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 &ethyl 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