J. Lippert J. and Sarholt-Kristensen L. Radiat. Efl. 1969 1, 130 Ortega R. F. in De Voe J. R. op. cdt. Special Publication 312 1 536.131 Ballaux C. Dams R. and Hoste J. Analyticu Chim. Acta 1969 47 397. 132 Gilbert €2. N. and Pronin V. A. Izv. Sib. Otd. Akad. Nauk SSSR No. 4 Ser. Khim. 133 Samadi A. A. and May S. Bull. SOC. Chem. Fr. 1969 10 3776. 134 Van Reenen T. J. and De Wet W. J. J . S. Afr. Chem. Inst. 1968,21 120. 135 Alimarin I. P. Miklishanskii A. Z. and Yakovlev Y . V. J. Radioanal. Chem. 1970 4, 136 Lobanov E. M. Mamadzhanov F. I. Chubarov L. B. Dokl. Akud. Nauk. Tadzh. SSR , 137 Neeb K. H. Stoeckert H. Braun R. and Bleich H. P. 2. Analyt. Chem. 1969 245, 138 Csada I. Schneer A. and Rausch H. Kozlern 1968 16 219. 139 Guinn V. P. and Bellanca S. C. in De Voe J. R. op. cit. Special Publication 312 1 93. 140 Das H. A. Hoede D. and Zonderhuis J. report RCN. 106. 141 Lee B.K. and Murphy G. Cuncev 1969,23 1315. 142 Nadkarni R. A. and Ehmann W. D. Radiochem. Radioa?aal. Lett. 1969,2 161. 143 Geisman J. R. Carey W. E. Gould W. A, and Alban E. K. J. Fd Sci. 1969,34 295. 144 Fer A. and Fourcy A. Nucl. Appl. 1969 6 360. 145 Bowen H. J. M. Analyst 1967 92 124. 146 Yamane Y. and Miyazaki M. Eisei Kagaku 1969 15 238. Analysis,’ Academic Press 1969 p. 163. Nucl. Appl. Technol. 1970 8 290. Appl. 1969,6 352. J. A. and Graber F. M. ApPl. Spectrosc. 1969,23 121. 312 1 475. Ser. Fiz-Mat. Nauk 1968 5 60. 33. Nauk 1969 2 168. 45. 1969 12 37. 233. 172 PIERCE Zoller W. H. and Gordon G. E. Arralyt. Chem. 1970 42 267. Brar S. S. Nelson D. M. Kanabrocki E. L. Moore C. E. Burnham C. D. and Hattori, D. M. in De Voe J. R. op. cit.Special Publication 312 1 43. Rancitelli L. A. BNWL-1051 (Part Z) 1969 p. 135. Loucks R. H. Winchester J. W. Matson W. R. and Tiffany M.A. in De Voe J. R., op. cit. Special Publication 312 1 36. Dudey N. D. Ross L. E. and Noshkin V. E. in De Voe J. R. op. cit. Special Publica-tion 312 1 55. Coleman R. F. Cripps F. H. Stimson A. and Scott H. D. Atomic Weapons Research Establishment Report 0-86 1967. Henke G. Arch. Tox. 1969 25 48. Williamson T. G. and Harrison W. W. in De Voe J. R. op. cit. Special Publication 312 1 283. Coleman R. F. J . Forensic Sci. SOC. 1968 8 32. Atalla L. T. and Lima F. W. Radiochem. Radioanal. Lett. 1970 3 13. Schlesinger H. L. Lukens H. R. and Settle D. M. in De Voe J. R. op. cit. Special Publication 312 1 265. Houtman J. P. W.and Turkstra J. ‘International Atomic Energy Agency Symposium on Radiochemical Methods of Analysis,’ Salzburg 1964. Lux F. and Braunstein L. 2. Andyt. Chem. 1966,221 235. A1 Kital R. A. Chan L. and Sayre E. V. in Van Beek G. W. Editor ‘Hajer Bin Humeid Investigations a t a Pre-Islamic Site in South Arabia,’ Johns Hopkins Press, 1969. Harbottle G. Archmometry 1970 12 23. Gordus A. A. Griffin J. B. and Wright G. A. COO 912-16 1968. Das H. A. Radiochem. Radioanal. Lett. 1969 1 289. Gibbons D. and Lawson D. in De Voe J. R. op. cit. Special Publication 312 1 226. Cappadona C. in De Voe J. R. op. cit. Special Publication 312 1 72. Gatz D. F. Dingle A. N. and Winchester J. W. J . Appl. Meteorol. 1969 8 229. Channell J. K. and Kruger P. in De Voe J. R. op. cit. Special Publication 312 1 81.Bando S. Tominaga H. Kawakami Y. and Senco M. Radioisotopes Tokyo 1969 18, Pillay K. K. S. Thomas C. C. Hart D. M. Didising D. and Thomas R. C. Nucl. Rewienska-Kosciukova B. Wezranowski E. Zarzecka E. Radwan M. and Palu-Mantel M. Gilat J. and Amiel S. in De Voe J. R. op. cit. Special Publication 312, Turkstra J. and De Wet W. J. J . S. Afr. Chem. Inst. 1969 22 1. Frana J. Vorbecky M. and Kristak J. Radiochem. Radioanal. Lett. 1969 1 41. Imamura M. Matsuda H. Horie K. and Honda M. Earth Planet Sci. Lett. 1969 6, Maslov I. A. Lukhnitskii V. A. Karnaukhova N. M. and Karaganova Q. I. Nucl. Ackermann R. J. Kojima M. Rauh €2. G. and Walters R. R. J . Chem. Thermodyn., Hopenfeld J. ANL 7520 (Part l) p. 163. Osawa H. Bunseki Kiki 1969 7 619.Greenwood R. C. and Reed J. H. IITRI-1193-53 Vols 1 and 2 Illinois Institute Tech. Rasmussen N. C. Hukai Y. Inouye T. and Orphan V. J. AFCRL-69-0071 US. Air Isenhour T. L. and Morrison G. H. Analyt. Chem. 1966,38 162. Garbrah B. W. and Whitley J. E. Ibid. 1967,39 345. Juna J. Konecny K. and Vobecky M. Colln Czech. Chem. Commun. 1969 34 1605. Elkady A. and Duffey D. Trans. Amer. Nucl. SOC. 1969 12 42. Lombard S. M. and Isenhour T. L. Analyt. Chem. 1969‘41 1113. Comar D. in Lenihan J. M. A. and Thomson S. J. Editors ‘Advances in Activation Basdekas D. L. US. Patent 3,436,538 1969. 488. AppI. Technol. 1970 8 73. chowicz L. Hutnik 1967,34 676. 1 482. 165. Appl. Technol. 1969 7 389. 1969 1 527. Res. Centre 1965. Force Cambridge Research Labs 1969. Analysis,’ Vol.I Academic Press 1969 p. 163. 147 148 149 160 151 162 153 154 165 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 18 RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS 173 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 21 1 212 213 214 215 216 217 218 219 220 22 1 222 223 224 225 226 227 228 229 230 231 Hanson A. O. Taschek R. F. and Williams J. H. Revs. Mod. Phys. 1949,21 635. Hillier M. Lomer P. D. Stark D. S. and Wood J. D. L. H. in Ebert H. G. Editor, ‘Proceedings of the Third Conference on Accelerator Targets Designed for the Produc-tion of Neutrons,’ Euratom Report EUR 3895 d-f-e 1967.Steele E. L. ‘Modem Trends in Activation Analysis,’ Proceedings of 1965 International Conference 1965 p. 102. Pierce T. B. Edwards J. W. and Haines K. Talanta 1968 15 1153. Oldham G. and Darrall K. G. Internat. J . Appl. Rad. Isotop. 1969,20 29. DeSoete D. Gijbels R. and Hoste J. in De Voe J. R. op. cit. Special Publication 312 Op De Beeck J. J. Radioanal. Chem. 1968 1 313. Morgan J. W. and Ehmann W. D. Analytica Chim. Acta 1970 49 287. Anders 0. U. and Briden D. W. Analyt. Chem. 1964 36 287. Lundgren F. A. and Nargolwalla S. S. Ibid. 1968,40 672. Hoste J. De Soete D. and Speecke A. Euratom Report EUR 3565e 1967. Coleman R. F. Analyst 1962,87 590. Nargolwalla S.S. Crambes M. R. and Suddueth J. E. Analytica Chim. Acta 1970 49, Stoll N. Wagner A. and Goedert L. Euratom Report EUR 3161f 3 1966. Tyou P. Revue de Metallurgie 1969 66 621. Gijbels R. Speecke A. and Hoste J. in De Voe J. R. op. cit. Special Publication 312, Doty W. H. Wood D. E. and Schneider E. L. J . Ass. Analyt. Chem. 1969 52 953. Kosta L. ‘Panel Meeting on New Approaches to Breeding for Improved Plant Protein, Bakes J. M. and Prapuolenis A. A. Radiochem. Radioanal. Lett. 1969 1 19. Tamura M. Radioisotopes Tokyo 1969 18 252. Ehmann W. D. and Morgan J. W. Science 1970 167 528. Schramel P. J. Radioanal. Chem. 1969 3 29. Przybylowicz E. P. Smith G. W. Suddueth J. E. and Nargolwalla S. S. Analyt. Tousset J, Condamin J. and Picon M. Method. Phys.Anal. 1968 4 202. Tover. P.. CEA-CONF 1410 1969. 2 699. 425. 2 1298. Roestanga Sweden,’ International Atomic Energy Agency Vienna 1969 p. 161. Chem. 1969 41 819. kichardson A. E. and Hahison A. Analyt. Golanski A. Report CEA R-3694 1969. Givens W. W. Mills W. R. and Caldwell, Nagai T. J . Nucl. Med. 1969 10 192. Nargolwalla S. S. Niewodniczanski J. and Senftle F. E. Report TID 25169 1967. Santos G. G.. Fite. L. E. Kuvkendall. W. Publication 312 2 929. 1969 12 510. Chem. 1969 41 1396. R. L. in De Voe J. R. op. cit. Special Suddueth J. E. Trans. Amer. Nucl. Soc., E. Wainerdi R. E. Bouma A. H. and Mineral Resources International Atomic Bryant W.. R. ‘Nuclear Teihniques -and Energy Agency Vienna 1969 p. 463. Publication 312 2 918. Rhodes J.R. Isotop. Radiat. Technol. 1969 6 359. Jervis R. E. Al-Shahristani H. and Nargolwalla S. S. in De Voe J. R. op. cit. Special Gorski L. Janczyszyn J. and Loska L. Radiochem. Radioanal. Lett. 1969 1 99. Pradzynski A. ‘Nuclear Techniques and Mineral Resources,’ International Atomic Fleischer A. A. in De Voe J. R. 09. cit. Special Publication 312.2 868. Bruninx E. in De Voe J. R. op. cit. Special Publication 312 2 860. Wilkniss P. E. in De Voe J. R. op. cit. Special Publication 312 2 874. Chamberlain M. J. Proc. Roy. Soc. Med. 1969 62 370. Thompson W. E. and Engesser F. C. Report USNRDLTR-861 1965. Martin T. G. Mathur S . C. and Morgan I. L. Internat. J . Appl. Rad. Isotop. 1964,15, Pierce T. B. Peck P. F. and Cuff D. R. A. J . Radioanal. Chem. 1970 4 305. Caldwell R.L. Mills W. R. and Givens W. W. ‘Nuclear Techniques and Mineral Energy Agency Vienna 1969 p. 451. 331. Resources,’ International Atomic Energy Agency Vienna 1969 p. 397 174 PIERCE ‘Radioactive Products 196911970.’ The Radiochemical Centre Amersham Bucks p. 180. Tatar J. Banyaszat 1968 101 287. Jeffery P. G. and Bakes J. M. Analyst 1967,92 151. Wing J. and Wahlgren M. A. Report ANL 7242 1966. - and - J . Radioanal. Chem. 1969 3 37. Starnes P. E. U.S. Patent Application 62488 1966. Ashe J. B. Berry P. F. and Rhodes J. R. in De Voe J. R. op. cit. Special Publication Downs W. E. and Davis M. W. Nucl. Agpl. Technol. 1969 7 466. Kuusi J. Nucl. Apfil. Technol. 1970 8 465. Porkopchik V. I. Bushkov A. P. and Bernardskii K. G. Atomn. Energ. 1969 27 161. Stewart R.E. I.S.A. Transactions 1967 6 200. Caldwell R. L. and Givens W. W. US. Patent 3,389,257 1968. Tolmie R. W. and Thompson C. J. ‘Nuclear Techniques and Mineral Resources,’ International Atomic Energy Agency Vienna 1969 489. Crandall J. L. Isotop. Rad. Technol. 1970 7 306. Ricci E. and Handley T. H. Analyt. Chem. 1970 42 378. Perkins R. W. Rancitelli L. A. Cooper J. A. and Brown R. E. Trans. Amer. Nucl. Senftle F. E. Duffey D. and Wiggins P. F. Marine Tech. SOC. J. 1969 3 9. Rook H. L. Schweikert E A. and Wainerdi R. L. in De Voe J. R. op. cit. Special Engelmann C. C.R. Hebd Sianc. Acad. Sci. Paris 1964,258 C 4279. Ricci E. and Hahn R. L. Analyt. Chem. 1967,39 794. Rook H. L. and Schweikert E. A. Ibid. 1969 41 958. Markowitz S. S. and Mahony J. D. Ibid.1962 34 329. Kim C. K. Radiochem. Radioanal. Lett. 1969 2 53. Nozaki T. Yatsurugi Y. and Akiyama N. J . Radioanal. Chem. 1970,4 87. Debrun J. L. Barrandon J. N. and Albert P. Bull. SOC. Chim. Fr. 1969 1011. Lamb J. F. Lee D. M. and Markowitz S. S. Andyt. Chem, 1970 42 209. Wohlleben K. and Schuster E. Rudiochim. Acta 1969 12 75. Rachmann J, and Biermann R. Solid State Commun. 1969 7 1771. Johansson A. Larsson B. Tibell G. and Ehrenberg L. ‘New Approaches to Breeding for Improved Plant Protein,’ International Atomic Energy Agency Vienna 1969, p. 169. Barrandon J. N. and Albert P. in De Voe J. R. op. cit. Special Publication 312 2, 794. Lamb J. F. Lee D. M. and Markowitz S. S. Analyt. Chem. 1970 42 212. Holm D. M. Sanders W. M. Briscoe W. L. and Parker J. L.in Polishuk P. Editor, Mackintosh W. D. and Davies J. A. Analyt. Chem. 1969 41 26A. Davies J. A. Denhartog J. Eriksson L. and Mayer J. W. Can. J . Phys. 1967,45,4053. Thompson D. A. Barber H. D. and Mackintosh W. D. A@@ Phys. Lett. 1969 14, Sippel R. F. Phys. Rev. 1959 115 1441. Patterson J. H. Turkevich A. L. and Franzgrote E. J . Geophys. Res. 1965 70 1311. Turkevitch A. L. Franzgrote E. and Patterson J. H. Science 1969 165 277. Pierce T. B. and Peck P. F. in Shdlis P. W. Editor ‘Proceedings of the S.A.C. Bewers J. M. and Flack F. C. Analyst 1969,94 7. Moller E. and Starfelt N. Report AE 237 1966. Vasil’ev S. S. Mikhailov G. I. Starchik L. P. and Konanykin L. V. Zav. Lab. 1969, Amsel G. and Samuel D. J . Phys. Chem. Solids 1962 23 1707. Wolicki E. A. and Knudson A.R. Int. J . Appl. Rad. Isotopes 1967 18 429. Pierce T. B. Peck P. F. and Cuff D. R. A. Nature 1966 211 66. Price P. B. and Bird J. R. Nucl. Instrum. Methods. 1969 69 277. Pierce T. B. Peck P. F. and Cuff D. R. A. Ibid. 1969 67 1. Cookson J. A. and Pilling F. D. Atomic Energy Research Establishment Report R6300, 312 2 913. SOC. 1970 13 63. Publication 312 2 768. ‘Nucleonics in Aerospace Plenum Press 1968 p. 306. 102. Conference Nottingham 1965,’ W. Heffer and Sons Ltd. 1965 p. 159. 35 299. 1970. 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 25 1 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 27 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 RECENT DEVELOPMENTS IN ACTIVATION ANALYSIS Ricci E.and Gibbons J. H. Trans. Amer. N w ~ . SOL 1969 12 509. Baker C. A. and Wood D. A. Atomic Energy Research Establishment Report R-5818, Debrun J. L. and Albert P. Bull. SOG. Cltim. Fr. 1969 1020. Davydov M. G. and Shcherbachenko V. A. At. Energ. (USSR) 1969,27 205. Hislop J. S. and Wood D. A. Atomic Energy Research Establishment Report 33-6165, Revel G. J . Radioanal. Chem. 1969 3 421. Engelmann C. and Loevuillet M. Bull. Soc. Chim. Fr. 1969 680. Lutz G. J. Analyt. Chem. 1970 42 531. Mackintosh W. D. and Jervis R. E. in De Voe J. R. op. cit. Special Publication 312, Kasymov A. K. and Masagutov V. S. Dokl. Akad. Nauk Uzb. S S R 1969,4 18. Sulin V. V. Trudy Vses Nauchno-Issled. Inst. Yadern. Geofiz. Geokhim 1968 244. Lutz G. J. and LaFleur P. D. Talanta 1969 16 1457. Wilkniss P. E. Radiochem. A d a 1969 11 138. Ohno S. Suzuki M. Sasajima K. and Iwata S. Analyst 1970 95 260. Shornikov S. I. Trudy Vses Nauchno-Issled. Inst. Yadern. Geofiz. Geokhim. 1968 274. Meijers P. and Aten A. H. W. Radiochem. Acta 1969 11 60. Berzin A. K. Vitozhents G. C. and Sulin V. V. Trudy Vses Nauchno-Issled. Inst. Yadern. Geofiz. Geokhim. 1968 266. Engelmann C. and JBrome D. Y. in Ebert H. G. Editor ‘Proceedings of the Seconq Conference on Practical Aspects of Activation Analysis with Charged Particles, Euratom report EUR 3896 d-f-e Brussels 1968 p. 119. Kodiri S. and Starchik L. P. Dokl. Akad. Nauk. Tadzh. SSR 1969 12 17. Veres A. and Pavlicsek I. Radioanal. Chem. 1969 3 25. Stutheit J. S. and Rampey W. P. Nucl. Instrum. Methods 1969 75 43. 175 1968. 1969. 2 835
ISSN:0300-9963
DOI:10.1039/AS9710100133
出版商:RSC
年代:1971
数据来源: RSC
|
4. |
Atomic absorption spectroscopy |
|
Selected Annual Reviews of the Analytical Sciences,
Volume 1,
Issue 1,
1971,
Page 177-234
P. Platt,
Preview
|
PDF (6370KB)
|
|
摘要:
Atomic Absorption Spectroscopy P. PLATT Colgate-Palmolive Limited Ordsall Law Salfovd Contents Introduction Nomenclature Sensitivity Limit of detection Noise level Resonance radiation Atomisation Atomiser Nebulisation Nebulisation efficiency Direct-injection burner Pre-mix system Separated flame Observation height Burner angle Resonance radiation detector Interference Matrix effect Ionisation buffer Releasing agent Nebuliser atomiser systems Monochromators Detector and read-out systems Automation Reagents for atomic absorption spectroscopy Interferences-Application of method Instrumentation Sources Experimental considerations 17 Applications The determination of macro components Elements determined in the nitrous oxide - acetylene flame Alkaline earth elements Aluminium Beryllium Boron Chromium Germanium Molybdenum Niobium Noble metals Rare earths Scandium Rhenium Silicon Tin Titanium Vanadium Zirconium Elements with analytical resonance lines between 190 and 230 nm Determinations by an indirect method Determination of mercury Determination of isotopic concentrations 17 ATOMIC ABSORPTION SPECTROSCOPY 179 I n trod uction This review is concerned with the developments that have taken place within the field of atomic absorption spectroscopy during the last 3 or 4 years.It is not aimed at the specialist in the field but rather is intended to provide the type of information required by the busy analyst who uses or has access to an atomic absorption spectrophotometer as just one of a number of instruments routinely used in the laboratory.He may perhaps use the technique regularly or only occasionally. It may also assist those who are debating whether or not an atomic absorption instrument can be justified in their own laboratory. Some idea of the rapid growth of the technique may be gained by noting that in 1966 there were between 1000 and 2000 atomic absorption spectrophotometers in use throughout the world1 whereas by 1969 this number had risen to more than 10000.2 West3 compares its impact on inorganic trace analysis to that of gas -liquid chromatography on organic analysis. Nevertheless despite it being now over 10 years since the first commercial unit was announced according to Reynolds* it is apparent that the technique is still essentially new to many laboratories.However atomic absorption spectroscopy must be accepted as being a very popular analytical tool in every field of activity because of the many advantages it can offer over other techniques. These advantages have already been well documented by various a ~ t h o r s . ~ ~ ~ ~ Whether or not all of the oft-claimed advantages are justified has been questioned by at least two prominent worker^.^-^ During the period covered by this review a number of books on the subject have been published and these are recommended for obtaining more detailed i n f o r m a t i ~ n . ~ ~ ~ - ~ ~ The House Journals of the various instrument manufacturers also provide much useful information and colour films on the theory and practice of atomic absorption spectroscopy have become available for free loan to qualified 0rganisations.l6,~~ In addition an excellent abstracting service dealing with all aspects of atomic absorption spectroscopy is now available.l* Nomenclature The definition of universally acceptable terms for use in atomic absorption spectroscopy is a subject that has recently received much active and necessary consideration by the Atomic Spectroscopy Group of the Society for Analytical Chemistry and by an I.U.P.A.C.Commission.18 The technique is still compara-tively new and there are obvious advantages to be gained by establishing an approved nomenclature as soon as possible. The ambiguity that can arise in the absence of such a nomenclature may be illustrated by the term ‘atomisation,’ which was until recently widely interpreted as the process that converts a liquid to a mist and not as a function producing atom^.^^^^^ The subject was given prominence by Powel121 when he considered terms selected with the analytical chemist in mind.A list of tentative proposals for the definition of terms used was proposed by the Atomic Spectroscopy Group and this was submitted for comment to the I.U.P.A.C. Commission to the Standard 180 PLATT Association of Australia and to several independent individuals. The following terms have been taken from the selected list and are suggested as the ones with which every analyst using the technique should be familiar. Under each term heading will be found any available comments and information taken from the literature concluding with the definition proposed by the Atomic Spectroscopy Group of the Society for Analytical Chemistry.It is hoped that in this way the importance of using a common set of defined terms may be encouraged, and hence facilitate the interchange of information and results. Considering their usefulness in making analytical comparisons it is not surprising that the terms ‘sensitivity’ and ‘detection limit’ have received the most publicity. Sensitivity Ramirez-Mufioz and Ulrich22 stated in 1966 that any change in the widely used ‘concentration for 1 per cent. absorption’ seemed unlikely. Many prominent workers including Elwell and GidleyJ6 Stupar and Dawson,lS Marshall and Schrenk23 and Slavinll have used the same definition. Ramirez-Muiioz Shifrin and Hell% defined sensitivity as the ratio between response obtained from an instrument and the concentration of the analyte in solution.Grant25 stated that the terns ‘sensitivity’ and ‘limit of detection’ had been used interchangeably and there was still no universal acceptance of definitions for these terms although there was a growing tendency to distinguish between them. He defined sensitivity as the ability to discern a small change in concen-tration of analyte at some specified concentration. Rendina26 complained of ‘a lack of definition’ and that conditions under which sensitivity measurements are made do not relate adequately to real life analytical situations. The latter complaint is a reasonable one in so much as it is true that the element in question is generally assumed to be the only solute present usually in water as the solvent.Also the quoted sensitivity figure may well have been determined by analysing a considerably more concentrated solution and then making a simple calculation to obtain the concentration corresponding to 1 per cent. absorption. With this in mind it will be readily appreciated that ‘limit of detection’ is a far more useful term for the analyst. It should be noted that ‘sensitivity’ is defined below as a direct measure of the slope of the calibration graph near zero concentration and therefore in common with the use of this definition in other branches of analytical chemistry it gives an indication of the rate of change of the observed signal with concentration.Proposed S.A.C. definition Sensitivity is the concentration in solution of the element to be determined that will produce a change compared to pure solvent, of 0.0044 absorbance units i.e. 1 per cent. absorption in the optical transmission of the atomic vapour at the wavelength of the radiation used. Limit of detection Ramirez-Muiioz and Ulrich22 suggested that this term may be defined as the concentration in parts per million of analyte giving a signal corresponding to th ATOMIC ABSORPTION SPECTROSCOPY 181 peak-to-peak noise obtained at 0 per cent. absorption level under some given operating conditions. These authors stressed that information concerning the operating conditions e.g. scale expansion and damping is necessary to properly evaluate and compare various systems.Kirkbright Semb and West2’ agreed with these latter observations and stated that it is necessary to agree upon a suitable definition of detection limit. While there is no official agreement for this limit an increasingly accepted definition is the concentration in water solution that gives a signal equal to twice the size of the background variability.28 In statistical terms, the concentration of an element at the detection limit can be determined with a coefficient of variation of 50 per cent. Marshall and S ~ h r e n k ~ ~ defined the limit of detection as that concentration of metal that produces an absorption equivalent to twice the magnitude of the fluctuation in the background at zero absorption. Slavinll used a similar definition and stated that for comparing analytical methods the concept of detection limit is more useful than sensitivity because it includes not only the factors that govern analytical sensitivity but also the factors that control the level of back-ground fluctuation in the analysis.Grant26 noted that there must be careful definition of the conditions under which the standard deviation was measured if the limit of detection value is to be meaningful. He defined the limit of detection as the smallest amount or concentra-tion of analyte that can be detected with certainty. Rains25 said the detection limit is not defined as readily as sensitivity and he defined it as the concentration in microgrammes per litre that produces an absorption signal equal to twice the magnitude of the fluctuations of the blank or background noise.RubeSka and Moldanl4 defined the detection limit as the concentration causing a deflection equal to three times the standard deviation of the fluctuations of the unabsorbed signal or twice the maximum noise level of the fluctuations. Assuming a normal distribution for the fluctuations both definitions are almost identical. The I.U.P.A.C. Commission recommended that three sigma should be used to give a minimum certainty of 95 per cent. at any form of distribution that can reasonably be anticipated. Cooke12 pointed out that detection limits of elements in atomic absorption, although less frequently quoted than sensitivities are nevertheless a better criterion of instrumental performance. He preferred to have detection limit in terms of standard deviation of the background noise and quoted what is essentially the proposed S.A.C.definition. Reynolds29 considered this definition to be more stringent and meaningful than the ‘twice the noise level’ one and made it clear that it is now more than ever necessary to ascertain exactly under what conditions and how published performance figures have been obtained and calculated. Proposed S.A.C. definition. The limit of detection is the minimum amount of an element that can be detected with a 95 per cent. certainty which is that amount of the element that gives a reading equal to twice the standard deviation of a series of at least ten determinations at or near blank level 182 PLAT" Noise level Proposed S.A.C. definition. Noise level is that concentration of the element to be determined that would give a signal equal to one fiftieth of the sum of twenty measurements taken as follows-The output of an atomic absorption spectrometer operating on a blank solution is recorded for ten time periods each of ten times the time constant of the instrument.The maximum displacements that occur to both sides of the median line in each of the ten periods are measured. These are the twenty measurements referred to above.3o Resonance radiation Proposed S.A.C. definition. Resonance radiation is the characteristic absorbed radiation that corresponds to the transfer of an electron from the ground state level to a higher energy level in the atom. Atomisation Stupar and Dawson19 defined this term as 'the production of atomic vapours.' Proposed S.A.C.definition. Atomisation is the process that converts the element to be determined or its compounds to an atomic vapour. Atomiser Proposed S.A.C. definition. An atomiser is the device usually a flame, used to produce and stabilise or maintain a population of free atoms. The I.U.P.A.C. Commission would accept this term only with great reluctance in view of its long common and somewhat unfortunate usage as a synonym for nebulizer. The S.A.C. committee however was of the opinion that it is the only term that follows naturally from atomisation. Nebulisation Stupar and Dawson19 defined this term simply as the production of aerosols. Proposed S.A.C. definition. Nebulisation is the process that converts a liquid to a mist.Nebulisation efficiency Proposed S.A.C. definition. Nebulisation efficiency is the ratio of the amount of sample reaching the atomiser to the total amount of sample entering the nebuliser. Djrect-injection burner Proposed S.A.C. definition. A direct-injection burner is one in which the liquid is nebulised directly into the flame. The flame obtained with such a burner is normally turbulent ATOMIC ABSORPTION SPECTROSCOPY 183 Pre-mix system Proposed S.A.C. definition. A pre-mix system is a sampling unit wherein the fuel oxidant gas and sample mist are mixed in a spray chamber before entering the flame. Flames obtained by using this system are normally laminar. Separated flame Proposed S.A.C. definition. A separated flame is one in which the diffusion combustion zone is so separated from the primary combustion zone as to enable the two zones to be observed independently.Observation height between the optical axis of the monochromator and the top of the burner. Proposed S.A.C. definition. The observation height is the vertical distance This term is sometimes referred to as burner height. Burner angle Proposed S.A.C. definition. The burner angle is the acute angle between the plane of the flame produced by a long path burner and the optical axis of the monochromat or. Resonance radiation detector Proposed S.A.C. definition. A resonance radiation detector is a selective detector in which atoms in an atomic vapour are excited by radiation from an external source and the intensity of the resulting fluorescence radiation is measured.Interference Stupar and Dawson19 defined this term as the discrepancy between the true and apparent concentrations of an element when estimated by comparing its absorption with that of a pure solution. Proposed S.A.C. definition. Interference is a general term for an effect that modifies the instrumental response to a particular concentration of the element to be determined. Matrix effect Proposed S.A.C. definition. A matrix effect is an interference caused by differences between the sample and a standard containing only the element to be determined and where appropriate a solvent. Ionisation buffer Proposed S.A.C. definition. An ionisation buffer is a spectroscopic buffer used to minimise or stabilise the ionisation of free atoms of the element to be determined 184 PLATT Releasing agent Proposed S.A.C.definition. A releasing agent is a spectroscopic buffer used to reduce interferences attributable to the formation of involatile compounds in the atomiser. I nst c u men t at i o n It is outside the scope of this paper to discuss the essential requirements of an atomic absorption spectrophotometer. These have already been more than adequately described in the original paper by Walsh5 and subsequently by a num-ber of well known authors in the field.334J0J1J3-15J9 Before discussing the particular developments of recent origin it may be of useful interest to indicate generally the points to be considered when choosing an atomic absorption instrument. According to Reynolds3l there are several ways of choosing such a unit.One method used more frequently than is desirable is that of selecting the most expensive unit within the laboratory budget and trusting that it will more than meet the essential requirements. He advises the chemist to give serious thought to the features he requires the instrument to possess. In the first place the characteristics and concentration levels of the elements to be determined should be considered together with the accuracy required. For example depending upon the level of the element present and the matrix in which it occurs the simplest equipment may allow the determination of elements such as magnesium and copper with the same level of sensitivity as the more expensive units. On the other hand, the determination of elements such as iron chromium and nickel or arsenic and selenium may require a higher performance monochromator and hence a more expensive instrument (see Sources).The possibility of using electrodeless discharge lamps for the determination of arsenic and selenium warrants serious consideration when choosing instrumentation with these two elements in Speed of operation is always important and must be given due consideration as must versatility. Most commercial instruments can be operated in the emission mode and offer a built-in scale expansion facility. Prices made it clear that to be of real value to the analyst an instrument must allow very ready interchange between all operating conditions for different elements e.g. lamps wavelengths gases and absorption - emission modes.It should also allow complete control over such instrumental parameters as burners, burner height and angle lamps scale expansion and read-out facilities. It should maintain high sensitivity for trace element determinations and high stability of zero and infinite absorbance levels for the determination of more major constituents and for long runs of routine analysis particularly when automatic sampling is being used. For the analysis of major constituents it is a big advantage according to Reynolds,31 to have an instrument capable of zero suppression as illustrated later in the determination of calcium although a number of workers have reported successful results without this facility (see Applications). Most commercial instruments use the single-beam optical system and for many applications this system is perfectly satisfactory.However the advantage ATOMIC ABSORPTION SPECTROSCOPY 185 of a double-beam system should be considered and weighed against the extra cost involved. Most are not genuine double-beam systems in contrast to those com-monly found in molecular spectroscopy because the reference beam does not pass through a reference flame. Hence a double-beam system does nothing to com-pensate for flame noise which with the advent of modern stable sources is the greatest producer of noise in atomic absorption measurement^.^^^^ It will however, compensate for source noise when present and this is an advantage in improving the detection limit and in compensating for zero drift during prolonged sample runs.This noise compensation will result in a higher signal-to-noise ratio which has a particular advantage when using scale expansion. However a double-beam system results in some loss of light energy and if the same slit-width is used more gain is necessary. More gain means more noise and hence a decrease in the signal-to-noise rati0.~~,~5 Therefore two opposite effects are operating which will in part at least cancel each other out. A single-beam system does not monitor source variations but does offer some advantages in that it allows the use of lower gain with modern stable sources and consequently allows a lower noise operation. This can mean better precision, detectability and signal-to-noise ratio.% Some instrumentation can have the output signal integrated over a fixed time peri0d.~6 This does not compensate noise but automatically ‘sums’ it and then feeds out the average as a noise-free reading.It is certainly a time saving con-venience but requires a drift free base-line otherwise results can be in error. In 1968 a most versatile instrument designed with a double monochromator system made its appearan~e.~’ The design allowed the simultaneous use of a resonant line and a nearby non-resonant line to compensate background or matrix interference. It also allows as one alternative the simultaneous use of two resonant absorbing lines of a particular element and when the strengths of two signals at their selected wavelength are compared the flame noise which is common to both is eliminated. The instrument also incorporates double-beam compensation of source variations.Elwell and Gidley6 have placed at the end of their book a tabulated insert on commercially available equipment. This gives the essential characteristics of six-teen instruments and although published in 1966 it is a most useful introduction to available instrumentation. Eardley and Mountford38 offer an excellent evaluation of seven commercial atomic absorption spectrometers with manufacturers comments regarding the evaluation results and Dawson and Bro~ghton~~ have published a guide to the selection of instruments for use in clinical biochemistry. Hence a considerable amount of information and advice is available that the analytical chemist can use to guide him when choosing an instrument. However the person responsible for doing the choosing is well advised to evaluate in practice his short list of instru-ments together with their various attachments.The evaluation should be carried out with samples peculiar to the particular laboratory and taking into account the experience and ability of the operators who will be using the equipment 186 PLAT" Reynoldss1 believes that the surest way of assessing the suitability of all equipment is to see it demonstrated by a competent operator under fair conditions that reasonably simulate those in which it is required for use. This should be in the actual laboratory where the instrument is to be used if this can be arranged but all manufacturers will arrange for the prospective purchaser to visit their specialist laboratories and analyse his own samples under fair conditions.Finally the appli-cations service and servicing facilities that are available from the manufacturers should be taken into account as this can be most valuable. The basic layout of all atomic absorption instruments is shown in block diagram form in Fig. 1. Recent developments relating to the various components will be considered in their instrumental sequence. t r Mono- Detector amplifier Source --sc- Atomiser = - - and I e Read-out chroma tor -Nebuliser cloud chamber Fig. 1. Basic layout of atomic absorption instruments Sources It must be said at the outset that hollow-cathode lamps are by far the most commonly used spectral sources and it is unlikely that this situation will change, at least for some time to come.It is in the field of lamp technology that the most dramatic improvements in instrumental performance have been made and there has been a steady increase in the number of elements for which reliable and efficient hollow-cathode lamps are available.29 Slavin and Slavin2 pointed out that in the early 1960's the chief problem with practical atomic absorption methods was associated with the difficulty of making useful light sources. By 1966 all of the elements except the alkalis and possibly mercury were able to be determined most effectively and commonly with hollow-cathode sources.28 Some lamps notably those for arsenic were regarded as inadequate for general use.40v41 Kahn28 also pointed out that although hollow-cathode lamps were available discharge lamps were best for the alkali metals.Slavin and Ringhardtz*O agreed with this conclusion. The advantage of discharge lamps lay in the fact that they were many times brighter than the hollow-cathode types and hence in th ATOMIC ABSORPTION SPECTROSCOPY 187 near infrared where the monochromator and detector are weakest and the alkali metals have their most sensitive resonance lines the discharge sources were considerably more efficient. Despite these difficulties the sources that were commonly available showed, in all but a few particularly difficult cases such long life times that it was almost impossible to gather statistical data on failure rates.42 These hollow-cathode lamps although satisfactory produced a rather low output intensity that necessi-tated the use of fairly wide monochromator slit settings.This resulted in a lack of sensitivity and selectivity and also in non-linear response curves. During operation the current in conventional hollow-cathode lamps is usually adjusted to the maximum consistent with the line width requirements. It is not in general increased beyond the point where the line width shows a signifi-cant increase because this is accompanied by self-absorption or self-reversal and this too may result in loss of sensitivity and selectivity and produce non-linear response cuves. For many elements lower currents were acceptable and provided quite satisfactory sharp line sources. However in certain circumstances an improve-ment was required and this was provided by the high intensity hollow-cathode lamp.43 As the name implies these lamps produced a big improvement in intensity over the conventional lamps and it was anticipated that they would permit much more linear sensitive and stable atomic absorption measurements to be made.The design of high intensity lamps provides for the vaporisation and excitation functions of the lamp to be separated by using the hollow cathode merely to produce the metal vapour which is then excited by an auxilliary discharge of about 500 mA between two electrodes in front of the cathode. This system allowed significant enhancement of the resonance line emission without appreciable increase in line width.& WilliP gives three circumstances in which distinct advantages were gained with high intensity lamps.The first is where a conventional lamp necessarily runs at a low current and where the presence of non-absorbing lines close to the resonance line would require the use of narrow slits to satisfactorily isolate this line. Under these conditions a better signal-to-noise ratio without loss of sensitivity may be obtained. The second is where there is an ionised metal or filler gas line so close to the resonance line that it cannot be separated by the monochromator and a tremendous improvement ensued by using a high intensity lamp to enhance preferentially the resonance line. The final use is in analyses that require the bright and somewhat noisy nitrous oxide - acetylene flame. This flame requires an intense source i.e. less amplifier gain to provide a noise-free analysis.C0balt,4~*~~ nickel4' and iron44 are examples of elements that were better determined by high intensity lamps although it was not expected that they would supersede the standard lamps for such elements as magnesium zinc lead calcium and copper.4s However high intensity lamps have a number of disadvantages31 in that an extra power supply is required to provide the high auxillary current they are costly have restricted life times and can only be produced for certain elements. 188 PLATT Considering the rather serious disadvantages involved it is not surprising that improvements to conventional lamps were sought. The re-designed conven-tional lamps that soon became available embodied the necessary improvements and were called high spectral output lamps.The rapidly replaced the high intensity lamps which no longer provided significant advantages in most Indeed it seems most unlikely that the four-electrode high intensity lamps will be used in the future at least in their present form and it is reported that of the two companies that originally produced them one has virtually abandoned their manufact~re.~~ The superior performance of the new two-electrode high spectral output lamps is because of the incorporation of shielded cathodes re-positioned and re-shaped electrode assemblies and better choice of filler gases. No extra power supply is required of course they are comparatively cheap and have a long life time. Several companies now give a lamp warranty for 5 ampere hours and under normal conditions of operation this means that 200 to 500 h are guaranteed.In actual fact most lamps far exceed this W a l ~ h ~ ~ notes that for most analyses there is little to be gained by using lamps of high intensity but if an increased intensity is required it can be obtained by (a) the pulsing technique of Dawson and EllisJS0 (b) high intensity hollow-cathode lamps or (c) electrodeless discharge lamps. Mannine9 was able to say in 1968 that greatly increased output from sodium and potassium hollow-cathode lamps has now enabled them to be recommended for atomic absorption work. This means that the vapour discharge lamps with their special power supply are no longer required for these two elements. RubidiumS1 and caesium do not have reliable hollow-cathode lamps and the excellent vapour dis-charge lamps that are available should be used for these two elements.A suitable hollow-cathode lamp is now available for mercury as well as an excellent vapour-discharge lamp. Reliable hollow-cathode lamps are also available for arsenic and selenium although the use of electrodeless discharge lamps3e should be considered by anyone interested in determining these elements and also tellurium. Reynolds31 in 1969 said that improvements in lamp technology have raised the performance of simple atomic absorption equipment to that of more expensive units. This can be seen in their capability of selecting the 248.33 nm from the 248.82 nm iron lines and by the separation with a simple monochromator of the 357.87 359.35 and 36063 nm lines of chromium. Perhaps the most valuable success according to Reynolds was the production of lead lamps that permit the reliable use of the very sensitive 217.0 nm line on relatively simple equipment.In fact by 1969 several commercial suppliers were offering a near complete range of hollow-cathode lamps including those for the lanthanides. Although some of these are rather expensive all are as bright for each element as an atomic absorp-tion spectrophotometer can ~ t i l i s e . ~ ~ ~ ~ Two modifications or extensions of the high intensity type of lamp have been reported. Bowman et al. described the isolation of atomic resonance lines by selective modulation.% In this technique the resonance lines of an elemen ATOMIC ABSORPTION SPECTROSCOPY 189 emitted by a d.c. operated high intensity lamp are selectively modulated by passing them through a pulsating cloud of atoms of the same element produced by an ax.operated hollow cathode. The resonance lines of the element to be determined are selectively modulated and measured by the ax. detector. None of the other lines is modulated and hence is not measured. This technique permits the use of much larger slit-widths and less exacting monochromators even simple filters may suffice. KoirtyohannM pointed out that a severe limitation of selectively modulated lamps was the extra power supply required for their operation. LoweS5 described an improved lamp of this type that did not require any extra electrode the cloud of atoms producing the selective modulation being obtained by a series of short, high current pulses superimposed on the steady direct current.Van Rensburg and Zeeman56 described the determination of gold platinum, palladium and rhodium by using a multi-element high intensity lamp with selective modulation. Butler and Brink57 prepared selectively modulated lamps for several different elements and found that the analytical curves were steeper and more linear than with conventional lamps. Also neither the band-width nor slit-width was critical even for line-rich elements such as iron and nickel. Dawson and Ellisso described a method to increase the intensity of hollow-cathode lamps. They superimposed a large pulsed current on to the low steady direct current and obtained up to several hundred fold gain in intensity with no deterioration in absorption signal.In this fashion the working life of the lamp was extended considerably and the shelf-life also appeared to be increased. Again an extra power supply is required. Neither of these modified high intensity sources has become commercially available although Koirtyohann5* considered that the latter technique50 might find use in atomic fluorescence measurements where the added intensity is so necessary. An interesting use of pulsed hollow-cathode lamps in a multi-channel atomic fluorescence device has in fact recently been described by Mitchell and Johanss0n.~8 A relatively large number of papers have appeared dealing with the use of multi-element hollow-cathode lamps. In common with the single-element lamps, changes in design have led to improved lamps with brighter emission improved stability and a longer life.59 One still obtains less light per element than with the single lamps but this should now be at an acceptable level.A good quality monochromator may be required to avoid spectral interference with some combi-nations of elements e.g. iron cobalt and nickel. A further disadvantage is the possibility of one element sputtering off before another. For example a sodium -potassium lamp may become predominantly a sodium lamp after a time. There is an obvious advantage when the final cost is less per element and the time saving by eliminating lamp changes from element to element should be considered, although it will not now be as important with modern instrumentation offering lamp ‘warming-up’ devices and hence instant use.In 1966 Slavinl said lamps were provided that combined several elements without loss of performance and that a large variety of such combinations woul 190 PLAT" eventually become available. Yet in 1968 Manningg was not so optimistic. His experience had not shown the multi-element lamps to have the previously expected economic and convenience advantages. He considered that although the con-venience might outweigh the disadvantages for some users the single-element lamp was recommended for most analyses. Jaworowski and Weberling6O found an occurence of apparent spectral interference in the use of multi-element lamps and suggested that a high dispersion monochromator should be used to eliminate the effect. Galassi and Hell*l used a multi-element lamp for barium calcium magne-sium and strontium and found interference with the calcium line at 422.7 nm by a secondary strontium line at 421.6 nm although the calcium line could be isolated by the use of a narrow slit-width.Analytical results compared favourably with those for a single-element calcium lamp. Frank et aLB1 studied the use of an iron hollow-cathode lamp as a multi-element source for magnesium manganese nickel, copper barium and silver. They concluded that an effective intermediate sensitivity was obtained. Heneagea2 found that when using a five-element lamp for chromium, cobalt copper manganese and nickel the difference between the results obtained using this or single-element lamps was usually very small. - I By 1969 Fernandez et ~ 1 . ~ ~ were able to report that multi-element lamps offered a reduction in lamp cost per element and although the emission intensity from the individual elements was not as great as that from single-element lamps, the lower intensity had little or no effect on performance.The sensitivities and detection limits obtained with various multi-element combinations were about equal to those obtained with single-element lamps. In 1969 Barnett and Kahnl7 found that a multi-element lamp for aluminium silicon calcium magnesium iron, copper and zinc gave a level of analytical performance with an ease and speed that indicated that this lamp suffered less by comparison with single-element lamps than had been supposed. Cooke12 suggested that difficulty may be experienced in selecting the required line when using a six-element lamp.A number of manufacturers offer multi-element lamps the calcium - magne-sium one in particular and as they axe now capable of a good performance they should certainly be given serious consideration when new or replacement lamps are required. A number of other hollow-cathode lamp modifications such as molten ~athodes,3~*~ magnetic controls4 and demountable typesap66 have been used by various workers but there is as yet no evidence of commercial development. Fassel et d.67 evaluated spectral continua as primary sources and the results obtained showed that these sources could be successfully used in atomic absorption spectroscopy. For thirty-two elements the sensitivities observed were either comparable to or exceeded those then obtained (1965 to 1966) with hollow-cathode sources.These workers suggested further refinements such as greater resolution in the spectrometer better continuum primary sources lower noise flames and e'lec-tronic integration of the absorption signal. DeGalan et ~ 1 . 6 8 concluded that with a good medium dispersion monochromator a continuous source offers several distinct advantages e g . the possibility of qualitative analysis simplicity of backgroun ATOMIC ABSORPTION SPECTROSCOPY 191 correction and low cost compensating for the higher priced high resolution mono-chromator. Also it yields detection limits that are approximately the same as those obtained with a hollow-cathode lamp. McGee and Winefordner@ used a continuum source in conjunction with an extended flame cell an argon - hydrogen -air flame and a medium dispersion monochromator.They found this system to be competitive with a hollow-cathode source and a typical air - acetylene flame system. Despite these apparent advantages the continuous source has not developed commercially. Kahn42 attributed this to the severe instrumental difficulties associated with it while Slavinll and Willis& pointed out that the calibration curves are so bent as to be useless at high absorbance and not much better even at fairly low absorbance values. Dagnall and West70 considered that the continuous source had only been used to demonstrate an academic point. The use of microwave-excited electrodeless discharge lamps (E.D.L.’s) as an alternative source to hollow-cathode lamps in atomic absorption spectroscopy has recently been propo~ed.3~~~~-73 This type of source has been known for many years and used mainly for studies of spectral structure.The reason for this present development is to be found in the search for more intense sources to produce better linearity in calibration curves and a higher signal-to-noise ratio. E.D.L.’s should be much less expensive to produce than high intensity or high spectral output hollow-cathode lamps which were developed for similar reasons.7o They do, however require a micro-wave generator and cavity which will add considerably to the cost of equipment. Since these initial publications conflicting views have been expressed about electrodeless lamps. Some workers have indicated that they can be used satisfac-torily while others have said they are much less stable than hollow-cathode lamps and cannot be recommended for atomic absorption spe~troscopy.7~ Dagnall and West70 in 1968 concluded that the application of electrodeless discharge lamps in atomic absorption spectroscopy is at present only in its early stages and many of the experimental parameters involved in their preparation and operation have not been fully examined.However from the results that have recently appeared in the literature these authors suggest that there is little doubt that E.D.L.’s will occupy an increasingly important position in the future. The lamps prepared in their laboratory were primarily used for atomic fluorescence spectroscopy. also in 1968 reported that a comparison of the intensity of electrode-less discharge lamps and hollow-cathode lamps showed the former to be more intense in nearly every case but owing to the greater line-width produced the absorption was never greater.The lower sensitivity obtained for the electrodeless lamps was attributed to self-reversal. In 1969 W~odward~~ reported that electrode-less discharge lamps did not have the same reliability as hollow-cathode lamps but, nevertheless they proved extremely useful for analysis by atomic absorption spectroscopy. In several instances low analytical sensitivity was obtained but electrodeless lamps had been prepared and applied regularly during the previous 9 months for the atomic absorption determination of over twenty elements in a wide variety of samples 192 PLATT Slavin and Slavin2 considered that the published literature tended to over-simplify the problems associated with electrodeless lamps and observed several practical problems although they confidently expected an eventual improvement.They found an important difficulty in that many of the lamps produced were very unstable with time and concluded that the utility of atomic fluorescence probably depends upon successful improvement of electrodeless discharge lamps. Double-beam instrumentation will be an advantage in accounting for base-line drift. Reynolds31 also points out that the very high output intensities make E.D.L’s ideal for atomic fluorescence measurements and they could even replace some of the less efficient hollow-cathode lamps for atomic absorption analysis.Headridge and Richardson7* compared electrodeless discharge lamps with hollow-cathode lamps for atomic absorption spectroscopy and concluded that for many determinations where the highest precision is not required electrodeless discharge lamps are satisfactory as light sources and are much cheaper to make than hollow-cathode lamps. As expected the detection limits are poorer for the electrodeless lamps. Browner Dagnall and West76 found that the performance of electrodeless lamps for lead mercury silver thallium and tin both modulated and unmodulated, was comparable or better than the respective hollow-cathode lamp run under optimised conditions. These workers concluded that previous work had shown the advantages of electrodeless lamps in the short wavelength regions for such elements as arsenic and selenium and there were other instances where they gave rise to considerable increases in sensitivity and a wider working range compared with hollow-cathode lamps.For the mercury electrodeless discharge lamp the better spectral profile of the emitted resonance line gave these advantages. Fisher and Hayward in a recent pape*2 said that probably the most frequent criticism of electrodeless discharge lamps was on the grounds of instability. They found that lamps run in a high intensity mode gave poor analytical sensitivity and when this was improved by reducing power instability increased and it was not possible to match the performance of hollow-cathode lamps. However by using a tunable cavity adjusted to minimise reflected power from the cavity these workers were able to improve lamp performance for arsenic selenium bismuth, gallium germanium thallium tellurium and aluminium lamps such that the sensitivity was better than that obtained with hollow-cathode lamps.This was achieved at the expense of lower intensity. They studied arsenic and selenium lamps in detail and concluded that the performance of electrodeless discharge lamps as atomic absorption sources compares favourably with that of hollow-cathode lamps for these elements in particular. Comparable stability better analytical sensitivity and better detection limits were obtained. Electrodeless discharge lamps retail for about half the price of hollow-cathode lamps but when the time and facilities are available they can be ‘home-made’ easier than can hollow-cathode lamps and for much less than the retail price.The commercially available lamps are now said to have a virtually unlimited shelf-life and shorter warm-up periods than the earlier lamps ATOMIC ABSORPTION SPECTROSCOPY 193 It appears that these sources will develop more for atomic fluorescence measurements but certainly anyone interested in determining elements with resonance lines in the ultraviolet region e.g. arsenic and selenium should consider their use for atomic absorption spectroscopy and compare them as far as possible with the improved hollow-cathode lamps that are now available for these elements. Nebuliser - atomiser systems To ensure efficient evaporation of the sample solution it must be introduced into the flame in a finely dispersed state.To carry out this operation pneumatic nebulisers are invariably used.13~14 In this system the supporting gas which is usually air or nitrous oxide aspirates the liquid sample and converts it to a mist. The mist may be either admitted directly into the flame as with total consumption burners or it may reach the flame via a cloud chamber where the larger droplets of liquid are deposited and rejected. The latter system is known as a pre-mix one and it allows the fine mist to be led through a burner slot of convenient length to produce a long path flame cell for absorption.ls The cloud chamber system is relatively inefficient in that only about 10 per cent. of the original sample volume sprayed reaches the flame but it has the advantage of sustaining a highly stable laminar flame the characteristics of which are influenced very little by the sample mist except when non-aqueous solutions are sprayed.12 Willis20 studied the mode of operation of the nebulising system and concluded that the sensitivity of the instrument and the effect of chemical interferences in the flame are critically dependent on the construction of the nebuliser and particu-larly on the rate of liquid uptake.Reynolds31 considered that at the present time (1969) the capabilities of all commercial nebulisers are very similar and it is in this sector that most workers recognise that improvements in the performance of atomic absorption equipment need to be made. RubeSka14 also expects further develop-ments in atomic absorption to depend on the improvement of methods for produc-ing atomic vapours.Rains25 suggests that the nebuliser - burner system is probably the most important component in any absorption measurement. Some workers have tried to avoid the inefficient waste of 90 per cent. of the solution when using pre-mix burnem2 Perhaps the most commonly tried method is the use of an ultrasonic nebuliser which will convert a larger fraction of the solution to the fine mist utilised in the pre-mix burner than will the common pneumatic type of nebuliser. Most workers have found that a smaller amount of sample will produce a given signal i.e. improved efficiency but that it is difficult to convert this into a real improvement in detection limits. Slavin and Slavin2 confirmed the general findings of Hoare et aL7' in that about one fifth of the sample up-take rate produced about the same absorption signal for some fifteen elements, but it was not possible to increase the uptake rate of the ultrasonic nebuliser to take advantage of this improvement in efficiency.West78 stated that in spite of the apparent advantages associated with ultra-sonic nebulisation the technique has not become widespread perhaps because o 194 PLATT the inconvenience of sample changing and clean-up of previously used systems. Several worker^^^-^^ improved the design to make sample changing more con-venient. Stupar and Dawsonl9 studied the theoretical and experiment+ aspects of the production of aerosols by both pneumatic and ultrasonic means. They concluded that unless operated at the high frequency (>ti00 kHz) and high power necessary to generate fine mist at sample flow-rates of 1 to 5 ml min-l the sensi-tivity and interference is worse than that obtained with a pneumatic nebuliser.Nevertheless at the time (1968) the ultrasonic type offered a practical means of increasing the efficiency with the possibility of reduced interference. However the usefulness of the ultrasonic nebulisers has remained remarkably disappointing2 and there appears to be no advantage in commercial development on the part of the instrument manufacturers although of course an ultrasonic generator may be purchased separately if so desired. Other methods used to improve the efficiency of the pre-mix system include heating the support gas before mixing and heating of the cloud chamber.Rawson81 and Riley and Taylors2 used the former technique and obtained increases in effi-ciency of up to 16 and 3 fold respectively. Heating of the cloud chamber has been used by a number of and at least one instrument manufacturer incorporates this system into commercially available equipment .86 Gray and Gallwas87 compared the use of a laminar flow high solids burner with a heated chamber laminar flow burner for the determination of serum calcium and con-cluded that the heated chamber type was less convenient for use with high solids but was nevertheless much more sensitive. Heating the support gas or the cloud chamber causes the sample mist to be dried and hence concentrated before entering the flame thus obtaining increased efficiency.Unfortunately according to WillisJM the increase is usually accompanied by some instability or drift in the readings and of carry-over or memory from one sample to the next. Three methods of introducing the sample into the flame without nebulisation have been reported. Venghiattisss described a method for the atomisation of solid samples. He did not claim that solid sampling would replace the conventional nebuliser type of sample introduction but showed that for some elements at least, the sensitivity attained was greater than with the conventional method. A solid sampling system is commercially available from one instrument manufacturers9 and has been successfully used for the determination of copper nickel gold silver, mercury lead and bismuth. The refractory elements e.g.aluminium vanadium, titanium silicon and the rare earths cannot at present be determined by this method. The sampling boat technique was described by Kahn et aL90 and was designed to get the maximum detection limit with small amounts of sample. The boat is constructed of tantalum and holds about 1 ml of sample which is dried by holding it near the flame and is then passed into the flame for atomisation. Quite dramatic improvements in detection limits were obtained for easily atomised elements such as arsenic zinc lead and mercury. Curry et aL9l applied the sample boat technique to the determination of thallium in biological material and found an increase i ATOMIC ABSORPTION SPECTROSCOPY 195 sensitivity of at least twenty five times over conventional methods ; sensitivities for lead and cadmium were also increased.The same workers reported that in contrast to conventional atomic absorption analytical methods for blood and urine, the results obtained with the tantalum boat are susceptible to inter-element interferences and calibration by the method of standard additions (see section dealing with ‘Experimental Considerations’) is essential for accurate quantitative results. However they also noted the advantages of rapid quantitative analysis and minimum sample pre-treatment which made the technique ideal for screening large numbers of samples for such toxic elements as thallium lead mercury and cadmium while using very small samples of blood and urine of 50 to 200 pl. The third technique adopted by a number of workers employed a wire to trans-port the sample into the flame.White92 supported the sample on a loop of platinum wire and mechanically introduced it into the flame; the atomic vapour thus pro-duced is directed into a nickel tube adsorption cell to increase the sensitivity and duration of the measurement. The principle application of this technique was the determination of lead in blood when an ear or a finger prick of 0.1 ml can be used rather than the more usual 5 ml extracted intravenously by syringe. Delvesg3 combined the techniques of Kahn and White and provided a rela-tively simple but accurate method in which the tantalum boat was replaced by a nickel micro-boat in association with an absorption tube as used by White. The concentration of lead in lo-$ samples of whole blood was accurately and rapidly determined.Along with the continued reign of the hollow-cathode lamp as the main primary source the flarne remains as the most commonly used atomiser. Price13 states that the combustion flame is one of the most convenient and efficient atom producing plasmas. Rubeska14 considered that the use of chemical flames has enjoyed the greatest popularity partly because of a carry-over from the emission technique and partly because of its simplicity and the ease with which it can be applied and controlled. The problem of atomisation according to W i l l i ~ ~ ~ is central to the whole atomic absorption technique and although several other techniques have been tried (see later) the use of a suitable flame still remains as the method favoured by the overwhelming majority of workers in the field.While the flame has certain disadvantages the equipment required is simple inexpensive , easy to use and well adapted to the rapid measurement of a series of different solutions so that it is unlikely that any other method of atomisation will replace it for the great bulk of analytical work. Slavinll also considered that the advantages of the flame are so compelling that it is unlikely any other technique will com-pletely replace it although he consideredzit to be the residual problems and limita-tions to a flame that have initiated a large number of research programmes of considerable interest. WestS made it clear that virtually the only atom reservoirs in use are flames although he said there are many factors indicating that other media may be more advantageous and these will undoubtedly come into use in the future.Apart from the different gas mixtures used there are two distinct types o 196 PLATT flame. In the first which is called the pre-mixed laminar flame and has a well defined inner cone or reaction the support gas converts the sample solution to a mist that passes into the cloud chamber where the larger droplets settle out and go to waste. The fuel gas may be added either before or after the chamber and the mixture of fuel gas support gas and sample mist proceed to the burner where it is burnt.94 Most workers agree that for atomic absorption work the laminar flow burner is ~ u p e r i o r ~ ~ J ~ * ~ ~ and it is significant that almost all commercial instruments currently produced use this type.The second type results from the fuel gas and the supporting gas not being mixed until the point at which they enter the flame and the sample solution is also introduced at this point. This one is called a total consumption system and is a combination nebuliser - burner. It is necessarily circular in design and cannot conveniently be used to provide a long absorption path for atomic a b s o r p t i ~ n . ~ ~ ~ ~ ~ At first sight one might expect total consumption burners to be more efficient and might question why pre-mix burners are used at all. However numerous examples from the literature have shown interferences to be present in the total consumption flame that are absent from pre-mixed flames2?31 With a total con-sumption burner a large proportion of the solution goes straight through un-evaporated cooling the flame and leading to poor sensitivity.Cowley et aLg5 indicated that a large part of the problem associated with turbulent flames results from disruption of the suitable chemical environment found in the well defined zones of the pre-mixed flame. Mossotti and Duggang6 found improvement in over-all performance when using a total consumption burner for emission measurements in which the gases were pre-mixed and produced a more laminar flame with well defined zones. Hence there appears to be no need for total consumption burners to be limited to turbulent flames. The total consumption burner is unlikely to flash-back as there is no large chamber containing a potentially explosive gas mixture but modem pre-mix burners are designed so that the chance of flash-back is minimal.At present the long path pre-mix burner is superior for atomic absorption purposes because it gives good sensitivity acceptable signal-to-noise ratios has remarkably little memory and excellent quantitative stability. The relatively low temperature of the flame(s) was one of its major disadvan-tages because it strictly limited the number of elements that could be determined.14 The air - propane flame at a temperature of about 1925 "C gives the best sensitivity for the alkali metals and for certain other elements forming compounds that are not thermally stable. The most commonly used and generally useful gas mixture is air - acetylene because the temperature and reducing conditions within this flame can be varied within fairly wide limits.For most routine determinations it is operated in a non-luminous manner at about 2300 "C while the luminous mode, providing reducing conditions is suitable for some of the elements forming refrac-tory oxides including tin barium chromium and molybden~m.~?~~J~?~7 However, a large number of elements forming refractory oxides require a much higher tem-perature than can be provided by acetylene burning in air. Higher temperature flames are provided by acetylene or hydrogen burning i ATOMIC ABSORPTION SPECTROSCOPY 197 oxygen and by acetylene burning in a mixture of oxygen and nitrogen. However, owing to the high burning velocity of these mixtures they are prone to flash-back and require special burners with narrow slots which easily become clogged by the crystallisation of salts or the formation of ~ 0 0 t .l ~ ~ ~ ~ Hence they are not as easy and safe to use as are air - acetylene flames. The problems involved in progressing to a high temperature flame were over-come most satisfactorily by the nitrous oxide - acetylene flame described by W i l l i ~ . ~ ~ It is predicted that this system will probably retain the privileged position of chemical flames in atomic absorption spectroscopy for at least a few years to c ~ m e . ~ ~ l ~ The development of the nitrous oxide - acetylene flame and suitable burners for its use resulted in the extension of the number of elements readily amenable to atomic absorption methods from thirty-five to about sixty-five which is approaching the theoretical maximum.loO It can thus be readily appreciated that the introduction of this flame was a major step forward in atomic absorption analysis.Nitrous oxide - acetylene has a low burning velocity similar to that of air -acetylene combined with a high temperature (of about 2900 "C). Measurement of the temperature was discussed by deGalan and Samaey,lo2 who obtained values agreeing well with literature data.lol WillisM and Rubeska14 state that the high temperature obtained is caused in part by the energy liberated by the decomposi-tion of the nitrous oxide. The burning velocity is limited by this decomposition reaction so that it is quite safe to work with the mixture when using standard pre-mix systems.Kirkbright et aZ.lo3 investigated the reactions that occur in the nitrous oxide - acetylene flame and concluded that these are more complex than can be explained by the reducing action of incandescent carbon particles. It appears that nitrous oxide decomposes to provide oxygen at the primary reaction zone and hence raises the temperature and burning velocity of the flame. Some of the nitrous oxide possibly via nitric oxide appears to react directly with fuel molecules in the primary zone to produce more -CN and -NH radicals than are produced in conventional hydrocarbon - air flames. A reducing atmosphere is thus provided to protect the metal atoms. Amos and Willis98 in describing the use of high temperature pre-mixed flames considered that the new flames apart from allowing the addition of some twenty-five metals to those already determinable possessed the further advantage of permitting the determination of the alkaline earth metals which are only partially atomised in cooler flames with higher sensitivity and greater freedom from chemi-cal interference.The higher temperature removes the need for the presence of lanthanum which was commonly added as a releasing agent when determining alkaline earth metals in the air - acetylene flame although it may still be used as an effective ionisation buffer in the latter flame. Potassium is more often used in this c a p a ~ i t y . ~ * J ~ It may still be true that air - acetylene is the most generally useful gas mix-ture but nitrous oxide - acetylene which has normally been considered most use-ful for elements forming refractory compounds is often found to have advantage 198 PLATT for a number of other elements normally determined in an air - acetylene flame and it is interesting to note that chemical interference is largely eradicated52 and better linearity of calibration data is often obtained.In general the hotter the flame the less chemical interferen~e.~ However because ionisation will be greater in the hotter flame the sensitivity will be less and whenever this is important a31 ionisation buffer should be added. In actual fact the 50-mm slot solid stainless-steel burner normally used for nitrous oxide - acetylene often gives adequate sensitivity when burning air - acetylene and for many of the analyses involving both gas mixtures it may not be necessary to change burner heads98 One of the worst faults of the nitrous oxide - acetylene burner head was its tendency to carbon up particularly when operated fuel rich.This fault has now been removed by re-designing the burner head. On the rare occasions when flash-backs do occur it is usually when the flame is being ignited or extinguished. Pampello5 described a method of avoiding flash-backs by using air as the oxidising gas during ignition and turning off. There is a difference of opinion among the instrument manufacturers about the necessity of this safety procedure some including it and others favouring direct ignition of nitrous oxide - acetylene. Should a flash-back occur modern instrumentation is designed so that no serious damage results.In 1966 Slavin et aZ.lo6 reporting recent experiences with the nitrous oxide -acetylene flame said that the technique had proved to be completely practical and many laboratories were utilising it on a routine basis for a number of elements. Nonetheless the method was still very new and much development work remained to be done. These workers studied chemical and ionisation effects and the effects of varying slot dimensions in relation to the determination of a number of elements including calcium barium silicon and titanium. Later in the same year NlanninglO7 listed the operating conditions together with sensitivities and detection limits for twenty-eight elements determined in the nitrous oxide - acetylene flame. The same author discussed effects caused by chemical interferences ionisation and flame noise.Walsh1°8 pointed out that it must not be expected that this new flame will completely eliminate chemical interferences in all analyses. West3 considered that the hotter flame had made the study and determination of elements such as aluminium molybdenum and niobium relatively easy and that further progress will no doubt improve quite considerably the sensitivities obtain-able. In 1968 Willis9* reviewed the features successes and remaining limitations of the nitrous oxide - acetylene flame. He traced the development from low tem-perature to high temperature flames and indicated that the nitrous oxide - acety-lene system had been adopted by the manufacturers of commercial equipment as the standard method of atomising metals that are not satisfactorily atomised in the air - acetylene flame.Willis discussed ionisation effects chemical interferences and inter-element enhancement effects in the use of high temperature flames in chemical analysis. He concluded that the development of high temperature flames for use in atomic absorption has not only provided a powerful technique fo ATOMIC ABSORPTION SPECTROSCOPY 199 the solution of analytical problems but has also led to a more critical consideration of the factors determining the performance of flames as atomising systems. Bowman and Willislo9 presented results showing the usefulness of the nitrous oxide - acetylene flame in the analysis of rocks oils minerals and steels. They studied several effects that need to be considered and concluded that any metal with an ionisation potential of below about 6.5 V for example aluminium and several of the rare earths will be affected to a significant extent by ionisation if this is not suppressed.Also the atomisation of some metals such as titanium is incomplete to an extent that is strongly dependent on the other substances present. In 1969, Shifrin et aZ.l1° reported recent results with a nitrous oxide burner showing detec-tion limits better than other values currently reported in the literature for several elements. These workers used hot and cold cloud chamber operation with aqueous and organic solvents. Also in 1969 Slavin and Slavin2 reported that the nitrous oxide - acetylene flame had made it possible to determine almost all metals with detection limits that are usually better than 1 pg ml-l.Finally a useful practical point is suggested by Julietti and Wilkinson.lll At the normal high flow-rate of nitrous oxide the expansion of gas in the regulator causes considerable cooling which may have an appreciable influence on the per-formance of the regulator. This was overcome by placing a 100-W light bulb a few centimetres above the regulator. Most atomic absorption work has been carried out with acetylene flames supported by air or nitrous oxide. Butler and Fulton112 and Fleming115 investi-gated the use of an acetylene flame supported by a mixture of air and nitrous oxide. This system allowed a continuously variable flame temperature between the lower air - acetylene and the higher nitrous oxide - acetylene.It was flexible and safe but the results seemed to indicate that for most applications there were no advantages significant enough to justify the additional complexity of gas mixings Acetylene burning in nitric oxide produces a flame that has a slower burning velocity and a slightly higher temperature than has nitrous oxide - acety-lene. However nitric oxide is relatively expensive and not readily available. Amos and WiUisg8 considered further investigation of this flame might be worth-while but Slavin et aZ.1°6 indicated no significant advantage in its use. Hence its price availability and its corrosive and toxic nature have ruled it out for atomic absorption spectroscopy. Temperatures higher even than the nitrous oxide -acetylene flame would probably limit their usefulness by the loss through ionisation of neutral atoms.s* Other pre-mixed flame types may however offer advantages for some ele-ments.The argon - entrained air - hydrogen flame has much the same characteris-tics as the air - hydrogen one and both are of use in determining elements having their resonance lines below 220 nm and when lower temperatures are required to avoid interference effects. Air - acetylene and air - propane flames absorb about 60 per cent. of the resonance radiation from elements like arsenic and selenium whereas argon and air supported hydrogen flames are considerably more trans-parent at the lower wavelengths required for these elements and only absor 200 PLATT about 15 per cent. of the radiation. A number of workers have recommended these flames particularly for arsenic and s e l e n i ~ m ~ ~ ~ - - ~ ~ ~ and for some other ele-m e n t ~ ~ ~ ' - ~ ~ ~ The sensitivity and detection limits for these elements are usefully increased but unfortunately owing to the lower flame temperature interferences are also liable to increase at the same time.A multi-slot burner head supporting three parallel flames provides the most satisfactory performance with these cooler flames presumably because the central flame is protected from disturbances to a greater degree than with a single-slot burner.116s118 Reynolds4 considered that tin is best determined with the air - hydrogen flame. Dagnall et ~ 2 . l ~ ~ showed the strongly reducing high temperature nitrous oxide - hydrogen flame to have con-siderable promise as an atom reservoir for a wide range of metals whereas Willis et found data that did not support this evidence and suggested only a limited usefulness for this flame compared with the nitrous oxide - acetylene system.The multi-slot burner was first described by Boling122 and has since been widely adopted as a burner capable of coping with solutions of high solids content. As already mentioned the protected inner flame is used for absorption and this leads to the marked reduction of flame noise. A more recent development that has found commercial application is the use of inert gas shielded or separated flames. Kirkbright et ~ 2 . l ~ ~ described an air -acetylene burner that was shielded from the atmosphere by a flow of nitrogen and the same workers also described a nitrous oxide - acetylene burner that was shielded by argon or nitrogen.12* The inert gas lifts off and separates the secondary diffusion zone leaving the primary and hottest zone for viewing with considerably less interference.Separation of the air - acetylene flame is particularly beneficial for elements with resonance lines below 200 nm where atmospheric oxygen entrained in the secondary zone of conventional flames absorbs strongly. In the determina-tion of arsenic and selenium improved detection limits result from the lower background ab~orpti0n.l~~ Separation of the nitrous oxide - acetylene flame pro-vides protection of the reducing atmosphere in the interconal zone and results in a larger population of ground state atoms of the refractory oxide forming elements.This is most useful in the determination of aluminium boron and silicon when a threefold increase in detection limit re~ults.1~~ During the period covered by this review Rubegka and Moldan have reported further investigations with long path absorption Considerable increases in sensitivity result and the best sensitivities are achieved under fuel rich condi-tions when the tube shields the flame gases from the oxidising atmosphere and separates the flame into primary and secondary reaction zones along the horizontal axis. Despite their almost exclusive use flames possess several disadvantage~,~~8~~~ and a number of workers have used various techniques for the flameless atomisa-tion of samples. Slavinll briefly discusses the non-flame methods none of which has as yet found any commercial application although it is possible that one or two may be offered as accessories in the future.L ' v o Y ~ ~ ~ used a lined graphite furnace electrically heated in an argon or nitroge ATOMIC ABSORPTION SPECTROSCOPY 201 atmosphere to produce atomic vapour from samples that are in either solution or powder form. Massrnannl3* also investigated the atomisation of samples in a graphite crucible heated in an argon atmosphere. West and Williams128 and Anderson et a1.l3l described an electrically heated carbon filament atom reservoir. This is a simple and efficient device which can readily be fitted or adapted for use with most commercial atomic absorption equipment and is capable of completing an analysis within 5 s and repeating it every 2 min with the filament having no memory effects and also being self-purging.These methods can all analyse very small samples with spectacular detection limits. Further work with high temperature furnaces has been described by a number of a ~ t h o r s . l ~ ~ - l ~ ~ Other flameless techniques used include 1asers,137$138 plasma t o r ~ h e s l ~ ~ - l ~ ~ and hot-wire amalgams or c o a t i n g ~ . l ~ J ~ ~ Monochromators The monochromator in an atomic absorption instrument does not fulfil quite the same function as it does in other types of spectrophotometer. As the source produces a series of monochromatic wavelengths the monochromator serves only to select the resonance line from unwanted radiation.52 Either a prism or a diffraction grating may be used as the monochromator, although commercial instrumentation now appears to favour the use of gratings.Reynolds31 discusses the general features and factors affecting the performance of monochromators and concludes that there is more to recommend a grating instru-ment than one utilising a prism. The same author also states that for most common metals a very moderate monochromator is quite suitable but for metals producing complex spectra or for which the sources possess poor output a better mono-chromator is necessary. Owing to the vast improvements in lamp technology the latter group of metals is few and diminishing in number. A new type of monochromator was described by Sullivan and W a l ~ h l ~ ~ and depends for its action on the absorption and re-emission of resonance radiation by the atomic vapour of the element being determined.The atomic vapour may be produced by the electrical heating of a small block of the particular metal or for metals of higher melting point and lower vapour pressure by cathodic sputtering as in a hollow-cathode lamp. Sullivan and W a l ~ h l ~ ~ discuss the applications of resonance monochromators and point out the favourable case of the Group I1 elements which have a resonance spectrum consisting of one line only. They mention the possible advantages in the design of instruments for the simultaneous determination of several elements. The same authors and also W i l l i ~ ~ ~ list the major advantages of these systems com-pared with dispersion types. The main advantages are that the effective resolution at about 0*001 nm enables the flame noise compared to the signal to be far less than that obtained with a dispersion monochromator where the slit-width is nearer 0.1 nm.A resonance monochromator is permanently tuned to the wavelength of the resonance lines it is designed to isolate so it can therefore withstand far more rigorous conditions and should be ideally suited to continuous routine use o 202 PLATT in-line systems. A conventional monochromator on the other hand is subject to thermal or mechanical drift. Constant calibration data may be facilitated by the over-all control that the monochromator has over the effective resolution this being independent of the resonance lines emitted by the source. Disadvantages include a limited life loss in sensitivity with increasing com-plexity of the resonance radiation spectrum and the possibility of having to use high apertures to achieve an adequate signal.A number of workers in the field have used resonance monochromators,~4~-1~0 but only one manufacturer has included them in commercially available equip-ment.151 Whether or not they become more generally accepted depends very largely on manufacturing Detector and read-out systems Almost all atomic absorption instruments are now fitted with photomultiplier detectors which enable the widest possible range of metals to be determined.44 Apart from the light of a particular wavelength originating from the source there will also be light of the same wavelength arising from the flame falling on the detector.It is necessary to distinguish between these two radiations as it is imperative that only that from the source is measured. This requirement is met by modulating the source radiation either mechanically or electronically and tuning the detector to the particular frequency of m~dulation.~~ Conventionally after detection the signal is amplified and displayed on a meter. Single-beam instruments have microammeters calibrated for transmission and absorbance measurements whereas double-beam instruments generally use null-point systems. Alternatively a chart recorder may be used to provide a permanent record and to display any signal fluctuations. In addition some type of recording device is essential for use with automated multi-sample handling.For atomic absorption spectroscopy it is convenient to use a logarithmic recorder that will automatically provide a read-out linear in absorbance units and hence providing the Beer - Lambert law is obeyed a linear concentration scale. It is even more convenient although no more accurate to use one of the digital concentration read-out or print-out devices offered by some manufacturers. If the Beer-Lambert law is not obeyed it is desirable to correct for the curvature a relatively difficult requirement. Kah1115~ described an instrument that provided a correction of the deviations from Beer’s law and a read-out directly in concentration. Integration techniques may be applied to the amplified signal either to pro-duce a steady noise-free r e a d i r ~ g l ~ ~ ~ or to accumulate low absorbance signals and recover them from the noise signal thus improving the signal-to-noise ratio.l= The majority of instruments have a scale expansion facility to increase small read-ings and improve the precision of measurement.For the determination of macro constituents an instrument with the ability to expand selected portions of the scale rather than the whole range from zero upwards is most useful. This is known as ‘zero suppression’ and when used in conjunction with an integrating system, gives most impressive results.36,1s ATOMIC ABSORPTION SPECTROSCOPY 203 W i l l i ~ ~ ~ considered there was no doubt that the growing demand for large numbers of routine analyses will lead to a widespread application of automated sampling techniques in conjunction with digital read-out and print-out facilities.To cope with the large amount of information resulting from automatic instrumen-tation computer techniques were described by Ramirez-Mufioz et ~ 1 . l ~ ~ More recently Wendt15s has described the application of a computer programme to handle all the calculations involved in converting percentage absorption readings into sample concentrations. Malakoff and c ~ - w o r k e r s ~ ~ ~ J ~ ~ have also described the application of computer techniques to atomic absorption methods. SlavinlG0 discussed the acquisition of data on a typewriter read-out which automatically punches paper tape for computer processing and this type of equipment has recently been announced comrner~ially.~~~ Automation According to Slavin ,11 the time required to introduce the prepared sample and record the signal is a small part of the analytical time and automation would not be a great practical advantage.Far more important is the time required to convert the absorption signal to actual concentration data and the time required to dilute the sample to the optimum analytical range. Slavin describes an apparatus for a fully automatic analysis that includes introducing diluting and mixing the sample , and the result in concentration terms is presented on a paper strip with sample identification numbers. Similar equipment to that described by Slavin is available commercially from at least one ma,nufacturer.lsl Dawson et aZ.162 described an automatic high speed scanning multi-channel instrument for the determination of sodium potassium calcium and magnesium in clinical samples by simultaneous emission and absorption measurements.Partial automation is readily available in the form of automatic sampling devices and automatic sampling plus dilution has been accomplished by connecting a Technicon AutoAnalyzer system to the nebuliser of an atomic absorption i n ~ t r u m e n t . l ~ - ~ ~ ~ A number of workers have recommended the use of automatic d i l u t e r ~ ~ ~ ~ J ~ ~ and syringesle8 to speed up solution preparation prior to atomic absorption measurement. When a manually operated instrument becomes overloaded with work the type of work involved will decide whether automation or a second manually operated instrument is the answer. Automation is worth considering only when the work consists of similar samples for a small number of elements.lel Reagents for atomic absorption spectroscopy With the rapid development and increasing popularity of the atomic absorption technique there was a need to consider the availability of reagents with a special degree of purity for use in the preparation of samples for analysis by atomic absorption spectroscopy.With this aim in mind a reagent sub-com-mittee of the Atomic Spectroscopy Group was set up in 1967 to approach the major reagent manufacturers. 204 PLATT The specifications of a list of reagents widely used as releasing agents and ionisation buffers and also inorganic and organic solvents was circulated to the suppliers. These reagents are used in large amounts relative to the concentration of elements actually being determined and need to be in a high state of purity with respect to those elements.Two further types of reagents were also considered. One was the supply of stock solutions of elements of accurately known strength and as such solutions are much diluted in use the accent was on accuracy rather than purity. The second consideration was for the provision of suitable chelating agents for the extraction of metals into organic solvents. Amongst the latter, ammonium pyrrolidine dithiocarbamate has been widely recommended for the extraction and concentration of heavy metals. Hopkin and Williams Ltd. have published a monograph on the use of ammonium pyrrolidine dithiocarbamate,leg Excellent mutual progress was made and it is now possible to purchase from B.D.H.Chemicals Ltd. and from Hopkin and Williams Ltd. a substantial range of the reagents mentioned above. Additionally Kodak Ltd. market in the United Kingdom a range of atomic absorption reagents from the Fischer Scientific Com-pany of America and Johnson Matthey offer their Spec. Pure Chemicals. For the analyst who prefers to make his own standard solutions or for those not commercially available Roth17* provides details for the preparation of 1000 p.p.m. solutions for sixty-five elements. Standard stock solutions (not lead) should be stored in polythene bottles and should be freshly diluted into standard working solutions each day. Experimental Considerations According to Ramirez-Muiio~,~~~ atomic absorption like most instrumental methods requires several conditions to be fulfilled to guarantee acceptable analy-tical results.These are (i) sufficient knowledge of the fundamentals by the analyst who has to plan new applications and realistically interpret the results; (ii) the availability of suitable equipment and (iii) a preliminary study of operating conditions for each analytical system including results of preliminary calibration data (linearity and sensitivity as well as a study of interferences both physical and chemical). As RubeSka and Moldanf4 point out detailed instruction manuals come with the equipment and it is only necessary to mention one or two general hints in this connection. The instrument should be placed on a solid ‘island type’ bench or trolley with an exhaust hood placed over the burner to efficiently remove the various toxic fumes.Sensitivity of the detection system may be altered by changing either the slit-width or the electrical gain setting. In general the broadest slit compatible with an admissible spectral band-pass is used or according to WalshF5 the slit-width used should be the maximum that will both permit isolation from all other lines to be achieved and give a maximum signal-to-noise ratio at the detector. Ramirez-MuiiozS points out that an operator facing the problem of increasing the instru-mental sensitivity to a maximum has to expect an increase of noise. Nois ATOMIC ABSORPTION SPECTROSCOPY 205 suppression devices or damping can then be used which corresponds to an increase in the time constant although this is not advisable when the sample size is limited.It is important when using wide slits to check that the signal caused by background does not result in a loss of absorption sensitivity. Price13 advises that although any continuous background signal is not recorded it is nevertheless good practice to keep it minimal otherwise the detector may become saturated. The use of highly luminous flames should therefore be avoided. Ramirez-Muiioz17* has dis-cussed the calculation of signal size and signal-to-noise ratio and the same author and co-workers have detailed the relationship between precision and sensitivity24 and between accuracy and More recently Ramirez-Muiioz has dis-cussed the application of sensitivity diagrams to atomic absorption spectroscopy.174 These diagrams have been applied in plane and space forms to facilitate the repre-sentation and comparison of the analytical behaviour of different elements in terms of sensitivity.They also aid in the study of single elements under diverse experimental conditions the performance of different instruments and in the variation of sensitivity in the presence of a releasing agent(s) under the same or differing experimental conditions. Limiting interference ratios can also be easily calculated with the help of these diagrams. It is necessary to clearly distinguish between sensitivity and limit of detection, and reference to the diagrammatic explanations given by Elwell and GidleyJ6 Slavinll or the Eel Bulletin36 is recommended. Detection limit is a valuable guide to instrument performance for a particular deterrnination.Grunder and B ~ e t t n e r l ~ ~ studied various instrumental parameters and com-pared different types of burners and hollow-cathode lamps from an analytical point of view the information obtained being used to study precision and sensiti-vity. They concluded that the analytical curve appears to be the single most important piece of information available to the analyst. The curve and its asso-ciated slope together with the signal-to-noise ratio can give information on how well the burner and the hollow-cathode lamp are functioning. Interferences Lewis176 indicated that despite their apparent simplicity atomic absorption methods require careful laboratory technique and a thorough knowledge of potential sources of error and their control.According to GrantF5 the most serious aspect of interferences is a lack of information about them. Simple methods for their correction can usually be devised once they are well understood. Spectral interference caused by the presentation of unwanted light to the detector may be caused by the spectrum lines of unwanted elements molecular bands and background from the flame or from major elements. With thehigh source resolution and modulation techniques used in modem equipment this type of interference is hardly ever encountered in atomic absorption.177 Some spectral line interferences have recently been observed but these are uncommon and can usually be avoided if the analyst is aware of their possible existence.178 Koirtyohann and Pickett17* have discussed the background interference resultin 206 PLATT from molecular absorption which is obtained when determining a number of elements in the presence of the alkali halides and alkaline earth elements.Earlier observations reported by other workers indicating light losses because of scattering by particles in the flame were not observed by Koirtyohann and Pickett and they considered that previous workers were actually measuring molecular absorption. They concluded that the number of specific spectral interferences found is small, and freedom from such interferences will continue to be one of the major advantages of the atomic absorption method. This is particularly true when using the hotter nitrous oxide - acetylene flame.14 RubeSka and Moldan14 and Kahnl80 consider that the question still exists as to what proportion of the light loss is caused by scatter rather than absorption.Kahn makes no effort to distinguish between them as the analytical effect of both phenomena is the same. Cooke and Price177 mention the mutual spectral interference encountered when sodium and potassium occur together caused by the intense background continua. This effect is nullified either by adding the correct amount of the second element to the standards when determining the first or by adding a large excess of the second element to both standards and sarnples if the sample content of this element is variable. A number of authors have discussed the use of non-absorbing lines to correct for background ab~orption,llJ~J8~ although Koirtyohann and Pickett17@ considered this to be unsatisfactory in many cases owing to the possible presence of structure in the background absorption and the fact that frequently no suitable non-resonant line is available.Kerberls2 has described a method to select non-absorbing wave-lengths by using the scanning accessory normally provided for flame emission work. Koirtyohann and Pi~kettl'~ used a continuous source to measure and correct for background absorption in the ultraviolet region. More recently Kahnlso has described a background compensation system in which the continuous light from a deuterium arc and that from a hollow-cathode lamp are alternately passed through the flame and a ratio taken to cancel out any background absorption.Kahn discusses other compensation methods and criticises the non-absorbing line method because some elements such as zinc cadmium and mercury have no suitable lines. The same author points out that the method of standard additions does not compensate for background and that the easiest method currently used is to have a blank solution adjusted to the same composition as the sample but without the element of interest. Unfortunately the preparation of such a blank is not always feasible. Kahn28 has listed the three major types of interference as chemical ionisation and bulk or matrix. The most common interference is chemical in nature arising from a failure to break the chemical bonds formed between the element being determined and other materials in the matrix solution during evaporation of the sample mist in the flame.Rains2%eported the necessity of optimising various para-meters to eliminate or control chemical interferences. These include oxidant and fuel flow-rates flame temperature the flame region used for the measurement and the sample medium. Parsons and VVinef0rdnerl8~ have provided an excellent dis-cussion about the optimisation of critical instrumental parameters. The ATOMIC ABSORPTION SPECTROSCOPY 207 considered the importance of the various parameters and indicated those that can be systematically varied to obtain optimum performance including fuel and oxidant flow-rate sample solution flow-rate temperature of cloud chamber height of measurement in the flame slit-width and flame path length. RohledeP described a method of providing constant and adjustable sample solution feed rates by using peristaltic or piston pumps.Cellier and S t a ~ e l ~ ~ discussed a statistical approach to the determination of optimum operating conditions that has several advantages over the traditional method of varying one factor at a time. As already pointed out the use of a higher temperature flame such as that provided by the nitrous oxide - acetylene mixture is a simple and efficient means of removing the majority of chemical interferences although it is still very neces-sary to optimise the other parameters mentioned by Rains. The use of hotter flames increases ionisation interference and this must be suppressed by the addition of relatively large amounts of an easily ionisable element such as sodium or potas-siurn,2,6$%J77J86 Conversely the presence in a particular sample of an easily ionisable metal will enhance the absorption of other elements and produce errone-ously high results for the element being measured52 unless the standards are properly compensated.The majority of workers stress the importance of adjusting the burner height or the region of measurement in the flame to obtain maximum absorption,~~20,34~1s3~1~~1~7~18~ but Reynolds,3l working with very low gas flow-rates, found it unnecessary to have a separate adjustment for burner height. He simply kept the support gas flow constant and adjusted the fuel flow-rate for maximum absorption. Elwell and Gidleys justify a careful investigation of cationic inter-ference with different fuel-to-oxidant ratios and heights of measurement because by so doing interferences can be minimised and sometimes eliminated.It is useful to remember that altering the burner angle relative to the optical path effectively alters the flame cell size and offers one means of coping with various concentrations of the required element.3~2e Whenever it becomes more expedient to use cooler flames the analyst must make a serious appraisal of all the possible interferences and effectively counteract them. In such cases chemical interference can be minirnised by the addition of releasing or chelating agents. When added in a sufficient amount these agents should restore the absorption of the required element to the value obtained in the absence of the interferent (s).Price13 advises that interference effects caused by other major constituents in the sample should be investigated one element at a time. Elwell and Gidley6 suggest the addition of various amounts of anions to a constant dilute solution of the required element. Whenever interference is observed, a corresponding amount of the offending anion should be added to all of the standard solutions. It may also be advisable to examine any matrix cations in 10,000 fold excess over the element to be determined. The possible mechanisms of chemical interference have been discussed by a number of workers,6~13~14~189-192 and RainsWi provides a useful table linking the required element with typical interferents and the type of releasing - chelatin 208 PLATT agent required together with the appropriate references to any work carried out in this respect.It is interesting to note that although lanthanum is more effective than glycerol or EDTA as a releasing agent for the determination of magnesium or calcium in the presence of aluminium if glycerol is used in conjunction with lanthanum the releasing action is considerably more efficient than with lanthanum itself. RubeSka and hLZanlg0 observed a similar improvement when using mixtures of lanthanum with EDTA or 8-hydroxyquinoline. Willis20 provides evidence to support the view that the principal factor in the atomisation of elements subject to chemical interference is the rate of evaporation of the particles formed from the droplets of solution in the flame. The atomisation efficiency is dependent both on droplet size and on the position in the flame at which the measurement is made and chemical interference is not always least in the part of the flame where the concentration of metal atoms is a maximum.Many of the conflicting data in the literature on chemical interferences arise from a lack of appreciation of these facts. Koirtyohann and Pickettlg3 reported a new type of interference in the nitrous oxide - acetylene flame from the presence of mineral acids in the sample solution. Enhancement of calcium zinc and aluminium absorptions in the presence of 0.3 to 1.0 M sulphuric phosphoric and perchloric acids or 1 to 2 per cent. of ammonium phosphate and sodium chloride were ob-served near the centre of the flame. Nitric acid produced a smaller enhancement, while hydrochloric acid and organic materials such as sucrose and glycerol produced no effect.The enhancements are not changed by variations in fuel or oxidant flow-rate but are reduced when the optical path is higher in the flame and dis-appear at about 15 mm above the primary reaction zone. In every case the en-hancement disppeared or was drastically reduced when the burner was turned at a right angle to the optical axis. The expected decrease in absorption because of reduced nebuliser efficiency was noted at higher concentrations of the matrix materials. Thus these workers concluded that a new type of interference was present involving the spatial distribution of the sample within the flame and although the cause is not yet known they postulated that it involved the rate of diffusion of the salt particles outwards from the centre of the flame.Sastri et aZ.lg4 investigated the effect of the metal - oxygen bond on the sensi-tivity of absorption measurements and concluded that for metals that form refrac-tory oxides in flames low sensitivities are obtained if the metals are used as simple or oxy-salts or even as complexes with ligands having an oxygen atom as the donor. The sensitivities of these metals can be enhanced if the metal - oxygen bond in solution can be avoided or reduced by using metallocenes or fluoro-complexes in aqueous or organic solvents. These workers extended their studies195s196 to deter-mine the r6le of mixed metal polynuclear oxygen-bonded species in solution reducing the sensitivity of the metal to be determined.Hartlage197 reported a depressive interference exhibited by various amines on the absorption signal of several metal ions in the air - acetylene flame. This inter-ference which is well illustrated by the presence of 6 per cent. of triethylamine in ethanol causing a 40 per cent. error in the determination of cobalt is apparentl ATOMIC ABSORPTION SPECTROSCOPY 209 explained by the metal - amine complexes not dissociating at the temperature of the air - acetylene flame. It may be avoided by using the nitrous oxide - acetylene flame. Pricelo4 in discussing the analysis of biological materials by atomic absorption spectroscopy mentions the possibility of using dry or wet ashing procedures when the interference effects are not understood or when a method free from such effects is required.Ashing may also be used to separate metals from an organic matrix or to effect a concentration step. The ashing procedure must be carefully chosen to avoid volatilisation of the required element and it should result in a purely inorganic solution which may then be compared directly with properly constituted inorganic standard solutions. This procedure can be used as a check against more rapid procedures involving only dilution or de-proteinisation. Bulk or matrix interferences are generally of a physical nature caused by the viscosity surface tension and vapour pressure of the solvent used in the sample solutions being different from that of the standards. According to Slavin,ll most workers have found it necessary to match the standards and samples with respect to the materials that are present in excess of 1 per cent.of the total solution. The match is usually made by adding salts to the standards to equalise the concentra-tion of major constituents. Ramirez-Muiiozlg8 has discussed the analysis of systems with high sodium contents and concluded that in some cases involving the use of laminar flow burner and heated cloud chamber the low interference ratios obtained in the presence of high sodium chloride concentrations enabled determinations to be made with uncompensated standards. At levels of interference higher than the limiting interference ratio standards should be compensated with equivalent concentrations of sodium ions. The presence of an organic solvent in the sample solution will affect the physical nature of the solution and cause enhanced ab~orption,~~s~~ thus making it necessary to compensate the standards ac~ording1y.l~~ The enhancing effect of a miscible organic solvent mixed with an aqueous sample solution may be utilised to increase the response from a required element,4Js9 although Rains25 points out that such an addition is of limited value because the increase in sensitivity is less than 5 fold and the element required is inevitably diluted.MitchelI2O0 has discussed the r6le of the solvent in flame photometry and indicates that the combustion of the solvent mixture will alter flame size flame temperature and flame background continuum. Hence it becomes necessary to re-adjust completely the instrument operating conditions.The direct nebulisation of an organic solvent solution is useful in the analysis of water insoluble materials such as lubricating oils,201 vegetable oils and and has been used for the determination of metals in phospholipids.206 Orren206 concludes his paper on the basic principles of atomic absorption spectroscopy by commenting that the technique provides a rapid sensitive and precise analytical method which while not interference free is much less subject to interference than are most other techniques of comparable sensitivity and scope. RubeBka and Moldanf4 indicate three ways of eliminating interference when it does occur the most suitable method from the andytical point of view being th 210 PLATT addition of buffer solutions to both the standards and samples.Alternatively it may be possible to prepare standards that imitate the sample composition or the method of standard additions may be used. There are two further ways of reducing interference. For most purposes the choice of wavelength is limited to a selection of the most absorbing line but according to Reynolds4 it is sometimes necessary to choose a less absorbing line to avoid the high background noise that can occur over certain wavelength ranges. RubeSka and Moldan14 also point out a number of cases when a line other than the commonly recommended one may be used to advantage. Also to avoid excessive dilutions it is often convenient to use a less sensitive line than the recommended 0ne.4969207-210 Solvent extraction frequently removes chemical as well as physical interferences and allows concentration of the required element whenever this is necessary.25 Extraction into an organic solvent after complexation with a suitable reagent is an effective method as the organic solvent is itself responsible for a further increase in the sensitivity of many elements over and above the concentration factor.lo4 Highly selective extraction is not necessary and is in fact not desirable.Separations should be between groups of elements rather than between single e l e m e n t ~ . l ~ ~ ~ ~ ~ l The most popular complexing reagent for this purpose is ammonium pyrrolidine dithiocarbonate (APDC) and dithizone has also been used extensively. Methyl iso-butyl ketone is the most commonly used solvent. Carbon tetrachloride and chloroform can be used but give considerably less enhancement and also introduce the possibility of forming poisonous gases on c ~ m b ~ ~ t i o n .~ ~ ~ ~ ~ J ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Although the latter disadvantage can be overcome it adds a somewhat more time consuming step.lsg The use of APDC makes possible the simultaneous extraction of approximately fourteen elements over a consider-able pH range.14 A number of authors have discussed the use of APDC for the determination of various elements including arsenic and se1enium169~212~213 as well as alternative complexing agents and solvent ~ y ~ t e m ~ . ~ 1 ~ 9 ~ ~ * ~ ~ ~ Munro200 has dis-cussed the effects of acidity and extraction ratio with reference to the APDC-copper complex and emphasises the need to equalise these factors for variable samples and standards.Campbel121e points out the need to saturate the organic solvent with water when adjusting the zero base-line and also the fact that organic solvent becomes a fuel when nebulised and a preliminary adjustment of the fuel -oxidant mixture should be made. Chau and Lum-Shue-Chan217 mention the serious errors involved if the equilibrium state is altered while using volumetric flasks for the extraction. Before leaving the subject of interferences it is convenient to mention two further points that have been reported to cause erratic and inconsistent results. Shepherd and Johnson218 noted the presence of acetone in acetylene cylinders causing an unstable and noisy flame particularly if the cylinder had been stored in a horizontal position.This was later ~ e r i f i e d ~ l ~ ~ ~ ~ ~ and the acetone removed by passing the acetylene stream through an activated charcoal column. The second undesirable effect was reported by Scheub and Stromsky221 and concerned the presence of dissolved air in distilled water. Air bubbles in the nebuliser syste ATOMIC ABSORPTION SPECTROSCOPY 21 1 apparently caused a steady decrease in sample uptake and hence a steady down-ward drift in absorption. The problem was overcome by de-aerating the distilled water under reduced pressure. Application of method Once the sample is in solution it is often only necessary to dilute the sample to the optimum analytical range which is usually 20 to 200 times the limit of detection although concentrations nearer the detection limit may be determined with less p r e c i s i ~ n .~ ~ J ~ ~ ~ ~ s ~ ~ If the dilution is such that the total sample content is less than 0.1 per cent. of the solution the standards will usually be satisfactory when they contain simply known amounts of the element of interest in the same solvent as the sample. Whenever the solids content is greater than 0.1 per cent. it is usually necessary to match the standards with respect to the principal sample constituents. The match need not be better than &25 per cent. in most cases. Standard solutions may of course be prepared containing more than one elementll and must also contain any releasing chelating or suppressing agents that may be necessary in the sample analysis. With the instrument conditions optimised the standard solutions are used for testing sensitivity limit of detection reproducibility and to prepare a calibra-tion curve.Ramirez-Muiio~~~~ lists five principal methods of calibration. To obtain a calibration curve absorbance obtained either directly or by calculation is plotted against concentration. Theoretically the resulting line should be straight but in actual practice it will for various reasons nearly always curve slightly towards the concentration axis0,11J3s25 This method is suitable for the majority of routine work provided that the number of standard solution readings is increased over the non-linear portion of the curve and standard solutions are measured regularly between samples to observe any changes that may occur for any reason during the total analysis time.Scale expansion should be used to measure very low concentrations near the detection limit. Under these conditions the relationship between percentage absorption and absorbance is reasonably linear and it is usually unnecessary to convert the scale expanded reading to absorbance before p10tting.l~~~~ When measuring very small absorbance values a recorder trace offers better precision. If a measurement with high precision is required as in macro determinations, two standard solutions with slightly higher and lower concentrations of the re-quired element than the sample solution are prepared. Scale expansion is used so that the lower standard assumes zero absorbance and the scale is concentrated on the particular range of interest.Certain instrumentation permits this so-called zero s~ppression.~~J~ It is not always feasible to prepare standards that match the sample solution and it may be advantageous to use a standard addition method as described by a number of authors.g~11~14~25~171 However this procedure should be used only as a last resort and then only for determining elements at trace level where a lower leve 212 PLAT" of accuracy may be acceptable. As pointed out in the section Experimental Con-siderations the method of additions does not differentiate between residual con-centrations and background absorbance. Having developed the basic method including calibration data samples of known composition e g . National and International Standards or chemically analysed samples should be checked against the calibration data and recovery experiments of known amounts of the required element will determine whether or not any interference effects have been completely over~ome.~J~ Ramirez-Muiioz222 has discussed the calculation of percentage recovery.Reproducibility and there-fore analytical accuracy should in the optimum concentration range be of the order of &1 per ~ e n t . ~ ? ~ ~ Applications Even the more recently published applications of atomic absorption spectro-scopy are far too numerous to be fully covered in this review. The ones discussed have been chosen either because they are somewhat novel or because they form part of the general developments that have taken place. The determination of macro components The major emphasis in the literature about atomic absorption has been its capability for trace level determinations and in fact as R ~ o n e y ~ ~ ~ points out the technique was specifically designed for this field.However many workers have reported the successful determination of metals present in high concentrations. Price13 indicates that the method can often be used for the determination of alloy-ing or comparatively major constituents because the reproducibility where the concentrations range is optimum should be about &-1 per cent. and this can be improved by replication. The analyst must of course be particularly careful when preparing the solution for analysis. When determining major constituents precision rather than sensitivity is the important factor; Kahn2% has indicated a procedure for obtaining optimum precision.It is necessary to observe the following three criteria (i) that the element required is present in the solution at a concentration within its optimum analytical range; (ii) the sample is sufficiently homogeneous so that a truly repre-sentative sample can be obtained; and (iii) the dilution of the sample does not contribute any error. Kahn discusses the effects of dilution and indicates that it can be avoided by reducing the optical path through the flame or by using a less sensitive resonance line although the latter must be carefully chosen. The same author also recommends the zero suppression technique when optimum precision at high concentration is required. Feldman et aZ.225 have reported the determination of major constituents by using an internal standard technique.These workers have analysed a wide range of materials and show the technique to be very accurate and to exhibit increased precision compared with the direct method ATOMIC ABSORPTION SPECTROSCOPY 213 The types of materials that have been analysed for major constituents by atomic absorption fall into two broad categories metallurgical and silicate types including cements. Sattur226 described the routine analysis of non-ferrous alloys and obtained good agreement with certificate or chemical analyses in many cases. Aqueous standards were used and no sample preparation other than acid dissolution was necessary. Sattur determined tin potassium copper bismuth lead antimony, zinc indium and cadmium all in concentrations of greater than 5 per cent.Copper was accurately determined in bronzes to concentrations as high as 83 per cent. Meddings and Kaiser227 determined nickel cobalt copper and iron and compared the precision with that obtained by wet chemical methods. By using equipment that included digital read-out and high brightness lamps these workers reported the coefficient of variation for routine analysis to be 0.3 to 0.8 per cent. for atomic absorption spectroscopy compared with 1 to 2 per cent. by wet chemical analysis. Ellis and Roger@ hawe determined lead and zinc in non-routine samples by using automatic dilution apparatus. The problem of sampling was overcome by success-ive grinding and rifBing. Welcher and Kriege228 developed specific procedures for the precise determination of major alloying elements in high temperature nickel-base alloys.These workers made a comprehensive study of techniques used for suppressing the numerous chemical interferences and found that for the most accurate analysis close matching of samples and standards with respect to both cation and anion concentrations was necessary. They preferred to extend the working range by rotating the burner rather than diluting the solutions. Working at the optimum concentration excellent agreement was obtained between the atomic absorption results and the National Bureau of Standards or wet chemical values for chromium aluminium molybdenum cobalt titanium iron tungsten, niobium tantalum and vanadium. Niobium and tantalum were determined after extraction into methyl iso-butyl ketone from an aqueous phase that was 10 M in hydrofluoric acid and 6~ in hydrochloric acid.The results quoted showed good precision for each element determined. Johns and Price229 described a comprehen-sive scheme for the analysis of brasses and bronzes in which the usual elements associated with these alloys were determined at levels from 0.1 to 10 per cent. The air - acetylene flame was suitable for the determination of iron lead manganese, nickel and zinc. Apart from viscosity differences introducing physical interference, no other interference was found from elements normally present in brasses and bronzes. When using the hotter nitrous oxide - acetylene flame ionisation inter-ferences were sometimes encountered that appeared as slight enhancements of aluminium and tin absorptions by copper.Both interferences were overcome by adding an equivalent amount of copper and solvent acid mixture to the standards. Whenever several elements present in a sample required the same dilution it was possible to use multi-element standards. Agreement with the certificate values was good and demonstrated the suitability of the recommended methods for routine analysis. discussed the application of atomic absorption spectro-scopy to the analysis of major constituents in alloys and steels. Zero suppressio 214 PLATT with scale expansion was used to achieve relative precisions of 0-1 to 0-3 per cent. The use of less sensitive lines and shorter flame path length was also explored. Cobb and Harrison231 determined aluminium in iron ores slags and refractory materials and achieved a reproducibility of k0.55 per cent.in the 20 to 70 per cent. range and the mean result agreed closely with the accepted chemical value. Langmyhr and Paus have published a series of papers concerned with the analysis of inorganic siliceous materials by atomic absorption spectroscopy and this work has been summarised by the same authors.232 A hydrofluoric acid decom-position technique was used which retained the silicon in solution. Primary standard solutions were prepared from highly purified reagents and details of potential interferences and the means of overcoming them are given. Good agreement with certificate values was obtained for silicon aluminium iron, magnesium calcium sodium and potassium and for titanium manganese chrom-ium and vanadium at somewhat lower levels.Langmyhr and PausSB have also provided some data on the intra-laboratory precision of atomic absorption analysis as applied to silicate rock. This precision can be classified as good to moderate and compares favourably with other methods. During the last 2 years a number of authors have described methods for the decomposition of silicates followed by a comprehensive analysis by using atomic absorption spectroscopy. B e r n a 9 found that a fluoboric - boric acid system provided a f avourable decomposition medium for the rapid and reliable determina-tion of silicon aluminium titanium and vanadium. The acid matrix used was found to eliminate the need for releasing compensating or complexing agents and it was not necessary to closely match the matrix composition to that of the un-known sample.Thus the use of single element standards for the analysis of complex materials is brought nearer. Hence whereas Langmyhr and Paus list interferences in several cases Bernas finds no inter-elemental effects. Van Loon and used lithium metaborate fusion and conducted extensive tests of possible inter-ferences. They found it necessary to correct for different types of interference in the determination of aluminium potassium sodium titanium magnesium calcium and manganese. No interference was found in the determination of iron and silicon. Similar fusion techniques using lithium borates followed by solution in nitric acid have been proposed by a number of other ~ o r k e r s .~ ~ ~ - - ~ ~ ~ In each case interferences were easily overcome and the precision was comparable with that obtainable by conventional methods of analysis. O r n a r ~ g ~ ~ ~ verified the results quoted by Van Loon but disagreed with Bernas over the use of cations which the latter worker considered to be a disadvantage. Karmie Galle242 developed a routine procedure for the determination of major constituents in geological samples by atomic absorption in which it was unnecessary to change from nitrous oxide-acetylene to air - acetylene to analyse non-refractory elements. The samples were obtained in solution by treatment with hydrofluoric acid followed by fusion with potassium pyrosulphate. CampbellM1 determined silicon (>5 per cent.) in alumi-nium alloys after treatment with sodium hydroxide and hydrogen peroxide.It wa ATOMIC ABSORPTION SPECTROSCOPY 215 necessary to prepare standards containing approximately the same amounts of aluminium and total solids as the sample solutions. The analysis of cement by atomic absorption spectroscopy has been discussed by a number of people; Crow et aLN3 found that a simple acid digestion step was sufficient sample preparation. The presence of silica affected the aluminium determination but this and other matrix interferences were compensated for by calibrating with National Bureau Standards similarly prepared. According to these workers cement chemists would probably accept errors of 1 per cent. for the determination of any oxide in cement except calcium.Satisfactory accuracy could be obtained for this element by using calibration curves but the determina-tion was limited by unacceptable precision and it was necessary to resort to the zero suppression technique to overcome this problem. Other workers have also recommended this technique for the determination of calcium in ~ e m e n t ~ ~ J ~ and very impressive results were obtained. Crow determined aluminium in cement with good precision and accuracy by using standard cement samples to prepare a standard curve. Capacho-Delgado and Manning2u obtained good precision for the determination of aluminium in cement by using aqueous standards containing about the same amount of calcium chloride and hydrochloric acid as was present in the sample solutions. Roos and PriceN6 determined calcium in cement without resorting to the zero suppression technique.They used an emission burner head with an absorption path of 1 cm to reduce the dilution factor but obtained a standard deviation of 0-7 per cent. which might not be acceptable when compared with standard gravimetric procedures giving nearer 0.2 per cent. The determina-tion of silicon also tended to lack precision. However one advantage of this procedure lay in its use of simple aqueous standards with the addition of lanthanum as a releasing agent. Elements determined in the nitrous oxide - acetylene flame Ramirez-MuiiozlsS recommended the use of the nitrous oxide - acetylene flame for thirty-three elements and in addition for the alkaline earth metals when high chemical interference is expected.Amos and Willisgs have determined the sensitivity at different wavelengths for twenty-seven elements many of which show none or only litttle absorption in the air - acetylene flame. Willisg4 lists sensi-tivities and detection limits for thirty-five elements and also indicates the likely applications for the determination of these elements. Shifrin et aZ.l10 reported recent results with a nitrous oxide - acetylene burner for twenty elements includ-ing seven that were normally determined in the air - acetylene flame but the analytical task was simplified by retaining the same flame throughout. Alkaline earth elements Amos and WilIisg8 called these the border-line elements. They can be deter-mined in an air - acetylene flame but generally show improved absorption in the hotter nitrous oxide - acetylene flame.Slavin et aZ.246 have illustrated the improve-ments shown when determining calcium or barium in the presence of phosphate 216 PLAT" and also point out the ease with which strontium in cement can be determined in the hotter flame compared with using the air - acetylene flame. In discussing the severe interference of phosphate on calcium Ulrich and Ramirez-Muiio~~~ indicated that the simple compensation of standards with phosphates leads to very small signals and it is better to use hot flames and to help the system with the action of some releasing agent (lanthanum) when massive amounts of phosphate are present. The lanthanum will also act as an ionisation buffer. The well known depressive effect of aluminium on calcium is overcome by using the nitrous oxide - acetylene flame together with lanthanum4 or by the addition of aluminium to the standards.25 Barium is remarkable for its poor sensitivity and is better determined in the nitrous oxide - acetylene flame.4J4 Kerber and BarnetF prefer to use an air -acetylene flame for the determination of barium in the absence of interferences, but when these are present or at lower concentrations the nitrous oxide flame is more convenient.At even lower concentrations these workers chose flame emis-sion with a nitrous oxide - acetylene flame as the best method. Ionisation must be suppressed by the addition of a suitable metal. Strontium is similar to calcium in its behaviour except that owing to the relatively low contents of strontium in natural materials the sample solutions cannot be diluted as much and the con-centration of interfering elements is thus likely to be higher than for calcium.By using a nitrous oxide - acetylene flame however chemical interference is efficiently overcome including that provided by iron(II1) .248 Magnesium is considerably more sensitive than calcium and its resonance line at 285.2 nm lies in the most favourable wavelength region. Hence high dilution can be used and this limits interferences. For example phosphates have at the most only a slight depressive effect. Interference caused by aluminium or silicate is not encountered when using the nitrous oxide - acetylene flame.14@9~250 The presence of alkali metals causes an enhancement interference when determining any alkaline earth metal in the hotter flame and the standards should be compensated accordingly or a large excess of the alkali metal added to both samples and standards.Klein and co-workers1@ adopted a more novel approach by coupling an auto-analyser to an atomic absorption instrument and simultaneously determining calcium and phosphate. An air - acetylene flame with lanthanum as the releasing agent was used but a nitrous oxide-acetylene flame could be used in such a system. Aluminium Until the advent of the nitrous oxide- acetylene flame aluminium was excluded from those elements determined by atomic absorption. At the present time aluminium is easily determined in this flame with few interferences. Potas-sium should be added to both samples and standards as an ionisation buffer.ll Ramakrishna et reported on the determination of aluminium and indicated that low concentrations could be determined accurately in the presence of various other ions and compounds.P a w l ~ k ~ ~ ~ determined aluminium in soil samples an ATOMIC ABSORPTION SPECTROSCOPY 217 obtained results that compared very well with those obtained by gravimetric analyses. Laflamme2= also reported the atomic absorption determination of alumi-nium in soils to be precise rapid and very useful. Van Loon254 obtained results agreeing well with those obtained by standard procedures in the analysis of high silica materials. The method used a hydrogen fluoride - sulphuric acid dissolution to remove the major component-silica-and is readily recommended.Dagnall et uZ.120 found the nitrous oxide - hydrogen flame showed considerable promise for the determination of aluminium and considered it worthy of further examination. Beryllium This element is barely detectable in the air - acetylene flame but according to Amos and Willis9* the use of a nitrous oxide - acetylene flame enables its deter-mination with a sensitivity approaching that for magnesium. Ramakrishna et ~ 1 . ~ ~ ~ reported a sensitivity of 0.025 p.p.m. and found a lack of interference from several ions and compounds. Manning255 confirmed the quoted sensitivity and found the addition of ionisation buffers to be unnecessary. B o k o ~ s k i ~ ~ ~ found the atomic absorption determination of beryllium in biological materials to be sufficiently rapid and sensitive for use as a monitoring procedure.provided a method for determining small amounts of beryllium in aluminium alloys. Boron Boron is not detected in an air - acetylene flame but is readily determined in aqueous solutions in the nitrous oxide - acetylene flame.ll Boron is not significantly ionised in the hotter flame and Manning255 found the sensitivity to be 35 pg ml-1 for 1 per cent. absorption with the detection limit at 6 pg m1-1 in fuel-rich flame with a 2 -nm slit-width. Bader and Brandenbe~-ger:~~ like Manning used ten times scale expansion to determine boron. They analysed biological materials and found a detection limit of 15 pg ml-l with a O.65-nm slit-width. Bader and Brandenberger determined sub-toxic levels of boron in serum and urine and solubilised tissue by a wet ashing procedure; dry ashing led to severe boron losses.H a r r i ~ ~ 5 ~ found it absolutely necessary to use the optimum conditions to obtain reproducible results when analysing low concentrations of boron flame conditions and burner height adjustment were both found to be critical. Owing to the low absorbance the method of standard additions was used by Harris to determine boron in potassium chloride. The particular isotope distribution in the sample made little difference to the final evaluation thus confirming earlier reports by other workers. Chromium It is well known that iron seriously interferes with the determination of chromium in an air - acetylene flame.260-262 Rarnirez-Mufioz and Roth262 con-sidered the suppression observed to be apparently caused by the formation o 218 PLATT iron - chromium compounds which axe hard to dissociate at the temperature of the air - acetylene flame.These workers suggested the use of a nitrous oxide - acetylene flame to help in overcoming the problem. Roos~~O found it necessary that the oxidation state in the sample solutions should be the same as that in the standards when using the air - acetylene system. Chromium(II1) gave considerably greater sensitivity than chromium(V1). Also in the cooler flame maximum absorbance and interference effects were found to be critically dependent on the flame gas mixture used and on the burner height. Investigations reported by Rooney and Pratt261 found that the nitrous oxide -acetylene flame overcame the inter-element effects and gas ratios were much less important and although there was still an effect from the presence of iron it was an enhancement rather than a suppression.It could be completely overcome by the addition of a small amount of iron to the final solutions. Wilson263 determined chromium in aluminium alloys by using a nitrous oxide - acetylene flame and found excellent agreement between the classical and the atomic absorption procedures. It was unnecessary to use standard solutions containing aluminium. Germanium Germanium shows some absorption in the fuel-rich air - acetylene flame but to attain really useful results a high temperature flame is required.ll Popham and Schrenk2G4 reported that by proper choice of experimental conditions germanium could be determined in a fuel rich nitrous oxide - acetylene flame with a detection limit of 0.5 p.p.m.and a sensitivity of 3 p.p.m. in a 50 per cent. aqueous acetone solution. This represented an approximately 2-fold enhancement over pure aqueous solutions. Interferences were more pronounced in the acetone medium. Kirkbright et aZ.,124 using separated flames obtained a significant improvement in the detection limit and extended the linear portion of the calibration graph. Molybdenum Molybdenum may be determined in either the fuel-rich air - acetylene or nitrous oxide - acetylene flames with about equal sensitivity. It is in fact one of the border-line Molybdenum is similar to chromium in that iron seriously interferes in an air - acetylene flame,14*260 and Kirkbright et aZ.266 recom-mend the use of a nitrous oxide - acetylene flame to remove interferences other than iron and then to compensate the standards and samples when necessary by the addition of iron.Absorbance with this flame is not as dependent on the flame composition as it is with an air - acetylene flame. Kirkbright et aZ.124 improved the detection limit and linearity of calibration by using separated flames. Ramakrishna et aZ.266 reported that most ions affect molybdenum absorption in the nitrous oxide supported flame some suppressing but most causing an enhancement in absorption. The method was made more selective by the addition of aluminium. Molybdenum has been determined in steel by using a nitrous oxide - acetylen ATOMIC ABSORPTION SPECTROSCOPY 219 flame and potassium sulphate as a buffer.267 It has also been determined by using an air - acetylene flame in fuels and lubricant^^^^^^^^ and in lake waters by using a nitrous oxide supported flame after complexation and extraction into an organic solvent .270 Butler and Mathew~~~l determined trace amounts of molybdenum in waters plant and silicate samples by using an air - acetylene flame after extraction into an organic solvent.Ure272 reported the determination of molybdenum in soil extracts by using conventional and separated nitrous oxide - acetylene flames. The absorption was enhanced by the presence of ammonia and therefore the molyb-denum was stripped from an organic complex by the addition of ammonia solution. With a 50-g soil sample a detection limit of better than 0.004 p.p.m.of molybdenum in the soil can be obtained. Kirkbright et recommend the use of a nitrous oxide - acetylene flame to achieve good sensitivity while avoiding the unpleasant, strongly luminous fuel-rich air - acetylene flame that can cause fatigue of the detection system. Niobium Manning255 reports a sensitivity of 20 p.p.m. and a detection limit of about 5 p.p.m. with potassium added at 1000 p.p.m. and a 10-fold scale expansion by using a nitrous oxide - acetylene flame. The niobium spectrum is very complex and Manning lists thirty-two ‘atom and ion lines,’ a number of which are un-resolved pairs. Noble metals Beamish et ~ 1 . ~ ’ ~ have reviewed the use of atomic absorption spectrochemical and X-ray fluorescence methods for the determination of the noble metals.The main difficulty in determining each of the noble metals is the extremely low concentration in which they are usually present in most materials.14 This necessitates the use of some concentration technique such as co-precipitation, cupellation ion-exchange or extraction into an organic solvent after complexation prior to the determination by flame methods. Gold palladium platinum rhodium ruthenium and iridium have all been determined in the air - acetylene flame by using a mufti-slot b ~ r n e r ~ 6 ~ ~ ~ 6 while reported determinations of osmium have nearly all used the nitrous oxide -acetylene Osolinski and Knight278 reported that the method for osmium was rapid precise and accurate. A sensitivity of 1 p.p.m.was attained and the method was applicable to both aqueous and non-aqueous solutions. The oxidation state of the osmium was not important. Fernandez2T7 found a 4-fold increase in sensitivity when using a nitrous oxide supported flame rather than air - acetylene for the determination of osmium, Atwell and Herbert279 compared the air - acetylene flame with the nitrous oxide - acetylene flame for the determination of rhodium and concluded that the serious interferences encountered with the former were essentially removed by using the hotter flame. Johns and Price2go recommended the nitrou 220 PUTT oxide - acetylene flame for the determination of rhodium to improve linearity of calibration and to overcome all interferences except the ionisation effect for which a compensating buffer is required.Aldous et aZ.281 described the preparation of electrodeless discharge lamps for palladium silver platinum and gold. Rare earths Amos and WillisQ8 pointed out that the atomic absorption technique with a high temperature flame should prove very valuable in the analysis of rare earth mixtures as no mutual interference was expected. The sensitivities for the heavy rare earths are considerably better than for light ones and it is necessary to include an ionisation buffer to obtain the best sensitivity. Jaworowski et aZ.282 list sensitivi-ties for rare earth elements determined in aqueous and organic solutions. These range from 1.5 p.p.m. in 80 per cent. of alcohol up to 100 p.p.m. in a purely aqueous medium. Kinnunen and Lindsjo283 found that the most serious interferences were caused by fluoride silica and aluminium the degree of interference depending on the element to be determined.Van Loon et aZ.28Q found aluminium to be the most serious interference and it was concluded that no releasing agent could completely eliminate its effect. It was however accounted for by adding 150 p.p.m. of alumi-nium to the standards and 1 per cent. of lanthanum eliminated all other commonly encountered interferences. The relative freedom of atomic absorption methods from matrix and spectral interferences makes the technique potentially useful in rare earth analysis although the poor sensitivity obtained for some elements precludes its use for a complete analysis in natural materials. Hingle et aZ2= found the detection limits for eight rare earths by using an emission technique and a separated nitrous oxide - acetylene flame to be equal or , superior to those obtainable by atomic absorption spectroscopy.Scandium Chau286 discussed the determination of trace amounts of scandium by atomic absorption in the nitrous oxide - acetylene flame. The flame gas mixture was found to be critical for optimum absorption. The presence of EDTA enhanced the absorption and a sensitivity of 0.06 p.p.m. was observed by using extraction into an oxine - butanol mixture. Rhenium Biechler and Long287 determined rhenium in a fuel-rich nitrous oxide -acetylene flame by using solvent extraction and obtained a sensitivity of 3.5 p.p.m. Molybdenum did not interfere at 7000 times the rhenium concentration.The burner height was adjusted for maximum response by using the highest standard, and it was ensured that the volumes of the aqueous and organic phases were the same for standards and samples ATOMIC ABSORPTION SPECTROSCOPY 221 Silicon Silicon is one of the less sensitive elements and optimum sensitivity requires careful adjustment of the instrument conditions. The fuel-rich nitrous oxide -acetylene flame used causes rapid carbon build-up around the burner slot unless one of the later re-designed burners is available. Price and Roos28s determined silicon in steel cast iron aluminium alloys and cement. They investigated the optimum operating conditions and discussed the effect of foreign ions and the preparation of standards and samples. The sensitivity was found to be 8 to 10 p.p.m.and the detection limit was 3 p.p.m. in aqueous solu-tion. McAuliffeBg reported a method for the determination of silicon in cast iron and steel. The results obtained were compared with certificate values and with those obtained by colorimetric analysis and showed the atomic absorption method to have the accuracy and repeatability necessary for control or referee analysis. Morrow and Dean290 used atomic absorption as a specific silicon detector for the gas-chromatographic determination of silylated aliphatic alcohols. Paralus~~~l described the determination of trace silicon in an organic matrix and Mario and GernerB2 determined silicon in a commercial hand lotion. Kirkbright et aZ.124 using a separated flame found that the high fuel flow required resulted in an instability at the ends of the burner slot when nitrogen shielding was used.The effect was not experienced with argon shielding and hence the latter gave a much more pronounced improvement in detection limit. The sensitivity and detection limit obtained were 2.5 and 0.24 p.p.m. respectively, and a calibration graph was linear from 5 to 200 p.p.m. Dagnall et aZ.293 compared the detection limits obtained for silicon by atomic absorption with those obtained by flame emission and atomic fluorescence. Absorption with a high intensity lamp or an electrodeless discharge lamp was 4 p.p.m. whereas emission and atomic fluorescence gave values of 20 and 5 p.p.m. respectively when using an argon-separated flame. Tin The most sensitive conditions for tin are obtained with an air - hydrogen flame in spite of its lower temperature but the cooler flame allows interferences to occur.The nitrous oxide-acetylene flame gives a greater freedom from any possible interference^.^^^^ Amos and Willisgs quote sensitivities of 2-5 p.p.m. in the hot flame compared with 1 p.p.m. in the cooler one. Interferences present in the determination of tin when using an air - hydrogen flame have been discussed by Capacho-Delgado and Manning294 and by Juliano and Harri~on.2~~ used the nitrous oxide - acetylene flame to eliminate the interferences during the determination of tin in tin ores and concentrates. He obtained a detection limit of 0.02 per cent. in the sample and the results were similar to those obtained by standard chemical methods.Shannon297 reported that the nitrous oxide - acetylene flame was unsatisfac-tory for the determination of tin in the presence of glycerol 222 PLAT" No satisfactory explanation appears to have been given for the high sensitivity obtained for tin when using the cooler air - hydrogen flame.l14 Titanium Amos and WillisS8 reported the determination of titanium in the nitrous oxide - acetylene flame with a sensitivity of 3.5 p.p.m. These workers also indicated the enhancement of titanium in the presence of hydrofluoric acid and hydrofluoric acid plus iron. They attribute this behaviour to the formation of a stable complex, fluoroacid which inhibits the formation of titanium oxide but is more readily decomposed at high temperatures to yield a higher population of titanium atoms.Nevertheless it was considered that further investigation was required before this element could be determined in a routine fashion. Headridge and H ~ b b a r d ~ ~ * determined titanium in steels permanent magnet alloys and cast iron by using a nitrous oxide - acetylene flame. Hydrofluoric acid was used as the solvent and the use of aqueous ethanolic solutions more than doubled the sensitivity. When the correct conditions were used there wasno interference from the other elements commonly found in these materials. The results obtained agreed well with certificate values for titanium contents of between 0.1 and 1.2 per cent. Kirkbright et aZ.lN used separated flames to determine tita-nium and found linear calibration curves from 10 to 250 p.p.m.The addition of potassium chloride gave an increase in absorbance at all concentrations of titanium and had no effect on the linearity of calibration. With potassium chloride present, the detection limit was 0.035 p.p.m. Mostyn and C~nningham~~~ avoided the use of hydrofluoric acid and used aqua regia as the solvent with 2000 pg ml-l of potassium chloride to obtain reliable analytical results that compared well with those obtained by chemical and S-ray fluorescence analysis on various alloys. Vanadium By using a high intensity lamp in conjunction with a nitrous oxide - acetylene flame Capacho-Delgado and Manning30° determined vanadium in steel and gas oils. An apparent compensating effect of sulphuric and phosphoric acids enabled results in good agreement with certificate values to be obtained in the analysis of steels.Gas oils were analysed by the method of standard additions after dilution in xylene and a detection limit of 0.05 pg ml-l was obtained. Sachdev et aLW1 found an enhancing effect by many potential interferents which may have been caused by competition in the formation of oxides. The same workers302 studied the r61e of mixed organic solvents and found that the addition of diethylene glycol and similar compounds increased the absorption by about 50 per cent. Hall et aLm3 reported that suitable standards were prepared for the determination of vanadium in steels simply by adding 1 per cent. of trivalent iron. Good agreement was obtained with certificate results. Kirkbright et aZ.124 obtained a detection limit of 0.04 p.p.m.in the presence of potassium chloride when using separated flames. Calibration curves were linea ATOMIC ABSORPTION SPECTROSCOPY 223 between 5 and 100 p.p.m. but curved towards the concentration axis between 100 and 200 p.p.m. Zirconium Amos and WillisB8 found a similar effect with zirconium in the presence of hydrofluoric acid and iron as for titanium. A detection limit of 5 pg ml-l has been reported.ll Slavinll found an enhancement with hydrochloric as well as with hydrofluoric acid. TyleSM found that potassium added as the sulphate entirely suppressed the absorption of zirconium instead of producing the expected enhance-ment. Tyler attributed this to the formation of zirconium sulphate and Slavinll found the expected enhancement when potassium was added as the chloride.Kirkbright et aZ.12* reported a detection limit of 0.24 p.p.m. in the argon-shielded flame and in the presence of potassium chloride. Calibration curves were linear between 500 and 1000 p.p.m. Elements with analytical resonance lines between 190 and 230 nm The air - acetylene flame strongly absorbs any radiation in the wavelength region between 190 and 230 nm in which arsenic antimony selenium tellurium, cadmium zinc and lead have their resonance lines. Strong light absorption by the flame leads to background instability and a reduction in the available energy reaching the detector. For these reasons and also because of difficulties in producing satisfactory sources little detailed information was available on the determination of elements such as arsenic and selenium until recently.However a number of ways of tackling the problems involved have now been described and suitable instrumentation has been made commercially available. The primary source intensity can be increased by the use of high spectral output hollow-cathode lamps or by using electrodeless discharge lamps. This allows the use of an air - acetylene flame to keep chemical interferences minimal, while still leaving ample energy to reach the detector. Alternatively the flame type may be changed to make it more transparent and hence absorb considerably less energy. The latter may be accomplished by using a cooler flame such as air-hydrogen or argon - hydrogen in which chemical interferences will be greater or by retaining the air - acetylene flame and separating it by nitrogen shielding.Dagnall et uZ.= have described the use of electrodeless lamps as sources for the determination of antimony arsenic and selenium. The intensity of the resonance lines was sufficient to allow the use of narrow slits and the stability was such that a 10-fold scale expansion could be used. Detection limits with different flame systems varied between 0.1 and 0.5 pg ml-l. Air - acetylene and the cooler more transparent nitrogen - hydrogen flames were used on single and triple-slot burners. For arsenic the latter flame absorbed 50 per cent. of the radiation on a single-slot burner but only 5 per cent. on a triple-slot one. This was attributed to the wider flame allowing all of the light from the lamp to pass through the centre of the flame.The cooler flame may give rise to chemical interference in whic 2 24 PLATT case it will be necessary to resort to a suitable separation technique such as solvent extraction or ion exchange. Menis and Rains306 determined arsenic in cast iron and in high purity selenium metal by using an electrodeless discharge lamp after solvent extraction followed by re-extraction into an aqueous solvent. Fisher and H a y ~ a r d ~ ~ studied the determination of arsenic and selenium by using electrode-less discharge lamps and found better detection limits and stability compared with the use of hollow-cathode lamps. Kahn and Schallis found considerable improvements in both sensitivity and detection limits for arsenic selenium and cadmium when using an argon - hydrogen flame and moderate improvements for zinc and lead,lf4 Pre-mixed air - hydrogen flames were also shown to produce better detection limits than air - acetylene for elements tellurium cadmium zinc and lead.ll8 Like Dagnall et ~ 1 .~ ~ these workers used a three-slot burner but in contrast they found little success when using nitrogen - hydrogen flames. The detection limit for arsenic was less than 0.05 pg ml-1 and for selenium less than 0.1 pg ml-l compared with 0.25 and 0.5 pg ml-1 respectively in the air - acetylene flame. Because of the background absorption from the air - acetylene flame at 217 nm the less sensitive 283.3-nm line has been preferred for the determination of lead and this gives a detection limit of about 0.03 pg ml-l.However with the air - acetylene flame the most sensitive lead line at 217 nm can be successfully used. As Kahn and Schallis point out the cooler flames produce at best improve-ment factors of only 2 or 3 and operators who are using air - acetylene are well advised to add hydrogen only if the degree of improvement is really necessary. They found no advantage in the air - hydrogen flame for the determination of antimony at the 217.6-nm line. This element is in fact determined easily in the air - acetylene flame.ll Cookella reported that a good hollow-cathode lamp was now available for arsenic. The lamp is of the high spectral output type which ensures that the intensity is sufficiently high to give a good signal-to-noise ratio in the final measure-ment.Cooke investigated air - acetylene and argon - hydrogen flames on a multi-slot burner and recommended the air - acetylene system to overcome interferences. He considered it desirable to use an instrument incorporating a silica prism mono-chromator as transmission and stray light characteristics are superior to diffrac-tion grating monochromator characteristics. In addition the dispersion of a prism is high at these short wavelengths thus allowing the use of relatively wide slit-widths while retaining adequate resolution. Kirkbright et uZ.123 determined arsenic and selenium in a nitrogen separated air - acetylene flame. This system offers good sensitivity and precision combined with relative freedom from chemical interferences. Kahn et ~ 1 . ~ 0 have reported greatly improved detection limits for arsenic selenium cadmium lead and zinc by using the sampling boat technique.Hill25 determined arsenic in steels iron ores and spelters in an argon - hydrogen flame following a rapid distillation procedure. Holak306 converted arsenic to arsine which was collected in a U-tube immersed in liquid nitrogen. The arsine was swep ATOMIC ABSORPTION SPECTROSCOPY 225 into the cloud chamber with nitrogen gas and determined in an air - acetylene flame on a three-slot burner. Ando et aL307 greatly increased the sensitivity for arsenic by using a nitrogen - hydrogen flame in conjunction with a Vycar long tube. A sensitivity of 0.006 p.p.m. was obtained. Chakrabarti308 found the air - acetylene flame to be more sensitive than air -hydrogen for the determination of selenium.A sensitivity limit of 0.72 p.p.m. in aqueous solutions was improved to 0.30 p.p.m. when extracted into methyl iso-butyl ketone as the diethyl-dithiocarbamate complex. The same author also in-vestigated the determination of tell~rium.30~ Barnett and Kahn310 described the determination of tellurium in steel by using the ‘Deuterium Background Corrector’ to compensate for the broad-band absorbance caused by the matrix material. Air -acetylene air - hydrogen and argon - hydrogen flames were examined and the first one selected because it showed good sensitivity and was likely to exhibit the fewest interferences. Taylor119 used the air - hydrogen flame to determine lead on a three-slot burner head. The determination of lead in methyl iso-butyl ketone was three times more sensitive than when using water as the solvent and the increased sensitivity at the expense of signal-to-noise ratio is illustrated for the 217-nm line.Slavin and SattuPl have noted a spectral interference of lead on antimony when using the primary antimony resonance line at 217.6 nm and a spectral slit-width of 0.7 nm. It is suggested that the resonance lines at 206.8 or 231.1 nm are used to avoid the interference. Determinations by an indirect method A number of indirect methods have been proposed whereby the component to be determined reacts quantitively with a metal that can subsequently be deter-mined by atomic absorption spectroscopy. However as Koirtyohann points these indirect methods must be used very cautiously because the specificity of the determination depends on the chemistry of the reactions and not on the final determinations.The methods can be quite valuable in some cases. Cations anions and organic materials have been determined indirectly most often following quantitative precipitation or complexation - extraction but some-times by utilising a quantitative interference effect. A brief selected summary is given in Table I. Before leaving the indirect methods a few points are worthy of further mention. Thorium cannot be determined directly by atomic absorption and so the indirect procedure described by Kirkbright et U Z . ~ ~ although complicated by many potential interferences does at least provide a means of determining thorium by this technique.Dunk et aL319 determined sulphate in textiles and found the atomic absorption results to compare reasonably well with volumetric and gravi-metric methods and to require less operator time. The determination of sugar in plant materials reported by Potter et aZ.,= gave results in close agreement with the existing Association of Official Agricultural Chemists method and had the advantage of simplicity. Ezel1325 found the atomic absorption determination o 226 PLATT chloride in plant liquors to be capable of high precision and to provide more accurate determinations than titrimetric methods. TABLE I Component required Metal used References Phosphate Phosphorus arsenic silicon Sulphur dioxide Sulphur dioxide Sulphate Nitrate Thioc yanate Fluoride Chlorine and chloride Ammonia Ammonia Thallium Thorium Phthalic acid Pentachlorophenol Sugar Non-ionic surfactants Molybdenum 312 Lead 317 Mercury 318 Barium 319 320 Copper 321 Copper 322 Magnesium zirconium or titanium 323 Silver 324 325 326 327 Molybdenum 328 Zirconium 329 Molybdenum 328 Molybdenum 330 Copper 331 Iron 332 Molybdenum 3 13-31 6 Copper 333 Molybdenum 334 Determination of mercury K;oirtyohann@ indicates that the use of static atomic vapours rather than flames allows us to expect big improvements in sensitivity and the available information for mercury permits an estimation of the general improvement to be expected.According to Koirtyohann about 5 pg of mercury are required to pro-duce 1 per cent. of absorption in the air - acetylene flame.In a cell of cross-section 0.25 cm2 the same absorption is produced by 0.000 05 pg of mercury in the static vapour. Brandenberger and Bade? described a dynamic vapour determination of mercury by using a commercial atomic absorption instrument. The mercury was amalgamated from solution on to a copper wire which was then electrically heated to vaporise the mercury in an absorption cell placed in the light beam of the spectrophotometer. The mercury vapour absorbed radiation while being pumped through the cell and the resulting absorption was plotted as a function of time by a recorder. The same workers later reported a static vapour approachs6 that was less complicated than the dynamic method and produced equivalent accuracy, sensitivity and detection limit.Between 10 and 200 ng of mercury was determined with an accuracy of &3 per cent. and an accuracy of &lo per cent. was obtained at 0.2 ng of mercury. ThillieF reported a similar procedure for the determination of mercury pollution in atmospheric air ATOMIC ABSORPTION SPECTROSCOPY 227 Determination of isotopic concentrations Atomic absorption has been used to determine isotopic concentrations of lithium lead and uranium. An attempt to determine boron isotopes was not successful .= K i r ~ h h o f ~ ~ measured the concentrations of lead-206 and lead-208 by using the absorption of the hyperfine components of the 405.8 and 283.3 nm lines. Known isotopic mixtures were incorporated into hollow cathodes which were used as primary radiation sources.The concentration of the individual isotopes was determined by measuring the absorption of isotopically pure vapours. BrimhaPo measured the concentrations of lead-206 lead-207 and lead-208 in solutions by using the hyperfine components of the 283.3-nm line. Three hollow cathodes each enriched in one of the isotopes were used and standard solutions of known isotopic composition were prepared. By using a standard atomic absorption system to measure the absorptions isotopic concentrations of an unknown were calculated by solving a set of three simultaneous equations. GolebS1 was able to determine uranium isotope ratios by using a hollow-cathode lamp as the absorption tube. Butler and SchroedersP2 reported the deter-mination of lithium isotope ratios by using a special instrument to enable the rapid and precise measurements of absorbance ratios.These atomic absorption procedures produced results in good agreement with mass spectrometric data. I thank the representatives of the various instrument manufacturers for their willingness to discuss various aspects and in particular Mr. W. J. Price for his valuable comments. I also thank my colleagues at Colgate-Palmolive Ltd. for their valuable assistance without which this work could not have been carried out. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 References Slavin W. Appl. Spectrosc. 1966 20 281. Slavin W. and Slavin S. Ibid. 1969 23 421. West T. S. Endeavour 1967,26 44. Reynolds R. J. Lab. Equip. Dig. 1968 6 79. Walsh A.Spectrochim. Acta 1955 7,‘108. Elwell W. T. and Gidley J. A. F. Atomic Absorption Spectrophotometry.’ Second Edition Pergamon Press 1966. Fassel V. A. and Becker D. A. Analyt. Chem. 1969,41 1522. Fassel V. A. Plenary Lecture International Atomic Absorption Spectroscopy Con-ference Sheffield July 1969. Winefordner J. D. y d Vickers T. J. Analyt. Chem. j970,42 206R. Ramfiez-Muiioz J . Atomic Absorption Spectroscopy Elsevier Publishing Company, 1968. Slavin W. ‘Atomic Absorption Spectroscopy,’ Interscience Publishers 1968. Cooke P. A. ‘Atomic Absorption Spectrophotometry,’ Second Edition Pye-Unicam Ltd., Price W. J. in Browning D. R. Editoor ‘Spectroscopy,’ McGraw-Hill 1969. RubeSka I. and Moldan B. ‘Atomic Absorption Spectrophotometry,’ Iliffe 1969 p. 154. Reynolds R.J. Aldous K. and Thompson K. C. ‘Atomic Absorption Spectroscopy-Pye-Unicam Ltd. Cambridge England. 1969. a Practical Guide.’ Charles Griffin 1970 228 PLATT 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 Perlun-Elmer Corporation Instrum News 19f9.20 15 Masek P R and Sutherland I Edztors Atomic Absorption and Flame Emission Stupar J and Dawson J B Appl Optzcs 1968,7 1351 Willis J B Spectrochzm Acta 1967 23A 811 Powell R J W ‘International Atomic Absorption Spectroscopy Conference,’ Sheffield, 1969 Ramirez-Mufioz J and Ulrich W F Flame Notes 1966 I(2) 33 Marshall D and Schrenk W G Spectroscopy Lett 1968 1 87 Ramfrez-Muiioz J Shifrin N and Hell A Mzcrochem J 1966 11 204 ‘Atomic Absorption Spectroscopy,’ Speczal Technzcal Publzcatzon No 443 American Rendma J Analyt Advances 1968 1 (2) Kxkbright G F Semb A and West T S Spectroscopy Lett 1968 1 (2) 97.Kahn H L J Chem Educ 1966 43A 7 Reynolds R J Metallurgza 1970 81 197 St John P A McCarthy W J and Winefordner J D Analyt Chem 1967,39 1495 Reynolds R J Lab Equzp Dzg 1969 7 75 Fisher A A and Hayward G C 21st Pittsburgh Conference on Analytical Chemistry Dagnall R M Thompson K C and West T S Atomzc Absorptzon Newslett 1967, Price W J Spectrovzszon 1966 No 15 2 Ramfrez-Muiioz J Amer Lab 1969 (August) Eel Bulletin No 7 1969 Muller R H Analyt Chem 1968,40 85A Eardley R P and Mountford A H Speczal Technzcal Publzcatzon No 59 British Dawson J B and Broughton P M G I Technzcal Bulletzn No 14 1968 Association Slavin W and Ringhardtz I Kern Teollzsuus 1966 11 985 Galassi M and Hell A Flame Notes 1966 1 (2) 28 Kahn H L J Chem Educ 1966,43 (2) A103 Sullivan J V and Walsh A Spectrochzm Acta 1965 21 721 Willis J B Rev Pure Appl Chem 1967 17 111 Cartwnght J S and Mannmg D C Atomzc Absorptzon Newslett 1966 5 114 Ure A M and Wtchell R I Spectrochzm Acta 1967 23B 79 Cartwright J S Sebens C and Slavm W Atomzc Absorptzon Newslett 1966 5 22 Price W J Spectrovzszon 1966 No 15 2 Manning R J Flame Notes 1968 3 ( l ) 11 Dawson J B and Ellis D J Spectrochzm Acta 1967,23A 565 Hussein A M I and Nickless G I InternationaI Atomic Absorption Spectroscopy Con-Cooke P A Process Bzochem 1967 2 (6) 27 Bowman J A Sullivan J V and Walsh A Spectrochzm Acta 1966 22 205 Kontyohann S R Atomzc Absorptzon NewsZett 1967 6 77 Lowe R M Spectrochzm Acta 1969 24B 191 Van Rensburg H C and Zeeman P B Analytzca Chzm Acta 1968,43 173 Butler L R P and Brink J A International Atomic Absorption Spectroscopy Mitchell D C and Johansson A Spectrochzm Acta 1970 25B 175 Fernandez F J Mamng D C and Vollmer J Atomzc Absorptzon Newslett 1969, Jaworowsh R J and Weberling R P Ibzd 1966 5 125 Frank C W Schrenk W G and Meloan C E Analyt Chem 1966 38 1005 Heneage P Atomzc Absorptzon Newslett 1966 5 67 Vollmer J Ibzd 1966 5 35 Schrenk W G and Meloan C E Spectrosc Lett 1968 1 237 Popham R E and Schrenk W G AppE Spectrosc 1968,22 192 Woodriff R A Wheeler G V and Ryder W A Ibzd 1968 22 348 Spectroscopy Abstracts,’ Science and Technology Agency Society for Testmg Materials Philadelphia 1968 and Applied Spectroscopy March 1970 6 117.Ceramic Research Association 1968 of Clinical Biochemists Scientific and Techmcal Communications ference Sheffield July 1969 Conference Sheffield July 1969 8 11 ATOMIC ABSORPTION SPECTROSCOPY 229 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 Fassel V. A. Mossotti V. G. Grossman W. E. L. and Kniseley R. N. Spectrochim. DeGalan L.McGee W. W. and Winefordner J. D. Analytica Chim. Acta 1967,37,436. McGee W. W. and Winefordner J. D. Ibid. 1967 37 429. Dagnall R. M. and West T. S. Appl. Optics 1968 7 1287. Dagnall R. M. Pribil R. and West T. S. Analyst 1968 93 281. Mansfield J. M. Bratzel M. P. Norgardon H. O. Knapp D. O. Zacha K. E. and Zacha K. E. Bratzel M. P. Winefordner J. D. and Mansfield J. M. Analyt. Chem., Headridge J. B. and Richardson J. Lab. Pract. 1970 19 372. Woodward C. Atomic Absorption Newslett. 1969 8 121. Browner R. F. Dagnall R. M. and West T. S. Analytica Chim. Acta 1969 45 163. Hoare H. C. Mostyn R. A. and Newland B. T. N. Ibid. 1968,40 181. West C. D. Analyt. Chem. 1968 40 253. Stupar J. and Dawson J. B. Atomic Absorption Newslett. 1969 8 38. Kirsten W.J. and Bertilsson G. 0. B. Analyt. Chem. 1966,38 648. Rawson R. A. G. Analyst 1966 91 630. Riley J. P. and Taylor D. Analytica Chim. Acta 1968 42 548. Veillon C. and Margoshes M. Spectrochim. Acta 1968 23B 553. Venghiattis A. A. AppL Optics 1968 7 1313. Hell A. Ulrich W. F. Shifrin N. and Ramirez-Muiloz J. Ibid. 1968,7 1317. Beckman Atomic Absoption Systems Bulletin 7134 p. 13. Gray J. M. and Gallwas G. E. Flame Notes 1967 2 2. Venghiattis A. A. Spectrochim. Acta 1967 23B 67. Instrum. News 1967 17 1. Kahn H. L. Peterson G. E. and Schallis J. E. Atomic Absorption Newslett. 1968,7 35. Curry A. S. Read J. F. and Knott A. R. Analyst 1969 94 744. White R. A, International Atomic Absorption Spectroscopy Conference Sheffield, Delves H. T. Analyst 1970 95 431 Willis J.B. Appl. Optics 1968 7 1295. Cowley T. G. Fassel V. A. and Kniseley R. N. Spectrochim. Acta 1968 23B 771. Mossotti V. H. ,and Duggan M. Appl. Ofltics 1968 7 1325. Price W. J. Efl. Wat. Treat. J. 1967,7 218. Amos M. D. and Willis J. B. Spectrochim. Acta 1966 22 1325. Willis J. B. Nature 1965 207 715. Amos M. D. Atomic Absorption Spectroscopy Symposium Cardiff April 1967. Willis J. B. Rasmuson J. O. Kniseley R. N. and Fassel V. A. Spectrochim. Acta, deGalan L. and Samaey G. F. International Atomic Absorption Spectroscopy Con-Kirkbright G. F. Peters M. K. and West T. S. Talanta 1967 14 789. Price W. J. Process Biochem. 1970 5 35. Pampel C. Atomic Absorption Newslett. 1967 6 133. Slavin W. Venghiattis A. and Manning D. C. Ibid. 1966 5 84. Manning D. C. Ibid.1966 5 127. Walsh A. J. N.Z. Inst. Chem. 1966 30 7. Bowman J. A. and Willis J. B. Analyt. Chem. 1967 39 1210. Shifrin N. Hell A. and Ramirez-Muiioz J. Appl. Spectrosc. 1969 23 365. Julietti R. J. and Wilkinson J. A. E. Ibid. 1969 23 173. Butler L. R. P. and Fulton A. Appl. Optics 1968 7 (lo) 2131. Fleming H. D. Sflectrochim. Acta 1967 23B 207. Kahn H. L. and Schallis J. E. Atomic Absorption Newslett. 1968 7 5. Dagnall R. M. Thompson K. C. and West T. S. Talanta 1968 15 677. Cooke P. A, International Atomic Absorption Spectroscopy Conference Sheffield, Ramirez-Mufioz J. Brace R. O. Claus R. W. and Shifrin N. Flame Notes 1967,2 30. Schallis J. E. and Kahn H. L. Atomic Absorption Newslett. 1968 7 75. Taylor J. H. Ibid. 1969 8 95. Acta 1966’22 347. Winefordner J.D. Spectrochim. Acta 1968 23B 389. 1968,40 1733. July 1969. 1968,22B 725. ference Sheffield July 1969. July 1969 PLATT Dagall R. M. Thompson K. C. and West T. S. Analyst 1968 93 153. Willis J. B. Fassel V. A. and Fiorino J. A. Spectrochzm. Acta 1969 24B 157. Bolmg E. A. Ibzd. 1966 22 425. Krrkbnght G. F. Sargent M. and West T. S. Atomzc Absorptzon Newslett. 1969 8 34. Kirkbright G. F. International Atomic Absorption Spectroscopy Conference Sheffield, RubeSka I. and Moldan B. Analyst 1968 93 148. - and - Appl. Optzcs 1968 7 (7) 1341. West T. S and Wdliams X. K Analytzca Chzm. Acta 1969 45 27, L’vov B. V. Spectrochzm. Acta 1969 24B 53. Massmann H. Ibzd. 1968,23B 215. Anderson R. G. Maines I. S. and West T. S International Atomic Absorption Spectro-Tomkms R.S. and Ercoli B. Appl. Optzcs 1967 6 (8) 1299. Woodriff R. and Stone R. W. Ibzd. 1968 7 (7) 1337. Woodriff R. Stone R. W. and Held A. M. Appl. Spectroscopy 1968,22 408. Woodriff R. and Ramelow G. Spectrochzm. Acta 1968,23B 665. Mornson G. H. and Talmi Y. Analyt. Chem 1970,42 809. Mossotti V. G. Laqua K. and Hagenah W. D. Spectrochzm. Acta 1967,2333 197. Karyakin A. V. and Kagarodov V. A. Zh. Analzt. Khzm 1968 23 930. Wendt R. H. and Fassel V. A. Analyt. Chem. 1966 38 337. Hoare H. C. and Mostyn R. A. Ibzd. 1967 39 1153. Vedlon C. and Margoshes M. Spectrochzm. Acta 1968,23B 503. Aldous K. M Dagnall R. M. Thompson K. C. and West T. S. Analytzca Chzm. Ada, Greenfield S. Smith P. B Breeze A. E. and Chdton N. M D. Ibzd 1968,41 385. Sullivan J. V.and Walsh A. Spectrochzm. Acta 1966,22 1843. Brandenberger H. Chzmza 1968,22 449. Sullivan J. V. and Walsh A. Appl. Optzcs 1968 7(7) 1271. Rawling B. S and Sullivan J. V. Mzneral Process. Extract. Metall. 1967 76 (c) 238. Boar P. L. and Sullivan J. V. Fuel 1967 46 47. Bowman J. A. Analytzca Chzm. Acta 1967 37 465. Johnson J. D. Yamasaki G. K and Burger J. C zn Grove E L and Perhns A J., Edztors Developments in Applied Spectroscopy Symposium,’ Vol IIIA Plenum Press 1969. Muller R. H. Analyt. Chem. 1967 39 99A. Kahn H. L. International Atomic Absorption Spectroscopy Conference Sheffield, Harrison W. W. and Berry I;. E. Analytzca Chzm Acta 1969 47 415. Evans Electroselenium Ltd. ‘Model 240 Brochure.’ Ramirez-Mufioz J Malakoff J. L and Aime C P. Analytzca Chzm.Acta 1966’36 328. Wendt R. H. Atomzc Absorptzon Newslett. 1968 7 28. Malakoff J. L Ramfrez-Muiioz J and Aime C. P Analytzca Chzm. Acta 1968,43 37. Malakoff J. L. Ramfrez-Muiioz J and Scott W. Z. Ibzd. 1968,42 515. Instrum News 1969,20 15. Slavin W. International Atomic Absorption Spectroscopy Conference Sheffield, Instrum. News 1968 19 6. Dawson J. B. Ellis D. J and Mdner R. Spectrochzm. Acta 1968,23B 695. Lacy J. Analyst 1965 90 65. Klein B Kaufman J. H andMorganstern S. Clzn. Chem 1967.13 388. Crawford L. R jun and Greweling T Appl Spectroscopy 1968,22 793. Ellis D J. and Rogers J Speclrovzszon 1967 17 2. Lewis R. R Atomzc Absorptzon Newslett. 1969 8 94. Kokot M. L and Butler L. R. P. Ibzd. 1969 8 92. Watson C. A. Monograph 74 Hopkin & Williams Ltd.Roth M. E. Flame Notes 1969 4 4. Ramfrez-Mufioz J. J. Forens. Scz. Soc. 1967 7 151. - Flame Notes 1967 2 54. - and - Talanta 1969 16 1467. July 1969. scopy Conference Sheffield July 1969. 1968 41 380. July 1969. July 1969. 230 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 17 ATOMIC ABSORPTION SPECTROSCOPY 231 173 Ramirez-Muiioz Micvochem. J. 1967 12 196. 174 - J. Talanta 1969 16 1037. 175 Grunder F. I. and Boettner E. A. in Grove E. L. and Perkins A. J. op. cit. p. 201. 176 Lewis L. L.Analyt. Chem. 1968,40 28A. 177 Cooke P. A. and Price W. J. Spectrovision 1966 16 7. 178 Fassel V. A. Rasmuson J. O. and Cowley T. G. Spectrochim. Acta 1968,23B 579. 179 Koirtyohann S. R. and Pickett E. E. Analyt. Chem. 1966,38 585. 180 Kahn H. L. Atomic Absorption Newslett. 1968 7 40. 181 Perry C. J. and Keyworth D. A. Can. Spectroscopy. 1967 12 47. 182 Kerber J. D. Atomic Absorption Newslett. 1967 6 131. 183 Parsons M. L. and Winefordner J. D. Appl. Spectrosc. 1967 21 368. 184 Rohleder H. A. International Atomic Absorption Spectroscopy Conference Sheffield, 185 Cellier K. M. and Stace H. C. T. Appl. Spectrosc. 1966 20 26. 186 Manning D. C. and Capacho-Delgado L. Analytica Chim. Acta 1966 36 312. 187 Brace R. O. Flame Notes 1966 1 2. 188 Ramirez-Muiioz J. in Grove E.L. and Perkins A. J. op. cit. p. 169. 189 Reynolds R. J. Wld Med. Instrumn 1969 7 10. 190 RubeSka I. and Moldan B. Analytica Chim. Acta 1967 37 421. 191 Halls D. J. and Townshend A. Ibid. 1966 36 278. 192 Popham R. E. and Schrenk W. G. in Grove E. L. and Perkins A. J. op. cit. p. 189. 193 Koirtyohann S. R. and Pickett E. E. Analyt. Chem. 1968,40 2068. 194 Sastri V. S. Chakrabarti C. L. and Willis D. E. Can. J. Chem. 1969,47 587. 196 - - and - International Atomic Absorption Spectroscopy Conference, 197 Hartlage F. R. jun. Analytica Chim. Ada 1967 39 273. 198 Ramirez-Muiioz J. Analyt. Chem. 1970,42 517. 199 Munro D. C. Appl. Spectrosc. 1968 22 199. 200 Mitchell D. G. Lab. Pract. 1967 16 587. 201 Kohlenberger D. W. Atomic Absorption Newslett. 1969 8 108.202 Guillaumin R. Revue Fr. Cps Gas 1969 16 497. 203 PrbvBt A. Atomic Absorption Newslett. 1966 5 13. 204 Piccolo B. and O’Connor R. T. J Amer. Oil Chem. SOC. 1968 45 789. 205 Montford B. and Cribbs S. C. Tuluntu 1969 16 1079. 206 Orren M. J. S.A. Chem. Processing 1968,3 36. 207 Peterson G. Atomic Absorption Newslett. 1966 5 117. 208 Margoshes M. Anulyt. Chem. 1967,39 1093. 209 Ramirez-MuEioz J. and Roth M. E. Flame Notes 1969 4 28. 210 - and - Ibid. 1969 4 62. 211 Takeuchi T. Suzuki M. and Yanagisawa M. Anulytica Chim. Acta 1966,36 258. 212 Mulford C. E. Atomic Absorfition Newslett. 1966 5 88. 213 Brooks R. R. Presley B. J. and Kaplan R. Talantu. 1967 14 809. 214 Montford B. Can. Spectrosc. 1968 13 126. 216 Husler J. W. and Cruft E. F. Analyt. Chem 1969,41 1688.216 Campbell D. J. Atomic Absorfition Newslett. 1967 6 49. 217 Chau Y. K. and Lum-Shue-Chan K. Anulytica Chim. Acta 1969,48 434. 218 Shepherd G. A. and Johnson A. J. Atomic Absorption Newslett. 1966,s; 142. 219 Manning D. C. Ibid. 1968 7 44. 220 Manning D. C. and Chabot H. Ibid. 1968 7 94. 221 Scheub W. H. and Stromsky C. J. Ibid. 1967,6 95. 222 Ramirez-Mufioz J. Flame Notes 1969 4 12. 223 Rooney R. C. Electron. Equip. News May 1967. 224 Kahn H. L. Atomic Absorption Newslett. 1967 6 61. 226 Feldman F. J. Blasi J. A. and Smith S. B. jun. Analyt. Chem. 1969,41 1095. 226 Sattur T. W. Atomic Absorption Newslett. 1966 5 37. 227 Meddings B. and Kaiser H. Ibid. 1967,6 28. 228 Welcher G. G. and Kriege H. Ibid. 1969 8 97. 229 Johns P. and Price W. J. Metallurgia 1970 81 75.July 1969. 195 - - and - Talanta 1969 16 1093. Sheffield July 1969 232 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 25 1 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 PLATT Welz B. Joint Symposium on Accurate Methods of Analysis for Major Constituents; S.A.C. & Koninklijke Nederlandse Chemishe Vereniging (Analyt. Sectn) London, April 1970. Cobb W. D. and Harrison T. S. Details as Ref. 230. Langmyhr F. J. and Paus P. E. Atomic Absor$tion Newslett. 1968 7 103. - and - Ibid. 1969 8 131. Bernas B. Analyt. Chem. 1968 40 1682. Van Loon J.C. and Parissis C. M. Analyst 1969 94 1057. Omang S. H. Analytica Chim. Acta 1969 46 225. Muter R. B. and Cockrell C. F. Ap$L Spectrosc. 1969 23 493. Medlin J. H. Suhr N. H. and Bodkin J. B. Atomic Absorption Newslett. 1969 8 25. Yule J. W. and Swanson G. A. Ibid. 1969 8 30. Boar P. L. and Ingram L. K. AnaZyst 1970 95 124. Campbell D. E. Analytica Chim. Acta 1969 46 31. Karmie Galle O. Appl. S$ectrosc. 1968 22 404. Crow R. F. Hime W. G. and Connolly J. D. J . Portland Cem. Ass. Res. Dev. Labs., Capacho-Delgado L. and Manning D. C. Analyst 1967,92 553. Roos J. H. T. and Price W. J. Analyst 1969 94 89. Slavin W. Venghiattis A. and Manning D. C. Atomic Absorption Newslett. 1966 5 84. Kerber J. D. and Barnett W. B. Ibid. 1969 8 113. Fulton A. and Butler L.R. P. S9ectrosc. Lett. 1968 1 317. Ramakrishna T. V. West P. W. and Robinson J. W. Analytica Chim. Acta 1968, Newsbitt R. W. Jbid. 1966 35 413. Ramakrishna T. V. West P. W. and Robinson J. W. Ibid. 1967 39 81. Pawluk S. Atomic Absorption Newslett. 1967 6 53. Laflamme Y. Ibid. 1967 6 70. Van Loon J. C. Ibid. 1968 7 3. Manning D. C. fbid. 1967 6 35. Bokowski D. L. Ibid. 1967 6 97. Peterson E. A. Ibid. 1969 8 53. Bade1 H. and Brandenberger H. Ibid. 1968 7 1. Harris R. Ibid. 1969 8 42. Roos J. T. H. International Atomic Absorption Spectroscopy Conference Sheffield, Rooney R. C. and Pratt C. G. International Atomic Absorption Spectroscopy Con-Ramirez-Mufioz J. and Roth M. E. Flame Notes 1968 3 2. Wilson L. Analytica Chim. Acta 1968 40 503. Popham R.E. and Schrenk W. G. Spectrochim Acta 1968,23B 543. Kirkbright G. F. Smith A. M. and West T. S. Analyst 1966 91 700. Ramakrishna T. V. West P. W. and Robinson J. W. Analytica Chim. Act@ 1969,41, Endo Y. Hata T. and Nakahara Y. JaFan Analyst 1969 18 878. Juliettu R. J. and Wilkinson J. A. E. Analyst 1968 93 797. Mostyn R. A. and Cunningham A. F. J . Inst. Pet. 1967 53 101. Chan Y . K. and Lum-Shue-Chau K. Analytica Chim. Acta 1969 48 205. Butler L. R. P. and Mathews P. M. Ibid. 1966 36 319. Ure A. M. International Atomic Absorption Spectroscopy Conference Sheffield July Kirkbright G. I?. Peters M. K. and West T. S. Analyst 1966 91 705. Reamish F. E. Lewis C. L. and Van Loon J. C. TaZanta 1969 16 1. Van Loon J. C. 2. Analyt. Chem. 1969,246 122. Fuhrman D. L. Atomic Absorption Newslett.1969 8 105. Fernandez F. Ibid. 1969 8 90. Osolinski T. W. and Knight N. H. A$pZ. Spectrosc. 1968 22 532. Atwell M. G. and Herbert J. Y. Ibid. 1969 23 480. Johns P. and Price W. J. Pittsburgh Conference on Analytical and Applied Spectro-1967 (May) 60. 40 350. July 1969. ference Sheffield July 1969. 437. 1969. scopy March 1970 ATOMIC ABSORPTION SPECTROSCOPY 233 281 282 283 284 285 286 287 288 289 290 29 1 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 32 1 322 32 3 324 325 326 327 328 329 330 33 1 332 Aldous K. M. Dagnall R. M. and West T. S. Analyst 1969 94 347.Jaworowski R. J. Weberling R. P. and Bracco D. J. Analytica Chim. Acta 1967 37, Kinnunen J. and Lindsjo O. Chemist Analyst 1967 56 25. Van Loon J. C. Aarden D. and Galbraith J. International Atomic Absorption Hingle D. N. Kirkbright G. F. and West T. S. Analyst 1969,94 864. Chau Y. K. Talanta 1968 15 421. Biechler D. G. and Long C. H. Atomic Absorption Newslett. 1969 8 56. Price W. J. and Roos J. T. H. Analyst 1968 93 709. McAuliffe J. J. Atomic Absorption Newslett. 1967 6 69. Morrow R. W. and Dean J. A. J . Chromatog. Sci. 1969 7 572. Paralusz C. M. Appl. Spectrosc. 1968 22 520. Mario E. and Gerner R. E. J . Pharm. Sci. 1968 57 1243. Dagnall R. M. Kirkbright G. F. West T. S. and Wood R. Analytica Chim. Acta 1969, Capacho-Delgado L. and Manning D. C. Spectrockim.A d a 1966,22 1505. Juliano P. O. and Harrison W. W. Analyt. Chem. 1970 42 84. Bowman J. A. Analytica Chim. Acta 1968,42 285. Shannon I. L. Caries Res. 1969 3 339. Headridge J. B. and Hubbard D. P. Analytica Chim. Acta 1967 37 151. Mostyn R. A. and Cunningham A. F. Atomic Absorption Newslett. 1967 6 86. Capacho-Delgado L. and Manning D. C. Ibid. 1966 5 1. Sachdev S. L. Robinson J. W. and West T. W. Analytica Chim. Acta 1967 37 12. and - Ibid. 1967 37 156. Hall G. Cochrane I. G. and Dorman R. W. International Atomic Absorption Spectro-scopy Conference Sheffield July 1969. Tyler J. B. Atomic Absorfltion Newslett. 1967 6 14. Menis O. and Rains T. C. Analyt. Chem. 1969 41 952. Holak W. Ibid. 1969’41 1713. Ando A. Suzuki M. Fuwa K. and Vallee B. Ibid. 1969 41 1974.Chakrabarti C. L. Analytica Chim. Acta 1969,42 379. - Ibid. 1967 39 293. Barnett W. B. and Kahn H. L. Atomic Absorption Newslett. 1969 8 21. Slavin S. and Sattur T. W. Ibid. 1968 7 99. Zaugg W. S. and Knox R. J. Analyt. Biochem. 1967 20 282. Zaugg W. S. Atomic Absorption Newslett. 1967 6 63. Kirkbright G. F. Smith A. M. and West T. S. Analyst 1967 92 411. Hurford T. R. and Boltz D. F. Analyt. Chem. 1968,40 379. Ramakrishna T. V. Robinson J. W. and West P. W. Analytica China. Acta 1969,45, Rose S. A. and Boltz D. F. Analytica Chim. Acta 1969 44 239. Jungreis E. and Anavi Z. Ibid. 1969 45 192. Dunk R. Mostyn R. A. and Hoare H. C. Atomic Absorption Newslett. 1969,8 79. Kadow A. and Rabban E. International Atomic Absorption Spectroscopy Conference, Yamamoto Y. Kumamaru T. Hayashi Y. Otani Y. Japan Analyst 1969 18 359. Danchik R. S. and Boltz D. F. Analyt. Chem. 1968 40 2215. Bond A. M. and O’Donnell T. A, Ibid. 1968 40 560. Reichel W. and Acs L. Analyt. Chem. 1969 41 1886. Ezell J. B. jun. Atomic Absorption Newslett. 1967 6 84. Bartels H. Ibid. 1967 6 132. Westerland-Helmerson U. Ibid. 1966 5 97. Danchik R. S. and Boltz D. F. Analyt. Lett. 1968 1 891. Bond A. M. and Willis J. B. Analyt. Ckem. 1968 40 2087. Kirkbright G. F. Rao A. P. and West T. S. Sflectrosc. Lett. 1969,2 69. Kamamaru T. Hayashi Y. Okamoto N. Tao E. and Yamamoto Y. Analytica Chim. Yamamoto Y. Kamamaru T. and Hayashi Y. Talanta 1967 14 611. 284. Spectroscopy Conference Sheffield July 1969. 47 407. - , 43. Sheffield July 1969. Acta 1966 35 524 234 PLATT Potter A. L. Ducay E. D. and McCready R. M. J . Ass. 08. Agric. Chem. 1968 51, Sheridan J. C. Lawu E. P. K. and Senbowski B. Z. Analyt. Chem. 1969 41 247. Brandenberger H. and Bader H. Atomic Absorption Newslett. 1967 6 101. - and - Ibid. 1968 7 53. Thilliez G. Ckim. Analyt. 1968 50 226. Goleb J. A. Anal. Chim. Acta 1966 36 130. Kirchhof H. Spectrochim. Acta 1969 24B 235. Brimhall W. H. Analyt. Chem. 1969 41 1349. Goleb J. A. Analytica Chim. Acta 1966 34 135. Butler L. R. P. and Schroeder W. W. International Atomic Absorption Spectroscopy 748. Conference Sheffield July 1969. 333 334 335 336 337 338 339 340 341 34
ISSN:0300-9963
DOI:10.1039/AS9710100177
出版商:RSC
年代:1971
数据来源: RSC
|
5. |
Catalytic methods in analytical chemistry |
|
Selected Annual Reviews of the Analytical Sciences,
Volume 1,
Issue 1,
1971,
Page 235-269
G. Svehla,
Preview
|
PDF (3919KB)
|
|
摘要:
Catalytic Methods in Analytical Chemistry G. SVEHLA Department of Analytical Chemistry Queen's University Belfast Contents Introduction Quantitative catalytic analysis Classification of the methods-Analytical classification -Chemical classification -Kinetical classification -Experimental classification Determination of metals-Aluminium -Barium -Beryllium -Bismuth -Calcium -Chromium -Cobalt -Copper -Germanium -Iron -Magnesium -Manganese -Mercury -Molybdenum -Nickel -Osmium -Ruthenium -Silver -Strontium -Thallium -Tungsten -Vanadium -Zinc 235 Determination of non-metallic substances and anions-Bromide -Cyanide -Hydrogen peroxide -Iodide -Oxygen -Phosphate -Selenium -Sulphide Catalytic and kinetic determination of organic substances-Amines -Ascorbic acid -0-Chloronitrobenzene -Complexing agents -Dihydric phenols -Esters -Ethanolamides -Ketones -Nap11 thols -Oxidative enzymes -Pesticides -Phenols -Serum enzymes and their substrates Qualitative catalytic analysis Application of catalytic waves in polarography Kinetochromic spectrophotometry Catalymetric titrations New techniques and instruments Theory of catalytic analysis reviews 23 CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 237 Introduction Catalytic or kinetic methods of chemical analysis are based on the well known fact that small amounts of catalysts can initiate or accelerate certain chemical reactions.Some reactions take place or are accelerated only in the presence of the catalyst and by the kinetic examination of the reaction the presence or absence of the catalyst substance can be ascertained by qualitative tests.As the amount of catalyst required for the initiation or acceleration of the reaction in usually very small these tests are very sensitive. The selectivity and specificity of such tests can vary according to the nature of the reactions involved; in some cases the catalytic action is specific in others a number of chemically related species e.g., transition metals show almost the same catalytic action and the selectivity of the tests is poor. The importance of qualitative tests made by catalytic methods is far surpassed by the importance of quantitative determinations which may offer cheap and sensitive alternatives to methods that require expensive instrumentation.These quantitative determinations are generally based on slow reactions proceeding in solution containing a suitable catalyst. In the absence of the catalyst the slow uncatalysed reaction proceeds alone; if the catalyst is added a new route for the over-all reaction in which the catalyst also takes part becomes available. The catalyst is then regenerated by another reaction that is the catalyst is involved in a reaction cycle. There are therefore two competing reactions proceeding at the same time the uncatalysed reaction and the catalysed one. The reaction rate depends on the contribution of these competing reactions to the over-all process. If the concentration of the catalyst is too low the share of the catalysed reaction in the over-all process becomes negligible and therefore there is no measureable change in the over-all reaction rate compared to that obtained in the absence of the catalyst.If on the other hand the concentration of the catalyst becomes so high that the contribution of the uncatalysed reaction to the over-all process becomes negligible and the catalytic reaction becomes predominant the rate of the over-all process becomes very high and somewhat insensitive to changes in the concentra-tion of the catalyst. Between these two extremes however there exists a con-centration range in which altering the concentration of the catalyst produces changes in the contribution of the two competing reactions to the over-all process. Accordingly the reaction rate will vary with the concentration of the catalyst.By measuring the reaction rate as a function of the catalyst concentration a calibra-tion curve can be obtained and used for practical analytical purposes. This review deals with the latest development of catalytic methods of analysis. The literature of the period January 1967 to June 1970 is covered. The earlier literature is well reviewed in the excellent monograph of Mark and Rechnit2.l Another monograph on the subject by Yatsimirskii2 contains selected procedures for the determination of certain ions and can also be used with advantage. Because of the importance of quantitative determinations these methods will be reviewed first and in most detail. This is followed by a short discussion of new Q 238 SVEHLA qualitative catalytic tests.The application of catalytic waves in polarography is followed by reviews of kinetochromic spectrophotometry and of catalymetric titrations. The review is concluded with accounts on new techniques and instru-ments and on the development of the theory of catalytic methods. Quantitative Catalytic Analysis Classification of the methods The literature of quantitative catalytic analysis is so vast and is growing so fast that it is very important to classify the various methods into separate groups. This classification can be made from various points of view. In this review the material is arranged so that it is most suitable for a practising analyst who wishes to choose a method for his particular purpose. This is based on the chemical nature of the analyte.The chemical classification is based on the nature of reactions involved in the method. A kinetical classification on the other hand is based on the principle involved in the measurement of the reaction rate while an experi-mental classification can be made according to the techniques used in measuring the concentration of the reactants. As the principles of quantitative catalytic methods can be elucidated most conveniently according to these principles a more detailed description of these classes is given. Analytical classification. From the practical point of view it is found convenient to classify the substances that can be determined by quantitative catalytic methods into three groups metals non-metallic inorganic substances and organic substances.Within each group substances are dealt with in alphabetical order. Chemical classification. This method of classification is based on the type of chemical reaction that is involved in the process. Thus we can distinguish between (a) redox (or electron-exchange) reactions (b) ligand-exchange reactions, (c) enzyme-catalysed reactions and (d) catalysed electrode reactions. Though the last of the classes might fall into category (a) it is logical to treat them separately as they are applied for quite special purposes. Redox reactions are by far the most common processes among catalytic reactions. In these reactions two redox couples take part. Considering homo-geneous redox systems only these can be characterised with the equilibria and and with the oxidation reduction potential El and E for systems 1 and 2 respec-tively.If El > E a reaction between Ox and Red will take place-n,Ox + n,Red + 2(m,n,-m2n,)Hf -+ %,Red + n,Ox + (m,n,-m,n,)H,O If lzl and/or n are higher than 1 more electrons have to be exchanged during the process. Such a process however cannot take place in one single step and so the mechanism of such processes includes a series of reactions with the transfer of Ox + 2m,H+ + n,e- =+ Red + m,H,O Ox + 2m,H+ + n,e- + Red + m,H, CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 239 one electron in each step. Some of these steps axe often slow. If a catalyst is present which reacts with one of the original reactants in such a way that an alternative reaction mechanism becomes available the over-all reaction rate increases considerably.This mechanism involves the product of the above reaction reacting with the other reactant one of the products of this process being the original form of the catalyst. Thus the catalyst is involved in a reaction cycle. As the catalytic reaction is also an oxidation - reduction reaction the catalyst itself must be a part of a redox system with an oxidation-reduction potential lying between those of the two reactant systems. This is the reason why the transition metal ions are so often found to be good catalysts of such reactions. Ligand-exchafige reactions represent a new so far rather neglected field of catalytic analysis. It is known that many ligand-exchange reactions between complexes MfL1 and M2L2 (where M1 and M2 represent metals L-l and L2 ligands) are slow-MIL1 + MaLB+MIL* + MaL1 The reason for this is that these over-all reactions take place as a series of dis-sociation and re-combination steps such as-MIL1 + MI+ L1 MaL8 +t M8 + L8 M1+ L* +MIL' M* + L1 + MaL1 Because of the high stability of the original complexes the over-all reaction is slow.Any of the free metal ions Mf or M2 or the ligands L1 or L2 will catalyse the reaction. If for example a small amount of the free ligand L1 is available the reactions L1 + MaL* 3 M4L1 + L* and L2 + MIL1 3 M1L2 + L1 will proceed and the over-all reaction rate increases. The ligand L1 is regenerated and a new reaction cycle can be initiated. Only a few of these systems have been investigated in the past but there is no doubt however that ligand-exchange reactions represent a potentially important field of catalytic analysis.Enzyme-catalysed reactiorts are frequently used in clinical food and pharma-ceutical laboratories mainly as a preliminary treatment for analysis. They can be used directly however for kinetic determination of the enzymes themselves or of the substrates activators or inhibitors. These reactions proceed in two steps. First the enzyme E and the substrate S form a transitional species ES in an equilibrium reaction-E + S t E S The transitional species then decomposes in a slow reaction to form the product P, and the enzyme is re-combined-ES+P+E The concentration of both the enzyme and substrate influence the reaction rate as do certain activators and inhibitors. Thus there are many possibilities for th 240 SVEHLA analytical applications of enzyme-catalysed reactions.One of the great advantages of enzyme-catalysed reactions is their specificity both towards the enzymes themselves and towards the activators and inhibitors. As enzymes of guaranteed purity can be purchased easily the importance of such reactions is growing steadily. CataZysed electrode reuctions known as the catalytic currents in polarography, are themselves redox processes but as they occur on the electrode surface and therefore need special equipment it is better that they are treated separately. To understand the nature of such processes let us consider the polarographic reduction of a catalyst (most often a metal ion) if present alone (with a suitable supporting electrolyte) If the potential applied is adequately chosen the reduction will proceed on the electrode giving rise to a diffusion current id which is governed by the Ilkovic equation.If however another substance A is present which re-oxidises the product of the above reaction C(m-n)+ + A +- Cm+ + B (where B is another reaction product) the current will increase considerably. (Note that neither A nor B is reduced or oxidised on the electrode at the given potential.) This catalytic current i, depends on the concentration of the catalyst in a rather complex way”; by its measurement the concentration of the catalyst can be determined by using calibration curves. It is not therefore the reaction rate that is measured and none of the reactions involved need be slow. The process can be considered however as the ‘catalytic’ reduction of A to B on the electrode and in the absence of C the reaction would not be thermodynamically possible.The product B is often hydrogen gas and therefore the role of the catalyst is the decrease of the overvoltage of hydrogen on the electrode (mainly the drop-ping mercury electrode). The role of the catalyst is in the initiation rather than the acceleration of the reaction. References to the above classes are frequently made in this review. Kinetical classification. Following the rate of any chemical reactions entails the measurement of two parameters concentration and time at certain intervals while keeping the temperature constant. The results with the aid of correlations of chemical kinetics can then be interpreted and used for analytical purposes.Although almost each author suggests an individual method of measur-ing the reaction rate these methods fall broadly into four groups; the initial rate method ; fixed Concentration methods ; fixed time methods and the simultaneous comparison method. The initial rate method is recommended by a number of authors for reactions where only the initial period of the reaction can be interpreted simply. This is the case if the product of the slow process is involved in a second slow reaction Le., consecutive reactions occur or if as the reaction product builds up the reaction is reversed Le. the reaction leads to an equilibrium. By plotting the concentration of one of the reactants or products as a function of time only the initial part of this Cm+ + ne- j C(m-n) CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 241 curve will be meaningful as regards evaluation (though it is necessary to take measurements over a longer period so as to reproduce the initial part of the curve in a proper way) By definition the slope of such curves taken at zero time is equal to the initial rate.This initial rate depends linearly on the concentration of the catalyst and the latter can be evaluated easily from a calibration graph. The weakness of the method lies in the difficulty in accurately drawing or measuring the initial slope. The fixed reaction time method involves the measurement of the concentra-tion of a reactant or product after a pre-determined lapse of time. If the order of the reaction and the rate constant are known these results can easily be inter-preted kinetically.However even if the kinetics of the reaction have not been investigated the results can be applied for analysis provided that all the experi-mental circumstances are kept constant and the only parameter varied during the experiments is the concentration of the catalyst. The calibration graph where the measured quantity (which is proportional to the concentration or the decrease or increase of concentration of a reactant or product) is plotted against the concentra-tion of the catalyst is a straight line in most cases. These calibration curves can be used in the analysis of unknown samples. The fixed concentration method is in many ways similar to the previous one, but instead of measuring the concentration after a fixed time the time is measured during which the concentration of a reactant or product changes to a pre-deter-mined value.These results are again easy to interpret kinetically if the order of the reaction and the rate constant are known; the integrated rate equations contain all of the parameters measured but again even if the kinetics of the reaction have not been investigated the method may be used for analytical purposes provided that all experimental circumstances are kept constant with the exception of the con-centration of the catalyst. Plotting the reciprocal values of these time values (the so-called reaction times) against the concentration of the catalyst a straight-line relationship is obtained in most cases and this calibration curve can be used in the analysis of unknown samples.One special form of the fixed concentration method is the application of Landolt reactions for analysis. As the method is based on the measurement of time it is often referred to as a ‘chronometric method.’ The method of simultaneous comparison is a simple technique that is suitable if the reaction is accompanied by a visible change of colour or optical density. A set of reagent solutions is prepared and placed into identical beakers or test-tubes. One of these contains the unknown sample while the others contain known amounts of the catalyst. With a special device a so-called ‘starter pipette,’ equal amounts of the other reagents are added to each of the solutions mixed and the colours of the solutions watched.As the time goes on the differences between colours of solutions with different amounts of catalyst in them becomes more and more apparent. The colour of the solution with the unknown concentration is compared to those of the others. The unknown concentration will be equal to the concentra-tion of the catalyst in the vessel that has a similar colour to that containing the sample. By repeated experiments in which the concentrations of the compariso 242 SVEHLA solutions are chosen according to the findings of the first one the concentration of the unknown solution can be reliably determined. The accuracy of the measure-ment is highly dependent on the reliability of the starter pipette and the skill of the operator. As quantitative measurements are not made the results are somewhat subjective and they cannot be interpreted kinetically.Experimental classification. As already mentioned catalytic analyses are based on the measurement of concentration and time. There is little variation in the methods for the measurement of time a stop-watch generally being used. There are several methods for the measurement of concentration and the instru-mentation differs substantially from one to another. The main methods of measuring the concentration are visual photometric potentiometric and gas volumetric although some special methods cannot be so classified. A visual method is the simplest and cheapest as it does not need any instru-ment. Naturally one cannot measure concentrations by the eye and therefore the visual method is restricted to certain but important special determinations.The method of simultaneous comparisons discussed above applies visual examination and provides a high degree of reliability. The Landolt reaction method uses a rapid colour change as the indication of the end of the period to be measured and this can be seen easily by the human eye. Some of the older catalytic techniques which are not included in the present review applied a visual technique when the time of complete decolourisation of a solution had to be measured. These methods, though simple offered only a very limited precision and have merely historical importance. The photometric method of monitoring the concentration of a reactant or product offers a high degree of precision. The solutions are mixed in an ordinary beaker or test-tube and the measurement of time is started at the same instant.The mixture is then poured into the cell of a photometer and its optical density measured. The main disadvantage of the method is that it is not easy to keep the temperature of the mixture constant which is an essential requirement of any catalytic method. The danger of heating the solution is especially high if it has to be irradiated by the light beam of the instrument without interruption (if for example the optical density is recorded continuously during this period). Spectro-photometers with thermostated cells can be used with advantage. The potentiometric method is based on the measurement of a reversible electrode potential in the solution. As defined by the Nernst equation the electrode potential depends on the concentration of a particular species.Its application is somewhat restricted as there are only a few really reversible electrodes available that respond instantaneously to changes in concentration. The hydrogen ion Concentration can be monitored easily with the glass electrode. As there are many slow oxidation - reduction reactions in which hydrogen ions are involved the potentiometric method might be the obvious choice especially if there is no visible change of colour occurring at the same time. The weakness of the potentiometric method lies in the high error built into such measurements because of the logarith CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 243 mic correlation between electrode potential and concentration.It is easy to show with a simple error calculation that if the electrode potential is measured with a precision of &l mV (or the pH monitored with a precision of &0*02) the relative error committed in the measurement of the concentration is about 4 per cent., which is quite high even for the purpose of catalytic analysis. The gas volumetric method is based on the measurement of volumes of gases evolved in the course of a reaction. This technique is frequently used in connection with processes involving the catalytic decomposition of hydrogen peroxide where the product is oxygen gas. This method is somewhat clumsy and is sensitive to changes in environmental temperature and pressure and of course is restricted to reactions where there is a gas among the products of the reaction.Besides these main types there are quite a few other methods suggested for the monitoring of the concentration of reagents or products. Fluorimetric monitoring is possible if one of the reagents or products emits fluorescence radiation when irradiated with an ultraviolet light. If saccharides are involved in the reaction, polarimetric monitoring can be used. The occurrence or disappearance of chemi-luminescence can be used if chemiluminescent materials are involved in the reac-tion. As these are good indicators of hydrogen peroxide in alkaline media they can also be used if hydrogen peroxide is one of the reactants. These methods however, are limited to certain substances and are applied only occasionally. Determination of metals The metals are listed in alphabetical order and the information on each method is presented in a condensed way including whenever available the classification of the technique according to the principles described above the reaction and the catalytic effect on which the method is based the concentration range within which the determination can be carried out information on the calibration curve and data on accuracy and precision and finally interferences and their elimination.When the original paper was not available the information is from the abstract. Aluminium. Continuing their work on the examination of the catalytic decarboxylation of oxaloacetic acid with various substituted pyridines Michaylova and Bontchev4 worked out a method for the determination of aluminium.A fixed reaction time method was used and the pressure of the carbon dioxide gas collected into a vessel is measured 35 minutes after the initiation of the reaction. The minimum amount of aluminium that can be determined is 06pg and 25-pg amounts of copper iron or zinc 250 pg of lead magnesium and cadmium and 4000pg of calcium can be tolerated. The calibration curve is constructed by plotting the pressure of carbon dioxide versus concentration of aluminium. Though not based strictly on catalytic action the kinetic method of Bognar and Pataky-Szabo5 for the determination of 0.1 to 1 pg of aluminium in 5 ml of solution has to be mentioned here. The slow reaction between aluminium and Pontachrome-Violet SW or Pontachrome-blue black R at pH 5 is monitored by examining the fluorescence of the product by a simultaneous comparison method.The error of the determination is reported to be below 10 per cent 244 SVEHLA Barium. A kinetic although not strictly catalytic method for the deter-mination of barium calcium magnesium and strontium was described by Pausch and Margerum.6 The ligand-exchange reaction between lead(I1) and the alkaline earth complexes of t~ans-l,2-diaminocyclohexane-N,N,N',N'-tetraacetate (CyDTA) was used. A photometric measurement of the optical density of the lead complex was coupled with an oscilloscopic observation the oscillograms were photographed and used for evaluation. The method is suitable for the simultaneous determination of the four alkaline earth ions. Beryllium. Two methods have been described recently for the determination of beryllium both based on the inhibition of enzyme-catalysed reactions.Towns-hend and Vaughan7 recommend the use of calf intestinal alkaline phosphatase, which catalyses the hydrolysis of phosphate esters. This hydrolysis is inhibited by 20 to 90-ng amounts of beryllium in 7 ml of solution. A fixed time method was used; the reaction is stopped by the addition of sodium hydroxide solution after 3 minutes followed by a spectrophotometric measurement. Interference from other metal ions is prevented by adding sodium diethyldithiocarbaminate to the mixture. Another method similar in principle was developed by Guilbault Sadar and Zimmer.8 Both acid phosphatase (working at pH 7) and alkaline phosphatase (pH 8) can be applied.The substrate used is umbelliferone phosphate which is cleaved by the enzyme to produce the highly fluorescent umbelligerone. The reaction rate is followed by measuring and recording the fluorescence activity as a function of time. In a blank experiment the blank background fluorescence is recorded. The percentage inhibition is calculated from the results and is plotted against the concentration of beryllium to obtain a calibration graph. Between 0.01 and 0.1 pg ml-1 of beryllium can be determined with an error of between +3.5 and -2-5 per cent. Larger amounts of fluoride chloride bromide iodide, phosphate sulphate dichromate aluminium cadmium copper lead lithium, magnesium manganese mercury nickel potassium silver and sodium can be tolerated.Bismuth interferes and can be determined in a similar way. Bismuth. The method of Guilbault Sadar and Zimmer,8 described for the determination of beryllium can also be used for the determination of 1 to 70 pg ml-1 of bismuth with an error of between -5 and +1 per cent. An initial rate method was described by Hargi~.~ The reaction is a combined ligand-exchange and redox process. Phosphate bismuth and molybdate ions react in a slow equilibrium process to form a product that is then reduced by ascorbic acid to form the blue 18-molybdo-bismuthophosphoric acid. The formation of the latter can be monitored by spectrometry and the initial rate expressed in change of absorbancy per minute is found to be proportional to the concentration of bismuth. For the determination of 1 to 5 pg ml-l of bismuth the error is between -1.2 and +0.9 per cent.while at the lowest concentration level at which deter-minations can still be made 0.04 pg ml-l the error is below 310 per cent. A great number of metal ions and anions do not interfere and these are tabulated in the original paper CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 246 Calcium. The kinetic method of Pausch and Margerum6 for the deter-mination of alkaline earth metals has already been mentioned in connection with barium. The method is suitable for the determination of calcium alone or in binary or ternary mixtures. Townshend and VaughanlO applied an interesting enzyme-catalysed method for the determination of 1 to 4 pg of calcium in 7 ml of solution. The apo-enzyme of calf-intestinal alkaline phosphatase can be activated in the presence of zinc by small amounts of calcium.p-Nitrophenyl phosphate is used as a substrate which is hydrolysed by the enzyme. A fixed reaction time method is applied the hydrolysis being quenched after 4 minutes by the addition of sodium hydroxide solution. The absorbance of the solution is measured at 410 nm and is plotted on the calibration curve against the concentration of calcium. The only disadvantage of the method is the fact that although the preparation of the apo-enzyme is simple it is some-what time-consuming and the product is not too stable. Chromium. Jasinskiene and Bilidiene recently described two methods for the determination of chromium. The first of thesell is based on the oxidation of indigocarmine with hydrogen peroxide in acid medium.The oxidation is catalysed by chromium(V1) ions. If 2,2’-bipyridil is added to the mixture as little as 6 ng ml-1 of chromium can be determined; in its absence the sensitivity is some-what lower. The extinction of the mixture is measured at 597 nrn and plotted as a function of time; the reaction rate is then determined from these plots. The calibration curve is constructed by plotting the reaction rate against the concen-tration of chromium. Copper interferes with the determination. Their second method12 utilises the catalytic action of chromium(V1) on the rate of the oxidation of methyl orange by hydrogen peroxide. Citric acid acts both as an activator and as a complexing agent. At pH 2.8 (obtained simply by the addition of hydro-chloric acid) between 14 alid 200 ng ml-l of chromium can be determined by a spectrophotometric monitoring of the reaction rate.The extinction at 490 nm is measured and plotted as a function of time; from these graphs the reaction rates can be determined. Calibration graphs are obtained by plotting the rate as a function of the concentration of chromium. The coefficient of variation of the method is below jC9 per cent. and about twenty common metals do not interfere. Other activators may be used instead of citric acid. Cobalt. Cobalt(I1) ions have a catalytic effect on the oxidation of various organic substances by hydrogen peroxide and all of the recent methods suggested for the determination of cobalt are based on this principle. Kucharkowski and DOgel3 applied the reaction between hydrogen peroxide and tiron at pH 10.3 to the determination of cobalt(I1).The reaction product has a slightly yellow colour and its concentration can be monitored by measuring the extinction of the mixture at 336 nm. The initial rate method can then be applied although more recently Kucharkowski14 has modified the technique by applying the fixed reaction time method. Between 0.6 and 10 ng ml-1 of cobalt can be determined although with a fixed reaction time of 24 hours the sensitivity can be extended to 0.03 ng ml-l 246 SVEHLA The method has successfully been applied to the determination of trace amounts of cobalt in molybdenum metal. Popa and Costache16 used 2,6,7-trihydroxy-9-phenylxanthen-3-one in ethanolic solution as a reagent to be oxidised by hydrogen peroxide and by measuring the concentration of the reagent by spectrophotometry at 460nm the initial rate method can be applied.The reaction takes place in alkaline medium whose pH is maintained by a sodium hydroxide - sodium borate buffer. The calibration graph obtained by plotting the initial rate against the concentration of cobalt is linear within the range of 0.08 to 0.64 ng ml-l and the error is less than &5 per cent. More recently Popa and Costache16 replaced their reagent by the more readily available Bordeaux S (C.I. Acid Red 27). Spectrophotometric monitoring is done at 496 nm. Between 3 and 16ng ml-l of cobalt can be determined with errors below 20 per cent. Kreingold and Bo~hevolnov~~ determined cobalt in germanium tetrachloride and trichlorosilane applying the reaction between hydrogen peroxide and salicylfluorone in an alkaline medium.The sensitivity of the method is 0-05 ng per 5 ml. A detailed study of interferences has also been made. The special merit of their paper is the fact that they recommend four different catalytic methods for the determination of traces of four metals in germanium tetrachloride and tri-chlorosilane samples. Copper. The catalytic action of copper on various slow redox reactions is well known and is the reason why so many methods have recently been proposed for its catalytic determination. Pall Svehla and ErdeylS determined copper by using a process based on the Landolt reaction between peroxidisulphate and iodide ions in acidic medium with sodium thiosulphate as delaying agent.In this visual, fixed concentration technique reaction times are measured and the reciprocal values are plotted against the concentration of the catalyst. For 1 to 100 pg ml-l of copper the calibration graph is linear and the error is below 10 per cent. The interference of iron(II1) can be eliminated by adding sodium fluoride to the mixture. The theoretical backgrounds have been described.19 In a series of papers Hessel-barth20-22 described the determination of copper based on its catalytic action on the reaction between iron(II1) ions and thiosulphate. Thiocyanate ions are used to produce the iron(II1) thiocyanate complex the colour of which can be examined visually when the method of simultaneous comparison is adapted. Over a certain level of acidity the results are independent of the concentration of hydrochloric acid in the solution.Between 0.1 and 100 pg ml-l of copper can be determined by the method. Although many ions interfere with the determination copper can easily be separated by cation exchanger22 or co-precipitated with mercury sulphide21 before analysis. Dittelzs based his method on the catalytic effect of copper on the reaction between hydrogen peroxide and quinol in the presence of pyridine at pH 7 which is maintained by a phosphate - tartrate buffer. The rate of the reaction is monitored by measuring the absorbance of the mixture at 470nm. Unknown samples are analysed with the aid of a calibration graph in which the rate is plotted against the concentration of copper.Up to 10 ng of copper can be determined with an error o CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 247 less than &5 per cent. the limit of detection being 4-2 ng under normal circum-stances although in the absence of tartrate ions this can be reduced to 049ng. Colovos and Papadopoulos24 have described an automatic method very similar to the above in which 2,4-diaminophenol was used as a reagent with hydrogen peroxide. The extinction at 500 nm is measured and reaction times are determined from the recorded charts on the basis of the fixed concentration method. Reci-procal reaction times increase linearly with copper concentration within the range of 0.2 to 12.0 pg ml-1. Tin molybdenum and iron interfere seriously with the determination. N e d ~ e d ~ ~ applied the reaction between hydrogen peroxide and the dipotassium salt of Congo Red (C.I.Direct Red 28) which is catalysed by copper. The concentration of the dye is monitored by spectrophotometric measurements at 501 nm. At pH 12 between 1 and 20 ng ml-l of copper can be determined in this way. Cerium cobalt silver vanadium thiocyanate hexacyanoferrate(II1) and EDTA interfere seriously. The interference by manganese can be eliminated by the addition of oxalate ion to the mixture. The method was adapted for the determination of copper in samples of technical potassium hydroxide lithium chloride caesium carbonate and rubidium sulphate. Heller and Guyon26 based their very sensitive method on the catalytic action of copper on the reduction of iso-propylmolybdate by ascorbic acid to molybdenum blue.The fixed time method is applied the extinction of the solution being measured at 750 nm after 60 minutes. At pH 1.85 between 50 and 300 p.p.b. of copper can be determined. The inter-ference of a number of ions can be overcome by a preliminary solvent extraction of copper. The inhibiting action of copper of an enzyme-catalysed reaction was applied by Guilbault Kramer and Hackley27 for the determination of 0.2 to 6 pg ml-l of copper. The hydrolysis of indol-3-yl-acetate by hyaluronidase can be followed by the fluorimetric measurement of the product of the hydrolysis. By recording the variation of the fluorescent intensity with time the initial rate method can be applied. A monochromatic ultraviolet radiation (395 nm) is used for excitation while the fluorescence is measured at 470 nm.The hydrolysis proceeds slower in a phosphate - citrate buffer of pH 6.4. The coefficient of variation of the determination is k2.3 per cent. For the determination of traces of copper in zinc metal Gyganok Cujko and Reznik28 applied a preliminary separation by co-precipitation of the copper wth a 0.02 per cent. amount of the zinc present in the sample by using hydrogen sulphide. After filtration washing and dissolution the catalytic determination of copper is carried out by applying the reaction between phosphomolybdate ions and thiourea in sulphuric acid medium. The extinction of the mixture is measured at 680nm at regular intervals the initial rate method being applied for evaluation. Between 7 and 33 x per cent.of copper in zinc metal or zinc sulphate can be determined with an error of less than 5 per cent. Kreingold and Bozhevolnov17 determined amounts as low as 5ng of copper in 5 ml of solution on the basis of its catalytic action on the dirnerization of lumocup-ferron [a-(4-dimethylaminobenzylidene) hippuric acid]. The method was applied to the analysis of copper in samples of high purity germanium tetrachloride and trichlorosilane 248 SVEHLA Germanium. The slow reaction between molybdate and iodide ions in acid medium which is catalysed by traces of germanium was applied for its deter-mination by Michalski and Ge10wa.~~ The concentration of iodine formed in the reaction can be measured by amperometry or biamperometry. The initial rate method can be applied for evaluation.The sensitivity under normal circumstances is 50 ng ml-l but with special precautions even 8 ng ml-l of germanium can be determined. The average error of the determination is 5 per cent. and ammonium, barium calcium potassium sodium and zinc ions (up to 0 . 0 5 ~ concentration) do not interfere. Iron. A number of new methods have been suggested for the determination of iron. Thompson and Svehla3* applied the perborate - iodide Landolt reaction with ascorbic acid as a delaying agent. At pH 4 maintained by an acetate buffer 1 to 10 pg ml-l of iron can be determined by measuring the reaction time that elapses from mixing the reagents to the occurrence of the Landolt effect. Reciprocal reaction times gave a linear calibration curve when plotted as a function of the concentration of iron.The deviations at the 95 per cent. level of significance were also calculated and plotted together with the calibration curve. Molybdenum and osmium interfere with the determination. The interference of molybdenum can be eliminated by masking it with tartrate ions. The theoretical backgrounds of the method were de~cribed.~~ Another Landolt reaction18 that is catalysed by iron ions is the slow reaction between peroxidisulphate and iodide ions in acid medium. Sodium thiosulphate is used as a delaying agent. The reaction time is measured between mixing the reagents and the liberation of iodine which occurs instan-taneously. Between 1 and 100 pg m1-1 of iron can be determined with an error of less than 10 per cent. The calibration curve made up by plotting reciprocal reac-tion times as a function of concentration of iron is linear.Copper interferes with the determination. The theoretical backgrounds of the method are discussed in a thesis.19 A fixed reaction time method was recommended by Orav and K ~ k k ~ ~ for the determination of as little as 0.1 pg ml-l or iron. The slow reaction between triethylene tetramine and hydrogen peroxide is used. The optimum pH is 10 which is maintained by a borate buffer. At the fixed reaction time of 10 minutes the catalytic reaction is stopped by acidifying with sulphuric acid and the amount of unreacted hydrogen peroxide is determined by iodimetric titration. In a calibra-tion graph the amount of hydrogen peroxide decomposed during the process is plotted as a function of catalyst concentration.The error of the method is high, but was kept below 28 per cent. in the cases examined by the authors. Kreingold and BozhevolnoP based their method on the catalytic effect of iron on the oxida-tion of Acid Chrome Dark Blue dye by hydrogen peroxide in acid medium. The reaction rate is followed photometrically; the initial rate when plotted against the concentration of iron gives a linear calibration curve. Amounts as low as 2 ng ml-I of iron can be determined. Antimony gallium molybdenum niobium, tin titanium and tungsten interfere with the determination. The method was applied for the determination of iron in lanthanum oxide when the coefficient o CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 249 variation was 18 per cent.Later Kreingold Bozhevolnov and Antonov3* recom-mended the oxidation of H-acid [4-amino-5-hydroxynaphthalene-2,7-disulphonic acid] by hydrogen peroxide in the presence of hydrochloric and acetic acids cata-lysed by traces of iron for the determination of the latter. The fixed times method was used the extinction of the solution being measured at 508 nm after 15 minutes. The calibration graph is constructed by plotting the difference of extinction between the test solution and a blank as a function of the concentration of iron. The method is sensitive for up to 5 ng of iron in the final solution. Alkali metals alkaline earth metals aluminium cobalt copper lanthanum manganese nickel and zirconium do not interfere. A third method of Kreingold and Bozhev~lnov~~ applies the reaction between Stilbexon [4,4'-bis(carboxymethyl)amino-stilbene-2,2'-disulphonic acid] and hydrogen peroxide in acid medium.The method was successfully applied for the determination of iron traces in high purity germanium tetrachloride and tri-chlorosilane. The limit of detection is 2.5 ng of iron in 5 ml of solution. Kharlamov, Dodin and Mantsevich35 based their method on the photo-oxidation of methyl orange by atmospheric oxygen in the presence of traces of iron. A fixed time method is applied with photometric measurement. The reagents are mixed at pH 2 (which is obtained by the addition of suitable amounts of oxalic sulphuric or perchloric acids) exposed for 10 minutes to ultraviolet radiation and before and after irradiation the extinction of the mixture is measured at 350 nm.The differ-ence of extinction is plotted as a function of catalyst concentration. 0.025 pg ml-1 of iron can be determined with an error not exceeding 3.4 per cent. Copper interferes. The method has successfully been applied for the determination of iron in nickel metal. Guilbault Kramer and Hackley2' applied the same enzyme-catalysed reaction for the determination of iron as the one described for copper. Between 0.2 and 12 pg ml-l of iron can be determined on the basis of its inhibiting action on the enzymatic hydrolysis of indol-3-yl-acetate. The coeficient of varia-tion of these determinations is better than 52.3 per cent. Though not based strictly on catalysis the photochemical kinetic method of Lukasiewicz and Fitz-gerald36 has also to be mentioned.They have determined 0.1 to 1 pg ml-1 amounts of iron by a spectrophotometric monitoring of the rate of the reduction of iron(II1) ions by oxalate in the presence of 1 ,lo-phenanthroline which not only complexes iron(I1) ions and so speeds up the reduction but is also an excellent spectro-photometric reagent for the monitoring of the process. The procedure requires irradiation by ultraviolet light or the reduction does not take place. The initial rate method is applied; the rate is directly proportional to the concentration of iron. Magnesium. The kinetic method of Pausch and Margerum,6 described in connection with barium and calcium can be applied for the determination of magnesium alone and in binary and ternary mixtures with other alkaline earth metals .Manganese. The catalytic method for the determination of manganese, based on the oxidation of ðyl aniline with potassium periodate has been know 250 SVEHLA for some time.37 Hadjiioannou and Kephalas3* automated the method by applying the instruments described by Malmstadt and Hadjiioannou3Q earlier. The method is based on the principle of fixed concentrations but both the measurements of extinction and time are made automatically and the results are recorded and printed out. Between 3 and 40-ng amounts of manganese can be determined with an error of less than 2 per cent. and a coefficient of variation of &l per cent. The method was successfully applied for the determination of manganese in natural waters. Janjic Milanovic and Celap40 used the oxidation of Alizarin S with hydro-gen peroxide in sulphuric acid medium.The reaction is catalysed by 2 to 10 ng ml-l of manganese. With spectrophotometric measurement of the extinction at 335 nm, both the fixed time method and the initial rate method can be applied though a procedure applying the method of simultaneous comparison is also described. The latter however is applicable only for higher amounts (0-1 to 1 pg ml-l of manganese. The influence of foreign ions has been investigated thoroughly and the kinetics of the reactions involved were also examined. Ditte123 recommended the use of the oxidation of diethylaniline by periodate ions. This reaction is catalysed by as little as 10 ng of manganese. The solution contains a phosphate -tartrate buffer of pH 6 that also acts as a complexing agent to eliminate the effect of some metals.The progress of the reaction is monitored by measuring its extinction at 471 nm. The initial rate method can be applied for evaluation. At the 10-ng level the coefficient of variation is 7.1 per cent. Kreingold and Bozhevol-determined 0-3ng of manganese in 5ml of solution on the basis of its catalytic effect on the reaction between hydrogen peroxide and lumomagneson [5-(5-chloro-2-hydroxy-3-sulphophenylazo)-barbituric acid]. Antimony bismuth, cobalt lanthanum lead magnesium nickel platinum silver thorium zinc and zirconium as well as citrate EDTA and phosphate interfere with the determination. Mercury. Mealor and Townshend41 used an enzyme-catalysed reaction, inhibited by traces of mercury for the determination of the latter within the to molar concentration range.The hydrolysis of sucrose by invertase proceeds with decreasing rate with increasing concentration of mercury. The fixed time method can be applied and the concentration of sucrose after a period of 60 minutes can be measured by polarimetry. A suitable buffer (of pH 4.0 or 5.5) has to be applied the pH being chosen according to other metals present. If silver is present a pH of 4 is the optimum level of acidity for there to be no inhibition by silver. In the absence of silver and some other interfering ions such as lead and zinc a pH of 5.5 should be used. The effects of various metal ions were examined. There is virtually no interference from cadmium copper lead silver uranium and zinc if present in 100 to 1000 fold amounts.More recently Townshend and V a ~ g h a n ~ ~ described a method based on a particularly interesting principle. The dehydrogenation of ethanol by nicotinamide adenine dinucleotide is catalysed by yeast alcohol dehydrogenase to yield acetaldehyde and the hydrogenated form of the nicotinamide adenine dinucleotide. The reaction leads to an equilibrium. The inverse process is also catalysed by the same enzyme. Metal ions such as mercury and silver inhibit both processes. The degrees of inhibition in both directions ar CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 251 much the same if mercury is present but there is an appreciable change however, in the behaviour of silver. On this basis a certain degree of selectivity can be obtained.The forward process is simpler and more practical and this is recom-mended as a basis for the determination of 2 to 20 ng of mercury. A fixed time method can be applied and the fluorescence of the hydrogenated nicotinamide adenine dinucleotide is measured both in a blank and in the test solution. The relative activities calculated from these results can be plotted as function of the catalyst concentration. Although the results obtained were reproducible in un-buffered solutions it is reported to be more advisable to work at constant pH values (8 for the forward and 5.6 for the reverse process). The main source of error, working at these concentration levels is the adsorption of mercury on to the walls of glassware. Soaking with a solution of dithizone in carbon tetrachloride removes the adsorbed mercury ions from the glass surface.Rodriguez and Pardue4' determined mercury by its inhibiting effect on the reaction between cerium(1V) and arsenic(III) catalysed by iodide and osmium. Details of the method will be described in connection with the determination of silver. Molybdenum. The catalytic action of molybdenum on redox reactions involving peroxy compounds has been known for some time. Thompson and Sveh1a3O based their method on the perborate - iodide Landolt reaction with ascorbic acid as a delaying agent. At pH 4 maintained by an acetate buffer 1 to 10 pg ml-l of molybdenum can be determined. In this fixed concentration method, reaction times between mixing the reagents and the appearance of iodine are measured.Reciprocal reaction times plotted as a function of the concentration of molybdenum give linear calibration graphs. The deviations around the calibration curve with a 95 per cent. level of significance were also calculated. Iron and osmium interfere with the determination. The interference of iron can be eliminated by adding EDTA to the solution but the effect of osmium remains unchanged. The theoretical backgrounds of the method have been discussed.31 The catalytic effect of molybdenum on the hydrogen peroxide - iodide reaction was utilised by Weisz Klockow and Ludwig.43 The novelty of their method lies in the interesting principle applied for the measurements. The authors used a potentiostat i e . an automatic titrimeter which can be used to maintain a pre-determined electrode potential in the cell by adding a reagent from an automatic burette.The volume of the reagent added is recorded as a function of time. If iodine and iodide are both present from the start the oxidation - reduction potential of the iodine - iodide couple can be measured in the cell. If some of the iodide is oxidised to iodine the potential will become more positive. To maintain the potential during the course of the reaction iodide has to be added. With a potentiostat this can be solved easily. Provided that both the iodide and iodine concentrations were large enough at the beginning the constancy of the electrode potential means that the concen-tration of iodide is kept constant during the measurements. By this the order of the reaction is reduced to zero (for a more detailed explanation the original paper should be consulted) and the reaction rate becomes constant.The latter will b 252 SVEHLA proportional to the rate of addition of iodine which is equal to the slope of the recorded curve. Thus a calibration curve can be drawn and samples containing 1 to 10 pg ml-l of molybdenum can be analysed with an error of less than -4 per cent. and those with 0 to 1 pg ml-l of molybdenum with an error of below -6 per cent. Lazarev44 applied a turbidimetric method for the kinetic determination of molybdenum on steel. The reaction applied is that which occurs between selenate and tin(I1) ions leading to the formation of selenium and which is catalysed by molybdenum. After the dissolution of the sample the solution is acidified with hydrochloric acid and gum acacia solution is added to protect the suspension of selenium.The reaction is started and after a fixed time of 30 minutes the extinction of the solution (caused by the turbidity in the mixture) is measured against a blank at 390 nm. In the final solution 1 mg of molybdenum can be determined with an error of -7 per cent. Oxidising agents and rhenium interfere. Nickel. Mealor and Town~hend*~ described the determination of 0.1 to 0.7 p.p.m. of nickel based on the decomposition of permanganate in an alkaline solu-tion that also contains 1-hydroxyethylidene-1,l-diphosphoric acid. A fixed concentration method with spectrophotometric measurement is applied and the time required for the extinction (at 600 nm) of the mixture to reach 0.300 against a water blank is measured.The calibration graphs are linear if the reciprocal reaction times are plotted as a function of the squares of concentrations of nickel. The measurement can be automated. Interference by cobalt copper iron and silver can be eliminated by removing these ions by solvent extraction. Phosphate and oxalate decrease the rate of reaction. Osmium. Osmium is a well known catalyst of redox reactions. Gregorowicz and S ~ w i n s k a ~ ~ applied its effect on the reaction between persulphate ions and 3-amino-4-hydroxy-benzenesulphonic acid for its catalytic determination. The concentration of the organic reagent can be monitored during the process by measuring the extinction of the mixture at 530 nm and results are evaluated on the principles of the initial rate method.Between 7 and 35 pg of osmium per 50 ml can be determined with an average error of 0.7 pg. Working at a pH of 2.2 and 20°C, cadmium mercury molybdenum silver tungsten vanadium zinc and zirconium do not interfere copper has a slight effect and iron oxidants and reductants have a marked interference on the rate of the reaction. The possibility of the deter-mination of osmium based on the reaction between cerium(1V) and arsenic(II1) ions was pointed out recently by Rodriguez and Pardue!' while Thompson and Svehla30 and Thompson3l reported its effect on the Landolt reaction between perborate and iodide ions with ascorbic acid as a delaying agent. None of these authors gives however detailed procedures for the determination of osmium.Ruthenium. Ottaway Fuller and Allan48 recommended the use of the slow oxidation of the tris(lJ0-phenanthroline) - iron(I1) complex by periodate cata-lysed by ruthenium for the determination of the latter. The extinction of th CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 253 iron(I1) complex can be measured at 505 nm. After a fixed time of 10 minutes at 25°C the extinction reading is taken. Its reciprocal value when plotted against the concentration of the catalyst gives a linear calibration curve within the region of to 2 x 1 0 - 8 ~ ruthenium. At higher concentrations the curve bends towards the concentration axis. The detection limit of ruthenium was found to be 10-lo M. A number of ions were investigated for interferences but only iridium, osmium and rhenium showed serious effects on the accuracy of ruthenium deter-minations.Silver. The catalytic action of silver on various redox reactions was utilised by a number of authors for its determination. Bontchev and Alexiev4O applied the oxidation of hydrochloric acid by cerium(1V) ions. This reaction is catalysed by as little as 0.06 pg ml-l of silver. The pH of the solution must be just over 2 nitric acid being used to keep the acidity at this level. A fixed reaction time of 20 minutes is allowed at 30"C and the extinction of the solution which is proportional to the concentration of cerium(1V) is measured at 420 nm. From these results a calibra-tion graph can be constructed. The error of the method is about 13 per cent. at 95 per cent. confidence limits.The method can be used for the determination of silver in high purity zinc and cadmium. Copper and manganese interfere with the determination. Alexiev and Bontchevso later examined the possibility of deter-mining silver on the basis of its catalytic effect on the reaction between perosi-disulphate ions and sulphanilic acid at pH 4.5. The product of the reaction has a yellow colour and its extinction can be measured at 420 nm even with a filter instrument. When using the initial rate method for evaluation 0 to 10 x M of silver can be determined with a linear calibration curve. Eontchev Alexiev and Dimitrovasl later modified the previous procedure by adding 2,2'-bipyridyl , which acts as an activator to the mixture. The sensitivity of the determination was increased by this modification to 4 x pg ml-l of silver.The coefficient of variation of the method is jz7.6 per cent. The inhibiting action of silver on some enzyme catalysed reactions was used for its determination by Mealor and Towns-hend4I and Townshend and Vaughan.42 The principles are identical to those described in connection with the determination of mercury. The sensitivities for silver are 0.2 pmole and 1 ng ml-1 respectively which is twice as good as those quoted for mercury. Rodriguez and Pardue4' utilised the inhibiting action of silver on the reaction between cerium(1V) and arsenic(II1) in the presence of iodide as catalyst. The inhibition is caused by the formation of silver iodide. The extinction of the solution caused by the presence of cerium(IV) can be measured at 407 nm and the results can be evaluated on the basis of the initial rate method.There are modifications of the method applying iodide alone or mixtures of iodide and os-mium as well as iodide and iodate as catalyst for determinations of silver in the con-centration range of 0 to 25 x 1 0 - 8 ~ . Menghani and Bakore52 examined the catalytic effect of silver on the reaction between peroxidisulphate ions and pinacol. Although their main task was to clarify the kinetics and mechanism of the reactions involved their findings can also be utilised for the catalytic determination of silver 254 SVEHLA Strontium. The method of Pausch and MargerumJ6 discussed in more detail in connection with barium can also be used for the determination of strontium alone or in binary or ternary mixtures with other alkaline earth metals.Thallium. The reaction between cerium(1V) ions and thallium(1) in acid medium is slow in the dark but the rate of reduction of cerium(IV) which can be followed by spectrophotometric measurements at 425 nm increases considerably in daylight. The phenomenon is interpreted by Schenk and B a ~ z e l l e ~ ~ as the catalysis of thallium on the photo-reduction of cerium(1V). Although their main concern was to accelerate reactions for the cerimetric titration of thallium(1) ions, they drew attention to the possibility of analysing mixtures of thallium(1) and other metals kinetically. The strong catalytic action of manganese was also noted. Tungsten. Hadjiioannou and Valkana54 applied the catalysis of the reaction between hydrogen peroxide and iodide for the determination of tungsten.The formation of iodine in acid medium can be monitored by photometric measure-ments. A fixed concentration method was applied and reaction times lying between 20 and 100 s were measured. Between 0.5 and 3 pg of tungsten can be determined with a coefficient of variation of between 1 and 2 per cent. Interferences of other ions were studied and the maximum amounts of these causing errors of less than 5 per cent. were determined and tabulated. Vanadium. The catalytic action of vanadium on redox reactions between halates and halides is well known. Thompson and S ~ e h l a ~ ~ utilised the Landolt reaction which is based upon the bromate - iodide reaction in acid solution by using ascorbic acid as a delaying agent.By applying a citric acid - sodium citrate buffer (pH 2.2) copper iron molybdenum and titanium which would otherwise interfere are also complexed and the determination of 0 to 10 pg ml-f of vanadium becomes specific only for this ion. The technique involves a fixed concentration method with measurement of the reaction time from mixing the reagents up to the appearance of the colour of iodine. Reciprocal reaction times result in a linear calibration curve when they are plotted against the concentration of vanadium. Another method based on a Landolt reaction was developed by Bognar and J e l l i ~ ~ e k . ~ ~ They apply the reaction between chlorate and chloride with tin(I1) chloride as a delaying agent o-Tolidine is used as an indicator for chlorine.As the reaction is slow the temperature should be kept at 40°C. For 1 to 1Opg of vanadium the coefficient of variation is below 5 per cent. Later Bognar and Sarosij' replaced tin(I1) chloride by hydrazine sulphate as a delaying agent. The temperature in this case has to be kept at 65°C. More recently Bognar and Jellinek,58 modifying their earlier method based on the bromate - bromide -ascorbic acid Landolt reaction recommended the use of dimethyl or tetraethyl-9-phenylene diamine as an indicator for the appearance of bromide. A highly sensitive method has recently been recommended by Fuller and O t t a ~ a y . ~ ~ The method utilises the oxidation of Bordeaux with bromate in acid medium. The rate of the reaction is rather sensitive to variations in acidity.A fixed concentratio CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 255 method was applied combined with the spectrophotometric monitoring of the concentration of Bordeaux at 515nm the time elapsing from the mixing of the reagents until the transmission had increased to 70 per cent. was measured. Between 0-005 and 0.2 pg ml-l amounts of vanadium can be determined with a standard deviation of 3 per cent. (at the 0.05 pg m1-1 level). The effects of thirty-seven cations anions and complexing agents were examined. Thiocyanate nitrite, iodide ruthenium and copper have a marked interference. The calibration curve, which is constructed by plotting reciprocal reaction times as a function of the concentration of the catalyst bends towards the concentration axis. Tanaka and Awata6O applied the reaction between 4-hydrazinobenzenesulphonic acid and chlorate ions at pH 2.5 to the determination of vanadium.After a fixed time the product of the reaction 4-diazobenzenesulphonic acid is coupled with l-naphthyl-amine and the concentration of the latter is determined spectrophotometrically at 530 nm. From 0.02 to 1 p.p.m. of vanadium can be determined. A preliminary extraction with 8-hydroxyquinoline at a pH of 4-45 into chloroform and re-extrac-tion at pH 10 into an ammonia - ammonium chloride buffer increases the selec-tivity. Iron can be masked with 1,2-diamino-cyclohexane-N,N,N',N'-tetraacetic acid but antimony bismuth silver and tin interfere seriously. Zinc. Townshend and Vaughan described two methods for the determination of zinc both based on enzyme-catalysed reactions.The first of these' is based on the same principle as described in connection with beryllium and with it 0.6 to 6 pg of zinc can be determined. The second methodlo is identical to that described for calcium; procedures for 6 to 65 and 65 to 650-ng amounts of zinc are given. Both methods are accurate and highly selective. Determination of non-metallic substances and anions Bromide. For the determination of as little as 3ng ml-l of bromide, Toropova and Tamarchenko61 recommended the use of potassium bromate in acid medium. On adding methyl orange to the solution its colour fades slowly because of the formation of bromine. The initial rate of decolourisation which can be determined spectrophotometrically at 490 nm is proportional to the amount of bromide present.Arsenite thiosulphate sulphide thiocyanate nitrite and iron interfere. Cyanide. The method of Guilbault Kraxner and Hackley,27 described in connection with copper can also be used for the determination of 0-1 to 4 pg ml-l amounts of cyanide with a coefficient of variation of 2-3 per cent. Hydrogen peroxide. A n interesting catalytic method for the determination of traces of hydrogen peroxide has been described by Krueger Warriner and Jaselkis.62 Xenon trioxide and t-butyl alcohol axe used as reagents in neutral solution. The reaction is preceded by an incubation period the length of which depends on the concentration of the catalyst. The ultraviolet absorption of xenon trioxide can be monitored at 200nm and from the inflexion of the extinction 256 SVEHLA vemw time curves which resemble a potentiometric titration curve the incubation period can be determined.Alternatively the half-life of xenon trioxide can be used for evaluation. From 36 to 360 parts per lo9 of hydrogen peroxide can be deter-mined with an error of below 10 per cent. Krause Zielinski and WeimannM as well as Krause Domka and Marciniecs4 described kinetic methods for the determination of traces of hydrogen peroxide based on the rate of decolourisation of certain dyes. These are not strictly speaking catalytic methods although the techniques used are related to those usually applied in catalytic determinations. Iodide. A number of new methods have been suggested for the catalytic microdetermination of iodide and a number of authors proposed modifications of older established techniques.Ballzos6 has utilised the iodide-catalysed oxidation of @,$'-tetramethyldiamino-diphenyl-methane with chloramine T. The formation of a reaction product can be monitored by spectrophotometry and the whole process can be automated. As little as 10-14-g amounts of iodide can be determined in 1 1 of drinking water. Toropova and TamarchenkoSg recommended the use of the reaction between arsenite and iodate which is catalysed by 5 ng ml-l or more of iodide. The rate of the reaction can be monitored by the amperometric measurement of the dif-fusion current of iodate. Proskuryakovas7 applied the oxidation of hexacyanofer-rate(I1) ions with nitrite to the determination of 0.02 to 0.1 pg ml-l of iodide while Bogurth and Schaeg68 made use of the oxidation of arsenite by manganese(II1) ions.A photometric monitoring is possible in both cases. Bognar and nag^^^ recom-mend a simultaneous comparison method for the determination of 0 to 1 pg ml-l amounts of iodide. The determination is based on the catalytic effect of iodide on the oxidation of 3,3 -dimethyl-naphthidine with hydrogen peroxide. A spectro-photometric monitoring is also possible. Ottaway Fuller and R o w ~ t o n ~ ~ re-examined the method originating from Iwasaki Utsumi and Ozawa.'l They pointed out the weaknesses of the procedure and suggested some important improvements. The well known method by Sandell and K ~ l t h o f f ~ ~ ~ ~ ~ based on the catalytic action of iodide on the reaction between cerium(1V) and arsenite was applied modified or automtaed by several auth~rs!~J*-~~ Oxygen.A method based on a combined redox and ligand-exchange reaction has been reported79 for the determination of dissolved oxygen in water. The method is strictly speaking not catalytic as oxygen is used up in the reaction but the rate of reaction can be monitored and the evaluation is made on a kinetic basis. Pentacyanocobaltate(I1) is oxidised first and the excess of the reagent is subjected to various ligand-exchange reactions. As little as 3 x lo-* moles per litre of dissolved oxygen can be determined. Phosphate. The method described by Crouch and Malmstadts0 is based on the fact that under certain conditions the initial rate of the formation of molyb-denum blue from phosphate molybdate and ascorbic acid is proportional to the concentration of phosphate.The formation of molybdenum blue can be monitore CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 257 spectrophotometrically at 650 nm. The reaction is quite fast readings having to be taken as soon as 10s after the mixing of the reagents. The method can be automated and applied to the routine determination of phosphate in serum. More recently, Javier Crouch and Malmstadtsl made some improvements and innovations in the instrumentation and their new procedure is now based on an automatic reading carried out 10 ms after mixing the reagents. Selenium. West and Ramakrishnas2 recommended the use of the reduction of methylene blue with sodium sulphide at pH 10-8 for the catalytic determination of selenium.If sodium sulphite is present the interference by polysulphide can be eliminated. Formaldehyde and disodium ethylenediamine tetraacetate must also be added making the method specific for selenium. The time for complete decolourisation is measured for various concentrations of selenium and is used to prepare the calibration curve. Between 0-1 and 1 pg of selenium can be determined by the method. Bognar and Sarosim recommended a Landolt reaction method based on the chlorate - chloride - hydrazine sulphate system by which 0.1 to 1 pg per 5 ml of selenium can be determined with a coefficient of variation of 10 per cent. The simultaneous comparison technique can also be applied. Larger amounts of some metals can be tolerated. Kawashima and TanakaM have determined 0 to 0*15 pg amounts of selenium by the catalytic reduction of 1,4,6,11-tetra-azo-naphthacene with phosphorous acid in hydrochloric acid medium.The reaction has to be started at 50"C and after a fixed time of 30 minutes the mixture is cooled to 0°C to stop the reaction. The extinction of the solution must be measured at 600nm within 30 s. Interferences were carefully examined and listed. More recently another method based on the oxidative coupling reaction of phenyl hydrazine fi-sulphonic acid with 1-naphthylamine has been reported.% The mixture should be kept at 50°C for 60 minutes when the reaction can be stopped by cooling the system to O"C and the extinction of the solution is measured at 525 nm. The extinction is directly proportional to the amount of selenium present.Potassium chlorate has to be present although its role in the reaction mechanism has not been studied. Sulphide. Babko and Maksimenkos* published a method for the deter-mination of 0.1 to 10 ng of sulphide. Iron(I1) ions reduce silver nitrate to silver metal in the presence of sulphide ions only. Lighting must be controlled during the reaction to avoid the photo-reduction of silver ions. The extinction of the colloidal silver solution can be measured at 530 nm and the calibration graph is prepared by plotting extinction values as a function of sulphide concentration. For 5ng of sulphide the coefficient of variation is about 4 per cent. Up to 50 pg of chloride do not interfere. Oxidising or reducing substances that interfere with the reaction must not be present.Cataltyic and kinetic determination of organic substances The principle of determining concentrations by kinetic measurements has long been extended to the analysis of organic substances. In these procedures the role o 258 SVEHLA the organic substance can only seldom be described as catalytic. The compound often takes part in the main reaction and is used up during the process. In some cases the organic species acts as an inhibitor or complexing agent and in this way influences the catalytic behaviour of a metal or enzyme. As the experimental techniques applied for the determination of organic substances by kinetic analysis are the same as those described earlier for inorganic systems this review also deals shortly with the kinetic (but not catalytic) determination of organic substances.The substances are dealt with in alphabetical order regardless of their structure or composition. A reviews7 on kinetic methods in organic analysis was published in 1967 and references to earlier works are available therein. Arnines. Shresta and DasS8 described a kinetic method for the determination of primary amines based on their reaction with salicylaldehyde. A Schiff base is formed during the reaction. After a fixed time the reaction can be stopped by the addition of acetic anhydride when the unreacted primary amines are acetylated. The Schiff base can then be titrated by perchloric acid in a non-aqueous medium. A special graphical method is used for the evaluation. Methods for the deter-mination of secondary and tertiary amines are also described.Ascorbic acid. Bognar and Jellineksg have described a kinetic method for the determinaiton of 10 to 100 pg ml-l amounts of ascorbic acid based on its delaying action on the bromate - bromide redox reaction. A linear calibration curve can be prepared by plotting reaction times as measured in other Landolt reactions against ascorbic acid. With the simultaneous comparison method the range of determin-able ascorbic acid is 5 to 50 pg ml-l although the possibility of determining as large amounts as lo00 pg ml-I is also outlined. o-Chloronitrobenzene. Legradig0 used a kinetic method for the detection and determination of o-chloronitrobenzene in the presence of $-chloronitrobenzene. The o-compound forms a red product when heated with hydroxylamine in an alkaline solution in ethanol; the extinction of the coloured product can be measured by spectrophotometry.Chlorodinitrobenzene reacts even in cold solution and can be determined in the presence of the two other compounds. Complexing agents. Mottola Haro and Freisergl determined various organic complexing agents such as cysteine EDTA and salicylic acid These substances inhibit the catalytic action of copper(I1) ions on the atmospheric oxidation of L-ascorbic acid. The rate of decomposition of ascorbic acid can be monitored by extinction measurements at 265nm. As the rate of oxidation depends on the pH of the solution which itself varies in a solution of ascorbic acid the use of a buffer (pH 7-4) is recommended. From 1 to 1Opg ml-1 amounts of the com-plexing agents can be determined.Dihydric phenols. Schenk and BrowB2 recommend the use of a kinetic method for the determination of either catechol or quinol in the presence of resorcinol. The method is based on the reaction between the free radical 2,2 CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 259 diphenyl-1-picrylhydrazil and the dihydric phenols. The reaction is started and the extinction of the free radical can be monitored at 540 nm. From the extinction versus time curves the initial rate can be determined and the concentration of phenols calculated. Ethanol is used as a solvent and daylight should be excluded. Esters. A dual temperature differential kinetic method was described by Munnellyg3 for the simultaneous determination of some acetate esters (methyl, 2-hydroxyethyl and phenyl acetate) by following the rate of second-order saponifi-cations at two temperatures 15" and 30°C.The graphical method of Hanna and Siggia,w a modified version of the initial rate method can be used for evaluation. Ethanolamides. Fatty alkanolamines can be saponified in alcoholic potas-sium hydroxide solutions when 2 moles of the weak base react with 1 mole of potassium hydroxide. This difference in the stoicheiometry was applieds5 to the determination of ethanolamides in mixtures with amines amine soaps and amide esters which behave in a different way. The unreacted amount of potassium hydroxide is determined after a certain time by titrating it with hydrochloric acid. Results are evaluated from a calibration curve where the amount of saponified sample is plotted against its concentration.The calibration curve is linear and the error of the determination of 10 to 75 mol per cent. of diethanolamide generally does not exceed a few per cent. Ketones. Toren and Gnuseg6 described a method for the determination of ketones in binary mixtures based on their different rates of reduction by hydroxyl-amine. The reactions can be followed by an automatic titrimeter (operated as a pH-stat) and the volume of titrant needed for the restoration of the original pH is recorded as a function of time. The precision of the method depends on the difference in the rate constants for the reduction of each ketone. If the two constants differ by a factor of 5 components of mole of total ketones can be determined with a coefficient of variation of 1 per cent.while if the factor is 3 the coefficient of variation increases to 3 per cent. Naphthols. For the kinetic determination of 1- or 2-naphthol Babking7 recommended that bromine be liberated from potassium bromate and potassium bromide in acid solution. This reacts with 1- or 2-naphthol present in the mixture, by forming a colourless brominated product. If methyl orange is present the excess of bromine will attack the dye and decolourise it. The time needed to decolourise the methyl orange is under controlled experimental conditions proportional to the amount of naphthol present. Up to 250 mg 1-1 of 1- or 2-naphthol can be deter-mined with an error of less than &5 per cent. Up to 1 g 1-l of chloride does not interfere but sulphate phenols and other organic substances that react with bromine do.Oxidative enzymes and their substrates. Monoamine and diamine oxidase can be determined with a kinetic methodg8 that is based on the reaction of fi-hydroxyphenyl-acetic acid with the enzyme to yield a fluorescent product 260 SVEHLA Hydrogen peroxide has to be present. The fluorescence of the product is monitored by excitation at 317 nm and measurement of the emission at 414 nm. The initial rate method can be applied to calculate results. The method can also be applied to the determination of the substrates of the enzymes such as cadaverine histamine, putrescine benzylamine tyramine and furfurylamine. The error of the method for the concentration ranges quoted in the paper is generally below -3 per cent.Pesticides. The hydrolysis of umbelliferone phosphate achieved by phos-phatase enzyme is inhibited by certain pesticides such as aldrin and heptachlor. This method was applieds to the determination of 5 to 1oOpg ml-l amounts of aldrine and 50 to 700pg ml-l of heptachlore. The method is identical to that discussed in connection with beryllium. Phenols. A method similar to that of Schenk and Browng2 for the deter-mination of dihydric phenols has been described.99 Phenol o-cresol $-cresol and 2,6-xylenol can be determined up to a concentration of 5 mmoles 1-1 with an error below 3 per cent. The accuracy of the determination of 3,4-xylenol and m-cresol was less satisfactory. Serum enzymes. Kinetic methods are frequently applied in clinical labora-tories for the determination of serum enzymes.SzaszlOO described a kinetic-photometric method for the determination of serum leucine aminopeptidase. Knedel and BottgerlOl use a fixed concentration method with photometric measure-ments for the determination of cholinesterase activity in serum. The optimum conditions for the determination of serum alkaline phosphatase were investigated by Hausamen Helger Rick and Gross.lo2 The well known method of Oliver103 was modifiedlo4 for the determination of serum creatine kinase applying commercially available reagents and a spectrophotometric measurement. Qualitative Catalytic Analysis Most of the quantitative methods reviewed above are suitable for qualitative testing also. There are however a few new techniques described specifically as qualitative tests that are reviewed here.Robinsonlo5 recommends a method for the identification of alcohols. First their 3,5-dinitrobenzoate esters are prepared and are then hydrolysed in alkaline medium. The rate of hydrolysis can be measured by spectrophotometry at 285 nm. From the rate of hydrolysis and the melting point of the ester twenty-seven different alcohols can be identified. A specific and sensitive test for arsenic was described by Krause and Slawek.106 The oxidation of indigo carmine by hydrogen peroxide is catalysed by arsenic and g of arsenic can be detected by the decolourisation of the dye in a concentration limit of 1 los. For the detection of barium Hamya and Townshendlo7 applied an induced precipitation method.This is based on the fact that under the circumstances described in the paper the formation of lead sulphate precipitate is strongly accelerated by the presence of as little as 0.4 pg of barium. Winterlo* reports on the identification (and semi-quantitative determination) o CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 261 catalase in plant extracts. The method is based on the gas volumetric measure-ments on the initial rate of the decomposition of hydrogen peroxide. LegradiBO recommends the method reviewed among the quantitative determinations of organic substances for the detection of o-chloronitrobenzene and related sub-stances. The detection of iodide can be achieved by using the same reaction as is used for its determinati~n.~~ For the identification of sugars Mikkelson and Robinsonfog recommend the spectrophotometric measurement of the rate of formation of oximes with hydroxylammonium chloride.Fifteen different sugars can be identified by the method. Application of Catalytic Waves in Polarography Some new quantitative methods recommended in the past 3 years involve the use of polarographic catalytic waves. Kolthoff and MaderllO observed two catalytic hydrogen waves in solutions containing cobalt(I1) or cobalt(II1) ions and mercapto-anilines. These waves could be used for the determination of cobalt and showed different characteristics to the Brdickalll waves. Later Mader and Kolthoff llC showed that in some cases adsorption phenomena occur in the polarography of similar systems. A catalytic reduction wave of nitrate occurs in acid solutions containing traces of uranium.This phenomenon can be used for the determination of uranium.l13 Although the current measured is not a linear function of the uranium concentration uranium can be determined with a coefficient of variation of &5 per cent. Toropova and Anisimova,ll* on the other hand dealt with solutions of 8-mercaptoquinolineJ and found that in acetate buffers this substance causes the evolution of catalytic hydrogen waves that can be used for its deter-mination. They extended the method for the determination of metals that react with 8-mercaptoquinone producing a smaller catalytic wave. The theory of kinetic currents has also been developing in the past 3 years. Pence Delmastro and Boomanlfs dealt with the determination of rate constants of very fast regeneration reactions at spherical electrodes.Delmastro116 later demonstrated that in practical cases the ratio of the kinetic current to the purely diffusion controlled current at a stationary spherical electrode is always less than two. Delmastro and Booman117 also dealt with polarographic kinetic currents of the first order. The kinetics of the adsorption of the reaction product in stationary electrode polarography has been examinedll* both theoretically and experi-mentally. Kinetochromic Spectrophotometry West and his co-workers have recently developed a technique that they call kinet ochromic spec trophot ometry . Certain ligand-exchange reactions in which zirconium ions take part seem to be catalysed by substoicheiometric amounts of certain anions.The product of these reactions is coloured and can be measured by spectrophotometry. However it is not the reaction rate that is monitored (and therefore these methods form a different class among catalytic techniques) but th 262 SVEHLA extinction of the final product is measured after the elapse of a certain minimum time when it remains constant. A closer examination of these processes showed that the role of the catalyst is to achieve a certain degree of depolymerisation of the hydrolysed zirconium polymers. The amount of depolymerised zirconyl ions is proportional to the amount of the catalyst ions present. As the complex is formed from the depolymerised zirconyl ions the extinction of the solution after sufficient time is allowed for the formation of the complex becomes proportional to the catalyst concentration.On this principle a method for the determination of 0.005 to 0.05 pg ml-l of fluoride (in the final volume) by using the xylenol orange -zirconium reaction has been reported.l19 Both sulphatef20 and fluoride121 have been determined by making use of the reaction between zirconium and methylthymol blue. Interferences were studied carefully and methods of their elimination were suggested. Catalymetric Titrations The principle of catalysis can also be applied for the end-point detection of titrations although it is not generally the catalyst that is determined. The catalyst is used as a titrant and until the equivalence point is reached it is used up by the titration reaction.The excess of the catalyst initiates a catalysed reaction that is accommpanied by a change of colour (or temperature or electrode potential), which indicates the end-point. As the catalyst is present in large concentrations, these catalysed reactions at the end-point are fast almost instantaneous. Follow-ing up earlier works on the subject Weisz and Janjic122 introduced catalymetric end-point detection into complexometric titrations. The oxidation of certain organic substances by hydrogen peroxide takes place only if free metal ions are present. On adding an excess of a complexing agent to a solution of metals the excess of the reagent can be back titrated with a solution of a metal by using the catalytic reactions mentioned for end-point detection. Weisz and Kiss123 used the same principle but instead of a visual method they measured the temperature of the mixture which increases rapidly when the end-point is reached.This thermo-metric-catalytic method was used earlierx2* for the end-point detection of precipita-tion titrations. Silver mercury and palladium can be titrated with potassium iodide by using the reaction between cerium(1V) and arsenic(II1) ions for end-point detection. The method can also be applied to the indirect determination of anions that react with these metals. Earlier a potentiometric measurement for end-point detection based on the same catalytic reaction had been reported.lZ5 Fedorova and Yatsimirskii126 determined palladium on the basis of its inhibition effect on the cerium(1V) - arsenic(II1) reaction catalysed by iodide.Bognar and Sarosil27 applied Landolt reactions for the end-point detection of titrations that involve silver or mercury as well as iodide as reactants. Mottola128 determined microgram amounts of anionopolycarboxylic acids and metals with manganese(I1) ions as titrant. The reaction between malachite green and periodate ions is used as the reaction indicating the end-point and the extinction of the mixture is measured CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 263 The method is entirely automatic and with the simple instrumentation described in the paper recorded titrigrams are obtained. applied the principle of differential thermometric end-point detection for catalymetric titration of cobalt, based on its catalytic action on the reaction between tartaric acid and hydrogen peroxide.Vajgand and GaaPo used an acetic acid medium for their catalytic thermometric titrations. Tertiary amines and salts of organic acids can be titrated with perchloric acid. Acetic anhydride must be present and does not react with the water that is produced by the titration reaction until free perchloric acid is present. The increase in temperature accompanying the reaction can be measured by a thermometer. Later the same principle was used for the determination of triethyl-amine brucine potassium acetate and dimethyl aniline.131 The procedure was automated at the same time. Tertiary amines and salts of organic acids have been determined by catalymetric titration in acetic acid by using a coulometrically generated tit rant .lSz New Techniques and Instruments There was considerable progress made in experimental techniques and instrument ation of catalytic analysis.Many authors recommend the use of computers for the evaluation of results. Thus CrouchlS described a small all-electronic device that computes the reciprocal reaction time for use in rate calculations. Parker Pardue and Willisla describe a miniature computer that can be used for catalytic ox kinetic analytical purposes. A larger model was recommended earlier.131 All of these were analogue computers. The application of digital computers in analytical chemistry including the solution of rate equations has been reviewed.136 Innovations in the experimental technique involving potentiometric techniques were made by several authors.An electrode capable of measuring urease enzyme activity has been recommended.la7 Guilbault Smith and Montal~ol~~ used an electrode sensitive to ammonium ions for studying the kinetics of deaminase enzyme systems. Sand and H ~ b e r l ~ ~ recommended the application of differential constant-current potentiometry for kinetic analysis. In this method the potential between two polarised cathodes or anodes is measured and recorded as a function of time. These graphs can be evaluated by using the initial rate technique. The potentiostatic methodg3 was discussed when dealing with the determination of molybdenum. Alexander and BarclaylQ0 described a new technique ‘current-cessation chronopotentiometry,’ and its analytical applications. An electrical pulse is passed through a coulometric cell containing certain metal ions and anions for 30 to 200 s and the potential of a working electrode is measured as a function of time.The time required for the potential of the working electrode to reach its maximum rate of decay is inversely proportional to the concentration of one of the substances in the cell. Olmsted and Nicholsonl*l recommended the use of a double potential step method for measuring rate constants of dimerization reactions. The method can also be applied for catalytic analyses 264 SVEHLA Interesting developments on the application of spectrophotometric methods to catalytic and kinetic analysis have also been made in the past 3 years. The automated fast reaction rate method in the millisecond range has already been reviewed in connection with the determination of phosphate.8l A semi-automatic instrument for the continuous measurement of concentrations has been reported.lg2 The core of the instrument is a spectrophotometer with a high stability and low noise.The auto-andyser was modified and adaptedlg3 for reaction rate measure-ments of multiple enzyme samples by continuous-flow analysis. The use of thermoanalytical methods in kinetic and catalytic analysis has been recommended by several authors. Besides the methods reviewed in connection with catalyrnetric titrations new techniques have been suggested for the measurement of reactions rates by thermoanalytical methods. These methods are interesting in that they can be used to measure the rate of reaction of substances in the solid (or molten) phase.Taylor and Watsonlg4 applied differential thermoanalysis to the measurement of relative reaction rates. Thermogravimetric analysis has been used for the determination of the rate constant and activation energy of various degradation rea~ti0ns.l~~ Sharp Sally and Went~orthl*~ examined and compared three methods of obtaining kinetic information from thermogravimetric curves. For the evaluation of data obtained by non-isothermal kinetic measurements, Kresze Kosbahn and WinMerlP7 described a method involving the use of a digital computer. The efficiency of mixing of reagents and the resulting instrumental delay effect on recorded kinetic curves have been examined,148 and the findings are relevant to methods involving both slow and fast reactions.Theory of Catalytic Analysis Reviews Many of the papers mentioned so far contain some discussion of the kinetics and reaction mechanism of the processes involved. There are a number of papers, however containing predominantly theoretical material in which the aim of the author is to examine a certain reaction or a principle regardless of its practical applications. Most of the theoretical papers deal with the kinetics and mechanism of a single reaction or a special group of reactions. Goldman and H a r g i ~ l ~ ~ studied the kinetics of the formation and reduction of phosphorous-bismuth dimeric heteropolymolybdate a reaction that can be applied for the determination of bismuth. Their results indicate that the heteropolymolyb-date complex contains eighteen atoms of molybdenum one atom of phosphorus and one atom of bismuth.The role of cobalt in catalytic reactions was studied by several authors. The oxidation of Tiron with hydrogen peroxide in the presence and absence of cobalt was examined,13J4 and a mechanism for the catalysed reaction was postulated. The rate constant was measured and used to explain the calibration curves obtained. The kinetics of oxidation of the cobalt(I1) bipyridyl complex by copper(I1) and iron(II1) perchlorates in anhydrous aceto-nitrile were studied by using polarographic measurements with a rotating platinum e1ectr0de.l~~ The rate equations of these processes were postulated and verified CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 265 Both reactions are first order with respect to the cobalt complex but second order with respect to iron(II1) while the order with respect to copper is 0.5.Heller and Guyon2s when reporting on their catalytic-spectrophotometric determination of copper made detailed kinetic measurements on the rate of the reduction of iso-polymolybdate with ascorbic acid both in the presence and absence of the catalyst. Reactions between more than thirty metals and traws-l,2-diaminocyclohexane N,N,N',N'-tetraacetate were in~estigatedl~l and the respective rate constants of the reactions were determined. Some ligand-exchange reactions between metals and between hydrogen ion and a metal were also examined. An interesting correlation between the atomic number of the element and rate constants within the series of lanthanides has been discovered enabling a differential kinetic method to be used for their determination in binary mixtures.The 'classical' reaction between cerium(1V) and arsenic(II1) ion catalysed by iodide has also been investigated.152 It is interesting to note that although this reaction is widely used in industrial and clinical laboratories for the determination of iodine the kinetics and mechanism of the reaction have not before been thoroughly examined and interpreted. A highly complex rate equation with no less than seven rate constants has been derived and verified by experimental results. The kinetics of the reaction between perborate and iodide ions catalysed by iron and molybdenum in acid medium has been in~estigated~l; the effects of osmium were also noted.The rate constants of the reactions between hydrogen peroxide and acid chrome dark blue, both in the presence and absence of iron as catalyst have been determined.% The kinetics of the oxidation which is catalysed by manganese of Alizarin S with hydrogen peroxide have been studied.40 This reaction proceeds in slightly alkaline solution. The kinetics of the ligand-exchange reaction between ethylenediamine tetraacetic acid and its nickel complex have been elucidated by using deuterated reagents and n.m.r. techniques.153 An exhaustive study of the catalytic properties of compounds of the platinum group lead to the discovery of new oxidation -reduction reactions that proceed at room temperature with a measurable rate and are catalysed by various of the compounds.lM The kinetics and mechanism of the reactions were investigated and possibilities of analytical applications outlined.Reactions involving vanadium were thoroughly studied by various groups of investigators; Fuller and O t t a ~ a y l ~ ~ investigated the oxidation of vanadium(1V) in acid medium by bromate. The rate equation contains not only a rate constant but also an equilibrium constant and while the reaction is first order with respect to vanadium(IV) a mixed order was found with respect to bromate. The oxidation of Bordeaux with bromate in acid medium was studied both in the presence156 and absence16' of vanadium. The uncatalysed reaction is a first-order process with respect to Bordeaux but the contribution of bromate to the rate cannot be explained so simply as the rate is influenced by the concentration of bromide that is inevitably present in bromate as a trace impurity.The process at the same time is autocatalytic for which the rate equation derived by the authors allows. The process catalysed by vanadium is more complex as it exhibits an induction period and the process is autocatalytic as well. The mechanism158 and the kinetics150 o 266 SVEHLA the oxidation of p-phenetidine with chlorate and the role of vanadium in these processes have been investigated. The role of vanadium is not merely that of an electron donor and acceptor (though it exists in both the tetravalent and penta-valent state in the reaction cycle) but it also forms a complex with charge transfer transition from which an arylamino radical is then formed.Bontchev and MladenowalG0 examined the kinetics of the oxidation of vanadium(1V) by chlorate and the effects of hydroxycarboxylic acids on the reaction rate. Thompson and S ~ e h l a ~ ~ and Thompson31 examined the kinetics of the oxidation of iodide with bromate in the presence and absence of vanadium. The same system was examined later by Bognar and JellineklG1 under somewhat different experimental conditions. The competing role of ascorbic acid with vanadium in the Landolt reaction was emphasised by them. While the contributions reviewed above were dealing with specified reactions, quite a number of theoretical papers tried to cover a broader field and to give explanations and theories of a more general nature. BontcheP2 dealt with reaction mechanisms in catalyitc analysis showing how the knowledge of the mechanism of a catalytic reaction can help in establishing the optimum experimental conditions and in developing new catalytic methods.The possibility of increasing the sensi-tivity of these procedures based on a knowledge of the reaction mechanism enhances the value of this paper. Svehlal@ reviewed the applications of Landolt reactions in quantitative analysis. Simple methods for the kinetic examination of the uncatalysed and catalysed reactions are outlined and a generally adaptable kinetic explanation has been given. On this basis an explanation of the particular shape of the calibration curve can be made. A general treatment on the selectivity, sensitivity and precision of Landolt reactions is also included.Catalytic methods using ligand-exchange reactions with masking agents and automatic rate measure-ments have been reviewed164 and those co-ordination chain reactions mentioned in the earlier part of this review were then first postulated. This field of catalytic analysis has hardly been explored yet the sensitivity and precision of these methods appears to be promising. The lack of selectivity of these techniques can partly be overcome by the application of a second complexing (masking) agent. Mottola165 reviewed the fast developing field of catalymetric titrations. The general theory presented in his paper is applicable to almost all existing methods and will no doubt help to find new applications of this technique. Lukasiewicz and Fitzgerald36 outlined the theory of their new technique called photochemical kinetic analysis which is distinguished by its high degree of selectivity.The reviews of Rechnitz166 and Guilbault16' on catalytic methods published among the annual reviews of AnalyticaE Chemistry can be read with interest. The author thanks Professor C. L. Wilson Mr. P. J. McKenna and Mrs. M. Scullion for assistance in the preparation of the manuscript CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 267 1 2 3 4 6 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 References Mark H. B. and Rechnitz G. A.‘Kinetics in Analytical Chemistry. Chemical Analysis,’ Yatsimirskii K. B. “Kinetic Methods of Analysis. International Series of Monographs Delahay P. and Stiehl G. J. J . Amer. Chem. Soc. 1952 74 3500. Michaylova V. and Bontchev P. R. Michrochim. Acta 1970 344. Bognar J. and Pataky-Szabo M. Ibid. 1969 221. Pausch J. B. and Margerum D. W. Analyt. Chem. 1969,41 226. Townshend A. and Vaughan A. Talanta 1969 16 929. Guilbault G. G. Sadar M. H. and Zimmer M. Analytica Chim. Acta 1969 44 361. Hargis L. G. Analyt. Chem. 1969 41 597. Townshend A. and Vaughan A. Talanta 1970,17 289. Jasinskiene E. and Bilidiene E. Zh. analit. Khim. 1967 22 741. - and - Ibid. 1968 23 143. Kucharkowski R. and Doge H. G. 2 . analyt. Chem. 1968,238 241. Kucharkowski R. Ibid. 1970,249 22.Popa G. and Costache D. Revue roum. Chim. 1067 12 963. - and- Ibzd. 1968 13 573. Kreingold S. U. and Bozhevolnov E. A. Trudy Kom. analit. Khim. 1968 16 194. Pall A. Svehla G. and Erdey L. Talanta 1970 17 211. Hrabeczy-Pall A. Ph.D. Thesis Technical University Budapest 1969. Hesselbarth H. Z. analyt. Chem. 1969,248 163. Dittel F. Ibid. 1967,229 193. Colovos G. and Papadopoulos C. Mikrochem. J. 1968 13 236. Nedved J. Chem. Listy 1968 62 698. Heller R. L. and Guyon J. C. Analyt. Chem. 1968,40 773. Guilbault G. G. Kramer D. N. and Hackley E. Analyt. Biochem. 1967 18 241. Cyganok L. P. Cujko V. T. and Reznik B. E. Zav. Lab. 1967,33 5. Michalski E. and Gelowa H, Chemia Analit (Warsaw) 1967 12 147. Thompson H. and Svehla G. Z. analyt. Chem. 1969,247 244. Thompson H.Ph.14. Thesis Queen’s University of Belfast 1970. Orav M. and Kokk K. Uchen. Zap. tcartu. gos. Univ. 1968 207. Kreingold S. U. and Bozhevolnov E. A. 21%. analit. Khim. 1967 22 218. Kreingold S. U. Bozhevolnov E. A. and Antonov V. N. Zav. Lab. 1968,34 260. Kharlamov I. P. Dodin E. I. and Mantsevich A. D. Zh. analzt. Khim. 1967 22 371. Lukasiewicz R. J. and Fitzgerald J. M. Anal. Lett. 1969 2 159. Kun E. J. Biol. Chem. 1947 170 509. Hadjiioannou T. P. and Kephalas T. A. Mikrochim. Acta 1969 1215. Malmstadt H. V. and Hadjiioannou S. I. Analyt. Chem. 1962 34 452. Janjic T. J. Milovanovic G. A. and Celap M. B. Ibid. 1970 42 27. Mealor D. and Townshend A. Talanta 1968 15 747. Townshend A. and Vaughan A. Ibid. 1970 17 299. Weisz H. Klockow D. and Ludwig H. Ibid.1969 16 921. Lazarev A. I. Zh. analit. Khim. 1967,22 1836. Mealor D. and Townshend A. Analytica Chim. Acta 1967 39 235. Gregorowicz Z. and Suwinska S. Mikrochim. Acta 1967 546. Rodriguez P. A. and Pardue H. L. Analyt. Chem. 1969,41 1376. Ottaway J. M. Fuller C. W. and Allan J. J. Analyst 1969 94 522. Bontchev P. R. and Alexiev A. A. Mikrochim. Acta 1968 875. Alexiev A. and Bontchev P. R. Ibid. 1970 13. Bontchev P. R. Alexiev A. and Dimitrova B. Talanta 1969 16 597. Menghani G. D. and Bakore G. V. Bull. Chhem. SOL Jafian 1968,41 2574. Schenk G. H. and Bazzelle W. E. Analyt. Chem. 1968 40 162. Hadjiioannou T. P. and Valkana C. G. Hem. Hron. 1967 32 89. Thompson H. and Svehla G, Microchem. J. 1968 13 576. Bognar J. and Jellinek O. Mikrochim. Acta 1967 193.Interscience Vol. 24 1968. of Analytical Chemistry,’ Pergamon Press Vol. 25 1966. - Ibid. 1969,248 167. - Ibid. 1969,248 289 268 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 SVEHLA Bognar J. and Sarosi Sz. Ibid. 1967 813. Bognar J. and Jellinek O. Ibid. 1969 366. Fuller C. W. and Ottaway J. M. Analyst 1970 95 41. Tanaka M. and Awata N. Analytica Chim. Acta 1967 39 485. Toropova V. F. and Tamarchenko L. M. Zh. analit. Khim. 1967,22 576. Krueger R. H. Warriner J. P. and Jaselkis B. Talanta 1968 15 741, Krause A. Zielinski S. and Weimann T. 2. Analyt. Chem. 1968,237 193. Krause A. Domka F. and Marciniec B. Chemia Analit. (Warsaw) 1968 13 883. Ballzo H. 2. Analyt. Chem. 1969 245 20. Toropova V. F. and Tamarchenko L.M. Zh. analit. Khim. 1967,22 234. Proskuryakova G. E. Ibid. 1967,22 802. Bogurth W. and Schaeg W. Mikrochim. Acta 1967 658. Bognar J. and Nagy L. Ibid. 1969 108. Ottaway J. M. Fuller C . W. and Rowston W. B. Analytica Chim. Acta 1969 45 541. Iwasaki I. Utsumi S. and Ozawa T. Bull. Chem. Soc. Japan 1953 26 108. Sandell E. B. and Kolthoff I. M. J . Amer. Chem. Soc. 1934 56 1426. - and - Mikrochim. Acta 1937 1 9. Knapp G. and Spitzy H. Talanta 1969 16 1353. - and - Ibid. 1969 16 1361. Appleby A. and Spillett R. E. U . K . Atomic Energy Authority Report RCC-M 1967, Comoy E. Revue fr. Etud. Clin. biol. 1967 12 189. Godin J. M. and Archimbaud M. Rapp. CEA No. 3145 1967. Margerum D. W. and Stehl R. H. Analyt. Chem. 1967,39 1351. Crouch S. R. and Malmstadt H.V. Ibid. 1967 39 1090. Javier A. C. Crouch S. R. and Malmstadt H. V. Ibid. 1969 41 239. West P. W. and Ramakrishna T. V. Ibid. 1968 40 966. Bognar J. and Sarosi Sz. Mikrochim. Acta 1969 361. Kawashima T. and Tanaka M. Analytica Chim. Acta 1968 40 137. Kawashima T. Nakano S. and Tanaka M. Ibid. 1970,49 443. Babko A. K. and Maksimenko T. S. Zh. analit. Khim. 1967 22 550. Legradi L. Magy. Kern. Lap. 1967,22 488. Shresta I. L. and Das M. N. Analytica Chim. Acta 1970 50 135. Bognar J. and Jellinek O. Mikrochim. Acta 1969 312. Legradi L. 2. analyt. Chem. 1968 237 426. Mottola H. A, Haro M. S. and Freiser H. Analyt. Chem. 1968 40 1263. Schenk G. H. and Brown D. J. Talanta 1967 14 257. Munnelly T. I. Analyt. Chem. 1968 40 1494. Hanna J. G. and Siggia S.Ibid. 1962 34 547. Lohman F. H. and Mulligan T. F. Ibid. 1969 41 243. Toren E. C. jun. and Gnuse M K. Analyt. Lett. 1968 1 295. Babkin M. P. Zh. analit. Khim. 1968,23 637. Guilbault G. G. Kuan S. S. and Brignac P. J. jun, Analytica Chim. Ada 1969, Kwiatkowska I. and Kwiatkowski E. Chemia Analit. (Warsaw) 1968 13 783. Szasz G. Amer. J . Clin. Path. 1967 47 607. Knedel M. and Bottger R. Klin. Wschr. 1967 45. 325. Hausamen T. V. Helger R. Rick W. and Gross W. Clin. Chim. Acta 1967 15 241. Oliver I. T. Biochem. J. 1955 61 116. Hess J. W. MacDonald R. P. Natho G. J. W. and Murdock K. J. Clin. Chem. 1967, Robinson J. R. Analyt. Chem. 1967 39 1178. Krause A. and Slawek J. 2. analyt. Chem. 1969 245 44. Hamya J. W. and Townshend A. Analytica Chim. Acta 1969 46 312.Winter E. 2. analyt. Chem. 1969 244 248. Mikkelson T. J. and Robinson J. R. J . Pharm. Sci. 1968 57 1180. Kolthoff I. M. and Mader P. Analyt. Chem. 1969 41 924. Brdicka R. Colln Czech. Chem. Commun. 1936 8 366. Mader P. and Kolthoff I. M. Analyt. Chem. 1969 41 932. Habashi F. and Thurston G. A, Ibid. 1967 39 242. p. 210, 47 503. 13 994. 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 11 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 CATALYTIC METHODS IN ANALYTICAL CHEMISTRY 269 Toropova V.F. and Anisimova L. A. Zh. analit. Khim. 1967 22 1264. Pence D. T. Delmastro J. R. and Booman G. L. Analyt Chem. 1969,41 737. Delmastro J. R. Ibid. 1969 41 747. Delmastro J. R. and Booman G. L. Ibid. 1969.41 1409. Hulbert M. H. and Shain I. Ibid. 1970,42 162. Cabello-Thomas M. L. and West T. S. Talanta 1969 16 781. Hems R. V. Kirkbright G. F. and West T. S. Ibid. 1969 16 789. - - and - Ibid. 1970 17 433. Weiiz H. and Janjic T. 2. analyt. Chem. 1967 227 1. Weisz H. and Kiss T. Ibid. 1970 249 302. Weisz H. Kiss T. and Klockow D. Ibzd. 1969 247 248. Weisz H. and Klockow D. Ibid. 1967 232 321. Fedorova T. I. and Yatsimirskii K.B. Zh. analit. Khim. 1967 22 283. Bognar J. and Sarosi Sz. Mzkrochczm. Acta 1969 463. Mottola H. A. Analyt. Chem. 1970,42 630. Sajo I. Talanta 1968 15 578. Vajgand V. J. and Gaal F. F. Ibid. 1967 14 345. Vajgand V. J. Kiss T. A. Gaal F. F. and Zsigrai I. J. Ibzd. 1968 15 699. Vajgand V. J. Gaal F. F. and Brusin S. S. Ibid. 1970 17 415. Crouch S. R. Analyt. Chem. 1969,41 880. Parker R. A. Pardue H. L. and Willis B. G. Ibid. 1970 42 56. James G. E. and Pardue H. L. Ibid. 1968 40 796. Childs C. W. Hallman P. S. and Perrin D. D. Talanta 1969 16 1119. Montalvo J. G. jun. Awalyt. Chem. 1969 41 2093. Guilbault G. G. Smith R. K. and Montalvo J. G. jun, Ibid. 1969 41 600. Sand J. R. and Huber S. O. Ibid. 1970 42 238. Alexander W. A. and Barclay 14. J. J. Electroanal. Chem. 1967 13 137. Olmstead M. L. and Nicholson R. S. Analyt. Chem. 1969 41 851. Weichselbaum T. E. Plumpe W. H. jun. Adams R. E. Hagerty J. C. and Mark, Brown H. H. and Ebner M. R. Clin. Chem. 1967 13 847. Taylor L. J. and Watson S. W. Analyt. Chem. 1970 42 297. Farre-Rius F. Huret J. Puyo M. and Guiochon G. Analytica Chim. Ada 1969 45, Sharp J. H. and Wentworth S. A. Analyt. Chem. 1969,41 2060. Kresze G. Kosbahn W. and Winkler J. 2. analyt. Chem. 1967 231 1. Caselli M. Cavaggioni A. and Papoff P. Talanta 1968 15 1335. Goldman H. D. and Hargis L. G. Analyt. Chem. 1969 41 490. Nemec I. Kies H. L. and Nemcova I. Analytica Chim. Acta 1970 49 541. Margerum D. W. Pausch J. B. Nyssen G. A. and Smith G. F. Analyt. Chem. 1969, Rodriguez P. A. and Pardue H. L. Ibid. 1969,41 1369. Can J. D. and Reilley C . N. Ibid. 1970 42 51. Yatsimirskii K. B. Kalinina V. E. Morozova R. P. and Federova T. L. Trudy zvanovsk. Khim-tekhnol. Inst. 1968 Jubilee No. 68. Fuller C. W. and Ottaway J. M. Analyst 1969 94 32. - and - Ibid. 1970 95 28. - and - Ibid. 1970 95 34. Bontchev P. R. and Jeliazkowa B. G. Mihrochim. Acta 1967 116. -and - Ibid. 1967 125. Bontchev P. R. and Mladenowa Z. Ibid. 1968 427. Bognar J. and Jellinek Ibid. 1969 318. Bontchev P. R. Talanta 1970 17 409. Svehla G. Analyst 1969 94 513. Margerum D. W. Stehl R. H. and Latterell J. J. Analyt. Chem. 1967 39 1346. Mottola H. A. Talanta 1969 16 1267. Rechnitz G. A. Analyt. Chem. 1968,40 455R. Guilbault G. G. Ibid. 1970 42 334R. H. B. jun. Ibid. 1969 41 725. 463. 41 233
ISSN:0300-9963
DOI:10.1039/AS9710100235
出版商:RSC
年代:1971
数据来源: RSC
|
|