摘要:

TH'E ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYEDITORIAL ADVISORY BOARD*Chairman: J. M. Ottaway (Glasgow)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)L. R. P. Butler (South Africa)E. A. M. F. Dahmen (The Netherlands)A. C. Docherty (Billingham)D. Dyrssen (Sweden)J. Hoste (Belgium)H. M. N. H. Irving (Leeds)M. T. Kelley (U.S.A.)W. Kemula (Poland)"J. H. Knox (Edinburgh)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)"H. J. Cluley (Wembley)'P. Gray (Leeds)H. W. Nurnberg (West Germany)E. Pungor (Hungary)D. 1. Rees (London)"R. Sawyer (London)P. H. Scholes (Sheffield)'W. H. C. Shaw (Greenford)S. Siggia (U.S.A.)"D. Simpson (Thorpe-le-Soken)A. A. Smales, O.B.E. (Harwell)*A.Townshend (Birmingham)A. WaIsh (Australia)T. S. West (Aberdeen)"J. Whitehead (Stockton-on- Tees)A. L. Wilson (Medmenham)P. Zuman (U.S.A.)"G. E. Penketh (Billingham)*Members of the Board serving on The Analyst Publications CommitteeREGIONAL ADVISORY EDITORSDr. J. Aggett, Department of Chemistry, University of Auckland, Private Bag, Auckland, NEW ZEALAND.Professor G. Ghersini, Laboratori CISE, Casella Postale 3986,201 00 Milano, ITALY.Professor L. Gierst, Universit6 Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50,Professor R. Herrmann, Abteilung fur Med. Physik., 63 Giessen, Schlangenzahl 29, W. GERMANY.Professor W. A. E. McBryde, Faculty of Science, University of Waterloo, Waterloo, Ontario, CANADA.Dr.W. Wayne Meinke, KMS Fusion Inc., 3941 Research Park Drive, P.O. Box 1567, Ann Arbor,Dr. 1. Rubeika, Geological Survey of Czechoslovakia, Kostelni 26, Praha 7, CZECHOSLOVAKIA.Professor J. RfiiiEka, Chemistry Department A, Technical University of Denmark, 2800 Lyngby,Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN,Dr. A. Strasheim. National Physical Research Laboratory. P.O. Box 395, Pretoria, SOUTH AFRICA.Bruxelles, BELGl UM.Mich. 481 06, U.S.A.DENMARK.Published by The Chemical SocietyEditorial: The Director of Publications, The Chemical Society, Burlington House,London, W1V OBN. Telephone 01 -734 9864. Telex No. 268001Advertisements: Advertisement Department, The Chemical Society, Burlington House, Piccadilly,London, W1 V OBN. Telephone 01 -734 9864Subscriptions (non-members) : The Chemical Society, Distribution Centre, Blackhorse Road,Letchworth, Herts., SG6 1 HNVolume 104 No 1236 March 1979@ The Chemical Society 197

ISSN:0003-2654    

DOI:10.1039/AN97904FX009

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

ANALAO 104 (1 236) 177-272 (1 979)ISSN 0003-2654March 1979 'THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTS177196201208224232241248258261265269Energy-dispersive X-ray Emission Analysis.Analysis of Steroids. Part XXXII. Determination of Allyloestrenol by Titri-metric, Polarographic and Gas-chromatographic Methods-Sandor Gorog,Anna Lauko and Zsofia SziklayA Review-W. C. CampbellDiffusion Assay by an Automated Procedure-J. W. Lightbown, R. A. Broadbridge,P. Isaacson, J. E. Sharpe and A. JonesMechanism of Atom Excitation i n Carbon Furnace Atomic-emission Spectro-metry-D. Littlejohn and J. M. OttawayDetermination of Chromium in Natural Waters and Sewage Effluents byAtomic-absorption Spectrophotometry Using an Air - Acetylene Flame-K.C. Thompson and K. WagstaffDetermination o f Selenium i n Soil Digests by Non-dispersive Atomic-fluorescence Spectrometry Using an Argon - Hydrogen Flame and theHydride Generation Technique-J. Azad, G. F. Kirkbright and R. D. SnookInvestigations on Reaction Mechanisms in the Determination o f Non-ionicSurfactants in Waters as Potassium Picrate Active Substances-L. Favretto,B. Stancher and F. TunisLimit of Detection in Analysis w i t h lon-selective Electrodes-Derek MidgleySHORT PAPERSSpectrophotometric Method for the Determination of Paraquat-M. Ganesan,S. Natesan and V. RanganathanTitrimetric Determination of Reducing Sugars with Copper( II) Sulphate-T. H. KhanDetermination o f Dimetridazole i n Pig and Poultry Feeds by High-performanceLiquid Chromatography-A.D. Jones, I. W. Burns and S. G. SellingsBook ReviewsSummaries of Papers in this Issue-Pages iv, v, viii, xPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class at New York, USA, Post OfficEuro-Standardsavailable fromBUREAU OF ANALYSEDSAMPLES LTD.Newham Hall, Newby,Middlesbrough, Cleveland TS8 9EA(Telephone: Middlesbrough 31 721 6)A range of over 30 samples in finely dividedform:Unalloyed SteelsAlloy SteelsCast IronsFerro-AlloysIron OresDustsThese samples have been analysed by 20laboratories from countries within the EEC.Full details on request.APPOINTMENT VACANTLancashire County CouncilCounty Laboratory - Assistant Analyst(Salary ~€4,461- &4,761 per annum + &312 perannum supplement)Applicants should have BSc.Honours (Chemistry) orM.R.I.C. Experience of Food Chemistry preferable.Application forms from Chief Executive/Clerk (Ref:41/PAT), County Hall, Preston, PR1 SXJ (Preston54868, Ext. 566) t o be returned by 22nd March 1979.BRITISH PATENT NO. 1 407 298Filtering ElementOwner desires commercial exploitation on reasonable terms by licenseor sale. Inquiries Fitzpatricks Chartered Patent Agents 14-1 8 CadoganStieet, Glasgow, G2 6QW ’and Warwick House, Warwick Court,London, WClR 5DJ.FOR SALEThe Analyst 1952-70; Journal of the Association ofPublic Analysts 1965-71 and Reports on the Progressof Applied Chemistry 1963-99.Telephone 01-204 6871 (evenings).THE QUEEN’S UNIVERSITYOF BELFASTMSc COURSE inANALYTICAL CHEMISTRYApplications are invited for admission to thisestablished 12 month full-time MSc coursewhich provides a comprehensive training inthe theory and practice of modern chemicaland instrumental methods of analysis. Appli-cants should normally possess an honoursdegree (or equivalent) in chemistry or cognatesubjects. Part-time courses are available.The Science Research Council has recognisedthe course for tenure of its Advanced CourseStudentships.A description booklet and application formscan be obtained from Professor D. ThorburnBurns, Dept. of Chemistry, Queen’s Universityof Belfast, Belfast, BT7 1 NN, Northern Ireland

ISSN:0003-2654    

DOI:10.1039/AN97904BX011

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

iv SUMMARIES OF PAPERS I N THIS ISSUE March, 197’9Summaries of Papers in this IssueEnergy- dispersive X-ray Emission AnalysisA ReviewSummary of ContentsIntroductionInstrumentationExcitationX-ray tubesRadioisotopesElectrons and protonsDetectorsElectronicsPre-amplifierAmplifierMulti-channel analyserDead timeData processingSpectral features and interpretationSum peaksEscape peaksDiffraction peaksAnomalous silicon, gold and argon peaksScatter peaksSpectral backgroundComparison of wavelength- and energy-dispersive systemsApplicationsAtmospheric particulatesWatersClinical and biochemicalRocks, ores and cementMetals and alloysCoal and petroleumOn-stream analysisOthersFuture developmentsKeywords : Review ; energy-dispersive X-ray emissiox analysisW.C. CAMPBELLImperial Chemical Industries Limited, Petrochemicals Division, Research andDevelopment Department, P.O. Box 90, Wilton, Middlesbrough, ClevelandT56 6 ; s .Analyst, 1979, 104, 177-195.Analysis of Steroids. Part XXXII. Determination of Allyloestrenolby Titrimetric, Polarographic and Gas-chromatographic MethodsA titrimeLiz method is described for the determination of allyloestrenoibased on methoxymercuration of its double bonds and titration of the aceticacid formed with standard sodium hydroxide solution. The relative standarddeviation of the method is 0.29%. The polarographic reduction of themercury addition compound on the dropping-mercury electrode is used forthe determination of allyloestrenol in a tablet formulation with a relativestandard deviation of 3.1 yo.A gas-chromatographic method with a relati\ estandard deviation of 1.5% is also described. The applicability of the:emethods to the determination of the stability of allyloestrenol and of itsdosage form is discussed.Keywords : Allyloestrenol determination ; titrimetry ; polarography ; gaschromatographySANDOR GORGG, ANNA LAUKO and ZSOFIA SZIKLAYChemical Works Gedeon Richter Ltd., P.O. Box 27, H-1475 Budapest, Hungary.Analyst, 1979, 104, 196-200March, 1.979 SUMMARIES O F PAPERS I N THIS ISSUEDiffusion Assay by an Automated ProcedureEquipment is described that allows cliffusion assays to be performed auto-matically in Petri dishes using the punch-hole technique.With a block ofsix dishes limits of error of approximately 3 2% can be obtained consistently.Various sources of systematic errors and their elimination are discussed.Keywovds : Autowation ; antibiotic assay ; di,ffusion assay ; systematic evvovsJ. W. LIGHTBOWN, R. A. BROADBRIDGE and P. ISAACSONNational Institute for Biological Standards and Control, Holly Hill, Hampstead,London, NW3 6RB.J. E. SHARPEDivision of Engineering, National Institute for Medical Research, Mill Hill, London,NW7 1AA.and A. JONESResearch Division, Keecham Pharmaceuticals, Worthing, West Sussex, BN14 3QH.Analyst, 1979, 104, 201-207.Mechanism of Atom Excitation in Carbon FurnaceAtomic-emission SpectrometryBy consideration of electronic and vibrational excitation temperatures andthe ionisation temperature, it is demonstrated that local thermal equilibrium(LTE) is established under the practical analytical conditions of interruptedgas flow in which commercial carbon furnace atomisers are used as emissionsources. The electron concentration is shown to be derived from thermionicemission from the carbon tube and calculated values of 5.2 x 10lo c r r 3 a t2558 K and 1.3 x 1 0 l 1 ~ m - ~ at 2766 K are reported.The processes thatcontribute to the establishment of LTE are considered in detail, and it issuggested that molecular collisions make the major contribution to atomicexcitation under all conditions, but that radiation absorption may besignificant when a monatomic gas is used as purge gas and when moleculesare present as impurities a t concentrations of only 0.01%.Keywovds ; Atom emission ; carbon furnace atomisation ; excitation mechanism ;electvon concentvation ; local thevmal equilibviumD.LITTLEJOHN and J. M. OTTAWAYDepartment of Pure and Applied Chemistry, University of Strathclyde, CathedralStreet, Glasgow, G1 1XL.Analyst, 1979, 104, 208-223.Determination of Chromium in Natural Waters and SewageEffluents by Atomic-absorption Spectrophotometry Using anAir - Acetylene FlameA simple method for the determination of chromium in natural waters andsewage final effluents by atomic-absorption spectrophotometry using anair - acetylene flame is described. The sample is concentrated by evapora-tion by a factor of five. Interference effects were minimised by the additionof ammonium perchlorate and were further reduced by working with a flameon the verge of luminosity rather than a distinctly luminous flame.Keywords : Chromium detevvnination ; atomic-absorption spectvophotometvy ;air - acetylene jlame ; natuval waters and sewage efluentsK. C. THOMPSON and K. WAGSTAFFSevern-Trent Water Authority, Malvern Regional Laboratory, 141 Church Street,Malvern, Worcestershire, TVR14 2AN.Analyst, 1979, 104, 224-231.

ISSN:0003-2654    

DOI:10.1039/AN97904FP013

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

...Vlll SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Selenium in Soil Digests by Non-dispersiveAtomic-fluorescence Spectrometry Using an Argon - HydrogenFlame and the Hydride Generation TechniqueMavch, 197.9The determination of selenium a t submicrogram levels by atomic-fluorescencespectrometry, based on the evolution of hydrogen selenide into an argon -hydrogen air-entrained flame, is described. Using a simple purpose-builtnon-dispersive atomic-fluorescence spectrometer a detection limit of 10 ng cm-3of selenium is obtained. The technique has been applied to the determina-tion of selenium in soil digests and experiments have been carried out inorder to study the interference of other elements on the determination.Procedures for the elimination of interferences from copper are recommended.Keywords : Selenium determination ; atomic-fluorescence spectrometry ; hydridegeneration ; soil digestsJ. AZAD, G.F. KIRKBRIGHT and R. D. SNOOKDepartment of Chemistry, Imperial College, London, SW7 2BP.Analyst, 1979, 104, 232-240.Investigations on Reaction Mechanisms in the Determinationof Non-ionic Surfactants in Waters as Potassium PicrateActive SubstancesThe two-phase extraction and spectrophotometric determination of poly-oxyethylene non-ionic surfactants in water a t trace levels is examined indetail by considering both monodisperse and polydisperse surfactants of thetype RO(CH,CH,O),H, where R = $)-tert-nonylphenyl and .M. is the degree ofpolymerisation. Potassium picrate is used as a reagent for the polyoxyethylenechain and 1,2-dichloroethane as an extracting phase.Monodisperse surfactants with n from 4 to 15 were isolated by liquid - solidabsorption chromatography.Their purity was checked by temperature-programmed gas - liquid chromatography. Their reactivity to the reagent isexplained qualitatively by considering the equilibria involved in the extrac-tion.Polydisperse surfactants with YZ (number-average degree of polymerisation)ranging from 3.3 to 21.5 are also considered and compared with other poly-disperse surfactants in which R = dodecyl. The concentration of these non-ionics in waters is conveniently expressed as potassium picrate active sub-stances (PPAS). It can be referred to the standard synthetic monodispersesurfactant RO(CH,CH,O),H, where R = dodecyl, which gives a spectrophoto-metric response acceptably near to that of the examined series of commercialsurfactants.Keywords : Polyoxyethylene alkylplaenyl ether non-ionic surfactant tracedetermination ; watev analysis ; spectrophotovlzetry ; potassium picrate ;reaction mechanismL.FAVRETTO, B. STANCHER and F. TUNISIstituto di Merceologia, Universita di Trieste, 34100 Trieste, ItalyAnalyst, 1979, 104, 241-24SUMMARIES OF PAPERS IN THIS ISSUE March, 1979Limit of Detection in Analysis with Ion- selective ElectrodesThe limit of detection in analysis with ion-selective electrodes is discussedand definitions that are based only on the deviation of an electrode’s cali-bration from the theoretical, and take no account of the random errors ofmeasurement, are shown to be inadequate.Equations are derived thatexpress the limit of detection in terms of the random error of measurementand the factors determining the deviation of the electrode response fromthe Nernstian value, i.e., reagent blanks, solubility products and inter-ferences. The equations enable one to predict ( a ) the degree of precision withwhich the e.m.f. has to be measured if an electrode is to attain a desired limitof detection in specified conditions or (b) whether changing the conditionsmight bring the desired limit of detection within reach of a given precisionof measurement. Practical examples with ion-selective electrodes justifythe proposed statistical treatment of limit of detection and demonstrate thatthe errors for electrodes operating in the non-Nernstian region are normallydistributed.Keywovds : Ion-seleciive electvodes ; flotentiometvy ; limit of detectionDEREK MIDGLEYCentral Electricity Research Laboratories, Kelvin Avenue, Leatherhead, Surrey,KT22 7SE.Analyst, 1979, 104, 248-257.Spectrophotometric Method for the Determination of ParaquatShort PaperKeywords : Pavaquat detevmination ; spectvophotometvyM.GANESAN, S. NATESAN and V. RANGANATHANDepartment of Chemistry, United Planters’ Association of Southern India, TeaResearch Station, Cinchona 642 106, India.Analyst, 1979, 104, 258-261.Titrimetric Determination of Reducing Sugars with Copper(I1)SulphateShovt PaPevKeywords : Reducing sugar detevmination ; coppev(II) sulphate Yeduction ;titvimetvyT. H. KHANDepartment of Industries (Chemical Directorate), 58, Dilkusha Commercial Area,Dacca-2, Bangladesh.Analyst, 1979, 104, 261-265.Determination of Dimetridazole in Pig and Poultry Feeds byHigh-performance Liquid ChromatographyShort PapevKeywovds : Dimetvidazole detevmhation ; animal feeds ; high-perfovmanceliquid chvomatogvaphyA. D. JONES, I. W. BURNS and S. 6. SELLINGSUnilever Research Laboratory, Colworth House, Sharnbrook, Bedfordshire,MK44 1LQ.Analyst, 1979, 104, 265-268.

ISSN:0003-2654    

DOI:10.1039/AN97904BP017

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

MARCH 1979 Vol. 104 No. 1236 The Analyst Energy-dispersive X-ray Emission Analysis A Review W. C. Campbell Imperial Chemical Industries Limited Petrochemicals Division Reseavch and Development Department, P.O. Box 90 Wilton Middlesbrough Cleveland TS6 8 J E Summary of Contents Introduction Instrumentation Excitation X-ray tubes Radioisotopes Electrons and protons Detectors Electronics Pre-amplifier Amplifier Multi-channel analyser Dead time Data processing Spectral features and interpretation Sum peaks Escape peaks Diffraction peaks Anomalous silicon gold and argon peaks Scatter peaks Spectral background Comparison of wavelength- and energy-dispersive systems Applications Atmospheric particulates Waters Clinical and biochemical Rocks ores and cement Metals and alloys Coal and petroleum On-stream analysis Others Future developments Keywords ; Review ; energy-dispersive X-ray emission awalysis Introduction In their review of X-ray fluorescence analysis in 1970 Carr-Brion and Paynel stated that resolutions of the order of 190 eV at 6 keV had been reported for semiconductor detectors and that better values were to be expected with improvement in detectors and associated electronics.Today detectors are routinely produced with resolutions in the 150-160-eV range and values as low as 140 eV can be attained. Energy dispersion using a lithium-drifted silicon Si(Li) detector was first introduced as a practical tool for X-ray spectrometry in the mid-l960~.~-~ Initially the major impact of the new technology was as an accessory on electron beam microprobes and scanning electron microscopes.6 It was not long however before dedicated X-ray spectrometers were being built around the new detectors.7-l2 Today energy-dispersive attachments are almost standard on electron microscopes and a number of manufacturers offer dedicated X-ray fluorescence equipment for qualitative and quantitative analysis.17 178 CAMPBELL ENERGY-‘DISPERSIVE X-RAY Analyst Vol. 104 It is intended to limit this review to those techniques which utilise the energy-dispersive properties of semiconductor detectors specifically the Si( Li) detector. There is some confusion over the terminology applied to this field of analytical chemistry. Detectors, such as the scintillation and flow proportional counters used in wavelength-dispersive X-ray fluorescence analysis are capable of limited energy resolution.It is therefore possible to consider these as energy-dispersive detectors and indeed use is made of this property in the application of “pulse-height analysis.” However the inability of these detectors to provide sufficient spectral resolution leads to their use in combination with diffraction crystals. Terms such as semiconductor solid-state non-dispersive and energy-dispersive have appeared associated with the Si(Li) detector. It is intended here to follow the guidelines given in the I UPA C Information BuZZetin on X-ray emission spectroscopy,13 for terminology and symbols. It has become the convention to express X-ray wavelengths in Angstroms when considering wavelength-dispersive systems and X-ray energy in kiloelectronvolts when considering energy-dispersive systems.The principal difference between energy-dispersive electron microscope attachments and energy-dispersive X-ray fluorescence analysers is in the mode of excitation used. The techniques share much of the data collection and data processing technology. It is therefore difficult to differentiate between the two systems in an absolute manner and some reference will be made to electron-excitation systems. For those new to the technique a number of books and descriptive articles are available with respect to both X-ray fluorescence analysis in genera11v14-20 and energy-dispersive X-ray fluorescence analysis in p a r t i c ~ l a r . 7 ~ ~ ~ ~ ~ ~ - ~ 0 Instrumentation Fig.1 shows the components of a typical energy-dispersive X-ray fluorescence analyser. The source of excitation shown is the X-ray tube but of course excitation of secondary X-rays can be achieved by using a variety of sources including electrons protons and other charged particles. The building blocks of the system are a source of excitation the sample compartment the solid-state Si(Li) detector the electronic package including pre-amplifier, amplifier and multi-channel analyser (M.C.A.) and the data processing package generally including computer with relevant software to convert the raw data into meaningful results. The over-all system efficiency of an energy-dispersive X-ray fluorescence instrument is a Sample changer Cont ro I assem b I y unit Fig.1. Typical energy-dispersive X-ray fluorescence spectrometer March 1979 EMISSION ANALYSIS. A REVIEW 179 function of a number of parameters including geometry mode of excitation excitation cross-section fluorescence yield and detector efficiency. Cothern et aL31 investigated the system efficiency for the situation where the source of excitation was broad-band X-rays. Excitation As in conventional wavelength-dispersive X-ray fluorescence analysis it is necessary to remove core electrons from the atoms of interest in order to produce the secondary fluorescent X-rays which are characteristic of the elements present in the sample. Normally X-rays, from an X-ray tube are used to fulfil this function and this is still true of most commercial energy-dispersive X-ray fluorescence spectrometers.The alternatives to the X-ray tube are electrons protons and radioisotopes all of which are capable of ejecting core electrons and all of which have various advantages and limitations. Jaklevics2 has compared the use of electrons charged particles (protons or alpha particles) and Xxays (continuous or mono-energetic) as excitation sources for energy-dispersive X-ray analysis. Electron excitation was shown to have much poorer limits of detection. The high continuous background found in electron-induced spectra increases the difficulty of determining low concentrations. Middleman and GelleF have demonstrated the improved peak to background ratios that can be achieved in the X-ray spectra from an electron microscope using X-ray excitation instead of the usual electron beam.Reldy et aE.34 have used muons to excite secondary X-rays. This produces a muonic X-ray spectrum in which the characteristic X-rays are raised in energy hence allowing the light elements to be determined more easily. X-ray Tube X-ray tubes on conventional wavelength-dispersive spectrometers use up to 3 kW of power and produce a high characteristic X-ray flux from the sample In energy-dispersive X-ray fluorescence systems all X-rays are incident simultaneously on the detector and because there is a finite counting capacity it is necessary to modify the X-ray tube output to reduce the secondary X-ray flux. There are basically two approaches to this problem. One is to reduce the power of the X-ray tube to around 10 W and the other to use a secondary target system.Fig. 2 shows the configurations typical for primary or direct excitation and secondary excitation. Specimen Specimen Optional I/’ \\ / / \ \ ,’ filter ; /X\ X’ 1 ;<; ’\ I ‘\ ‘\ \ n Secondary target X-ray tube Detector Detector (a) (b) Fig. 2. X-ray excitation (a) primary or direct; ( b ) secondary. The low-power tube used for primary excitation emits a broad band of X-ray energies from just below the tube potential. Used directly this has the advantage of exciting a wide range of elements but unfortunately produces a high spectral background due to scattering of the radiation by the sample. The use of a filter placed between the tube and the sample attenuates the X-rays and can be used to moderate the primary broad-band radiation to approximate to a mono-energetic source.A proper choice of filter will increase the sensitivity for particular elements while reducing the spectral background over a given region 180 CAMPBELL ENERGY-DISPERSIVE X-RAY Analyst Vol. 104 In the secondary-excitation geometry X-rays from a high-power tube impinge on a chosen target material. The target material itself is’ induced to emit characteristic and mainly monochromatic X-rays which are made to fall upon the sample. This produces a well monochromated beam that gives rise to a lower spectral background contribution. Gedcke et aZ.35 compared the detection limits that could be obtained with the two geometries for a variety of elements and concluded that over the 5-30-keV range there was little to choose between the two techniques.However in a similar study Artz and Short36 found that the best sensitivity was achieved in the atomic number range 35-60 with direct radiation from a tungsten X-ray tube but that secondary excitation with properly chosen targets produced a higher sensitivity for other elements. In a study of the detection limits for light elements in a “synthetic rock mixture” and NBS orchard leaves Anselmo37 found that secondary excitation was generally superior. Investigations of this nature are complicated by a dependence on system geometry. Both primary- and secondary-excitation systems are available on commercial spectro-meters. The primary-excitation system with the optional filter appears to offer a more flexible approach in terms of exciting a large number of elements simultaneously or effici-ently exciting just a few.The secondary target geometry probably gives somewhat better sensitivity over narrow ranges necessitating the use of a variety of targets to cover the X-ray spectrum of interest. Counting times on energy-dispersive X-ray fluorescence instru-ments are typically 100-1000 s owing to the low count-rate capability of the detection and amplification system. This places a high stability requirement both long and short term, on the output from the X-ray tube. Low-power generators are more stable than high-power generators and this favours the use of primary excitation. Van Espen and ad am^^^ have described a technique to compensate for the fluctuating X-ray intensity from a high-power system by making use of a reference signal from a thin metal wire.Detection limits are a function of counting time excitation conditions spectrometer geometry matrix and atomic number. For X-ray excitation t’hey are typically less than 1 p.p.m. for the transition elements and about 0.1% for sodium. The use of pulsed X-ray beams is advocated as a means of increasing the counting-rate capacity of energy-dispersive ~ y s t e m s . ~ ~ ~ * ~ Basically the system operates by turning off the X-ray tube as soon as an event is detected by the spectrometer. Pulse pile-up events are virtually eliminated and counting rates are considerably increased. Polarised X-rays can be obtained by scattering from a suitable material such as boron carbide or from a synchrotron source.Pola,rised radiation is not scattered isotropically. Therefore if a detector is placed in the plane of polarisation and a t right-angles to the incident beam the scatter signal reaching the detector is considerably reduced. This results in a lowering of the spectral background and therefore an improvement in detection l i m i t ~ . ~ l - ~ ~ When using the X-ray tube and polariser it is necessary to use high power to compensate for energy loss on polarisation. R y ~ n ~ ~ found improvements in detection limits of up to 4.5 times for the elements from potassium to strontium in NBS orchard leaves when compared with direct excitation. Radioisotopes Most radioisotopes decay with the emission of X-rays y-rays or a-particles or a combina-tion of these and they are therefore suitable for use as excitation sources.Typical radio-isotope sources used in energy-dispersive X-ray fluorescence work include iron-55 cadmium-109 americium-241 cobalt-57 and gadoliniuni-153 for photon excitation and polonium-210 for a-particle excitation. It is possible to produce broad-band excitation sources utilising radioisotopes. p-Emitters such as promethium-147 or tritium when mixed with a suitable target material produce an X-ray continuum. However as the conversion efficiency is low high activity levels must be employed to obtain good counting rates. A description of the available photon-emitting radioisotopes and a discussion on the design and construction of new sources were given by Leonowich et al 47 A major constraint on the use of radioisotopes is that the emission should be as simple as possible.Where a large number of photon energies are emitted elastic and inelastic scatter from the sample will give rise to a complicated X-ray spectrum. Where high-energy y-rays are emitted high backgrounds can occur owing to detector Compton escape. Hence, to cover the spectral region of interest several simple emitters are preferable March 1979 EMISSION ANALYSIS. A REVIEW 181 The use of an or-emitter such as polonium-210 together with a windowless detector has been shown to allow the detection of oxygen and fluorine by energy-dispersive X-ray Curium-244 a photon and particle source has been used for the excitation of the light elements in rocks by Fran~grote.~g In a study of the limits of detection that can be attained using radioisotope excitation Spatz and Lieserso found that given the correct choice of isotope values were almost as good as those found using X-ray tube excitation.Electrons and Protons The use of electrons to excite characteristic sample X-rays is the basis of electron-microprobe analysis and it was in this field that energy-dispersive X-ray emission first found appli~ation.~~ The topic has been extensively dealt with by a number of worker^.^^-^^ The major disadvantage in the use of electron excitation is the high spectral background produced by electron deceleration. The use of protons to excite characteristic X-rays is attractive from a number of view-points. The ionisation cross-section shows a marked increase with decreasing atomic number thus increasing the sensitivity in this region.Protons (and other charged particles) do not produce the intense continuum found with electrons and the X-ray spectral back-ground is therefore small. However proton-induced X-ray emission or PIXE when applied to conventional “thick” samples can in fact exhibit a high spectral background. This is caused by the production of energetic photoelectrons in the sample that give rise to an X-ray continuum and for this reason PIXE has found most application in “thin” sample analysis. The need to obtain the use of a charged-particle accelerator is an obvious draw-back limiting the production of standard equipment. The principles of the technique have been described57258 and two review papers have a~peared599~O along with application studies in water analysis,57~61 biological material57262 and airborne parti~ulates.~3 A description of the design requirements of the experimental apparatus and a thorough evaluation of the qualitative and quantitative nature of the technique have been given by Johansson et aP4 Reuter and Lurlo65 investigated the application of proton excitation to “thick” samples and compared it with the use of electron excitation.Proton excitation produced sensi-tivities higher by factors of 2 or 3 when applied to low-alloy steels. It has been demon-strated66 that the sample depth probed by protons is smaller and more consistent across the elemental range than it is with X-ray excitation. Ahlberg and AdamsB7 compared the use of proton with X-ray excitation in the analysis of air particulate matter.PIXE was shown to be more sensitive for most elements. However the inhomogeneities found on air particu-late filters caused problems due to the small sampling area in the PIXE technique. Detectors Most X-ray detectors have some capacity to resolve photons in terms of their energy but none perform this task so well as the solid-state detectors. It was chiefly the advent of the solid-state detector with its associated pulse-processing electronics which gave rise to the technique of energy-dispersive X-ray fluorescence. Fig. 3 shows a diagrammatic repre-sentation of a solid-state Si(Li) detector. The detector is a diode consisting of a cylindrical piece of p-type silicon doped with lithium to increase the electrical resistivity. A Schottky barrier contact at the front of the detector produces the p - i - n diode and the application of a reverse bias voltage typically 1000 V depletes the diode of free charge carriers.The dimensions of the detector vary with the application but are typically 4-16 mm in diameter and 3-5mm thick. An X-ray photon entering the diode causes ionisation of the silicon and produces a number of “electron-hole pairs,” the number of which is proportional to the energy of the incident X-ray photon. The applied voltage sweeps the charge from the diode and it is collected at a charge sensitive pre-amplifier. The use of lithium-drifted silicon arises from two factors. The ionisation energy or band gap is small enough at 1.1 eV to produce sufficient charge carriers for good statistical definition of the pulse size but is high enough to prevent thermal-electron excitation being significant.The detector is normally contained in a vacuum behind a thin beryllium window. Both the detector and pre-amplifier are held at liquid-nitrogen temperature to reduce lithium migration and minimise electronic noise. A useful review of the design of solid-state detectors and thei 182 CAMPBELL ENERGY-DISPERSIVE X-RAY Analyst Vol. 104 application to X-ray spectrometry has been given by Heath,68 while Keith and L o ~ r n i s ~ ~ have critically examined the use of the detector with respect to energy calibration detector efficiency and detector phenomena such as escape and sum peaks. Gold - _-_-I Schottky barrier contact - n-type silicon Intrinsic or active region Gold -+, Fig.3. Silicon(lit1iiuin) detector. The efficiency of a solid-state detector in recording an event is a function of such para-meters as area and thickness of the active region dead layer and contact material and thickness of the entrance window. At low X-ray energies (low atomic number) much of the intensity is attenuated by the beryllium window. At high X-ray energies the efficiency is dependent on the detector thickness. Between 4 and 20 keV the detector efficiency exceeds 90% but falls off above and below this energy range.e9 The fundamental limitation to the resolving power of the Si(Li) detector lies in the statistical fluctuation in the number of electron-hole pairs produced by a given X-ray energy. It is customary to express the resolution of an energy-dispersive X-ray system as the full width at half maximum (FWHM) of a peak in the energy spectrum normally that of manganese KM.The FWHM contains contributions from noise sources other than the detector and in particular from the pre-amplifier. At low energies electronic noise is greater than that associated with statistical fluctuations. The detector resolution is normally expressed as FWHM (eV) = 2.3552/FE~ where E is the energy of the X-ray F the Fano factor and E the average energy to create an electron-hole pair. The Fano factor is related to the fractional amount of total energy absorbed resulting in the production of electron-hole pairs and is normally assumed to be about 0.15.68 The total system resolution including the contribution from electronic noise is expressed -as Typically the values of FWHM at 5.9 keV (manganese Ka) are in the range 150-180 eV.The Si(Li) solid-state detector finds application in other fields including X-ray diff rac-tion70~71 and astronomical spectroscopy.72 At photon energies above about 50 keV it is usual to employ another solid-state detector namely the lithium-drifted germanium diode. Electronics The electronic package used to process the 'output from the detector consists of the pre-amplifier amplifier pile-up rejector and multi-channel analyser. It is beyond the scope of this review to describe these components in. detail but they will be discussed briefly for completeness. A more detailed explanation of the design and operation of these components can be found elsewhere.l2P2 March 1979 EMISSION ANALYSIS.A REVIEW 183 Pre - amplifier The function of the pre-amplifier is to convert the charge pulse from the detector into a voltage signal while still retaining the proportionality to the incident X-ray photon energy and adding as little electronic noise as possible. The operation generally involves a form of current integration utilising a cooled field-eff ect transistor (FET) and electronic band-pass filters controlled by electronic shaping time constants. There are a number of pre-amplifier types each using a different technique to minimise the noise contribution. These include continuous optical feedba~k,‘~ drain feedback,7*975 modified-resistive feedback,76 pulsed-optical feedback77 and dynamic-charge re~toration.~~ Amplifier The function of the amplifiers175 is to convert and amplify the signals from the pre-amplifier in such a manner as to make them suitable for presentation to the multi-channel analyser.This function is achieved by “pulse shaping” techniques to attempt to obtain the optimum in energy resolution and counting-rate performance. A large shaping time constant produces optimum energy resolution at the expense of counting-rate performance and a small constant produces the reverse effect. T k important characteristics of a good amplifier are sophisticated pulse shaping good correlation between pulse height and energy and stability of gain and base line to changes in tempera-ture and counting rate. The use of long shaping time constants to obtain optimum energy resolution increases the probability of pulse overlap or “pulse pile-up.” This gives rise to two undesirable spectral features.Pulses are lost from the full energy peaks and a pulse pile-up continuum extends from just above the full energy peak for all peaks in the spectrum giving rise to complicated spectra. A further complication is that the overlap increases with counting rate and the pile-up loss is non-linear with counting rate. These problems can be almost completely overcome by the use of a “pulse pile-up rejector” system.79 This operates by inspecting the time interval between successive pulses from the pre-amplifier and denies entry to those signals where overlap is detected so that the system has a dead time associated with it. The pile-up rejector eliminates all but those sum peaks that are within the pulse width of the pre-amplifier and the intrinsic charge collection time of the detector.The use of a pulsed-excitation source will markedly reduce the problems of pulse overlap.39 Mu1 ti -channel Analy ser The MCA performs the function of sorting the pulses from the amplifier in terms of pulse amplitude and placing these in a “memory” composed of voltage windows or channels. The first stage of the MCA the analogue to digital converter (ADC) allows the incoming pulse to charge a capacitor which is then discharged at a constant current. The time of discharge is proportional to the pulse amplitude and this is used to gate on a constant frequency oscillator to produce a number of pulses. This “number” can then be related to a specific address or channel number.During this process the system will exhibit a dead time during which pulses cannot be accepted and this must be accounted for. There must be sufficient channels available in the MCA memory to span the energy range of interest with good coverage. Normally 1024 channels are used each with the capacity to store lo6 counts. This is important in order to allow good statistical precision when analysing trace amounts in the presence of major components. The information stored in the MCA memory is therefore a histogram of number of counts veisus channel number or after proper Cali-bration energy. The important features of the MCA12 are channel number varying linearly with energy uniform channel widths low dead time live-time clock adequately compensating for dead time and sufficient channels to cover the required energy range.Clearly these must be weighed against each other. Dead Time It is imperative to employ adequate dead-time compensation for quantitative analy~is.’~9~~ Normally the correction procedure employs the concept of “live time.” A clock measures the actual data accumulation time and ignores those intervals when the system is dead. This generally gives adequate compensation of up to about lo4 counts s-l and above this level pulse pile-up effects can cause problems 184 CAMPBELL ENERGY-1)ISPERSIVE X-RAY Analyst Vol. 104 From the foregoing discussion it is obvious that each parameter must be considered in the light of the requirements and limitations of the others.Typically counting rates of about lo3 counts s-1 can be achieved with amplifier shaping constants of between 4 and lops, giving rise to little peak shift or resolution degradation. Data Processing The raw data accumulated by the MCA must be processed to obtain useful information. For simple qualitative analysis the spectral peaks must be identified. More important to make use of a spectral feature for quantitative analysis the peak area spectral background, overlap effects and inter-element absorption and enhancement effects must be considered. I t has become customary to interface the MCA to a mini-computer or micro-processor. Data stored by the MCA are passed to the computer processed according to the user’s program and the final data printed out. An obvious progression from this situation is the use of the computer to control spectrometer parameters such as X-ray tube current and voltage filter or secondary target material and sample changer.The coupling of a computer to an energy-dispersive X-ray fluorescence system and the design of suitable software has been discussed by Keenan.8l Most commercial manufacturers offer systems with either partial or full computer control. Making use of a computer to help identify the elements giving rise to an energy-dispersive X-ray spectrum is a relatively simple matter. By storing data such as spectral line energies and relative intensities of lines from the same element the computer can be used to produce K L and M line markers with which spectral identification can be made. Positive identifi-cation is made by the presence of two X-ray lines in the correct intensity ratio for a particular element.In quantitative analysis there are certain features in the energy-dispersive spectra that complicate quantitative analysis when compared with wavelength-dispersive X-ray analysis. The poorer energy resolution increases the number of spectral overlaps and the peak to background ratio is normally poorer making the operation of background subtraction much more important. Fortunately the phenomena (of higher order reflections from the diffraction crystal can be ignored. The need to correct for inter-element effects both absorption and enhancement exists with all X-ray fluorescence systems. The problems of background subtraction and peak overlap effects are inter-related and a number of papers have dealt with these t ~ p i ~ s .~ ~ ~ ~ ~ - ~ ~ There are principally four methods used to calculate the background under a peak of interest shape fitting interpolation blank subtraction and the regressed constant method. Peak stripping utilising library spectra, overlap factors generated peak shapes or a combination of these is generally used to over-come overlap effects. RusP gives an excellent review of the methods applicable to back-ground subtraction and peak overlaps and lists a number of typical computer programs in BASIC that can be applied to the interpretation of energy-dispersive X-ray spectra. The high backgrounds encountered in electron-excited spectra make it imperative to use effective background calculation techniques.*8~S9 In a n attempt to remove the dependence on efficient background subtraction Nielsons7 has described a method of direct peak analysis based on a method proposed originally by C o ~ e l l .~ ~ No background subtraction is attempted and only a partial peak area is used. The precision is claimed to be acceptable for many routine applications. Statham919g2 has proposed a procedure for deconvolution and back-ground subtraction by suppressing the background using a digital filter and then proceeding to a conventional least-squares fit. The procedures developed for conventional X-ray analysis are equally applicable to energy-dispersive X-ray spectra for the correction of inter-element effects. There are basically two approaches the empirical and the theoretical.The empirical approach proposed by Lucas-Tooth and C O - W O ~ ~ ~ ~ S ~ ~ ~ ~ * - ~ ~ employs a number of standards to determine influence coefficients. The theoretical approach or “ fu ndamental-parameters” met hodgs-loo ut ilises known values of absorption coefficients and fluorescence yields as well as instrumental parameters. The specific application of these correction procedures after modification to energy-dispersive X-ray fluorescence analysis has been discussed by a number of workers.101,102 Neilsonlo3 has described a matrix correction program based on the measurement of the scatte March 1979 EMISSION ANALYSIS. A REVIEW 185 of the characteristic primary radiation. Shen and Russlo4 have produced a modified version of Stephenson’sgg fundamental-parameter model for application to energy-dispersive X-ray spectra.The results obtained indicate that when used carefully this approach can produce acceptable results for many applications without resort to conventional standardisation but that the method should not be used for work where high accuracy and precision are required. Spectral Features and Interpretation In the X-ray spectra obtained with the energy-dispersive X-ray fluorescence spectro-meter the major features are the characteristic X-ray lines from the elements present in the sample being analysed. Unfortunately these are not the only spectral features encountered. A number of other phenomena can give rise to features that must be recognised so as to avoid confusion with the characteristic X-ray lines or “full-energy” peaks as they are known.Sum Peaks Pulse pile-up effects were discussed above and it is inevitable with conventional systems that a certain amount of pulse pile-up will occur. Pulse pile-up increases with increasing counting rate and its effect is to produce sum peaks in the spectra which are essentially the sum of two X-rays being detected simultaneously. Normally sum peaks will appear only for the most intense features of a spectrum. For example in the X-ray spectrum of a steel sample it is probable that sum peaks will occur for the intense iron Ka and KP lines. These will appear in the spectrum at energies corresponding to 2Ka 2K/3 and Ka + KP. Elimination of sum peaks can be accomplished by either reducing the counting rate or using a filter between the sample and detector to remove a portion of the spectrum.Escape Peaks The escape peak phenomenon is caused by the escape of a silicon Ka X-ray from the intrinsic region of the detector. Generally an X-ray entering the detector transfers all of its energy to ionisation in this region. This results in the production of silicon Ka X-rays, which are normally contained within the active region producing further ionisation. How-ever where the probability of silicon Ka X-ray production is high less than 10 keV incoming X-ray energy a significant number of these photons can escape from the intrinsic region. The energy deposited in the detector will therefore be the energy of the incoming photon less the energy of the escaping X-ray 1.74 keV. This will give rise to a spectral feature at full energy minus 1.74 keV.The ratio of the escape peak to the full energy peak decreases with increasing atomic number and is normally about 1 200 for chlorine and 1 1000 for iron. This feature is characteristic of the detector independent of the sample and cannot be eliminated from the spectrum. Diffraction Peaks A sample of a highly ordered or crystalline nature can give rise to a peak in the spectrum by diffraction of the primary X-ray beam. A diffraction peak will appear when the incident X-ray energy and the spectrometer geometry are such that Bragg’s law is satisfied. The diffraction peak can be identified by altering the sample to detector or X-ray tube distance, thus changing the angle and causing the peak to appear at a different energy.This pheno-menon is normally seen where a continuum source of X-rays is used. The probability of satisfying the Bragg condition with mono-energetic excitation is small. The diffraction peak is generally broad and irregular in shape. Anomalous Silicon Gold and Argon Peaks The anomalous silicon and gold X-ray peaks are produced by the interaction of secondary X-rays with the detector dead layer and the gold Schottky barrier. X-rays interacting in these regions produce silicon and gold X-rays which have a probability of reaching the intrinsic region of the detector and being recorded as discrete pulses. Instances of anomalous silicon and gold peaks are fortunately rare; however where a high secondary X-ray flux is incident on the detector with energy between 2 and 3 keV the anomalous silicon peak ma 186 CAMPBELL ENERGY-DISPERSIVE X-RAY Analyst Vol.104 appear. In electron microprobe analysis the appearance of gold absorption edges in the spectral background can complicate background calculations.105 If an air path exists between the X-ray tube sample and detector then argon present in the air can be excited giving rise to secondary argon Ka X-rays at 2.96 keV. The occurrence of anomalous argon peaks is easily avoided by the use of either vacuum or helium paths in the spectrometer. Scatter Peaks There are two types of X-ray scatter coherent (Rayleigh) and incoherent (Compton). Peaks in the X-ray spectrum are generated by scatter of the characteristic lines in the primary-excitation radiation by the sample.Coherent scatter occurs without loss of energy and therefore produces a peak at the energy of the characteristic primary radiation and with a similar band width. Incoherent scatter occurs with energy loss and therefore, produces a spectral peak at a lower energy than that of the primary-excitation radiation and with a fairly broad band width. The ratio of coherent to incoherent scatter peak intensities is a direct function of atomic number. Spectral Background The spectral background continuum is determined by the nature of the excitation system. With direct or primary X-ray tube excitation the spectral background is fairly high and is due mainly to scattering of the primary bremsstrahlung. Other contributions to the back-ground are incomplete charge collection in the detector and sample-generated bremsstrahlung.With a near monochromatic secondary target or filtered X-ray excitation the background can be significantly reduced and is mainly associated with incomplete detector charge collection. With charged-particle excitation the background is due mainly to sample-generated bremsstrahlung. Cooperlos has discussed the features to be found in the energy-dispersive X-ray spectrum and proposed means of identification. Keith and loo mi^^^ have described the origin of some of the spectral features and proposed computer programs to correct the X-ray spectra for such phenomena as escape and sum peaks. The use of computer-generated K L and M line markers enables many of the above-men tioned spectral features to be distinguished easily from full-energy peaks.Comparison of Wavelength- and Energy-dispersive Systems A number of parameters must be considered when comparing energy-dispersive and wavelength-dispersive X-ray fluorescence equipment. These include resolution counting-rate capacity spectral interference peak to background ratio and the time cost and con-venience when carrying out particular analytical procedures. A comparison between energy-dispersive and wavelength-dispersive X-ray spectrometers for electron microscopes has been made by Malissa et aZ.107 but this is not completely relevant to the present discussion. Walingalos has discussed the advantages and limitations of the energy-dispersive X-ray technique in both fluorescence and diffraction in comparison with conventional or wave-length-dispersive systems.He concluded that energy-dispersive X-ray fluorescence analysis would find its major applications in qualitative and semi-quantitative analysis. However, this viewpoint seems dated when compared with more recent publication^^^ and the large number of papers appearing on quantitative analytical applications. Techniques for the analysis of samples of air pollutants have been. compared by Gilfrich et aZ.lo9 They investi-gated the use of various excitation sources such as X-rays a-particles radioisotopes and protons when used on energy-dispersive X-ray equipment and compared detection limits with those found using a wavelength-dispersive spectrometer. Detection limits were found to be comparable when the X-ray tube excited-energy dispersive system was used.Dewolfs et a1 .l10 compared a high-powered secondary-target energy-dispersive X-ray spectrometer with a commercial wavelength-dispersive X-ra.y spectrometer in terms of energy resolution, spectral interferences intensity and peak to background ratio sensitivity reproducibility and precision. The eventual conclusion was that the energy-dispersive technique was advantageous when applied to an analytical sj tuation requiring multi-element analysis with limited precision. In a study of energy- and wavelength-dispersive systems on the electro March 1979 EMISSION ANALYSIS. A REVIEW 187 micro-probe Dunham and Wilkinsonlll found that the techniques compared well in terms of accuracy and precision but that the energy-dispersive system produced poorer limits of detection while being faster and more convenient to use.The resolution of energy-dispersive X-ray systems is poor in comparison with wavelength-dispersive systems with the exception of the K lines of the heavy elements with atomic number greater than 45. This produces more spectral overlaps often requires the use of mathematical fitting procedures or forces the use of less sensitive lines. The counting-rate capacity is poorer for the solid-state Si(Li) detector and the associated electronics and this means that the sensitivity is poorer when only the number of counts per second measured is considered. This is at least partially balanced by the fact that the multi-element capacity of energy-dispersive systems allows much longer counting times to be tolerated.Peak to background ratios are generally poorer for energy-dispersive systems placing more emphasis on efficient background subtraction. Qualitative analysis is considerably simplified and less time consuming using the energy-dispersive system especially when computer-generated markers are used. It is difficult to emphasise sufficiently the benefits of accumulating data simultaneously. For quantitative analysis the energy-dispersive technique can produce good multi-element data with reasonable precision ; however the counting statistics usually produce poorer limits of detection than is the case with wavelength-dispersive spectrometers. Inter-element absorption and enhancement effects should be similar given similar spectro-meter geometries and excitation conditions.The time taken for a particular analysis favours wavelength-dispersive systems when only a few elements are being quantified but favours energy-dispersive systems for multi-element analysis. Energy-dispersive X-ray spectro-meters are somewhat cheaper than sequential wavelength-dispersive spectrometers and much cheaper than simultaneous wavelength-dispersive spectrometers. The energy-dispersive spectrometer is usually smaller and less bulky (especially true of low-power generator systems) and can be initially simpler to use for the non-X-ray spectroscopist. The newer energy-dispersive X-ray fluorescence technique can be seen as a competitor to the wavelength-dispersive spectrometer or as a complement to it depending on the particular analytical requirements of a given situation.For many applications the ultimate in precision is not required and the energy-dispersive system can provide an adequate versatile and relatively inexpensive solution. When fast qualitative analysis is important then the energy-dispersive system must be carefully considered. When very high precision work is required the simultaneous or sequential wavelength-dispersive spectrometer must still be the instrument of choice. When multi-element analysis is required on a limited budget then the energy-dispersive system may appear advantageous when compared with a simultaneous wavelength-dispersive spectrometer. In some analytical situations the choice will be complicated by the need to satisfy a number of conflicting criteria and here the energy-dispersive system can be seen as a complement to the wavelength-dispersive system.Applications The last few years have seen a large increase in the number of publications dealing with the application of the energy-dispersive X-ray emission technique to practical analytical problems. For convenience these will be detailed here in terms of the various fields of interest with which the papers have dealt. All of the applications have the solid-state Si(Li) detector in common but excitation of the characteristic sample X-rays may be by a variety of means as already detailed. Atmospheric Particulates Energy-dispersive X-ray fluorescence analysis has attracted considerable interest for the analysis of airborne particulate matter.l12 There are a number of reasons for this popularity, including its potential multi-element capability non-destructive nature and the advantages gained when working with thin samples.A number of papers have dealt with the problems to be overcome before acceptable multi-element analyses can be 0btained,ll~-~~8 while others describe the results found in various locations and their implication^.^^^-^^^ Air particulates are normally collected by passing a known volume of air through a filter. This results in “thin-specimen,’ samples and these possess a number of advantages over conventional “infinitely-thick” samples. Inter-element effects are eliminated or at worst 188 CAMPBELL ENERGY-:DISPERSIVE X-RAY Analyst Vol. 104 reduced considerably peak to background ratio is normally increased particle size effects are simplified and linear calibrations are generally found over wide ranges.The filter must be as thin as possible to reduce absorption and enhancement effects and contain no contami-nants heavier than sodium. Cellulose fibre or membrane filters are popular although weighing can be a problem owing to hygroscopicity. Using radioisotope excitation Rhodes113 and Rhodes et al.l14 have investigated particle-size distribution sampling and standardisation with respect to thin-filter samples. Adams and co-workers,116-118 using secondary-target X-ray tube excitation have considered a number of potential problems including sample homogeneity and position background subtraction and spectral overlap effects. Generator instability was compensated for by placing a thin zirconium wire below the sample surface and normalising all spectral features to the zirconium Ka X-radiation.Adams and Van Grieken115 have also demonstrated that absorption effects cannot be ignored for elements of low atomic number and correction procedures are necessary for both the air particulate matrix and the filter material. The procedure normally adopted to correct for filter attenuation involves the measurement of front to back intensity ratios of secondary X-rays from the material deposited on the filters.122 Ahlberg and ad am^^^ demonstrated the extra sensitivity that can be obtained using proton excitation as compared with X-ray excitation, especially for the lighter elements. The use of 5-MeV protons to excite secondary X-rays from filter samples has been described by Pilotte et aZ.121 A number of elements including sulphur chlorine and potassium were determined with sufficient sensitivity to allow observa-tion of time variations in elemental concentrations.The variation of elemental composition across the particle size range is of interest with respect to the determination of the source of the particulate matter. Jaklevic and co-worker~~~3J~~ have described a “dichotomous” sampler which is designed to separate airborne particulates above and below 2.4 pm particle diameter. The two particle-size ranges were collected separately and subjected to energy-dispersive X-ray fluorescence analysis using photon excitation. The use of energy-dispersive X-ray analysis on a scanning electron microscope for the characterisation of atmospheric particulate matter was described by Butler et al.lZ5 Specific particles can be chosen for investigation thus allowing a correlation to be made between particle size and elemental composition.The limitation will of course be the lower sensitivity that is available. Calibration standards prepared by vacuum deposition on thin films are commercially available from Micro Matter Co. Seattle and are most useful as an aid to the quantification of energy-dispersive X-ray fluorescence data of filter samples. Waters Information can be sought on one or both of these phases or a total figure may be sufficient. For particulate matter filtration produces thin samples similar to those obtained with atmospheric particu-lates and much of the above discussion applies equally.In most instances dissolved solids are present at concentrations below those directly observable using the energy-dispersive X-ray technique and some form of pre-concent ration is usually employed. In a study of the analysis of sediments and particulate matter in sea water Vanderstappen and Van Grieken126 concluded that a 0.4-,urn Nuclepore polycarbonate filter was optimum for filtration. Samples from the North Sea arid Mediterranean were analysed for a number of elements at the parts per billion (lo9) level with acceptable accuracy and precision. The use of trace-element precipitation by complexing with the chelating agent ammonium tet ramet h ylenedithiocarbamate (ammonium-1 - p yrrolidine dit hiocarbamate APDC) followed by filtration is recommended for the determination of dissolved solids in natural waters.1279128 APDC forms insoluble complexes with about 30 transition metals but does not complex the alkali metals or the alkaline earth metals.Where only the dissolved solids are of interest, particulate matter must first be removed by filtration or centrifugation. Elder et al.128 reported the analysis of a number of elements at the parts per billion level by APDC precipita-tion and filtration through a membrane filter while Pradzynski et ~ z . 1 ~ ~ determined uranium, molybdenum and thorium at the 1 p.p.b. level. Electro-deposition has been advocated as a pre-concentration technique to determine reducible metals in aqueous media by wavelength-dispersive X-ray fluore~cence.~~~ Boslett et a2.130 have proposed a similar technique for use with energy-dispersive apparatus.Zinc copper and nickel were determined at the 2-100 Waters can contain both dissolved solids and particulate matter March 1979 EMISSION ANALYSIS. A REVIEW 189 p.p.b. level using potentiostatic electro-deposition on to a graphite rod producing a thin metallic film. Special cylindrical monochromators were necessary to reduce scatter from the graphite rod and fairly long deposition times were required. Carlton and Russl3l described the use of ion-exchange resin loaded filter-papers to pre-concentrate trace elements. An automated sample collection and preparation modple was presented and sensitivities obtained were in the parts per billion range. Van Grieken et ~ 1 . l ~ ~ have shown that Chelex-100 ion-exchange membranes can be used to pre-concentrate trace elements prior to energy dispersive X-ray fluorescence analysis provided that the water samples contain only modest concentrations of alkali and alkaline earth metals.The use of chelating ion exchangers based on cellulose was described by Burba et ~ 1 . l ~ ~ and illustrated by the determination of uranium in natural waters down to 0.3 p.p.b. A simple sample-spotting procedure has been proposed by Smits and Van Grieken.134 About 1.5 ml of the sample is placed on a cellulose filter and held in position with a wax ring. The water is evaporated using an unheated air stream from below. The choice of the pre-concentration technique depends on the elements of interest the analysis time restrictions and sensitivity requirements.The sample-spotting procedure is probably the simplest and most widely applicable but could give problems at high concentra-tions where crystallisation can occur. Birks and G i l f r i ~ h l ~ ~ have evaluated seven typical energy-dispersive X-ray fluorescence instruments with respect to the determination of trace elements in polluted waters. All were considered capable of measuring elemental concentra-tions at levels appropriate to the problem. Claimed precisions are 15-20% at the 50-100 p.p.b. level. Clinical and Biochemical In many clinical and biochemical studies it is necessary to determine a number of elements in small samples of blood urine tissue etc. When only one element is determined at a time problems can be caused in terms of sample size and analysis time.The multi-element capacity of energy-dispersive X-ray fluorescence is therefore attractive in this field. How-ever its relatively poor sensitivity compared with say atomic-absorption spectrometry, requires that more effort be devoted to sample preparation. The application of the energy-dispersive X-ray technique in this area has been discussed by a number of w o r k e r ~ . ~ ~ ~ - - 1 ~ ~ The determination of trace elements in whole blood plasma and serum has attracted considerable interest. Bearse et ~ 1 . l ~ ~ using a plasma-ashing preparation technique followed by proton-induced X-ray emission determined iron copper zinc selenium and rubidium in 0.1-ml samples of blood. Detection limits for elements with atomic numbers 2545 were shown to be between 0.1 and 1 p.p.m.which should be useful for many applications. Holynska and Marko~iczl~~ also determined selenium in whole blood (and tissue) but used wet ashing and coprecipitation followed by low-powered X-ray tube excitation. Levels as low as 100 p.p.b. were determined. The use of proton excitation allowed the determination of selenium in blood serum at the 10 p.p.b. le~el.l4~ Freeze-drying and pelletising have been used145 to prepare samples of whole blood and plasma for the determination of copper zinc, bromine and rubidium by energy-dispersive X-ray fluorescence using secondary target X-ray excitation with detection limits in the 100-400 p.p.b. range. Knoth et aZ.146 suggested the use of total X-ray reflectionf4’ on a sample support as a means of reducing the spectral background in the determination of copper and iron in blood serum.A 10-pl sample was dried on a support of optically flat silica glass and the incident X-ray beam adjusted to strike the support a t a very low angle such that it is “totally reflected.” Agarwal et ~ 1 . l ~ ~ applied energy-dispersive X-ray fluorescence with secondary target X-ray excitation to the analysis of copper zinc and lead in urine. They suggested the use of an ion-exchange chelating resin as a pre-concentration step and also used yttrium as an internal standard. The method is limited to those elements chelated by the resin. Inductively-coupled plasma-emission spectroscopy has a similar multi-element capacity to energy-dispersive X-ray fluorescence. An evaluation of both techniques for biological work has been carried out by Irons et aZ.149 with reference to sensitivity precision and accuracy.They conclude that the choice of method depends on the sample type plasma emission having advantages for fluids and X-ray fluorescence for solids. The need to apply inter-element corrections in X-ray work was emphasised hence increasing the computer size requirement relative to plasma-emission spectroscopy 190 CAMPBELL ENERGY-DISPERSIVE X-RAY Analyst Vol. 104 Rocks Ores and Cement Cooper and Sch10fkel~~ have described the application of energy-dispersive X-ray fluorescence with direct X-ray tube excitation to rock and ore analysis. Samples of this type normally receive minimal preparation prior to analysis. They are generally ground and pressed into disc form.StanddYdisation is achieved by using similar standard reference materials. Corrections for inter-element effects and peak overlaps must be applied to obtain acceptable results. The general utility of energy-dispersive X-ray fluorescence analysis to geochemical specimens has been demonstrated by Giauque et al.15l Twenty-six trace and two major elements were determined in reference materials use being made of the relationship between the intensity of the incoherent scattered radiation specimen mass absorption coefficient and spectral background intensity to correct for inter-element effects. Other workers have undertaken the analysis of platinum152 and copper153 in ores. Hebert and Bowman154 described a special spectrometer for the analysis of the light elements in rocks which possessed a sensitivity of 5 p.p.m.for sodium. In the analysis of cement-type materials both major and minor elements are determined, although the most important are aluminium silicon calcium and iron. Energy-dispersive X-ray fluorescence with primary X-ray tube excitation has been used to determine these four and eight other elements in Portland cement by Cooper and ~o-workers.15~J56 Once again inter-element correction procedures were shown to improve the quality of the analytical results. Carr-Brion et a1.l5’ have described an on-stream energy-dispersive X-ray fluorescence analyser for the determination of the four main elements of interest in cement raw meal. Results appear to be sufficient for raw meal feed control requirements despite the need to analyse the raw meal “unground.” Metals and Alloys The use of energy-dispersive X-ray emission as an alloy-sorting technique has been des~ribed.l58J~~ Alloys can be quickly characterised by their spectral features in a non-destructive fashion using a portable analyser with minimum or no sample preparation.Janssens et aZ.fsO have proposed a high-precision method for the determination of manganese in ferromanganese. Radioisotope excitation (using cadmium-109) with careful control of the instrumental conditions and standardisation produced a relative standard deviation of 0.274,. The application of the energy-dispersive X-ray technique to multi-element analysis of nickel-alloys has been investigated by Verbeke et al.,lal who found that acceptable results coald be obtained only after applying inter-element corrections using a fundamental-parameters approach.The analysis of thin nickel- gold films has been examined by Franken,162 who found that energy-dispersive X-ray fluorescence produced better data than conventional X-ray diffraction because it was less sensitive to structural effects. Coal and Petroleum The multi-element analysis of coal coke and fly-ash materials by energy-dispersive X-ray fluorescence has been shown to provide acceptable data after application of an inter-element correction routine using multiple standards.163 Lloyd and Francis164 compared the results obtained for the determination of sulphur in coal by conventional ASTM procedures with energy-dispersive X-ray fluorescence results.They concluded that the instrumental tech-nique with proper matrix corrections was close to meeting the ASTM standards of accuracy and precision but was far superior in terms of speed of analysis and convenience. A different approach to the analysis of coal has been taken by De Kalb and Fassel,ls5 who minimised the inter-element effects by converting the powdered coal into a thin film using the Chungl66 technique and obtained good results without inter-element correction procedures. The application of PIXE to coal analysis has been described by Cronch et aZ.167 Yousif and Al-ShahristanP8 applied energy-dispersive X-ray emission with radio-isotope excitation (using iron-55) to the determination of sulphur and vanadium in crude oils. At the concentration levels found in crude oil it was necessary to correct the vanadium data for the presence of sulphur.Vanadium was determined down to 2 p.p.m. and sulphur to 0.03%. Tellerlsg has described an immersion probe designed to determine sulphur in fuel oils or lead in refinery products by the simultaneous measurement of scattered and transmitted low-energy X-rays March 1979 EMISSION ANALYSIS. A REVIEW 19L On-stream Analysis Energy-dispersive X-ray fluorescence possesses a number of attractive features for on-stream analysis when compared with wavelength-dispersive equipment. The use of high take-off angles and broad-beam geometry helps lessen the effects of surface roughness and a well designed system is relatively insensitive to temperature variations. Very short path lengths can be achieved thus increasing light-element sensitivity.The ability to use a variety of excitation systems increases the over-all system flexibility. The advantages and limitations of energy-dispersive X-ray emission and diffraction equipment in process control situations have been discussed by Carr-Brion170 and Kawatra and Da1t0n.l~~ On-stream systems have been described for the process control of cement raw-meal powder157 and finished ~ i n t e r . l ~ ~ Others Applications of the energy-dispersive technique have been both numerous and varied. Cobalt and molybdenum have been determined in hydrodesulphurisation c a t a l y ~ t s l ~ ~ 9174 noble metals in automotive exhaust catalysts,175 silver in photographic materials,176 thorium in optical uranium in aqueous media,17* technetium in nuclear fuel processing waste,l79 copper and silver in Roman silver coins180 and in paper additives.181 Other uses include the analysis of sedimentary pollutantsls2 and agricultural wastes,ls3 ion-exchange s t ~ d i e s l ~ ~ ~ ~ ~ and process control.ls6 Hansonls7 proposed the use of energy-dispersive X-ray fluorescence as a technique to determine the authenticity of art objects.The use of the technique to determine particle sizes has been described by Tominaga et aZ.ls8 while Kawamoto et aZ.lss have designed a milli-analyser to investigate small particles of the order of 1 mm. Future Developments It is unlikely that any dramatic improvement in detector resolution will be made in the immediate future. Small resolution improvements although welcome will not significantly affect the over-all system performance.Improvements in the quality of the detector can, however produce a lowering of the spectral background which would be of considerable benefit. Improvements to the stability of the X-ray generator especially for high-powered systems, would be welcome. The use of pulsed X-ray sources now becoming commercially available, should help to improve the counting-rate capability of the system but it is not clear whether the pulse-processing electronics are sufficiently sophisticated to allow full benefit to be gained. With many commercial systems the determination of the heavy elements can present problems. The typical X-ray generator is designed to produce a maximum of 50 kV and this does not efficiently excite the K lines of elements with 2 m 50-60,91 while the L lines are subject to a variety of possible spectral overlaps.A move toward generators capable of producing 60-80 kV would be welcome. The most significant advances in the near future will lie in the area of computer software. The power of the dedicated mini-computer especially when coupled to interactive disc drives has still to be fully utilised. The identification of spectral features is primarily the work of the instrument operator though generally aided by the computer with such devices as K L and M markers. The extraction of quantitative data from energy-dispersive X-ray spectra using the fundamental parameters approach will improve to a high degree of sophistication.190 The techniques of deconvolution and inter-element correction though now well documented require further development.Finally although a large number of publications have appeared dealing with applications of the technique many more studies are necessary to define the best fields of application, especially with respect to conventional infinitely thick samples. There appears to be no successor to the solid-state Si(Li) detector on the horizon. This task will become more and more computer dependent. References 1. 2. 3. Carr-Brion K. G. and Payne K. W. Analyst 1970 95 977. Elad E. Nucl. Instrum. Meth. 1965 37 327. Elad E. and Nakamura M. Nucl. Instrum. Meth. 1966 41 161 192 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. CAMPBELL ENERGY-DISPERSIVE X-RAY Analyst Vol. 104 Bowman H. R. Hyde E. K. Thompson S. G. and Jared R. C. Science N . Y . 1966 151 562. Elad E. and Nakamura M. I.E.E.E. Trans. Nucl. Sci. 1967. 14 523. “Energy Dispersive X-Ray Analysis X-Ray and Electron Probe Analysis,” Special Publication Heinrich K. F. J. Barrett C. S. Newkirk J. B. and Ruud C. O. Editors Advances in X-Ray Heath R. L. Adv. X-Ray Analysis 1972 15 1. Giauque R. D. and Jaklevic J. M. Adv. X-Ruy Analysis 1972 15 164. Dyer G. R. Gedcke D A. and Harris T. R. Adv. X-Ray Analysis 1972 15 228.Jaklevic J. M. Giauque R. D. Malone D. F. and Searles W. L. Adv. X-Ray Analysis 1972 15, Gedcke D A. X-Ray Spectrom. 1972 1 129. “Appendices on Provisional Nomenclature Symbols Terminology and Conventions-Number 54,” Bertin E. P. “Principles and Practice of X-Ray Spectrometric Analysis,” Plenum Press New Liebhafsky H. A. Pfeiffer H. G. Winslow E. H. and Zemany P. D. “X-Rays Electrons and Jenkins R. and De Vries J. L. “Practical X-Ray Spectrometry,” Macmillan London 1975. Birks I,. S. and Gilfrich J. V. Analyt. Chem 1976 48 273R. De Vries J . L. “X-Ray Fluorescence Spectrometry-Review of Literature,” Seventh Edition, Jenkins R. “An Introduction to X-Ray Spectrometry,” Heyden London 1974. Bertin E. P. “Introduction to X-Ray Spectrometric Analysis,” Plenum New York 1978.Woldseth R. “X-Ray Energy Spectrometry,” Kevex Corp. Burlinghame Calif. 1973. Schunemann D. Miner. Slov. 1973 5 569. Friant A. Onde Elect. 1976 56 69. Russ J. C. X-Ray Spectrom. 1972 1 119. Friant A. Gras R. Lorin A. and Saliou C. Bull. Instrumn Nucl. 1971 42 34. Porter D. E. and Woldseth R. Analyt. Chem. 1973 45 604A. Lister D. B. “Application of Energy-Dispersive X-Ray Fluorescence,” AID 75437 Instrument Gedcke D. A. and Elad E. “Proceedings of the 6th International Conference on X-Ray Optics Goulding F. S . and Jaklevic J . M. A . Rev. N’ucl. Sci. 1973 23 45. Russ J. C. Shen R. B. and Jenkins R. “Energy Dispersive X-Ray Analysis of Materials Principles Cothern C. R. Manuel H. L. and Millette R. J. X-Ray Spectrom. 1974 3 53. Jaklevic J. M. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich.1976,” Middleman L. M. and Geller J. D. “Proceedings of the 9th Annual Scanning Electron Microscope Reldy J. J. Hutson R. L. Daniel H. and Springer K. Analyt. Chem. 1978 50 40. Gedcke D. A. Elad E. and Denee P. B. X-Ray Spectrom. 1977 6 21. Artz B. E. and Short M. A. “Proceedings of :ERDA X- and Gamma-Ray Symposium Ann Arbor, Anselmo V. C. Report Lawrence Livermore Laboratory UCID 17317 1976. Van Espen P. J. and Adams F. C. Analyt. Chem. 1976 48 1823. Jakievic J. M. Goulding F. S. and Landis D. A. I.E.E.E. Trans. Nucl. Sci. 1972 NS19 392. Jaklevic J. M. Landis D. A. and Goulding F. S. Report Lawrence Berkeley Laboratory LBL Howell R. H. Pickles W. L. and Cate J. L. Adv. X-Ray Analysis 1974 18 265. Kaufman L.and Camp D. C. Adv. X-Ray Analysis 1974 18 247. Dzubay T. G. Jarret. B. V. and Jaklevic J. M. Nucl. Instrum. Meth. 1974 115 297. Howell R. H. and Pickles W. L. Nucl. Instmm. Meth. 1974 120 187. Ryon R. W. Report Lawrence Livermore Laboratory UCRL 78063 1976. Aiginger H. Wobrauschek P. and Brauner C. Nucl. Instrum. Meth. 1974 120 541. Leonowich J. Pandian S. and Preiss I. L. J . Radioanalyt. Chem. 1977 40 175. Robertson R. Nucl. Instrum. Meth. 1977 142 121. Franzgrote E. Adv. X-Ray Analysis 1972 15 388. Spatz R. and Lieser K. H. 2. Analyt. Chem. 1977 288 267. Fitzgerald R. Keil K. and Heinrich K. F. J Science N . Y . 1968 159 528. Gedcke D. A. in Holt D. B. Muir M. D. Grant P. R. and Boswarva I. M. Editors “Quantita-Birks L. S. “Electron Microprobe Analysis,” Interscience New York 1963.Waldl E. Wolfermann H. Rusovic N. and Warlimont H. Analyt. Chem. 1975 47 1017. Desborough G. A. and Heidel R. H. APPl. Spectrosc. 1973 27 456. Reed S. J. B. and Ware N. G. X-Ray Spectvom. 1973 2 69. Walter R. L. Willis R. D. Gutknecht W. F. and Joyce J. M. Analyt. Chem. 1974 46 843. Johansson G. Akselsson R. Bohgard M. Carlsson L. E. Hansson H. C. Lannefors H. and Johansson S. A. E. and Johansson T. B. Nwcl. Instrum. Meth. 1976 137 473. STP 485 American Society for Testing and Materials Philadelphia Pa. 1971. Analysis 1972 Volume 15. 266. I UPAC Information Bulletin 1976. York 1970. Analytical Chemistry,” Wiley New York 19’72. Philips Eindhoven 1976. Society of America 1975. and Microanalysis Osaka Japan 1971,” University of Tokyo Press Toyko 1972.and Experiments,” Edax International Prairie View Ill. 1978. NTIS Springfield Va. 1977 p. 1. Symposium Chicago 1976,” Scanning Electron Microsc. 1976 9 Part 1 171. Mich. 1976,” NTIS Springfield Va. 1977 p. 7. 4248 1978. tive Scanning Electron Microscopy,” Academic Press New York and London 1974 Ch. 12. Malmqvist K. Proc. Analyt. Div. Chem. SOL 1978 15 24 March 1979 EMISSION ANALYSIS. A REVIEW 193 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115.116. 117. 118. 119. Folkmann F. J . Phys. E 1975 8 429. Lochmiiller C. H. Galbraith J. W. and Walter R. L. Analyt. Chem. 1974 46 440. Mangelson N. F. Hill M. W. Nielson K. K. and Ryder J . F. Nucl. Instrum. Meth. 1977 142, Malmqvist K. Akselsson R. A. and Johansson G. Proc. Analyt. Div. Chem. Soc. 1978 15 13. Johansson T. B. Van Grieken R. E. Nelson J . W. and Winchester J . W. Analyt. Chem. 1975, Reuter W. and Lurlo A. Analyt. Chem. 1977 49 1737. hhlberg M. S. Nucl. Instrum. Meth. 1977 146 465. Ahlberg M. S. and Adams F. C. X-Ray Spectrom. 1978 7 73. Heath R. L. Adv. X-Ray Analysis 1972 15 1. Keith H. D. and Loomis T. C. X-Ray Spectrom. 1976 5 93. Martin G. W. and Klein A. S. Adv. X-Pay Analysis 1972 15 254. Laine E. Lahteenmaki I. and Kantola M. X-Ray Spectrom.1972 1 93. Holt S. S. Space Sci. Instrumn 1976 2 205. Elad E. “Energy Dispersive X-Ray Analysis X-Ray and Electron Probe Analysis,” Special Publication STP 485 American Society for Testing and Materials Philadelphia Pa. 1971 p. 57. Goulding F. S. Walton J. T. and Pehl R. H. Report Lawrence Berkeley Laboratory UCRL 19377 1977. Walter F. J. “Energy Dispersive X-Ray Analysis X-Ray and Electron Probe Analysis,” Special Publication STP 485 American Society for Testing and Materials Philadelphia Pa. 1971 p. 82. Elad E. I.E.E.E. Trans. Nucl. Sci. 1972 19 403. Landis D. A. Goulding I;. S. Pehl R. H. and Walton J. T. I.E.E.E. Trans. Nucl. Sci. 1971, Gedcke D. A. Elad E. and Dyer G. R. “Proceedings of the 6th National Conference on Electron Reed S. J. B.J . Phys. E 1972 5 997. Reed S. J. B. J . Phys. E 1972 5 994. Keenan J. A, Int. Lab. 1975 MaylJune 49. Fiori C. E. Myklebust R. L. Heinrich K. F. J. and Yabowitz H. Analyt. Chem. 1976 48 172. Connelly A. L. and Black W. W. Nucl. Instrum. Meth. 1970 82 141. Smith D. G. W. Gold C. M. and Tomlinson D. A. X-Ray Spectrom. 1975 4 149. Geiss R. H. and Hwang T. C. X-Ray Spectrom. 1975 4 196. Russ J. C. X-Ray Spectrom. 1977 6 37. Nielson K. K. X-Ray Spectrom. 1978 7 15. Comins N. R. and Thirlwall. J. T. X-Ray Spectrom. 1978 7 92. Bauer R. and Rick R. X-Ray Spectrom. 1978 7 63. Covell D. F. Analyt. Chem. 1959 31 1785. Statham P. J. Analyt. Chem. 1977 49 2149. Statham P. J. X-Ray Spectrom. 1978 7 132. Lucas-Tooth H. J. and Price B. J. Metallurgia 1961 64 149. Lucas-Tooth H.J. and Pyne C. Adv. X-Ray Analysis 1964 7 523. Traill K. J. and Lachance G. R. Geol. Surv. Can. Pap. 1965 No. 64. Traill R. J. and Lachance G. R. Can. Spectrosc. 1966 11 (3) 63. Rasberry S. D. and Heinrich K. F. J. Analyt. Chem. 1974 46 81. Criss J. W. and Birks L. S. Analyt. Chem. 1968 40 1080. Stephenson A. Analyt. Chem. 1971 43 1761. de Jongh W. K. X-Ray Spectrom. 1973 2 151, Russ J. C. Sandborg A. O. Barnhart M. W. Soderquist C. E. Lichtinger R. W. and Walsh, C. J. Adv. X-Ray Analysis 1973 16 284. Ong P. S. Cheng E. L. and Sroka G. Adv. X-Ray Analysis 1974 17 269. Nielson K. K. Analyt. Chem. 1977 49 641. Shen R. B. and RUSS J . C. X-Ray Spectrom. 1977 6 56. Brombach J. D. X-Ray Spectrom. 1978 7 81. Cooper J . A. Am. Lab. 1976 Nov. 35. Malissa H.Grasserbauer M. and Hoke E. Microchim. Acta 1974 5 465. Walinga J . .‘:Proceedings of the 8th Philips X-Ray Analytical Conference Birmingham England, Gilfrich J. V. Burkhalter P. G. and Birks L. S. Analyt. Chem. 1973 45 2002. Dewolfs K. De Neve R. and Adams F. Analytica Chim. Acta 1976 75 47. Dunham A. C. and Wilkinson F. C. F. X-Ray Spectrom. 1978 7 50. Dzubay T. G. Editor “X-Ray Fluorescence Analysis of Environmental Samples,” Ann Arbor Rhodes J. R. Am. Lab. 1973 July 57. Khodes. J . R. Pradzynski A. H. and Sieberg R. D. Air Qual. Instrumn 1974 2 1. Adams F. C. and Van Grieken R. E. Analyt. Chem. 1975 47 1767. Adams F. C. and \’an Espen P. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Van Espen P. and Adams F. C. Analytica Chim. Acta 1974 75 61. Van Espen Y.Nullens H. and Adams F. C. 2. Analyt. Chem. 1977 285 215. Rhodes J. R. Pradzynski A. H. Hunter C. B. Payne J. S. and Lindgren J. L. Envir. Sci. 133. 47 855. 18 115. Probe Analysis,” Electron Probe Analysis Society of America New York 1971 p. 5.4. 1972,” Philips Eindhoven 1973. Science Publishers Ann Arbor Mich. 1977. Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 182. Technol. 1972 6 922 194 CAMPBELL ENERGU-DISPERSIVE X-RAY Analyst VoE. 104 120. 121. 122. 123. 124. Hammerle R. H. and Pierson W. R. Envir. Sci. Technol. 1975 9 1058. Pilotte J . O. Nelson J. W. and Winchester, J. W. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 161. Davis D. W. Reynolds R. L. Tsou G. C. and Zafonte L.Analyt. Chem. 1977 49 1990. Jaklevic J. M. Loo B. W. and Goulding F. S. Report Lawrence Berkeley Laboratory LBL 4834, 1976. Loo B. W. Jaklevic J . M. and Goulding F. S. in Liu B. Y. H. Editor “Fine Particles Aerosol Generation Measurement Sampling and Analysis,” Academic Press New York London 1976, p. 311. Butler J. D. MacMurdo S. D. and Stewart C. J. Int. J . Environ. Stud. 1976 9 93. Vanderstappen M. and Van Grieken R. 2. d4nalyt. Chem. 1976 282 25. Pradzynski A. H. Henry R. E. and Draper E. L. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 175. Elder J. F. Perry S. K. and Brady F. P. Envir. Sci. Technol. 1975 9 1039. Vassos B. H. Hirsch R. F. and Letterman W. Analyt. Chem. 1973 45 792. Boslett J. A.Towns R. L. R. Megargle R. G. Pearson K. H. and Furnas T. C. Jr. Analyt. Carlton D. T. and Russ J. C. X-Ray Spectrom. 1976 5 172. Van Grieken R. E. Bresseleers C. M. and Vanderborght B. M. Analyt. Chem. 1977 49 1326. Burba P. Gleitsmann B. and Lieser K. H. 2. Analyt. Chem. 1978 289 28. Smits J. and Van Grieken R. Analytica Chtim Acta 1977 88 97. Birks L. S. and Gilfrich J. V. Appl. Spectrosc. 1978 32 204. Cox H. L. Jr. and Ong P. S. Med. Phys. 11977 4 99. Kaufman L. Nelson J. Price D. Shames D. and Wilson C. J. I.E.E.E. Trans. Nucl. SCi., 1973 20 402. Mangelson N. F. Allison G. E. Christensen J. J. Eatough D. J. Hill M. W. Izatt R. M. and Nielson K. K. “Proceedings of the 2nd International Symposium on Trace Element Metabolism in Animals 1973,” University Park Press Baltimore Md.1974 p. 439. Reuter F. W. and Reynolds W. L. Adv. Exp. Med. Biol. 1974 48 621. Kaufman L. Deconinck F. Camp D. C. Voegele A. L. Friesen R. D. and Nelson J. A., “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS, Springfield Va. 1977 p. 118. Zombola R. R. Kitos P. A. and Bearse R. C. Analyt. Chem. 1977 49 2203. Bearse R. C. Close D. A. Malanify J. J. and Umbarger C. J. Analyt. Chem. 1974 46 499. Holynska B. and Markowicz A. Radiochem. Radioanalyt. Lett. 1977 31 165. Berti M. Buso G. Colautti P. Moschini G. Stievano B. M. and Tregnaghi C. Analyt. Chem., Stump I. G. Carruthers J. D’Auria J . M. Applegarth D. A. and Davidson A. G. F. Clin. Knoth J. Schwenke H. Marten R. and Glauer J. J . Clin. Chem. Clin. Biochem. 1977 15 537.Wobrauschek P. and Aiginger H. Analyt. Chem. 1975 47 852. Agarwal M. Bennett R. B. Stump I. G. and D’Auria J. M. Analyt. Chem. 1975 47 924. Irons R. D. Schenk E. A. and Giauque R. D. Clin. Chem. 1976 22 2018. Cooper J. A. and Schlofke D. B. Skillings .Min. Rev. 1976 65 1. Giauque R. D. Garrett R. B. and Goda L. Y. Analyt. Chem. 1977 49 62. Coombes R. J. Chow A. and Flint R. W. Analytica Chim. Acta 1977 91 273. Shenberg C. Ben Haim A. and Amiel S. Analyt. Chem. 1973 45 1804. Hebert A. J. and Bowman H. R. Report Lawrence Berkeley Laboratory LBL 3268 1975. Cooper J. A. Wheeler B. D. Bartell D. M. and Gedcke D. A. Adv. X-Ray Analysis 1976,19, Cooper J. A. Wheeler B. D. and Bartell D. M. Cem. Technol. 1976 March/April 68. Carr-Brion K. G. Kipping P. J. New R. Nutter J.C. and Thomlinson F. I. Wld Cem. Technol., Pradzynski A. H. and Draper E. L. Jr. “Proceedings of the 9th Symposium N.D.E. San Antonio, Harrison P. E. and Kenna B. T. Talanta 1972 19 810. Janssens R. Maenhaut W. and Hoste J. Analytica Chim. Acta 1975 76 37. Verbeke P. Nullens H. and Adams F. Proc. Analyt. Div. Chem. Soc. 1978 15 18. Franken P. E. C. Thin Solid Films 1976 31 337. Cooper J. A. Wheeler B. D. Wolfe G. J. Bartell D. M. and Schlafke D. B. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977, p. 169. Lloyd W. G. and Francis H. E. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 166. De Kalb E. L. and Fassel V. A. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich.1976,” NTIS Springfield Va. 1977 p. 209. Chung F. H. Adv. X-Ray Analysis 1976 19 181. Cronch S. M. Ehmann W. D. Laumer H. W. and Cabbard F. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 190. Yousif A. N. and Al-Shahristani H. Int. J . Appl. Radiat. Isotopes 1977 28 759. Teller S. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich. 1976,” 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. Chem. 1977 49 1734. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 166. 157. 158. 159. 160. 161. 162. 163. 1977 49 1313. Biochem. 1977 10 127. 213. 1977 8 123.1973,” Southwest Research Institute San Antonio Texas 1973 p. 47. 164. 165. 166. 167. 168. 169. NTIS Springfield Va. 1977 p. 194 March 1979 EMISSION ANALYSIS. A REVIEW 195 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. Carr-Brion K. G. X-Ray Spectrom. 1973 2 63. Kawatra S. K. and Dalton J. L. Canmet Rep, 1977 77 21. Grobstuck P. Gumprich M. Ihlefeldt J. Kopineck H. J. and Tappe W. Stahl Eisen 1977 97, Labrecque J. J. J. Radioanalyt. Ckem. 1977 41 127. Labrecque J. J. Preiss I. L. and Pandian S. “Proceedings of ERDA X- and Gamma-Rap Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 173. Elgart M. F. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich.1976,” NTIS Springfield Va. 1977 p. 227. Ehn E. X-Ray Spectrom. 1973 2 27. Keenan J. A. Appl. Spectrosc. 1975 29 63. Bertrand C. C. and Linn T. A. Jr. Analyt. Chem. 1972 44 383. Metcalf S. G. Analytica Chim. Acta 1977 93 297. Klockenkamper R. and Hasler K. 2. Analyt. Chem. 1978 289 346. Buchnea A. McNelles L. A. Sinclair A. H. and Hewitt J. S. “Proceedings of ERDA X- and Wogman N. A, Rieck H. G. and Kosorok J . R. Nucl. Instrum. Meth. 1975 128 561. Reuter F. W. Hautala E. Randall J. M. Friedman M. and Masri M. S. “Proceedings of the 2nd International Conference on Nuclear Methods of Environmental Research 1974,” NTIS, Springfield Va. 1974 p. 168. 1106. Gamma-Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 242. Nucci J. F. and Stearns R. L. Soil SGZ’. 1977 123 264. Mucci J. F. and Stearns R. L. J. Chem. Educ. 1975 52 750. Delmastro A. M. “Proceedings of ERDA X- and Gamma-Ray Symposium Ann Arbor Mich., 1976,” NTIS Springfield Va. 1977 p. 215. Hanson V. F. AppZ. Spectrosc. 1973 27 309. Tominaga H. Enomoto S. Senoo M. and Tachikawa N. “Proceedings of ERDA X- and Gamma-Kawamoto A. Hirao 0, Kashiwakura J . and Gohshi Y. “Proceedings of ERDA X- and Gamma-Hurley R. G. and Goss R. L. X-Ray Spectrom. 1978 7 70. Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 211. Ray Symposium Ann Arbor Mich. 1976,” NTIS Springfield Va. 1977 p. 263. Received July 6th 1978 Accepted October 5th 197

ISSN:0003-2654    

DOI:10.1039/AN9790400177

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

196 Analyst, March, 1979, Vol. 104, pp. 196-200 Analysis of Steroids Part XXXII." Polarographic and Gas-chromatographic Methods Sandor Gorog, Anna Lauk6 and Zsofia Sziklay Determination of Allyloestrenol by Titrimetric, Chemical Works Gedeon Richter Ltd., P.O. Box 27, H-1476 Budapest, Hungary A titrimetric method is described for the determination of allyloestrenol based on methoxymercuration of its double bonds and titration of the acetic acid formed with standard sodium hydroxide solution. The relative standard deviation of the method is 0.29%. The polarographic reduction of the mercury addition compound on the dropping-mercury electrode is used for the determination of allyloestrenol in a tablet formulation with a relative standard deviation of 3.1 %. A gas-chromatographic method with a relative standard deviation of 1.5% is also described.The applicability of these methods to the determination of the stability of allyloestrenol and of its dosage form is discussed. Keywords : A llyloestvenol determination ; titrimetry ; polarography ; gas chromatography The analysis of allyloestrenol (17~-allyloestr-4-en-17-01) and its tablet formulation sets serious problems for the analyst. The molecule of this synthetic progestogen is one of the least substituted of the steroid drugs, its functional groups being two isolated double bonds and a tertiary hydroxyl group. In addition the drug, both in bulk and in formulations, is very sensitive to atmospheric oxygen and, therefore, any reliable analytical method should enable the unchanged material to be determined in the presence of its oxidative degradation products.Little information is available in the literature on the determination of allyloestrenol. Fokkens and Polderman1 described its examination by thin-layer chromatography, including the detection of the degradation products. Ganshirt and Poldennan2 developed this method into a quantitative colorimetric procedure using aqueous sulphuric acid for the colour development after thin-layer chromatographic separation and spot elution. This method enabled an assay of the stability of solid dosage forms to be carried out. Another colori- metric method was described by Kato3 using aluminium chloride as the reagent. Cavina et aL4 reported on the column-chromatographic determination of the material using flame- ionisation detection for monitoring of the eluent.The gas-chromatographic separation and gas-chromatographic - mass-spectrometric investigation of allyloestrenol in biological fluids was described by Brooks and Middletich.5 The purpose of this study was to develop a reliable titrimetric method for the determina- tion without the use of standards and to develop rapid polarographic and gas-chromato- graphic methods for examining the stability of the drug when in tablet form. Experimental Reagents and Apparatus without further purification. All materials and the solvent (methanol) were of analytical-reagent grade and were used Titrimetric method is prepared. Mercury(II) acetate solution, 0.1 M in methwzol. Sodium nitrate solution, saturated solution ixb methanol.Sodium bromide solution, 4 M in water. Sodium hydroxide solation, 0.1 M in water. medium using phenolphthalein as indicator. This should be used the same day as it Standardised against oxalic acid in a methanolic * For Part XXXI of this series, see Analyst, 1978, 103, 346.GOROG, L A U K ~ AND SZIKLAY Polarographic method Radelkis (Budapest) OH-105 d.c. - a.c. polarograph. &lercury(II) acetate solution, 0.1 M in methanol containing 1 ml 1-1 of acetic acid. Thymol solution, 0.1% m/V in methanol. Sodium hydroxide solution, 1 M in water. 197 Gas chromatographic method integrator. Materials investigated Bulk allyloestrenol (crude and purified samples) and other steroids were products of Chemical Works Gedeon Richter Ltd., Budapest. Turinal (Richter) tablets contain 5 mg of allyloestrenol per tablet.Hewlett-Packard 7620 gas chromatograph with jame-ionisation detector and 3380 reporting Internal standard. Dehydroepiandrosterone (3/3-hydroxyandrost-5-en-17-one). Procedures Titrimetric method An accurately weighed sample of allyloestrenol (approximately 0.12 g) is dissolved in 10 ml of methanol, 20 ml of 0.1 M mercury(I1) acetate solution and 5 ml of saturated sodium nitrate solution are added and the solution is allowed to stand for 30 min. Then 2 ml of 4~ sodium bromide solution are added and the slightly turbid solution is titrated with 0.1 M sodium hydroxide solution to a phenolphthalein end-point. A blank titration is carried out in the same manner but omitting the allyloestrenol, and this titration value (less than 0.2 ml) is subtracted from the above titre.The equivalent mass is half of the relative molecular mass (150.2). Polarographic method Fifteen Turinal tablets are finely powdered and a portion of the powder equivalent to about 60 mg of allyloestrenol is accurately weighed. It is triturated with 20 ml of methanol for 30min and the solution is filtered through a filter-paper into a 25-ml calibrated flask, washed with small portions of methanol and the filtrate diluted to volume with methanol and mixed. A 10-ml volume of this solution is transferred into a 25-ml calibrated flask, 5 ml of 0.1 M mercury(I1) acetate solution and 2 ml of saturated sodium nitrate solution are added and the solution is allowed to stand for 30 min. Then 2.5 ml of 1 M sodium hydroxide solution are added and the solution is diluted to volume with methanol.A 10-ml portion of the yellow, turbid solution is transferred into a polarographic cell, 1 ml of 0.1% thymol solution is added and the cell is carefully de-aerated with oxygen-free nitrogen. The solution polarographed in the ax. mode from -1 V, using dropping-mercury and mercury-pool electrodes as the working and reference electrodes, respectively. A standard solution is prepared by dissolving an accurately weighed amount of standard allyloestrenol (about 25 mg) in 10 ml of methanol in a 25-ml calibrated flask and treating the solution as described above beginning at ". . . 5 ml of 0.1 M mercury(I1) acetate solution and 2 ml o f . . ." The allyloestrenol content of the tablets is calculated from the peak currents of sample and standard in the usual way.Gas-chromatographic method One accurately weighed tablet is placed in a 25-ml calibrated flask, 26ml of methanol are added and the tablet is disintegrated by vigorous shaking. The flask is shaken for a further 15 min, 2 ml of internal standard solution containing 10 mg of dehydroepiandro- sterone dissolved in methanol are added and the volume of the solution is adjusted to the mark with methanol. The solution is mixed thoroughly and allowed to stand until the tablet base has settled. A 2-pl aliquot of the clear supernatant liquid is injected into the gas chromatograph. A glass column, 6 f t x 3mm i.d., packed with Anakrom ABS, 90-100 mesh, coated with 1% of OV-101 was used. The column temperature was 220°C, the198 GOROG, L A U K ~ AND SZIKLAY: ANALYSIS A?aalyst, Vol.102 vaporiser zone temperature 250 "C and the detector temperature 250 "C. The carrier gas (nitrogen) flow-rate was 30 ml min-l. The retention times of allyloestrenol and dehydro- epiandrosterone were 5.3 and 6.6 min, respectively. Results and Discussion Titrimetric method Titrimetric methods are seldom used in the analytical chemistry of steroids6 The favourable results obtained in this laboratory for the titrimetric analysis of steroidal double b0nds,7-~ ethyny1,lO keto,ll phenolic hydroxyl12J3 and epoxide14 groups, 16, 17-diols15 and 21-acyloxycorticosteroid~~~ encouraged us to try to develop such a method for the deter- mination of allyloestrenol. Further, titrimet ric methods in general have the advantage that no standard sample is necessary for the analysis; this is particularly important in the instance of allyloestrenol as it is one of the 1es:j stable steroid drug materials.Our first experiments in which we aimed to base the determination on the titration of the double bonds by our earlier described catalytic bromination m e t h ~ d , ~ failed to give acceptable results. However, good results were obtained by the addition of mercury( 11) acetate and methanol to the double bonds, complexing the excess of mercury(I1) acetate with sodium bromide and titration of the liberated acetic acid. OH OH + 2Hg(OCOCH3)2 + 2CH30H 1 O-C-CHs ll + 2CH3COO H 0 The principle of this method was described by Martin17 and developed to its generally accepted form by Johnson and Fletcher.l* On the basis of the papers published on the application of this method and references in standard monographs,lS2l this method appears to be applicable mainly to olefins with three, or at least two, hydrogen atoms in the cis position attached to the double bond.As double bonds of this type are usually not present in steroids, no indication of the use of this method for steroids has been found in the literature. According to our measurements 2 mol of mercury(I1) acetate are consumed by 1 mol of allyloestrenol, indicating that both double bonds are involved in the reaction. If the reaction is carried out at 0 "C 1 mol is consumed instantaneously while the reaction with the second mole takes place at a measurable rate (1.15 mol after a reaction time of 2 min).The first mole of mercury(I1) acetate undoubtedly reacts with the allylic double bond. The reaction between the A4 double bond and.niercury(I1) acetate is sluggish even at room temperature and a reaction time of more than 1 h would be necessary to complete the reaction at 25 "C. The use of sodium nitrate as catalyst considerably decreases the reaction time. The use of longer reaction times does not influence the results. By using the described method a freshly prepared standard allyloestrenol sample was found to be 100.21 yo pure. The relative standard deviation (eight parallel determinations) was 0.29%. Similarly good results were obtained with oestr-4-en-l7p-01 and oestr-4-en-17- one where naturally only 1 mol of the reagent was consumed. Several other steroids were also tested in order to obtain data for determining the selectivity of the method.It has been found that the most frequently occurring double bonds in steroid hormones do not react at all (6-ene-3p-hydroxy, 4-ene-3-ket0, 1,4-diene- 3-keto, 16-ene-20-ket0, 9-ene or ll-ene derivatives) and hence the method can be regarded as being fairly specific. The most serious interference was caused by the 17a-ethynyl group and the free phenolic hydroxyl group of oestrogens. However, no stoicheiometric amountsMarch, 1979 OF STEROIDS. PART XXXII 199 of acetic acid were formed in these instances (about 2.5 equivalents in the first group and 1-2 equivalents, depending on the reaction time, in the second group). The results obtained suggest that the titrimetric method is accurate and precise enough for it to be used for the assay of bulk allyloestrenol.The relatively large sample size obviously precludes the possibility of its use for the assay of tablets. Polarographic method The polarographic determination of olefinic unsaturation based on methoxymercuration was described by Fleet and Jee.22s23 The present method is a slightly modified application of the above method to the determination of allyloestrenol in tablets. Rectilinear calibration graphs were obtained over the range of allyloestrenol concentration of 5 x The peak potential veys'szcs the mercury-pool electrode was found to be -1.47 V. Antioxidants such as tocopherol do not interfere in the assay. As the polarographic method is based on the same reaction scheme, the remarks made on the selectivity of the volumetric method apply also to this method.to 5 x 10-3 moll-l. Gas-chromatographic method As the molecule of allyloestrenol has few substituent and stable functional groups, the gas-chromatographic determination can be carried out without derivatisation and at not too high a temperature. Use of the internal standard technique affords a suitable method for the assay of single tablets containing allyloestrenol. A comparison of the results for the allyloestrenol content of one batch of tablets as deter- mined by the polarographic and gas-chromatographic methods is shown in Table I. TABLE I DETERMINATION OF ALLYLOESTRENOL IN TURINAL TABLETS Each tablet had a nominal allyloestrenol content of 5 mg.Allyloestrenol content per tabletlmg r I Polarographic Gas-chromatographic 4.79, 5.05, 4.86 5.15, 4.76, 4.96 method method 5.01, 5.13, 4.91 5.04, 5.01, 6.08 Mean .. .. . . .. . . 4.93 Standard deviation .. . . 0.153 Relative standard deviation . . 3.1% 5.03 0.076 1.6% Examination of the Stability of Allyloestrenol When allyloestrenol as such or its tablet dosage form is exposed to air, polar degradation products appear in its thin-layer chromatogram.lP2 We examined the three methods to see whether they could measure allyloestrenol selectively in the presence of the degradation products. The oxidative degradation product can easily be separated from a highly contaminated sample on the basis of its extremely low solubility in hexane. Analysis by thin-layer chromatography showed that it was a mixture of several components, presumably peroxides or even more highly oxidised derivatives.When tested by the titrimetric method an allyloestranol content of about 50% was found for this material, indicating that only one of the double bonds (presumably that in the A4 position) is attacked during the autoxidation. From these results we conclude that the excellent accuracy and precision of the titrimetric method make it suitable for the deter- mination of bulk allyloestrenol but it must be remembered that a completely decomposed material shows a virtual content of 50%. The polarographic assay based on the same reaction is suitable for the rapid testing of tablets but neither method is suitable for examining the stability of samples.The isolated degradation product does not give gas-chromatographic peaks under the recommended The gas-chromatographic procedure appears to be suitable for this purpose.200 GOROG, LAUKC) AND SZIKLAY conditions or at higher temperatures, indicating that its components are further decomposed or polymerised in the vaporiser zone. Gas chromatography is therefore suitable for deter- mining the active ingredient content of tablets that have been stored under various con- ditions and hence this method was used in testing the stability of Turinal tablets. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. References Fokkens, J., and Polderman, J., Pharm. Weekbl., 1961, 96, 657. Ganshirt, H. G., and Polderman, J., J . Chromat., 1954, 16, 510.Kato, K., Chem. Pharm. Bull., Tokyo, 1964, 12, 578 and 824. Cavina, G., Moretti, G., Mollica, A., and Antonini, R., J . Chromat., 1971, 60, 179. Brooks, C. J . W., and Middletich, B. S., Clinica Chim. Actu, 1971, 34, 145. Gorog, S., and SzAsz, Gy., “Analysis of Steroid Hormone Drugs,” Elsevier, Amsterdam, 1978, Gorog, S., Acta Chim. Hung., 1966, 47, 1. Gorog, S., Acta Chim. Hung., 1966, 48, 121. CsizCr, E., and Giirijg, S., J . Chromat., 1973, 76, 502. Gorog. S., Acta Chim. Hung., 1966, 47, 7. Gorog, S., and TomcsAnyi, L., Acta Chim. Hung., 1966, 47, 121. Gorog, S., and Foldes, V., Acta Chim. Hung., 1!366, 48, 249. Gorag, S., Acta Chim. Hung., 1967, 51, 221. CsizCr, E., Giirog, S., and Gyenes, I., Acta Chinz. Hung., 1972, 73, 175. CsizCr, E., Giirdg, S., and SzCn, T., Mikrochim. Acta, 1970, 966. Gorog, S., J. Pharm. Pharmac., 1969, 21, 46s. Martin, R. W., Analyt. Chem., 1949, 21, 921. Johnson, J. B., and Fletcher, J. P., Analyt. Chem., 1959, 31, 1563. Cheronis, N. D., and Ma, T. S., “Organic Fuiictional Group Analysis,” John Wiley, New York, Polgk, A., and Jungnickel, J. L., iiz Mitchell, J . , Kolthoff, I. M., Proskauer, E. S., and Weissberger, Gyenes, I., “Titrationen in nichtwassrigen Medien,” Enke Verlag, Stuttgart, 1970, pp. 592-594. Fleet, B., and Jee, R. D., Talanta, 1969, 16, 1561. Fleet, B., and Jee, R. D., J . Electroanalyt. Chern., 1970, 25, 397. pp. 261-262. 1964, pp. 374-375, 525-527. A., Editors, “Organic Analysis,” Volume 111, Interscience, New York, 1956, pp. 301-310. NOTE-References 7, 8, 9, 10, 11, 12, 13, 14, 15 and 16 are to Parts 1, IT‘, XXIII, 11, 111, v, VI, XIX, Received August 7th, 1978 Accepted September 27th, 1978 XV and XI1 of this series, respectively.

ISSN:0003-2654    

DOI:10.1039/AN9790400196

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Analyst March 1979 Vol. 104 pp. 201-207 201 Diffusion Assay by an Automated Procedure J. W. Lightbown R. A. Broadbridge and P. Isaacson" J. E. Sharpef-and A. Jones National Institute for Biological Standards and Control Holly Hill Havnpstead London N W 3 6RB Division of Engineering National Institute for Medical Research Mill Hill London NW7 1AA Research Division Beechana Pharmaceuticals Worthing West Sztssex BN14 8QH Equipment is described that allows diffusion assays to be performed auto-matically in Petri dishes using the punch-hole technique. With a block of six dishes limits of error of approximately &.2% can be obtained consistently. Various sources of systematic errors and their elimination are discussed. Keywords Automation ; mtibiotic assay ; diffusion assay ; systematic errors A systematic programme of automation of antibiotic diffusion assay was started some years ago by the Antibiotics Division National Institute for Biological Standards and Control (NIBSC) (then Division of Biological Standards National Institute for Medical Research).The assay consists basically of three stages pouring of plates; application of assay solutions; and measurement of zones of inhibition. It was considered that the third stage had the greatest content of subjective involvement of the operator and was the stage that could most easily allow the introduction of an operator bias in the measurement of potency. Most laboratories at that time measured the diameter of zones of inhibition in millimetres using some form of magnification either by projection of an enlarged image on to a screen or indirect magnification of the zone and scale by a lens system.The use of a television camera in conjunction with a graticule was developed by Tatum and Lightbownl and has been used successfully for more than 10 years to quantitate the areas of zones of inhibition on Petri dishes carrying six zones. The six areas were measured consecutively and recorded in arbitrary units. This method removed the need for the operator to make a subjective decision regarding the position of the edge of each zone and allowed zone area measurements to be made that were independent of the operator. It was found however that the performance of the image analyser could deteriorate under certain conditions in such a way that a bias was introduced by the machine which could, for example result in the first of the six zones measured on a dish being found to be too large or small.The fault was eliminated electronically but as it could apparently develop again progressively any possible effects of this bias were eliminated by reading the replicate dishes in groups of six as a Latin square so that any effects of the bias were systematically and evenly distributed to all responses in the assay. Early in 1969 the Beecham laboratory started measuring inhibition zones by means of a commercial image analyser with a graticule adapted for Petri dishes. The automation of the second stage of the assay i.e. the application of the solutions to the dishes was developed subsequently and jointly by the two laboratories and in the course of the work Quantimet Petriscoyes Models 60 and 720 were used for zone measurements (the Optomax recently available from Micromeasurements Ltd.has also been used). A decision to base the assay on the use of plastic Petri dishes rather than large square plates which were currently used by NIBSC had been taken during the development of the image analyser device for measuring the zones. Various advantages accrued from the use of such dishes. (i) They are disposable thus avoiding problems arising from contamina-tion of assay plates by antibiotics such as occurs with neomycin and glass surfaces. Experience has shown in one laboratory that sufficient neomycin can be adsorbed on to the glass surface of an assay plate so that after washing three times and dry heat sterilisation, zones of inhibition were still produced by the residual antibiotic on the glass.Steaming in * Present address National Biological Standards Laboratory Canberra Australia. Present address Division of Mechanical Engineering Queen Mary College London El 4NS 202 LIGHBOWN et a1 DIFFUSION Analyst Vol. 104 dilute acid was necessary to clean the plates. Penicillinase was also adsorbed on to glass plates and resisted washing and dry heat sterilisation thus interfering with subsequent assays of penicillin. (ii) The dishes are readily obtainable with a uniform plane surface. (iii) Bias from edge or corner effects is minimised as each zone is equidistant from the edge. (iv) Automation is more readily applied to a circular array than to a Latin square.(v) Automation of pouring is more easily arranged for Petri dishes. A number of possible procedures were tried viz. application of fish-spine beads and subsequent filling with solutions delivery of precise volumes of assay solutions to the surface of the agar (fixed volume and several concentrations or fixed concentration and several volumes) injection of assay solutions into the agar layer assay solutions solidified with agar in tubing then extruded and cut into plugs and punching holes in the agar layer with subsequent delivery of assay solutions into the holes. As a result of these preliminary experiments and in consideration of the wide variation in the nature of possible assay solutions particularly variations caused by differences in surface tension it was decided to concentrate on the development of a system based on holes punched in the agar.The size of the zone of inhibition that develops from the hole containing antibiotic solution is affected by a number of factors which may vary from hole to hole size of hole diameter and surface area of agar to solution interface; seal of agar to surface of dish; volume of assay solution added; and period of diffusion before the zone of inhibition is defined (critical time). Various procedures for punching six holes per dish were examined using a single punch and rotating the dish in six steps; punching six holes simultaneously; drilling the holes with a twist drill; and removing the agar plug whole by suction by blowing and by fragmentation and suction. Probably the most difficult problem was the removal of the plug without disturbing the seal between the agar and the dish.If this seal was disturbed the assay solution spread in the interface between agar and dish producing distorted or enlarged zones of inhibition. Various means of delivering the assay solutions into the holes were considered and examined using single or multiple devices; a minimum precision of &0.2% was considered desirable for repeated deliveries of the chosen volume. Delivering the six volumes simul-taneously had the advantage that variation in diffusion time between the assay solutions was eliminated and this design was therefore adopted. Using Petri dishes with a nominal diameter of 90 mm an apparatus was constructed that would cut six holes (6 mm in diameter) in the contained agar remove the plugs and intro-duce equal volumes of six assay solutions simultaneously into the six holes.Consideration was given to the application of the assay solution. Description and Use of Apparatus The apparatus2 shown in Fig. 1 consists of two units the punch head (right) and the dispensing head (left). Both units move vertically enabling the punches and pipette tips, both of which are plastic and disposable to perform the various operations. The assay dish together with the reservoirs for the test and standard solutions are transported horizontally underneath the heads on a carriage. A container for the spent agar plugs is placed under the cutting head. The apparatus normally performs automatically the sequence of operations but can be controlled manually if required.Operation is electro-pneumatic and vertical movements of the punch and dispenser heads are independently adjustable. The dispensed volume can readily be adjusted by means of a calibrated micrometer. Likewise the action of the punches is adjusted to suit a wide range of gel types and thicknesses. The sequence of operation begins with the operator placing a dish prepared for assay on to the right-hand position of the carriage. The carriage moves to the right until the dish is under the punch head and the reservoirs of test and standard are under the dispensing head. The heads move vertically down the dispenser pipette tips are immersed in the solutions in the reservoirs and the punches penetrate the agar to the required depth.The dispenser pistons then draw a pre-determined amount of solution into each of the pipette tips while the punching assembly rotates slightly to free the agar plugs. Both heads are retracted the punches carrying with them the plugs removed from the agar. The carriage returns to the left until the holes in the assay agar are under th Iiig. 1. Automatic bioassay machine March 1979 ASSAY BY AN AUTOMATED PROCEDURE 203 dispenser pipettes. The dispenser head descends the pipettes discharging the solution into the holes in the agar while the plugs remaining in the punches are ejected into the container below the punch head. The dispenser head retracts and the punched and filled assay dish is removed by the operator. The complete cycle takes approximately 15 s.Dispensed volumes of 50 and of 70 p1 are currently being used. The apparatus can be adjusted to deliver a maximum of approximately 100 pl. The dispensing unit is constructed in such a way that the tips are not completely emptied at each operation. Experience has shown that it may not be necessary to replace pipette tips or to sterilise the punches between assays of different antibiotics. In certain conditions however it has been found necessary to treat the punch tips with alcohol e g . when changing from Psezcdomonas aerzcginosa to BnciZZus szdh%s as assay organism. The plastic pipette and punch tips are readily replaceable, if this is necessary. Evaluation of the Apparatus Variations in volumes delivered into the six holes were measured initially by collecting repeated simultaneous deliveries of an aqueous solution of albumen labelled with iodine-131 and measuring the activity in each sample.With the prototype machine (three micro-meters) and using pipette tips of various sources of manufacture coefficients of variation for a single delivery point ranged from 1.0 to 4.0%. Measurement of the movement of the six pistons showed that these were all well within the 0.1-0.2~0 target. Measurement of the changes in pressure within a plastic tip during the cycle of filling and emptying showed that the changes were complex and that surface tension was a major factor affecting the volume delivered. Variations were noted in the size of orifice of the pipette tip and in the effective-ness of the seal at the cone joint of the tip.It was important to adjust the movement of the pipette head so that the tips at filling and delivery were as close as possible to the base of the dish. When the base of the dish was not plane and parallel to the six tips then variable interference with the flow of liquids occurred either at the filling or emptying stages. The adjustments had to ensure that the tips were immersed as soon as possible during delivery and in addition the operator had to be careful to observe that liquid remaining on the out-side of the pipette tips between emptying and re-filling did not vary greatly. By careful choice of tips control of cleanliness and care in fitting it was readily possible to obtain coefficients of variation of approximately 1% for the volumes delivered from the six tips, as is shown in Table I.This variation was greater than hoped but could not be improved and it was considered that by means of replication effects of the error could be reduced. Variation between the mean volumes delivered by the six pipettes was found to be less than 0.5% by mass. The prototype machine was designed with three micrometer adjust-ments in order to allow any measured bias affecting the six zones (in a circular pattern) arising from any source to be removed by the application of an opposite bias to the volumes delivered at the six points. Experience proved that this was not practicable because bias arising independently of the machine procedures was not necessarily constant from day to day. For this reason the final engineering design ensured a uniform displacement of the six pistons and a single micrometer head allowed all six to be varied equally and simultaneously.The results given in Table I were obtained with this design of equipment and it is seen that the variation between the mean volumes delivered by each pipette was approximately 0.2%. Variations in the size of punched holes were examined by using dishes containing nutrient TABLE I VARIATIONS I N DELIVERY OF PIPETTES USING WATER Positions 1 2 3 4 5 6 r A 1 Mean mass of 20 volumes delivered from pipette/mg . . . . . . 51.59 51.55 51.71 51.52 51.60 51.61 Coefficient of variation . . . . 1.26 0.029 1.32 1.02 1.37 1.46 Grand mean masslmg . . . . . . 51.60 Range of means 1-6 . . . . . . 99.90-100.2% of grand mean. Range of individual deliveries for all positions 50.1-53.2 mg.. 204 LIGHTBOWN et al. DIFFUSION Analyst Voi. 104 1 4901 1 - 10 950 - 10900 - 10 850 - 10 800 w-0 (0 f Punch position Fig. 2. 0 . . 0 Area of punch holes in arbitrary units; mean of 12 plates & 2 standard deviations of the mean for areas of punch holes. @-* Area of inhibition zones in arbitrary units; mean of 18 plates. Systematic bias originating from punches. agar flooding the plates with water and measuring the cross-sectional areas of the six holes by means of the plate reading device. The results shown in Fig. 2 were obtained using the prototype assay machine. On the same day rtnd with the same conditions 18 plates were prepared and filled with solutions for a tetracycline assay (3 + 3 design).The solutions were applied to three blocks each of six dishes in a Latin square within each block so that within a block each solution was delivered once into a hole produced by each cutter as shown in Table 11. TABLE I1 TEST FORMAT Identification of punch hole* f I h Petri dish 1 2 3 4 5 6 1 SH TM SL TH SM TL 2 TM SL TH SM TL SH 3 SL TH SM TL SH TM 4 TH SM TL SH TM SL 5 SM TL SH TM SL TH 6 TL SH TM SL TH SM * Solution identity SH = standard high ; SM = standard medium ; SL = standard low; TH = test high; TM = test medium; TL = test low. After normal incubation the zones were measured. The difference in response between columns represents the effects of differences in the six punch positions. Fig. 2 shows that the variation in size of the zones produced around holes from different punches follows closely the variation in the size of those holes.If holes from one punch are always used for the same assay solution then with the pattern of distribution of assay solutions used in the two laboratories (1 SH 2 TM 3 SL 4 TH 5 SM 6 TL) the two dose response lines for standard and test would be expected to be biased as shown in Fig. 3 with the possible introduction of non-parallelism and curvature. Assays carried out on a proposed standard preparation using the machine were each composed of 36 dishes. The assay design was 3 + 3 in six blocks of six dishes per assay. The dishes within a block had a fixed relationship between the six pipette and punch stations and the six assay solutions one block of six dishes for each of the six relationships shown in Table I1 for the six Petri dishes.Analysis of separate blocks of six dishes showed invalidities of curvature and non-parallelism within a number of blocks but when the 36 dishes were taken together as was intended the assays were usually statistically valid. Results from six such assays are reported in Table 111. Valid assays of a high degree of precision could be obtained with this assay system but the procedure was very cumbersome in performance and statistical analysis March 1979 ASSAY BY AN AUTOMATED PROCEDURE 205 % 4 a 0 t + Log dose Fig. 3. Effect of systematic bias on assay under conditions shown in Fig. 2. Arrows indicate direction of bias. One of the two laboratories developed the use of the automated apparatus in the form of two independent units one to punch holes and the second to dispense the assay solutions.This procedure had the advantage that it allowed dishes to be punched at a more rapid rate than could be accommodated by the dispensing unit. In this way one punch unit could supply dishes for several dispensing units. After punching the holes the dishes were mixed so that at subsequent dispensing a particular punch position was not tied to a particular assay solution. The randomisation thus introduced at this stage had the advantage of reducing or removing the bias which resulted from the punches. Assays performed in this way were usually free from invalidities of curvature or non-parallelism but there was an expected reduction in precision (see Table IV).The precision obtained with 10 dishes per assay was however adequate for the purpose required. A higher precision could be obtained by increasing the number of dishes but the use of the linked units together with the systematic rotation of the solutions is more economical if the higher precision is required. Later assays with linked units have used a total of six Petri dishes per assay instead of 36, with one dish for each of the six positions (punch hole relative to solutions). With this arrangement valid assays with confidence limits of &2 to 3% are regularly obtained (Table V) with a number of different antibiotics. In this way the precision obtained with a unit of six Petri dishes (36 zones) is better than was previously obtained using a large plate with a 6 x 6 or 8 x 8 Latin square arrangement and applying the solutions by means of fish-spine beads.The greater part of the effort in the assay goes into the preparation of the solutions and TABLE I11 -4SSAY OF CHLORTETRACYCLINE PROPOSED STANDARD USING PUNCH AND PIPETTE UNITS LINKED 3 + 3 design. Six blocks of six dishes per assay. Each solution was delivered in turn by each station to six dishes within one assay. Fiducial limits 1 1 1.024 1.012 to 1.035 2 1.023 1.010 to 1.036 Day Assay Potency ratio (P = 0.95) 2 3 1.023 1.006 to 1.041 4 1.022 1.005 to 1.040 3 5 1.039" 1.025 to 1.053 6 1.016* 1.000 to 1.033 * Significant non-parallelism 206 LIGHBOWN et all. DIFFUSION Analyst Vol. 104 TABLE IV AUTOMATED BIOASSAY UNIT IN USE FOR ROUTINE ASSAY USING PUNCH AND PIPETTE UNITS SEPARATE Ten dishes employed per assay.Departure Penicillin from Quadratic Difference of Limits content/ parallelism curvature curvature (9 = 0.95), Antibiotic Assay pgmg-l F F F % Ampicillin trihydrate . . A (1) 843 0.02 1.08 1.76 1 837 0.94 0.26 1.86 5 0.63 0.81 0.38 5-6 0.09 1.83 0.03 3-4 A (2) 838 B" (3) 858 Flucloxacillin-syrup preparations . . A 36.3 0.02 0.41 1.29 3-4 B 35.1 0.02 0.88 0.03 5 c 35.9 0.05 1.82 0.02 4-5 dishes and there is a temptation to use a large number of replicate dishes in the assay by means of automation in order to obtain an even higher degree of precision. This tempta-tion should be resisted; if greater precision is necessary it should be obtained by repetition of completely independent assays.Under our own conditions there is little to be gained by using more than six dishes in a single assay with the linked apparatus. Statistical analysis commonly used to determine confidence limits from the internal evidence of the assay is in any event of doubtful significance in such automated procedures, where the random error has been reduced to a very low level by a systematic distribution of known errors to all doses. The residual errolr can be reduced to an extent such that the usual analysis of variance may invite false conclusions. TABLE V ASSAYS WITH PUNCH AND PIPETTE UNITS LINKED Antibiotic Neoymcin Gentamicin Tetracycline Oxytetracycline Tobramycin Streptomycin Erythromycin Amikacin Rolitetracycline Spectinomycin Doxyc ycline Nystatin .. Potency/ U mg-l . . 95.8* . . 649 . . 970 . . 102.6* . . 104.9* 771 . . 933 . . 926 . . 780 . . 98.2* . . 874 . . 5075 Departure from linearity P ~ 0 . 0 5 0.05 >P >0.01 P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 0.05 >P >0.01 P >0.05 P ~ 0 . 0 5 0.05 >P >0.01 Departure from parallelism P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 P >0.05 0.05 >P >0.01 P >0.05 P >0.05 Fiducial limits (P = 0.95) 93.5 to 98.2 626 to 672 942 to 999 100.5 to 104.9 102.7 to 107.2 746 to 797 914 to 952 906 to 947 762 to 798 97.4 to 98.9 855 to 893 5045 to 5263 Limits, 2.5 3.6 2.9 2.2 2.2 3.3 2.1 2.3 2.3 0.7 2.2 2.1 % * Potency expressed as percentage of claim on label.Discussion The apparatus described has been in constant use for 6 years in the two laboratories one concerned mainly with assays where a high precision was desirable e.g. in calibration of new standards and the other which had the additional need for a high throughput of less precise assays. The performance of the machines has gradually improved over this period as general operating experience has been gained both with the prototype and commercial machines. They have been operating over the past 3 or 4 years constantly and routinely with no significant operational failures producing results in line with those described in the tables. This has involved the use of a wide range of different assay media and solvents as can be seen from Table V.Perhaps the most important lesson to be learnt was the practical impossibility of reducin March 1979 ASSAY BY AN AUTOMATED PROCEDURE 207 the systematic errors to an insignificant level. It is perhaps surprising that a systematic variation in the size of the punch holes within a single dish of only approximately 0.05 mm can produce a zone error of approximately 2%; when converted into potency error this would become approximately 3%. The systematic variation in size of holes shown in Fig. 2 is difficult to explain. It is of interest that a similar circular variation was found when the holes were cut automatically and successively with a single punch. It is possible that a rapid change in the gel properties commences following the first rupture of the agar surface and is progressive during the time period necessary to complete the punching of the six holes.Although the volumetric displacement of the six dispenser pistons was probably within 0.1% of target (represented by an error of the piston movement of 0.01 mm or an error in diameter of the pistons of 0.001 mm) the six volumes delivered varied by up to 4%. How-ever the dispensed volume from a single pipette was consistent to within approximately l yo. Hence the combined errors arising from the punching and the dispensing can be considered in two parts; systematic errors related to each station are 2 and lyo respectively and the random errors for the punches and pipettes approximately 1% in each instance.If no attempt is made to eliminate the deterministic errors then an inaccuracy of up to 4% may be introduced and not revealed by statistical analysis which may however demonstrate apparent invalidities arising from these deterministic errors. If the deterministic errors are systematically eliminated as described only the random errors remain and a precision of approximately 2% may be achieved. Alternatively when pre-punched plates positioned randomly under the dispenser are used, larger random errors are generated although the dispenser deterministic errors are also retained; in this instance traditional statistical analysis tends to produce valid estimates. There is a narrow dividing line between these two situations arising from the use of the machine in the two modes (punch and pipette linked or separate) and any particular assay may fall into either category.Measurement of true machine errors is extremely difficult and therefore seldom recognised, let alone quantitatively assessed. It is however likely that there are many similar situations in automated biological assays where repeated multiple volume measurements and repetitive quantitative physical observations are made. These effects can only be observed in experi-ments specifically designed to show them. Once the deterministic and random errors within a particular assay have been quantified it is possible by careful design of the assay systematically to remove the former. However in weighing solution preparation etc. this will not usually be possible.Little is gained by using more than the minimum number of replicates; for example in the assay described where there are six positions it is possible to remove all deterministic errors by allowing each punch position to be filled in turn from a different dispenser or with a different solution. Thus only six dishes are required to elimi-nate all deterministic machine errors and assays that previously required four large 12 x 12 in plates can be replaced with six Petri dishes. The pipette tips on the apparatus are those commonly used manually in clinical and analytical laboratories for repetitive delivery of small volumes of reagents. It is likely that the variability obtained with manual operation will be much greater than that shown in Table I. Thanks are due to the staff of the Worthing Research Division laboratories during the long development and proving period of the equipment. Research Engineers Ltd. Orsman Road London N1 5KD have been associated with the further development of the prototype machine to the commercial stage. References 1. 2. Tatum F. and Lightbown J. W. Demonstration Meeting of the Society for Applied Bacteriology, Imperial College London 1968. Sharpe J. E. Lightbown J. W. and Jones A. in “Second International Symposium on Rapid Methods and Automation in Microbiology Cambridge 1976,” Learned Information (Europe) Ltd., Oxford 1976 p. 63. Received August 16th 1978 Accepted October 25th. 197

ISSN:0003-2654    

DOI:10.1039/AN9790400201

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

208 Analyst March 1979 Vol. 104 $9. 208-223 Mechanism of Atom Excitation in Carbon Furnace Atomic-emission Spectrometry D. Littlejohn and J. M. Ottaway Department of Pure and Applied Chemistry Universit,y of Strathclyde Cathedral Street Glasgow G I 1XL By consideration of electronic and vibrational excitation temperatures and the ionisation temperature it is demonstrated that local thermal equilibrium (LTE) is established under the practical analytical conditions of interrupted gas flow in which commercial carbon furnace atomisers are used as emission sources. The electron concentration is shown to1 be derived from thermionic emission from the carbon tube and calculated values of 5.2 x 1O1O cm-3 at 2558 K and 1.3 x 101l cm-3 at 2766 K are reported. The processes that contribute to the establishment of LTE are considered in detail and it is suggested that molecular collisions make the major contribution to atomic excitation under all conditions but that radiation absorption may be significant when a monatomic gas is used as purge gas and when molecules are present as impurities at concentrations of only 0.01 7,.Keywords Atom emission ; carbon furnace atomisation ; excitation mechanism ; electifon concentration ; local thermal equilibrium The original introduction of electrothermal atomisers now used extensively in analytical atomic-absorption and -emission spectrometry can be traced back to the work of A. S. King in 1905 and 1908.132 He designed a resistively heated electrothermal atomiser in order to measure emission spectra of atoms arid molecules in a source which was free from the electrical action of the current carrying vapour of an arc or spark and where the complicated and often unknown chemical reactions of a combustion flame could be avoided.The measurement of atomic- and molecular-emission spectra generated by thermal energy alone in a source closely approximating to thermodynamic equilibrium rendered the King furnace of great value in the study of a number of fundamental spectroscopic proce~ses.l-~ The King furnace has been used more recently in a number of other fundamental studies. R. B. King and co-workers used furnace-emission intensity measurements at known temperatures to calculate relative gf values for a number of atomic species,6 and determined the distribution of CN molecules amongst the vibrational and rotational energy levels in a calculation of the relative vibration transition probabilities of the CN violet bands.' In studies of the vapour pressure and heat of sublimation of graphite Brewer and co-~orkers*9~ measured the emission spectrum of C in a King furnace and calculated the dissociation energy of gaseous C,.During their studies the vapour-phase temperature was measured using zirconium line-reversal at 473.95 nm. For a tube-wall temperature of 2973 K they obtained a vapour-phase temperature of 2963 5 20 K which was also in agreement with the C Swan band rotational temperat~re.~ The existence of thermal equilibrium within the electrically heated furnace was thus confirmed. Despite the information derived from such studies of the King furnace no applications of this type of atomiser in analytical emission spectrometry were reported at that time.The first analytical applications of a modified atomiser incorporating arc atomisation were reported only in 1959 and involved atomic-absorption measurements.10 Since then and notably in recent years the development of electrothermal atomisers for use in atomic-absorption analysis has advanced rapidly,ll but it is only since 1975 that the possibility of using a furnace as an analytical emission source has been advanced12913 and as yet only a limited number of analytical procedures have been described.14-lg Commercially available furnace atomisers are resistively heated like the King furnace, and it might therefore be expected that they would also act as thermal emission sources.However differences in the design and operation of modern atomisers compared with the King furnace may cause deviations from thermal equilibrium during the analysis time. There has to date been no evaluation for a furnace of the relative importance of the various excitation processes known to populate atomic energy states. The King furnace wa LITTLE JOHN AND OTTAWAY 209 generally enclosed and sealed and either operated at high pressures up to 16 atm,3 or at pressures below 1 atm.5-9 Samples introduced into the furnace were spread in bulk along the hot section of the tube often fusing to the carbon surface for the lifetime of the tube. The furnace was usually operated at the desired temperature for 10-15 min to allow equi-librium to be attained before emission measurements were taken over a further period of several minutes8 In modern analytical carbon furnace atomic-absorption and -emission spectrometry however transient signals are measured when both the tube temperature and the concentration of atoms are changing rapidly.Under these conditions there is a net transport of energy and mass through the system and an inhomogeneous temperature distribution exists. The chemical species in the furnace never reach over-all equilibrium as in the King furnace and atomisation is carried out under non-isothermal conditions. However if the rates of transport and temperature change are small compared with the rate at which energy is partitioned over the different degrees of freedom a state of equi-librium can be established for each volume element of the furnace in each small time interval.Each volume element at any instant can then be assigned a local temperature and local thermal equilibrium (LTE) can be achieved for each volume element in each small interval of time. A knowledge of the vapour temperature and how this relates to the change in wall temperature is of fundamental importance in emission spectrometry as the intensity of the analytical signal is related directly to the temperature which controls (under LTE) the distribution of the analyte species amongst the various energy levels. To understand the processes that control this distribution it is necessary to ascertain both the magnitude of the apparent source temperature and the effect of the temperature gradient on the measured emission intensity.In most instances the influence of the temperature gradient will be a function of the distribution of atoms along the tube at any instant of time. In contrast, in atomic-absorption spectrometry where resonance lines are employed for most analyses under normal experimental conditions the magnitude of the vapour temperature and the severity of the temperature gradient are less significant spectroscopically provided that the furnace temperature is high enough to maintain the atomic vapour. However the vapour temperature will affect the absorption (and emission) signals because of the tempera-ture dependence of physical parameters such as the degree of dissociation of molecular species and the residence time of atoms in the furnace.To investigate these processes and to assess whether LTE is established during the time required for the measurement of atomic-absorption signals a number of have compared vapour-phase temperatures with tube-wall temperatures measured at different times during atomisation. In most instances two-line atomic-absorption methods have been applied to estimate the vapour temperature from electronic excitation temperature calculations and these and other methods have been reviewed recently.20-22 The results of these investigations show a marked lack of agreement on the relationship between vapour and furnace-wall temperatures which is significant if the occurrence of LTE is to be con-firmed. Adams and Kirkbright23 reported that the excitation temperature of indium in a Perkin-Elmer HGA-2000 furnace increased with increase in the furnace-wall temperature, achieved a maximum similar to the temperature of the wall at the time of peak absorbance and finally decreased to a value that was up to 700 K lower than the final wall temperature.More recently Sturgeon and Chakrabarti20 found that excitation temperatures calculated from indium gallium iron and tin absorption signals in a Perkin-Elmer HGA-2100 furnace and a Varian Techtron CRA-63 atomiser rose to a maximum that occurred a t a different time to the absorption maximum The excitation temperature varied with the volatility of the thermometric species and was always lower than the tube-wall temperature by as much as 1300 K (indium). In contrast Matou~ek~~ reported vapour-phase temperatures in a Varian CRA-63 mini-furnace that were 300 K higher than the wall temperature over most of the absorption signal.He used a nickel two-line atomic-absorption procedure and this discrepancy was traced subsequently to the use of erroneous gf literature values.22 When corrected values were applied the average vapour temperature was found to lag behind the maximum wall temperature by only 50-100 K.22 Van den Broek et aZ.22 have suggested that the differences observed in the literature vapour temperature calculations for different elements can often be attributed to random and systematic errors in the methods used. From measurements of the vapour temperatures of both a Perkin-Elmer HGA-2100 furnac 210 LITTLEJOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION AnaZyst VoZ.104 and a Varian CRA-63 mini-furnace using the same nickel line-pair as Matousek,= they proposed22 a model for heat transfer that indicates that in the absence of convective flow through a furnace the gas temperature will follow the wall temperature of the heated furnace to within a few degrees. Electronic excitation temperatures based on emission measurements have also been reported and show closer agreement with the furnace-wall temperature than those based on atomic-absorption methods. Alder et aZ.26 calculated the electronic excitation temperature of iron by a two-line atomic-emission procedure using a Perkin-Elmer HGA-70 furnace. The temperature was calculated for the time of the peak emission signal and was found to be of the order of 2450 K when the temperature of the furnace wall was 2420 K.Hutton2s employed iron two-line atomic-emission and “slope techniques”21 as well as two-line atomic absorption procedures involving indium and gallium to measure the vapour temperature in argon nitrogen and helium in a Perkin-Elmer HGA-72 furnace at the time of maximum atomic-emission or -absorption signals. He concluded that the vapour temperature lagged behind the wall temperature by up to 250 K and that the difference was greatest in helium and least in argon. A subsequent investigation using iron electronic excitation temperature measurements indicatedz1 that the vapour temperatures in argon nitrogen and helium were only 80-150 K lower than the wall temperature at the tube centre throughout the duration of the emission signal.Temperatures measured using the same procedure with an HGA-2200 furnace operated with maximum power heating and temperature control have also indicated2’ close agreement between the vapour-phase temperature and the furnace-wall temperature at the centre of the tube. Although some of the above results are conflicting the balance of recent experimental data supports the view that LTE does exist during carbon furnace atomisation. The supporting evidence is however derived solely from studies of atomic species. In order to confirm the existence of LTE more conclusively we have extended the application of emission measurements to molecular and ionic species and present electronic and vibrational excitation temperatures and ionisation temperatures obtained from a detailed study using the Perkin-Elmer HGA-72 and -74 furnace atomisers.Our results support the view that a thermal process or processes seem adequate to explain analyte emission intensities. 21v25 The mechanism responsible for the establishment of a Boltzmann distribution of energy under different experimental conditions has also been examined. The relative importance of electron and molecular collisions and radiative processes are assessed in a kinetic study of sodium atom excitation and an explanation is offered for the observation21 that similar atomic-emission intensities are obtained in argon and nitrogen furnace gases. In addition the electron concentration during furnace atomisation has been calculated using Saha’s equation and the result shown to be consistent with the hypothesis that electrons are generated from thermionic emission from the graphite surface.Experimental Reagents Standard solutions were prepared from reagents of the highest purity available and distilled water was used at all times. Stock solutions of each element (1000 pg ml-1) were prepared by dissolving the appropriate amount of sulphate or nitrate salt in distilled water with the addition of sufficient nitric or sulphuric acid to give a final acid concentration of 10-2~. Working solutions of the required concentration were prepared from the stock standard solutions as required. Research-grade argon (99.996% purity) was used as the furnace purge gas. Apparatus Two instrumental systems were used for the measurements reported in this paper.A Perkin-Elmer HGA-72 carbon furnace atomiser was mounted in a Perkin-Elmer 306 atomic-absorption/emission spectrometer and coupled to a Servoscribe RE 541.20 strip-chart recorder. This was used for the measurement of (;) nickel atomic emission for calcula-tion of the electronic excitation temperature and (ii) barium calcium europium strontium and ytterbium atom and ion emission for calculation of the ionisation temperature and electron concentration. The operation of this system for measurement of atomic emissio March 1979 I N CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY 211 has been described previously.12.21 Standard HGA-72 graphite tubes were used. Solutions of the required concentration were transferred to the centre of the carbon tube using a 50-pl Oxford micropipette and were then dried for 40 s at 373 K and atomised for 10-13 s at maximum power which gave a final temperature of approximately 2700 K under argon gas-stop conditions.The PE 306 monochromator slit was set to give a band pass of 0.2 nm in the ultraviolet and 0.4 nm in the visible region and emission signals were recorded at a chart speed of 2 cm s-l. A Perkin-Elmer HGA-74 carbon furnace atomiser mounted in a Perkin-Elmer 360 atomic-absorption/emission spectrometer and coupled to the Servoscribe RE 541.20 strip-chart recorder was used to measure the Av = 0 series CN band emission of the B2C+ -+ X2C+ violet system for calculation of the vibrational excitation temperature. Standard HGA-74 graphite tubes were used. The operation of this system for measurement of furnace-emission signals is similar to that of the HGA-2200 which has been described previou~ly.~~ A small volume of nitrogen was introduced into the furnace with the argon purge gas to allow formation of CN molecules.This was achieved either by mixing 1-2% nitrogen with the argon supply or by operating the furnace with the end windows removed to allow ingress of atmospheric nitrogen. Both methods gave signals of sufficient magnitude and stability to permit the measurement of CN emission over the time required to scan the wavelength region of interest from 382 to 390nm. The HGA-74 was operated at maximum power under a continuous argon gas flow for 90 s. A maximum tube temprature of 2700 K was obtained within 8 s and remained constant for the duration of the measurement period.The CN band emission was scanned using the PE 360 wavelength drive at a rate of 5 nm min-l. The spectrometer slit width was set to give a spectral band pass of 0.2 nm with reduced slit height and the emission signals were recorded at a chart speed of 1 cm min-l. For measurements of atomic and ionic emission the spectrometers were adjusted to the required wavelength using the appropriate hollow-cathode lamp. In all instances suitable choice of slit height and aperture stops reduced the amount of tube-wall radiation directly entering the monochromator to a minimum. Residual background emission signals were recorded under the same conditions as the atomic- and ionic-emission measurements. The background emission at the time of the maximum of the combined analyte and background signal was then subtracted to give the net maximum atom- or ion-emission signal.The methods employed in the calculation of the spectroscopic temperatures presented in this paper are described in the Results and Discussion section. The values obtained are compared with the tube-wall temperatures measured with an Ircon series 1100 automatic optical pyrometer the design and application of which have been described elsewhere.21 Results and Discussion The processes reponsible for the excitation and de-excitation of atoms ions and molecules in arcs sparks flames and other emission sources have been well characterised by many workers (see for example references 28-34 and the literature cited therein). Excitation mechanisms have been classified28 into radiative processes collisions with electrons atoms and molecules and chemical reactions and in any specific source one or some combination of these processes may make a significant contribution.The relative importance of these processes in stimulating atomic emission from a carbon or tungsten tube atomiser has not been considered previously. If as appears likely from the earlier work cited above local thermal equilibrium is confirmed for electrothermal atomisation then one or more of these processes must be responsible for establishing LTE. It seems unlikely therefore that the chemical processes that often produce suprathennal atomic emission in combustion flames will be important. However photoreactions involving dissociation and excitation by continuum radiation from the walls and two-body exothermic exchange reactions of gaseous carbon with molecular oxides could conceivably take place under particular conditions.Similarly although the furnace system does not exhibit the electrical action of highly ionised vapours the concentration of electrons and their signifi-cance in the excitation of atoms has yet to be investigated. In order to eliminate these possible but unlikely excitation processes it is important to obtain conclusive evidence of the existence of local thermal equilibrium in the furnace during the initial few seconds of Atoms produced in a graphite furnace exist in a chemically inactive system 212 LITTLEJOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION Analyst VoZ. 104 atomisation and to calculate the concentration of electrons in the furnace atmosphere in the same period.Local Thermal Equilibrium For a gas to be in complete thermodynamic equilibrium it is required that the various gaseous components of the system be in equilibrium mutually and with respect to the surroundings. This will normally prevail only in an enclosure whose walls and interior have a uniform temperature with respect to radiation and internal and although a situation closely approximating to this can exist under certain conditions for the King furnace,7s9 modern commercial electrothermal atomisers do not fulfil this criterion. How-ever as previously mentioned local thermal equilibrium can be established within a furnace during the normal atomisation period.This will be characterised for each volume element of the source by the following c~nditions~ls~~ (a) the velocity distribution of all species in all energy levels satisfies Maxwell’s equation; (b) for each chemical species the relative population of energy levels conforms to Boltzmann’s distribution law; (c) the degree of ionisation of each species is described by Saha’s equation; and (d) the radiation density is consistent with Planck’s law. Local thermal equilibrium can be shown to exist if the same value of temperature deter-mines each of the above conditions at the same time in each volume element. In real sources with concentration and thermal gradients it is impossible to consider each volume element individually and a collective measurement is obtained.This usually involves the measure-ment of the combined radiation emission from each volume element and the value of the resulting apparent temperature depends on the distribution of atoms along the temperature gradient of the s o u r ~ e . ~ ~ ~ ~ ~ As most of the atoms and molecules in an electrothermal tube furnace have similar mass and atmospheric pressure is normally used translational energy will be partitioned by collisions almost instantaneously and it can therefore be assumed that the velocity distri-bution will be given by Maxwell’s formula. The experimental determination of the trans-lational (or kinetic) temperature is possible from Doppler half-widths but is known to be fairly inaccurate3’ and is not considered further here. The likelihood of LTE existing in the graphite tube furnace is enhanced by the presence of the tube-wall enclosure which is the source of all energy subsequently transferred to the vapour phase.In most emission sources condition (d) above concerning Planck’s radiation density is not fulfilled and deviations from LTE can exist because radiation emitted by vapour species is not compensated for by absorption of an equal amount of radiation from the surroundings. In many sources where the molecular concentration is high and collisional processes of excitation and de-excitation predominate the effect of this outward radiation loss is minimal. Departures from L‘TE through radiative dis-equilibrium have, however been observed for hydrogen - argon and for a low current d.c. arc operated in an atomic gas (see p.126 of reference 32). In a previously reported s t ~ d y 3 ~ we have shown that the spectral distribution of energy from a graphite tube atomiser closely fulfils the requirements of Planck’s radiation law. A temperature of 2553 K was calculated from the spectral distribution of the graphite tube continuum when the wall temperature as measured by the optical pyrometer was 2603 K at the tube centre and a temperature of 2534 K was obtained from the intensity of wall radiation scattered by the components of the vapour phase when the wall temperature was measured at 2573 K. In this investigation we considered LTE criteria (b) and (c) above by comparing electronic and vibrational excitation temperatures and the ionisation temperature with corresponding tube-wall temperatures in order to ascertain whether LTE holds during the initial few seconds of atomisation when both atomic-absorption and atomic-emission signals are recorded.Excitation ternperatu~es31~32s37~~~ The wavelength-integrated emission intensity measured by a spectrometer when an atom or molecule undergoes a radiative transition from an energy level Ex to a lower energy state Ey can be expressed by hc L I = KAz,Nx x - x - hzv 47 March 1979 IN CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY 213 where I is the intensity over the total line width K is a machine constant Ax is the Einstein transition probability Nx is the concentration of atoms or molecules in the upper energy level A,* is the wavelength of the transition L is the source length in the direction of viewing and h and c have their usual meanings and values.If the species in question are thermally distributed amongst the various electronic vibra-tional and rotational levels Nx can be replaced with where k is Boltzmann's constant g is the statistical weight of the upper energy level N is the total species concentration in all states and B(T) is the partition function or the state sum41 and is defined as where the subscripts denote the excited states and zero the ground state. Combining equations (1) and (2) gives which can be rearranged to In ("> = In (K'N,) - - E X kT * * gxA xv where K' covers all of the constants in equation (4) including the partition function B(T), which is approximately constant for the species of interest over the temperature range discussed here.42 The value of ln(K'Nt) will vary during the atomisation cycle of a furnace because N,, the total concentration of atoms in the tube varies but will be constant for all lines of an element at any specific instant.Hence by measuring the relative intensity of a series of spectral lines at different times during atomisation it is possible to calculate T the electronic excitation temperature at each moment from the slope of In (Lk,) - against E, if the energy levels are populated in accordance with the Boltzmann equation. The atomic-emission signals of the nickel lines listed in Table I were measured in duplicate with respect to time using the HGA-72/PE 306 system. A 50-pl aliquot of a 0.5 pg ml-l nickel solution was used in each instance.At this concentration self-absorption effects were observed to be negligible. Correction factors for the slight variation in the spectro-meter spectral response at different wavelengths were applied to the mean of the recorded intensities. The slope temperatures were then calculated as illustrated in Fig. l ( a ) at various times during atomisation from the least squares calculation of the slope of the graph TABLE 1 NICKEL LINES EMPLOYED FOR THE MEASUREMENT OF THE ELECTRONIC EXCITATION TEMPERATURE BY THE SLOPE METHOD43 Wavelength h/nm Energy levels/eV log, gA / A 305.08 4.088-0.025 5.35 341.48 3.655-0.025 5.24 342.37 3.832-0.2 12 4.72 343.36 3.635-0.025 4.72 344.63 3.705-0.109 4.98 349.30 3.657-0.109 5.0 214 LITTLE JOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION AnaZyst VoZ.104 3.0 3.4 3.8 4.2 €,,lev V + Es- E,/eV Variation of the intensity of carbon furnace emission signals with excitation and ionisation energy in a calculation of (a) nickel electronic excitation temperature (HGA-72) wall temperature a t the centre 2 670 K and slope temperature 2 568 f 143 K; (b) CN vibrational excitation temperature (HGA-74) wall temperature a t the centre 2 700 K and slope temperature 2 491 & 143 K ; and (G) ionisation temperature (HGA-72) wall temperature at the centre 2 700 K and slope temperature 2 603 jl 174 K. Measurements made in argon at maximum power (999 units). Electronic and ionisation temperature calculated from atomic and ionic emission intensities recorded with inter-rupted gas flow at 4 and 6 s respectively from the start of atomisation.Fig. 1. of In (”-> against E,. The gJ values employed were those described by C o r l i s ~ ~ ~ as the best recommended values for the nickel lines in question. g J m The calculated temperatures at various times up to -9 s during atomisation in Fig. 2 are slightly lower than the equivalent wall temperatures at the centre of the tube owing to the temperature gradient along the carbon surface and are in agreement with our previous studies using iron as the thermometric species in Perkin-Elmer HGA-7221 and HGA-220027 furnaces. The error bars in the observed temperatures were obtained by the method of least squares and represent one standard deviation of the points from the straight line, taking into account variations in sample introduction signal measurement and errors in the 6 + 8 9 ’ Tirnels Fig.2. Variation with time oi the wall temperature a t the centre of an HGA-72 graphite tube atomiser (-) with the electronic excitation temperature (@) calculated from the nickel atomic emission. Atomisation a t maximum power (999 units) in argon with interrupted gas flow March 1979 I N CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY 215 gxA values. Close agreement between tube-wall and electronic excitation temperature was also reported by Van den Broek et aZ.,22 who used a nickel two-line atomic-absorption procedure to calculate the vapour temperatures of a Varian CRA-63 mini-furnace and a Perkin-Elmer HGA-2100 atomiser. When molecules are used as the thermometric species the excitation temperature can be calculated from rotational and vibrational spectra.This requires the replacement of the transition probability Ax# in equation (4) by the expression 167~3 1 . # A = - x- x s 3 4 & gx where E, is the permittivity of a vacuum and S is the line strength which by definition is the square of the electric dipole transition moment.41 Rotational lines of certain molecules, such as OH and CN are nearly always observed as impurities in emission sources but high resolution is required to resolve the rotational s t r u c t ~ r e . ~ ~ ~ ~ * Vibrational bands of un-resolved rotational character can however be used easily to calculate the vibrational excitation temperature. In this instance the line strength S is replaced by the vibrational transition probability pvsvI calculation of which requires a knowledge of the dependence of the electronic transition moment on internuclear distance.46 Equation (4) can then be expressed in the form which on rearrangement (7) becomes where K" includes all of the constants in the previous equations and Ev2 is the energy of the upper vibrational level of the transition of interest.Emission spectra of a number of molecules formed during carbon furnace atomisation were described in detail by Hutton et aZ.46 Of those molecules which are easily observed CN seemed the most suitable and equation (8) was applied to the vibrational spectra of the Av = 0 sequence of the B2C+ -+ X2C+ band system given in Table 11. The CN emission was measured using an HGA-74 atomiser operated at maximum power with argon gas flow, using the wavelength drive of the spectrometer (PE 360) to scan the bands in the 385-390-nm region.As with the vibrational spectra of many diatomic molecules the bands overlap strongly and it was necessary to extrapolate the tail of each band to the maximum of the adjacent band to subtract the overlapping background intensity and obtain the net emission intensity of each band. With these reduced band-head intensities a graph of In ?$) zle~syszcs Eva was constructed as illustrated in Fig. l ( b ) and the vibrational excita-tion temperature calculated. A period of approximately 90s was necessary to scan the wavelength region required, TABLE I1 BANDS OF THE CN VIOLET B2 C+ -+ X 2 C+ SYSTEM EMPLOYED FOR THE MEASUREMENT O F THE VIBRATIONAL EXCITATION TEMPERATURE BY THE SLOPE METHOD4' Wavelength/ Energy levels/ Relative vibrational Band nm eV transition probabilities 010 388.34 3.198-0.0 1000 1,1 387.14 3.462-0.253 880 282 386.19 3.720-0.503 790 3,3 386.47 3.973-0.750 74 216 LITTLEJOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION Analyst VoE.104 and to maintain reproducible and measurable CN emission over this time nitrogen impurities were continuously introduced at a constant rate into the furnace with the argon gas flow by operating the atomiser without the end windows to allow ingress of air. As the furnace was also operated at constant temperature the CN bands were therefore measured under equilibrium conditions. The residence time of the nitrogen in the furnace is however, relatively short and the establishment of equilibrium conditions therefore needs to be rapid.Temperatures calculated by this procedure give little information about changes in the vapour temperature during the initial few seconds of atoniisation normally used for analysis, but do give an indication of the influence of tube-temperature gradient on the vapour-phase temperature. Because the measured emission intensity is a combination of the photons emitted from each small volume element of the furnace and the intensity from each section is dependent on its temperature and atom or molecule content the deviation of the apparent vapour temperature from that of the carbon wall at the tube centre will depend on the distribution of species along the tube-temperature gradient.Under the conditions normally used for analytical atomic-emission studies i.e. interrupted gas flow (or gas stop) the concentration of atoms etc. will always be greatest at the tube centre but for the procedure used for the measurement of CN an almost even distribution will exist along the tube. The vibrational excitation temperature of 2491 & 143 K calculated from the slope in Fig. l ( b ) is about 200 K lower than the wall temperature at the tube centre. Although this is a greater deviation than that observed for the nickel atorn measurements the difference is relatively small considering that the ends of the tube and the carbon cones will be of the order of lo00 K or lower when the centre of the tube is at 2700 K. This suggests that an apparent temperature gradient of only a few hundred degrees from the centre to a few millimetres from the ends of the carbon tube will be effective in determining the intensity of atomic or molecular emission from an electrothermal atomiser depending on the concentration gradient therein.This argument will be developed elsewhere47 in a consideration of tube design and temperature on the intensity of atomic-emission signals. As far as the present work is concerned measurements of the vibrational excitation temperature of CN indicate a thermal population of energy levels and support the contention that LTE is attained during carbon furnace atomisation. Ionisation tempe~atu~e~7~4~~@ is represented by Saha's equation for the ionisation equilibrium between the atoms and ions of one element where N, Ni and N are the concentrations of atoms ions and electrons respectively, B(T)u and B(T) are the partition functions of atom and ion V is the ionisation potential, m is the mass of the electron and the other terms have their usual meanings and values.The intensity of the ionic emission can be expressed in a similar manner to that for atomic emission and combination of equation (4) for each species with equation (9) yields the relationship where subscripts s and t refer to the upper and lower energy levels of the ionic transition. Equation (10) requires that the emission intensities of an atom and ion line be measured for a series of elements or line pairs. At any instant in time during atomisation the ternpera-ture and hence 1.5 1nT will be constant and a graph of the left-hand side of equation (10) veYsus V + Es - E for each element or line pair will be a straight line of slope +l/kT, from which the ionisation temperature can be calculated.Emission signals were measured with respect to time in duplicate for each of the atom and ion line pairs shown in Table 111, using the HGA-72/PE 306 system at maximum furnace power and with interrupted argo March 1979 Element Eu Yb Ca Crt Sr Ba IN CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY TABLE I11 ATOM AND ION LINE PAIRS EMPLOYED FOR THE CALCULATION OF THE IONISATION TEMPERATURE49 Ionisation Wavelength/ Energy levels/ potential/ Line nm eV eV gA x 10-8/s I 321.281 3.858-0.0 5.67 9.6 I1 420.505 2.949-0.0 3.2 I 377.010 5.433-2.144 6.20 8.6 I1 369.4 19 3.356-0.0 0.74 I 430.253 4.781-1.899 6.11 7.1 I1 393.367 3.152-0.0 0.91 I 445.673 4.683-1.899 6.11 7.5 I1 393.367 3.152-0.0 0.91 I 496.226 4.346-1.848 5.69 4.8 I1 407.771 3.04 1-0.0 0.66 I 611.076 3.219-1.190 5.21 5.2 I1 614.172 2.723-0.704 0.38 21’7 Concentration of solution*/ pg ml-I 5 50 40 40 10 20 * Injections of 50 p1.gas flow. Similar spectrometer conditions were applied for the measurement of the atom and ion signals of each pair. Wavelengths were chosen to give emission intensities of a similar order of magnitude for both species at concentrations giving negligible self-absorption. Atom and ion lines of similar wavelength were employed where possible to minimise errors in the correction for variations in spectrometer spectral response.The ionisation temperature was calculated at various times during atomisation as illustrated in Fig. l ( c ) and the collated results shown in Fig. 3 compared reasonably well with the tube-wall temperature at the corresponding time. The error bars signify the random errors for the derived temperatures as calculated with a programmable calculator by applying the method of least squares. The errors appear to be acceptable for the procedure con-sidering the number of measurements contributing to each calculation. 2 600 2 200 -1 ’ 6ooe$-- 3 4 5 6 j 8- 9 I 0 1’1 1; Time/s Variation with time of the wall temperature at the centre of an HGA-72 graphite tube atomiser (-) with the ionisation tempera-ture (*) calculated from the atomic and ionic emission of calcium, barium strontium europium and ytterbium.Atomisation a t maximum power (999 units) in argon with interrupted gas flow. Fig. 3. Electron concentration Once the ionisation temperature at any point during atomisation has been calculated, the abscissa of the above graph of equation (10) can be used to calculate the value of N, 218 LITTLE JOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION Analyst VoZ. 104 the natural level of electrons in the furnace at the same time. The electron concentration was found to range from 5.2 x 1O1O ~m-~ at 2558 K to 1.3 x loll ~ m - ~ at 2766 K. The errors in the electron concentrations at each point during atomisation were calculated from a knowledge of the error in the slope (and hence the abscissa) and the temperature.The errors were found to vary between hO.9 and k l . 9 orders of magnitude for all the points considered. This appears to be acceptable for the procedure used considering the degree of extrapolation required to obtain the abscissa and the dependence of the calculation of electron concentration on temperature (see p. 173 of reference 31). No comparative values for the electron concentration in electrothermal atomisers have been reported. It is unlikely that furnace electrons will be produced through the ionisation of vapour phase carbon species as the ionisation potentials of CN C C, C, etc. are of the order of 11 eV or higher.50 Graphite however has a comparatively low thermionic work function and an indication of the concentration of electrons produced by thermionic emission as the tube surface is heated is given by the expression51 2 (2nm,kT) 5 -w N = .. . . (11) where W is the work function for carbon 4.6 eV. At 2760 K the value of 2.7 x 101 cm-3 obtained from equation (11) compares reasonably well with the value obtained above from Saha's equation and is within experimental error. Although trace impurities of sodium, potassium and other easily ionised elements present in the furnace material or purge gas will add to the partial pressure of electrons the concentration will be much smaller (10s cm-,) than that produced by thermionic emission,2o which hence seems the only viable source of electrons. Conclusion From the preceding discussions it appears that conditions closely approximating LTE are achieved in a carbon furnace during atomisation excitation and ionisation of atomic and molecular species.The evidence presented above appears conclusive at least for the presently available commercial systems investigated. Although the measurement of thermal emission from a carbon furnace precludes excitation by suprathermal chemiluminescence mechanisms confirmation of a Boltzmann distribution of energy does not in itself identify the relative contributions of the remaining processes in establishing thermal equilibrium. To obtain information on the most likely mechanism or mechanisms it is necessary to consider the kinetics and practical rate of each process under normal atomisation conditions. Atom Excitation Mechanism The competing processes that are likely to excite metal atoms (or ions )electronically in a graphite tube atomiser are therefore (a) collisions with electrons involving transfer of the translational energy from electrons (b) collisions with molecules with transfer of the vibra-tional and rotational energy of molecules (c) discrete absorption of radiation from the tube-wall continuum.These processes can be expressed in the form of equations as follows: M+e- +M*+e- . . . . (124 M + X Y + M * + X Y . . (12b) M + h v +M* . . (124 where e represents an electron M and M* a metal atom in the ground and excited states, respectively XY a molecule where X may or may not be the same as Y and hv is a photon of discrete energy. The excitation rate equations for these processes can then be described respectively as Ratee = ke[e-][M] .. . . (13a March 1979 IN CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY 219 Rate, = kxY[XY][M] . . . . (13b) where k is the rate coefficient or constant for each process and k includes a term accounting for the radiation density. It is generally accepted that monatomic species show low efficiency in the electronic excitation of metal atoms,% which would be expected from classical mechanics and this process is not considered further in this discussion. However owing to their much smaller mass and larger mean velocity under thermal conditions electrons are expected to be considerably more efficient than inert gas atoms in the (de-)population of atomic electronic states that lie several electronvolts above the ground state% and so have been considered.Gilmore et aZ.,29 in a review of atomic and molecular excitation mechanisms tabulated rate coefficients for excitation de-excitation and excitation energy transfer for a number of reactions but in general investigations involving metal atom species have been limited to the alkali metals. For all the processes described above expressions were available that allowed calculation of the rate coefficients for the excitation of the sodium resonance line at 589.00 nm. These calculations are used in the following discussion as an illustrative example, using a typical furnace temperature of 2500 K. As the concentration of sodium is a constant whichever process is considered values of Rate/[Na] will be calculated in each instance and compared. Electron collisions T by averaging over a Maxwellian distribution of electron energy giving28 The excitation rate constant ke can be obtained under thermal conditions at temperature where E is the energy of the excited state in this instance 2.1 eV for sodium and uTHB is the effective cross-section at energies greater than the threshold required for excitation and is of the order of 2.3 x 10-15 crn2.5 Substitution for the other constants in equation (14) and T = 2500 K gives a value for he of 4.4 x cm3 s-l.An alternative expression for ke was derived by Gilmore et based on the integration of Zapes~chnyi’s~~ cross-sections over a Maxwell velocity distribution for several temperatures and fitting the data for sodium by an Arrhenius function of temperature The resulting equation was ke = 6 x 10-10To.6exp (-2y) ____ which was shown to fit well over a wide temperature range from 1500 to 20000 K.With this relationship a value of 3.6 x cm3 s-l is obtained for ke at 2500 K in reasonably close agreement to the alternative value above. Taking a value for ke of 4.4 x cm3 s-l and the value of the electron concentration of 3.1 x 1011 CM-~ calculated as produced by thermionic emission at 2500 K equation (13a) gives a value for Ratee/[Na] of 14 s-1. The alternative value of ke would result in a smaller value of Rate,/Na and the above value may be considered as a possible maximum. Molecular collisions The concentration of molecules such as N, O, CN C, etc. present in a graphite tube atomiser during the production and excitation of metal atoms depends greatly on the conditions of operation such as inert gas used and temperature and the design of the furnace.Two conditions will therefore be considered that can be taken as the extremes of the maximum and minimum molecular content normally encountered in practical analysis. When an atomiser is operated in a nitrogen atmosphere the concentration of molecules will be totally dominated by N at a level of about 3 x lo1* molecules ~ r n - ~ at 1 atm and 2500 K. Mental1 et aZ.53 have calculated the rate coefficient for excitation of sodium b 220 LITTLEJOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION APtaZyst Vol. 104 nitrogen to be approximately k, 10-lo cm3 s-l att temperatures between 2 100 and 2800 K. These two figures combined give a value of Ratel\~,(~~~.)[Na] from equation (13b) of 3 x lo8 s-l.When a monatomic gas such as argon is used as purge gas in an enclosed furnace system like the IL 555 atomiser the molecular concentration arises from nitrogen and oxygen present as impurities in the gas and from desorption from the furnace material at elevated temperatures. The argon normally used in this laboratory contains 0.004% impurity, which if it is assumed to be exclusively 0 and N, gives about 1014 molecules cm-3. In addition at room temperature a level of at least 1014 molecules cm- of nitrogen and oxygen will be adsorbed on the surface of the atomiser when opened to the air.54 On desorption, during atomisation this will add a further 5 x 1014 molecules ~ m - ~ to the furnace volume. At 2500 K the partial pressure of C species will be approximately 10-7 atm,55 giving 3 x loll C molecules ~ m - ~ .This is increased at 3000 K. to 3 x lo1* molecules ~ m - - ~ atm) and only at this temperature would the concentration approach that of the nitrogen and oxygen. As optimum atomic-emission signals are normally measured for sodium and many other elements at temperatures less than 3000 K,56 it is unlikely that excitation by C will be significant unless the rate constant for such excitation is very large. The fact that LTE exists at temperatures a t which the concentration of C species is very low suggests that this process cannot be the major excitation medhanism. A value for the rate constant for the excitation of sodium by oxygen was not available. The minimum concentration of molecules under these conditions is therefore expressed exclusively as N and is assigned a value of 5 x 1014 ~ m - ~ .The collision cross-section of 0 at 2000 K (given on p. 48 of reference 28) suggests that oxygen will be at least as effective as nitrogen as a molecular species capable of exciting metal atoms. The rate of excitation at this minimum concentration of molecules expressed as nitrogen and given by equation (13b) RateN2,,,,,,/[Na] will be 5 x lo4 s-l. Photon absorption Unlike other emission sources a graphite or tungsten tube atomiser will fulfil the con-ditions of "detailed balance" for radiative processes irrespective of the concentration of analyte present as loss of photons through atomic ionic or molecular emission will be balanced by absorption of the tube-wall black-body radiation at discrete energies.The rate constant for radiational excitation kRAD can be expressed as (reference 28, P. 64) where f is the oscillator strength for absorption and PA is the spectral volume density of the radiation field. For a graphite tube atorniser PA rn PA the black-body radiation density which is given by Wien's approximation of Planck's law39 as 87Thc PAb = - x e!xp A5 The oscillator strength f can be expressed in terms of the Einstein transition probability for emission A, by (reference 28 p. 17) Substitution of equations (17) and (18) for PA andf into equation expression for kR,D : . . (18) (16) produces a simplifie March 1979 IN CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY 221 The magnitude of (g2/gl)A2 is of the order of 108 s,49 giving for the sodium resonance line at 589.0 nm a value of RateaAD/[Na] approaching 5.8 x 103 s-1 at 2500 K.Using data available in the literature it has been possible to calculate the reaction rates for processes (12a) (12b) and (1%) for the sodium resonance transition at 589.00 nm. Similar data for all the processes do not seem to be available for any other atom or transition. The reaction rates for the sodium transition at 2500 K are compared in Table IV and additional values computed for 2 100 and 2 800 K are given. These temperatures represent the extremes of the temperature interval over which the nitrogen rate coefficient of 1O1O cm3 s-l can be applied. A number of conclusions can be derived from these figures regarding the mechanism of sodium excitation and atomic excitation generally in a carbon furnace atomiser.When nitrogen is used as the furnace purge gas excitation and de-excitation of all species will be dominated by N molecular collisions. Although the temperature dependence of electron concentration greatly increases the rate of electron excitation from 2100 to 2800 K it is unlikely that electron collisions will be a significant excitation process at the temperatures at which emission signals are normally measured (less than 3000 K) irrespective of the pressure or composition of the furnace gas. (C) When a monatomic gas such as argon is used as purge gas at temperatures approach-ing 3000 K in an enclosed furnace the contribution of photon absorption can become similar to that of nitrogen molecular impurities.In open furnaces the contribution of photon absorption is less significant than in a closed furnace owing to the increased ingress of atmospheric nitrogen. A lack of detailed information regarding the temperature variation of the N collision-rate constant leads to an unexpected trend of reduced rate with increased temperature (Table IV) for nitrogen collisions. This trend may in fact be erroneous but it will not affect the conclusions drawn above over the most useful temperature range of presently available carbon furnace atomisers. (A) (B) (D) TABLE IV RATES OF REACTIONS FOR PHYSICAL PROCESSES OF SODIUM EXCITATION AT DIFFERENT TEMPERATURES OF 2 100,2500 AND 2800 K Rate/ [Na]/s-l r n 7 Process 2100 K 2500 K 2800 K Electron collisions .. 3.08 x 1.39 x lo1 4.43 x 102 Nitrogen collisions maximum . . 3.5 x lo8 3.0 x los 2.6 x 108 Nitrogen collisions minimum . . -5 x lo4 -5 x 104 -5 x 104 Photon absorption . . 9.0 x 102 5.8 x 103 1.7 x 104 Although the excitation of atomic and molecular species in nitrogen can be up to lo4 times faster than in argon (owing to the greater concentration of N,) similar emission intensities are measured in both gases at the same furnace temperature.21 This is explained by the principle of “detailed balance” that applies to all processes in conditions of local thermal equilibrium. “Detailed balance” states that the total number of atoms or other species leaving an excited state per second by any process (collision radiation etc.) just equals the number arriving in that state per second by the exact reverse process.34 Hence, in nitrogen the de-excitation rate will be also about lo4 times faster than in argon.In carbon furnace atomic-emission spectrometry reduction of the molecular concentration from 100% nitrogen to approximately O.Olyo does not appear to disturb conditions of local thermal equilibrium as indicated previously by a comparison of atomic-emission intensities in argon and nitrogen gases. In contrast departures from LTE arise in a d.c. arc operated in an inert gas with less than 1% molecular impurity and infrathermal emission is en-countered in combustion flames of low molecular concentration (see pp. 126131 of reference 32). This can be understood from a comparison of the number of molecular collisions that a metal atom suffers during the average residence time in the observation zone of each source.At atmospheric pressure the number of collisions between sodium and nitroge 222 LITTLEJOHN AND OTTAWAY MECHANISM OF ATOM EXCITATION Analyst VoZ. 104 is about 109 s-1 at 2500 KS (assuming 100% nitrogen). The residence time of a metal atom in an air - acetylene flame or d.c. arc is of the order of millisecond^,^^ and therefore each sodium atom will take part in 105-106 collisions in this time. At 2000 K at least one out of every 105 colliding molecules has an energy of about 3 eV,34 and the collision frequency therefore seems sufficient to establish an equilibrium population of the excited state. At lower nitrogen concentrations this is less likely to be true. However the residence time of metal atoms in a HGA-72 furnace operated with interrupted gas flow is about 1 s,57 and therefore even at nitrogen concentrations as low as O .O l ~ o sodium and other atoms will still make 105-106 collisions with nitrogen molecules while present in the furnace atmosphere. The existence of LTE under such conditions is therefore likely even without the contribution of tube-wall radiation. It is also interesting to note that when the molecular concentration, in a d.c. arc or combustion flame falls below the level at which molecular collisions dominate all other processes departures from thermal populations arise through radiative disequi-librium as photons emitted by atoms are not balanced by the absorption of photons at an equal rate from the radiation field.In a furnace atomiser the semi-enclosure of the tube wall which radiates as a black body ensures that radiative disequilibrium does not occur and LTE is maintained. Similar conclusions to the above will apply to most furnace tube atomisers operated under interrupted gas flow conditions including those made from other materials such as tungsten. The thermionic work function for tungsten 4.52 eV,50 is similar to that of carbon and will give a similar electron concentration and Kirchoff’s law2* ensures that an approximate black body radiation density will exist in the tungstein-tube atomiser. The conclusions reached above may not necessarily apply under conditions of convective gas flow where vapour-phase temperatures may differ substantially from tube-wall temperatures.22 The work described in this paper indicates that the graphite furnace is unique amongst emission sources in that local thermal equilibrium exists under normal working conditions.In achieving this state molecular collisions appear to make the major contribution but radiation absorption may be of minor importance when argon is used as purge gas. Although some of the fundamental properties of the King furnace have been known for many years, the possibility of applying the technique of furnace emission more widely in analytical chemistry is only now being investigated in detail. While the carbon furnace remains a relatively cool source compared with other currently available emission sources the signal to background ratios signal stabilities and atom residence times are such that very low detection limits have already been achieved for a wide range of elements,27@ and spectral interferences are much reduced in complexity.The fundamental origin of the observed emission signals established in this paper will contribute to the development of the technique, and to considerations of the most suitable design of atomisation source. The authors thank the Salters’ Company for the award of a Scholarship to D.L. and The Royal Society for the award of a research grant to J.M.O. for the purchase of the HGA-72 atomiser. The gift of the HGA-74/PE 360 system by the British Steel Corporation, Ravenscraig Works and helpful discussions with many colleagues particularly C. Th. J. Alkemade are also gratefully acknowledged.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References King A. S. Astrophys. J. 1905 21 236. King A. S. Astrophys. J. 1908 27 353. King A. S. Astrophys. J. 1912 35 183. King A. S. Astrophys. J. 1913 37 239. King A. S. Astrophys. J. 1922 56 318. King R. B. Parnes B. R. Davis M. H. and Olslen K. H. J. Opt. Soc. Am. 1955 45 350. Floyd A. L. and King R. B. J. Opt. SOC. Am. 1955 45 249. Brewer L. Gilles P. W. and Jenkins F. A. J . Chem. Phys. 1948 16 797. Brewer L. Hicks W. T. and Krikorian 0. H. J. Chem. Phys. 1962 36 182. L’vov B. V. J. Eng. Phys. (USSR) 1959 2 44. Fuller C. W. “Electrothermal Atomization for Atomic Absorption Spectrometry,” AnabyticaE Ottaway J . M. and Shaw F. Analyst 1975 100 438. Molnar C . J. Chuang F.S. and Winefordner J. D. Spectrochim. Acta 1975 30B 183. Shaw F. and Ottaway J. M. Analyt. Lett. 1975 8 911. Sciences Monographs No. 4 Chemical Society L,ondon 1977 March 1979 IN CARBON FURNACE ATOMIC-EMISSION SPECTROMETRY 223 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31, 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. Epstein M. S. Rains T. C. and O’Haver T. C. Appl. Spectrosc. 1976 30 324. Littlejohn D. and Ottaway J. M. Analyst 1977 102 393. Hutton R. C. Ottaway J. M. Rains T. C. and Epstein M. S. Analyst 1966 102 429. Ebdon L. Hutton R. C. and Ottaway J. M. Analytica Chim. Acta 1978 96 63. Epstein M.S. Rains T. C. Brady T. J. Moody J. R. and Barnes I. L. Analyt. Chem. 1978, Sturgeon R. E. and Chakrabarti C. L. Spectrochim. Acta 1977 32B 231. Littlejohn D. and Ottaway J. M. Analyst 1978 103 595. Van den Broek W. M. G. T. de Galan L. Matousek J. P. and Czobik E. J. Analytica Chim. Adams M. J. and Kirkbright G. F. Analytica Chim. Ada 1976 84 79. Matousek J . P. “Proceedings of the 17th Colloquium Spectroscopicum Internationale Florence, Alder J. F. Samuel A. J. and Snook R. D. Spectvochim. Acta 1976 31B 509. Hutton R. C. PhD Thesis University of Strathclyde 1977. Littlejohn D. and Ottaway J. M. Analytica Chim. Acta 1978 98 279. Alkemade C . Th. J . and Zeegers P. J. Th. in Winefordner J. D. Editor “Spectrochemical Methods Quantitative Analysis of Atoms and Molecules,” John Wiley New York 1971, Gilmore F.R. Bauer E. and McGowan J. W. J . Quant. Spectrosc. Radiat. Transfer 1969 9 157. Caller A. B. Appl. Opt. Suppl. Chem. Lasers 1965 145. Boumans P. W. J. M. “Theory of Spectrochemical Excitation,” Hilger and Watts London 1966. Boumans P. W. J. M. in Groves E. L. Editor “Analytical Emission Spectroscopy Part 11,” Boumans P. W. J. M. and De Boer F. J. Spectrochim. Acta B in the press. Alkemade C. Th. J . “Proceedings of the 10th Colloquium Spectroscopicum Internationale Mary-Reif I. Fassel V. A. and Kniseley R. N. Spectrochim. Acta 1973 28B 105. Reif I. Fassel V. A. and Kniseley R. N. Sfiectrochim. Acta 1974 29B 79. D e Galan L. and Winefordner J. D. J . Quant. Spectrosc. Radiat. Transfer 1967 7 703. Hooymayers H. P.and Alkemade C. Th. J. J . Quant. Spectrosc. Radiat. Transfer 1966 6 501, Littlejohn D. and Ottaway J . M. Analyst 1977 102 553. Winefordner J. D. McGee W. W. Mansfield J. M. Parsons M. L. and Zacha K. E. Analytic@ Thorne A. P. “Spectrophysics,” Chapman and Hall London 1974 Ch. 9 and 11. De Galan L. Smith R. and Winefordner J. D. Spectrochim. Acta 1968 23B 521. Corliss C. H. J . Res. Natn. Bur. Stand. 1965 69A 87. Smit J. A. Physica 1946 12 683. , Nicholls R. W. Proc. Phys. Soc. 1956 69 51. Hutton R. C. Ottaway J. M. Epstein M. S. and Rains T. C. Analyst 1977 102 658. Littlejohn D. and Ottaway J. M. Analyst submitted for publication. Mavrodineanu R. and Boiteux H. “Flame Spectroscopy,” John Wiley New York 1965. Corliss C. H. and B:zmann W. R. “Experimental Transition Probabilities for Spectral Lines of Weast R. C. Editor “Handbook of Chemistry and Physics,” 56th Edition Chemical Rubber Page F. M. and Woolley P. E. Combust. Flame 1974 23 121. Zapesochnyi I. P. High Temp. 1967 5 6. Mentall J. E. Krause H. F. and Fite W. L. Discwss. Faraday SOL 1967 44 157. Ehrlich G. J . Phys. Chem. 1956 60 1388. Palmer H. B. and Shelef M. in Walker P. L. Jr. Editor “Chemistry and Physics of Carbon, Littlejohn D. and Ottaway J. M. Analytica Chim. Acta in the press. Sturgeon R. E. Chakrabarti C. L. and Bertels P. C. Analyt. Chem. 1975 47 1250. Littlejohn D. and Ottaway J. M. Analyst 1978 103 662. 50 874. Acta 1978 100 121. 1973,” Volume I Adam Hilger Bristol 1975 p. 57. of Analysis. Ch. 1. Marcel Dekker New York 1972 Ch. 6. land 1962,” Spartan Washington D.C. 1963 p. 143. 847 and 912. Chint. Acta 1966 36 25. Seventy Elements Nut. Bur. Stand. Monogr. No. 53 1962. Company Cleveland Ohio 1976. Volume 4,” Edward Arnold London 1968 p. 85. Received June 12th 1978 Accepted October llth 197

ISSN:0003-2654    

DOI:10.1039/AN9790400208

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

224 Analyst, March, 1979, Vol. 104, pp. 224-231 Determination of Chromium in Natural Waters and Sewage Effluents by Atomic-absorption Spectrophotometry Using an Air = Acetylene Flame K. C. Thompson and K. Wagstaff Severn-Trent Water Authority, Malvern Regional LaboGvatory, 141 Church Street, Malvern Worcestershire WR14 2AN A simple method for the determination of chromium in natural waters and sewage final effluents by atomic-absorption spectrophotometry using an air - acetylene flame is described. The sample is concentrated by evapora- tion by a factor of five. Interference effects were minimised by the addition of ammonium perchlorate and were further reduced by working with a flame on the verge of luminosity rather than a distinctly luminous flame. Keywovds : Chromium determination ; atomic-abs orption spectrophotometry ; aiv - acetylene pame ; .Pzat.ural waters and sewage efluents The World Health Organization European Standard1 quotes a limit for chromium(V1) in potable waters of 0.05 pg ml-1, and a European Economic Community Directive,2 con- cerning the quality required of surface waters intended for abstraction of drinking water, quotes a total chromium limit of 0.05 pg ml-l.There is a requirement for a routine atomic- absorption spectrophotometric method suitable for the determination of chromium at these levels in a wide range of natural waters and sewage effluents. Ideally, the method should have a sample preparation stage that will allow subsequent analysis for other toxic metals of interest. The proposed method should have a detection limit of about 0.005 pg ml-l of chromium.Most manufacturers of atomic-absorption spectrophotometers quote detection limits for chromium in pure solution, when calculated as 4.65 times the within-batch standard deviation of the blank,3 of between 0.007 and 0.01 pg ml-l in the luminous air - acetylene flame. Under these flame conditions, inter-element effects for chromium are severe:s5 and it appears that some form of pre-concentration and also some method of minimising potential inter-element effects are necessary. The commonly used solvent-extraction technique using ammonium t et rame t hylenedit hiocarbam a te - 4-met hylpent an-2-one is very sensitive but suffers from several disadvantage^.^,^ A pre-treatment step utilising acid digestion is required to break down any insoluble or or,ganically bound chromium.A potassium permanganate oxidation step is necessary to convert chromium( 111) into chromium(V1). Reduction of the excess of permanganate and careful pH adjustment must be made before the extraction is performed. The pH adjustment step is critical if co-extraction of manganese with subsequent emulsion formation is to be avoided. Overnight standing of the extract prior to nebulisation is recommended.s This paper describes a simple atomic-absorption spectrophotometric method that utilises a concentration by evaporation technique, in which ammonium perchlorate is incorporated into the sample solution in order to minimise inter-element effects from the sample matrix. The determination is carried out in an air - acetylene flame.Experimental Apparatus Atomic-absorptiort spectrophotometer. A Varian Techtron 1200 fitted with a standard high-solids air - acetylene burner (titanium burner grid) and a corrosion-resistant nebuliser was used. Borosilicate glass beakem and test-tGbes. Tall-form 100-ml beakers with a spout, and engraved at the 5 ml level. These beakers were initially boiled in 50% V/V hydrochloric acid (36% m/m) and were reserved for this work. Calibrated (0.1 m1) 10-ml tubes with ground-glass stoppers (Exelo Ltd.), which were initially soaked in 50% V/V hydrochloric acid and regularly cleaned using laboratory detergent, were used.THOMPSON AND WAGSTAFF 225 Reagents Hydrochloric acid, 25% V/V. Dilute 250 ml (&2 ml) of hydrochloric acid (36% m/m) (analytical-reagent grade) to 1 1 ( 5 2 ml) with de-ionised water.Ammonium perchlorate solution, 10% m/V. Dissolve 50 g (&O.l g) of ammonium perchlorate (Fisons Ltd.) in about 450 ml of de-ionised water and dilute to 500 ml (&l ml) with de-ionised water. Caution-Ammonium perchlorate is a potentially hazardous chemical and any solution spillage should be dealt with immediately in order to avoid any subsequent fire risk. Hydrogen peroxide, 6% mfm. Aluminium oxide anti-bumping granules. BDH Chemicals. These granules were boiled with nitric acid, washed with de-ionised water and dried prior to use. Standard chromizm(lI1) chloride solution (1 000 pg ml-l of chromium). BDH Chemicals. Standard potassium dichromate solution (1 000 pg ml-l of chromium).Hopkin and Williams. BDH Chemicals, analytical-reagent grade (20 volume). Optimisation of the Method Choice of flame and flame conditions It is well known that chromium determinations in the air - acetylene flame are prone to inter-element effect^^^^^*-^^ and that the chromium sensitivity is dependent on the oxidation state of the c h r ~ m i u m . ~ J ~ J ~ These problems can be avoided by using the dinitrogen oxide - acetylene flame but the use of this flame results in a decrease in the detection limit of approximately 4-&fold compared with the luminous air - acetylene flame. Many labora- tories prefer to avoid the routine use of the dinitrogen oxide-acetylene flame for their standard toxic metal analyses and for these reasons a method utilising the air - acetylene flame was developed.Inter-element effects in the air - acetylene flame can be minimised by setting the acetylene flow so that a non-luminous flame is 0btained.~,~J0 However, this results in a significantly decreased detection limit compared with the luminous-flame conditions that are normally used for this determination. Previous work14 has shown that under luminous-flame con- ditions chromium calibration graphs over the range 0-20 pg ml-l exhibited inflexions and well defined maxima. The effect was especially pronounced for chromium(II1) solutions; it appeared to depend upon the age of the solutions and was also observed in solutions con- taining 10% V/V nitric acid (70% m/m). For this study, the acetylene flow was set so that a flame on the verge of luminosity was obtained.The chromium characteristic concentration under these conditions (0.10 pg ml-l) was just over twice that observed in the luminous flame (0.043 pg ml-l). Choice of interference suppressor Various reagents have been recommended for minimising inter-element effects in the atomic-absorption spectrophotometric determination of chromium in the air - acetylene flame. These include lanthanum chloride,15 sodium sulphate,s ammonium ~ h l o r i d e , ~ ~ ~ ammonium bifluoride4~lo and quinolin-8-01.~~~~ The incorporation of 5 000 pg ml-l of lanthanum (as the chloride) in all the solutions was found not to overcome many inter- element effects and resulted in a significant background-absorption signal. The addition of 5000 pg ml-l of sodium sulphate actually enhanced the suppression caused by 1000 pg ml-l of calcium and magnesium on a 10 pg ml-l chromium solution.* Recently, the addition of ammonium perchlorate has been recommended for minimising inter-element effects of elements other than chr~mium.l~-~~ It was decided, therefore, to compare the addition of ammonium chloride with the addition of ammonium perchlorate to various synthetic solutions so that the final solutions contained 2% m/V of the added salt, as recommended by BarnesQ for ammonium chloride.Ammonium salts, as expected, were found not to increase significantly the background-absorption signal. Table I shows that an interference suppressant is essential and that ammonium perchlorate is better than ammonium chloride. All the results in Table I were obtained using a flame on the verge of luminosity.If the acetylene flow was increased so as to obtain the maximum chromium response (a distinctly luminous flame), inter-element effects were significantly increased in all instances. Ammonium perchlorate, at a concentration of 2% m/V in the final nebulised solution, was used in all further work. Increasing the ammonium perchlorate concentration226 THOMPSON AND WAGSTAFF: DETERM~NATION OF CHROMIUM IN Analyst, VoZ. 104 COMPARISON OF AMMONIUM CHLORIDE AND AMMONIUM PERCHLORATE TABLE 1 AS INTERFERENCE SUPPRESSORS OF INTER-ELEMENT EFFECTS All solutions contained 2 pg ml-l of chromium(1 [I) and 10% V / B hydrochloric acid (36 yo nz/m). Sequential background correction was applied. Relative signal with 2% m/ V of added suppressant r \ A Interfering substance Concentration* / Signal with no Ammonium Ammonium added p g ml-l suppressant added chloride perchlorate None .... .. - 100 100 100 :e} 74 92.5 95.3 Ca (as C1) .. SO, (as H,SO,)' ' .. Na (as Cl) .. .. 480 79.5 85.9 90.1 4 000 1000) ca (as Cl) . . .. Mg (as Cl) . . .. 2 000 . . . . 200} 85 Ca (asCl) .. .. Fe (as Cl) . Mg (as Cl) . . . . 200 89.9 98.4 M g (as Cl) . . .. 1000 66.2 92.7 100.8 should be divided by 6 when related to the analyte solution. * The concentrations shown represent the final concentrations in the nebulised solution and to 3% m/V did not appear to offer any further significant reduction in the inter-element effects. Instrzcmental operating conditions Table I1 gives the optimised instrumental operating conditions.Automatic background correction at the 357.9-nm chromium line is not to be recommended, as balancing the hydrogen and chromium lamp intensities at this wavelength is not easy and a severely degraded chromium detection limit is normally observed. Sequential background correction using the lead 357.3-nm non-resonance line wits used and found to be satisfactory. Table I11 gives the typical background-absorption signals (expressed as a chromium concentration) from the main matrix elements and from some typical samples. It can be seen that the presence of sulphate significantly enhances thie background absorption from calcium and magnesium. However, this table shows that the background-absorption signals for most natural-water and sewage-effluent samples are irelatively small.TABL:E I1 OPTIMISED INSTRUMENTAL OPERATING CONDITIONS Wavelength . . . . . . . . . . . . 357.9 nm Background correction wavelength . . . . 357.3 nm (Pb) Slit width . . .. . . . . .. . . 0.5nm Airflow .. .. .. . . . . . . As recommended in handbook Acetylene flow . . . . .. .. . . Flame on verge of luminosity (no yellow luminosity visible) Distance from top of burner grid to pos:ition where the grid just intercepted the light beam (0.01 absorbance) . . .. .. . . . . 3.5mm Integration period . . . . . . . . . . 3 s Wash solution . . . . . . . . . . 3% V / V hydrochloric acid (36% 44March, 1979 WATERS AND EFFLUENTS BY AAS USING AN AIR - ACETYLENE FLAME TABLE I11 TYPICAL BACKGROUND-ABSORPTION SIGNALS USING THE 357.3-nm LEAD NON-RESONANCE LINE All solutions contained 2% m/V ammonium perchlorate and 10% V / V hydrochloric acid (36% m/m).227 Substance added or Concentration/ sample p g ml-l Ca (as C1) . . .. 10000 .. 10 000 9 600) Ca (as C1) SO, (as H2So4)' * .. Na (as C1) . . .. 10000 SO, (as 'H2S04) .. Mg (as C1) .. .. 10000 Mg (as Cl) . . . . 10 000 .. 10000 9 GOO} Na (as C1) . . SO, (as H,S04) .. 9 600) SO, (as H2S04) .. 19 200 Tap water 1* . . River water 2* River water 3* .. Sewage effluent 6* . . ' '} See Table V Background-absorption signal/pg ml-1 of chromium 0.09 0.22 0.03 0.035 0.17 0.26 <0.01 <0.002 0.007 0 0.009 0 0.006 0 * The background-absorption signals were measured after the samples were concentrated by evaporating to one fifth of volume. The observed signals were then divided by 5.Comerratvation techniqzle A five-fold concentration step of 50ml to 10 ml was found to result in an acceptable detection limit (less than 0.005 pg rnl-l), tolerable inter-element effects and very low back- ground absorption. The time for the evaporation step was approximately 1.5 h. Hydrochloric acid was used rather than nitric acid, as previous work had shown that the background-absorption signal at 357.3 nm from 10000 pg ml-l of calcium (as the chloride) was increased approximately four-fold in the presence of 10% V/V nitric acid (70% m/m), but was unaffected by the presence of 10% V/V hydrochloric acid (36% m/m). Also, a small, but significant, negative background-absorption signal had been observed with strong nitric acid solutions at 357.9 nm. However, nitric acid is the preferred acid for other types of samples (e.g., sewage sludge) .20 Method NOTE- This method should not be used with trade wastes or any unknown samples for safety reasons.alternative method is given below (see Alternative Method). An Volumes (50 & 0.5ml) of the samples, standards and blanks were placed in 100-ml borosilicate glass beakers and 4 ml (&O.l ml) of 25% V/V hydrochloric acid, 2 ml (50.05 ml) of 10% m/V ammonium perchlorate solution and some aluminium oxide anti-bumping granules were added. The beakers were then placed on a hot-plate with the temperature set such that gentle simmering occurred. The evaporation step was carried out in a fume cupboard and all normal safety precautions were observed. When the volume of the solution had decreased to 20 ml (-&5 ml), 0.5 ml (&0.05 ml) of hydrogen peroxide (6% m/m) was then added. This ensured that any chromium(V1) would be converted into chromium(III).21 The evaporation was then continued until the final volume was about 5 ml (rf 1 ml) and this took approximately 1.5 h.The solutions were allowed to cool and the contents transferred into the 10-ml calibrated borosilicate glass tubes. The beakers were228 THOMPSON AND WAGSTAFF: DETERMINATION OF CHROMIUM IN Analyst, VoZ, 104 carefully washed out using three approximately 1.5-ml washes with de-ionised water from a wash-bottle with a very fine nozzle. The contents of the tube were then diluted to volume and shaken and any suspended matter was allowed to settle prior to nebulisation.The final acid concentration during the evaporation step, in conjunction with the addition of hydrogen peroxide, should ensure adequate digestion of particulate and organically bound chromium in natural waters and effluents. The solutions in the beakers do not boil dry if the temperature of the hot-plate is carefully set because the presence of the ammonium perchlorate significantly increases the boiling-point of the liquid, and hence decreases the rate of evaporation as the solution approaches dryness. It is not normal practice to heat solutions of perchlorate in the presence of organic matter, but it should be stressed that each beaker contains only 200 mg of ammonium perchlorate (equivalent to 170 mg of perchloric acid). The method is applicable only to river samples, potable waters and sewage final efluents and the solution in the beaker should not boil dry.In order to attempt to evaluate the risk. of explosion, 10 ml of an industrial digested sludge containing 6% of dry solids with levels of over 3000 pg g-l of copper and zinc were added to three beakers, diluted to 50ml and1 carried through the procedure. When the evaporation stage was nearly completed the temperature of the hot-plate was increased to the maximum and the solutions in the beakers were allowed to boil dry. Although violent spitting and some deflagration were observed, no explosive reaction occurred. Numerous similar tests with 50 ml of 5000 and 10000 pg ml-l glucose solutions and 50 ml of a 1000 pg ml-l glucose solution containing 0.5 nil of vegetable oil yielded similar results.It is generally agreed that carbohydrates, and vegetable oils and fats constitute a particular hazard in the presence of perchloric acid.22 The use of 10-ml calibrated stoppered tubes rather than 10-ml calibrated flasks was considered accurate enough for this type of analysis and facilitated solution transfer and storage of large numbers of samples. Some laboratories will prefer to avoid boiling down a solution containing ammonium perchlorate and an alternative method was tested. This involved using a larger initial volume so that the ammonium perchlorate could be incorporated after the evaporation stage and allow efficient washing of the beaker used for the evaporation. Caution-During the whole course of this study no violent reactions were observed, but if the technique was scaled up the risk would almost certainly increase.Alternative Method Volumes (100 -+ 1 ml) of the samples, standards and blanks were placed in 150-ml boro- silicate beakers and 8 ml (rfi0.2 ml) of 25% V/V hydrochloric acid and some aluminium oxide anti-bumping granules were added. The beakers were then placed on a hot-plate and gently simmered until the solution volume had decreased to 40 ml (&lo ml), then 1 ml (kO.1 ml) of hydrogen peroxide (6% m/m) was added. The evaporation was continued until the final volume was about 8 ml (&2 ml). The solutions were allowed to cool, 4 ml (kO.1 ml) of 10% m/V ammonium perchlorate then added and the contents transferred into 20-ml calibrated borosilicate glass tubes.The beakers were carefully washed using four approximately 2-ml washes with de-ionised water from a wash-bottle with a very fine nozzle. The contents of the tube were then diluted to volume and shaken and any suspended matter was allowed to settle prior to nebulisation. All the results quoted in this paper were obtained by using the original method except for some precision measurements. No significant blanks were observed but prior to this work base-line noise and drift were observed during chromium determinations on humid days. Water droplets could be seen in the tube connecting the air supply from the air compressor water trap to the atomic- absorption spectrophotometer, This probleni was overcome by inserting an additional water trap in the air line. Inter-element Effects Table IV (in addition to Table I) shows the effect of various substances on the chromium response.It can be seen that for typical levels of the main matrix elements in natural waters and sewage effluents, the method would appear to be satisfactory. Aliquots of tap water, sample 1 (see Table V), were spiked with equal amounts (0.04 and 0.4pgml-1) ofMarch, 1979 WATERS AND EFFLUENTS BY AAS USING AN AIR - ACETYLENE FLAME TABLE IV EFFECT OF VARIOUS SUBSTANCES ON THE CHROMIUM RESPONSE All solutions contained 2% m/V ammonium perchlorate and 10% V/V hydrochloric acid (36% m/m). Instrumental conditions as in Table 11. Concentration of Chromium interfering substance? / concentration*/ Interfering substance* pg ml-l pg ml-l Relative signal 2 100.0 2 98.7 2 000 200} 2 96.6 100 None - Ca Mg Ca Mg PO, (as P) Ca 100 1000 Mg c u Zn Ca Mg Fe 200 None 1 100.0 500 1 100.0 Ca 100 Mg SiO, 500 100) 1 96.1 Ca 50 Mg Detergent (Mannoxol) 500 100) 1 99.6 Ca 500 Mg so* 500 loo} 1 100.8 Ca Mg PO, (as P) 50 Ca 5001 1000 loo} 2 95.0 loo} 2 97.4 229 Mg NH, (as N) NO, (as N) Ca SiO, Ca Zn Mn Mg Mg c 1 99.6 500 100 1000 100 99.7 97.1 * All cations were added as chlorides.SO, was added as H,SO,, SiO, as sodium silicate, t The concentrations shown represent the final concentrations in the nebulised solution phosphates as (NH,),HPO, and NH, and NO, as NH,NO,. and should be divided by 5 when related to the analyte solution. chromium( 111) and chromium(V1) and taken through the procedure, and no significant difference in response for the two oxidation states was detected.Results The technique was initially tested using a sample of tap water with a moderate total230 THOMPSON AND WAGSTAFF : DETERMINATION OF CHROMIUM IN Analyst, Vol. 104 ANALYTICAL RESULTS ON SAMPLES USED, PRIOR TO CONCENTRATION BY EVAPORATION Sample No. 1 2 3 4 5 6 7 Total hardness Source (CaCO,) Tap water . . . . . . 238 River water . . .. . . 516 River water . . .. . . 538 River water . . .. .. 327 Sewage works final effluent . . 563 Sewage works final effluent . . 593 Contentlpg ml-l - h Total Mg Na Fe 12 11 (0.03 43 35 0.38 51 460 0.20 15 25 1.7 24 44 0.65 27 78 0.41 1 Conductivity/ SO,a- C1- pScm-l 50 18 405 302 34 986 403 800 2 900 119 37 587 348 39 1093 296 94 1431 hardness value (see Table V). A large sample of the tap water was acidified with hydro- chloric acid to pH 2.5 and split into four aliquots, three of which were then spiked with 0.02, 0.2 and 0.4 pg ml-l of chromium(II1). The resulting samples were then analysed in duplicate over nine days and Table VI gives the results of these analyses.The alternative method, in which the ammonium perchlorate was added after the evaporation stage, showed within- batch standard deviations comparable to those in the original method (see Table VII). TYPICAL Sample Unknown standard TABLE VI PERFORMANCE DATA FOR TAP WATER, SAMPLE NO. 1, ANALYSED IN DUPLICATE OVER NINE DAYS Added chromium/ Mean chromium Standard deviation/ pg ml-l result/pg ml-l pg ml-l I (0.2 pg ml-l) . . .. Tap water .. .. Tap water . . .. 0.02 Tap water .. ..0.2 - Tap water .. .. 0.4 Blank . . .. .. - 0.197 0.001 3 0.0204 0.200 0.396 - 0.004 94 0.000 93 0.001 2 0.004 37 0.01 11 0.000 92* * The within-batch standard deviation (s) of the blank was calculated using the nine sets of paired blank r e s ~ l t s . ~ Hence, the 4.65 s detection limit is 0.0043 pg ml-l. Recovery tests using the original method were then carried out using three river-water samples and two sewage works final effluents. These samples were selected as they exhibited high conductivities and high sulphate levels and would be expected to give a good indication of maximum potential inter-element effects. Table V gives the main matrix element concentrations for these samples. Table VIII shows the recoveries obtained after the addition of 0.04 and 0.4pgml-l of chromium(II1) to these samples, and also shows some recoveries obtained in the absence of ammonium perchlorate.The natural chromium level in the samples, except for sample 2, was below the detection limit. Sample 2 was found to contain 0.015 pug ml-l of chromium. Using the electrothermal atomisation technique utilising the standard additions method of calibration, a chromium level of 0.016 pg m1-l was found in this sample. It can also be TABLE VII COMPARISON OF ORIGINAL AND ALTERNATIVE METHODS Sample Standard . . * . Tap water . . .. Within-batch standard deviationlpg ml-l (9 degrees of freedom) p g ml-L method method 0.2 0.002 0 0.001 8 0.2 0.002 2 0.001 9 - Added chromium/ Origmal AlternativeMarch, 1979 WATERS AND EFFLUENTS BY AAS USING AN AIR - ACETYLENE FLAME 231 seen from Table VIII that the proposed method would appear to be satisfactory for natural water and sewage final effluent analysis and that in the absence of ammonium perchlorate poor recoveries were observed, especially if the flame conditions were set for maximum response (Le., a distinctly luminous flame).TABLE VIII RECOVERY TEST RESULTS Chromium recoveredl pg ml-1 f A Flame on verge of luminosity (Table 11) Sample Sample No. River water . . 2 River water . . 3 River water . . 4 Sewage final effluent . . 5 Sewage final effluent .. 6 Chromium added (0.04 pg ml;l) plus ammonium perchlorate 0.040 0.037 1 0.039 6 0.0386 0.038 2 Chromium added (0.4 pg ml-l) plus ammonium perchlorate 0.380 0.379 0.404 0.387 0.370 7 ~- Chromium added (0.4 pg ml-l) and no ammonium perchlorate 0.293 0.296 0.332 0.312 0.306 Luminous flame with chromium added (0.4 p g ml-l) and no ammonium perchlorate 0.188 0.184 0.230 0.243 0.246 Conclusions The concentration by evaporation technique with the addition of ammonium perchlorate and hydrochloric acid would appear to be a rapid pre-concentration technique for the atomic-absorption spectrophotometric analysis of total chromium in natural waters and sewage final effluents using the air - acetylene flame.A detection limit of 0.0043 pg ml-l (4.65 s) was obtained and inter-element effects were considered acceptable. The proposed technique has the additional advantage that other elements can also be determined. The authors thank Mr. W. F. Lester, Director of Scientific Services, Severn-Trent Water Authority, for permission to publish this work.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References “World Health Organization European Standards for Drinking Water,” Second Edition, 1970. EEC Directive, Oflcial Journal of the European Communities, 75/440/EEC, July 1975. Wilson, A. L., and Cheeseman, R. V., “Manual on Analytical Quality Control for the Water Industry,” Water Research Centre Technical Report TR66, Water Research Centre, Medmenham, 1978. Rawa, J. A., and Henn, E. L., Am. Lab., 1977, 9 (8), 31. Thompson, K. C., and Reynolds, R. J ., “Atomic Absorption, Fluorescence and Flame Emission Spectroscopy, A Practical Approach,” Griffin, London, 1978. Midgett, M. R., and Fishman, M. J., Atom. Absorption Newsl., 1967, 6, 128. Gilbert, T. R., and Clay, A. M., Analytica Chim. Acta, 1973, 67, 289. Hurlbut, J. A., and Chriswell, C. D., Analyt. Chem., 1971, 43, 465. Barnes, L., Analyt. Chem., 1966, 38, 1083. Purushottam, A., Naidu, P. P., and Lal, S. S., Talanta, 1973, 20, 631. Ottaway, J. M., and Pradhan, N. K., Talanta, 1973, 20, 927. Kodama, M., Shimizu, S., Sato, M., and Tominaga, T., Analyt. Lett., 1977, 10, 591. Kraft, G., Lindenberger, D., and Beck, H., 2. Analyt. Chem., 1976, 282, 119. Thompson, K. C., Analyst, 1978, 103, 1258. Tenny, A. M., Instrum. News, 1967, 18, 14. Burke, K. E., and Albright, C. H., Dev. Appl. Spectrosc., 1970, 8, 33. Oguro, H., Nippon Kagaku Kaishi, 1976, 7 , 1409. Oguro, H., Nippon Kagaku Kaishi, 1977, 8, 225. Oguro, H., Nippon Kagaku Kaishi, 1977, 8, 218. Thompson, K. C., and Wagstaff, K., “Development of a Technique for the Analysis of Certain Toxic Metals in Sewage Sludges,” Internal Report No. TP116/ML19, Severn-Trent Water Authority , Malvern . Vogel, A. I., “A Text Book of Macro and Semimicro Qualitative Inorganic Analysis,” Longmans, Green, London, 1955. Gorsuch, T. T., “The Destruction of Organic Matter,” Pergamon Press, Oxford, 1970. Received September 7th, 1978 Accepted October 17th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9790400224

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

232 Analyst, March, 1979, Vol. 104, pp. 232-240 Determination of Selenium in Soil Digests by Non -dispersive Atomic-FI uorescence Spectrometry Using an Argon = Hydrogen Flame and the Hydride Generation Technique J. Azad, G. F. Kirkbright and R. D. Snook Department of Chemistry, Imperial College, London SW7 2BP The determination of selenium at submicrogram levels by atomic-fluorescence spectrometry, based on the evolution of hydrogen selenide into an argon - hydrogen air-entrained flame, is described. Using a simple purpose-built non-dispersive atomic-fluorescence spectrometer a detection limit of 10 ng cm-3 of selenium is obtained. The technique has been applied to the determina- tion of selenium in soil digests and experiments have been carried out in order to study the interference of other elements on the determination.Procedures for the elimination of interferences from copper are recommended. Keywords Selenium determination ; atoulzic-fluoracence spectrometry ; lzydride generation ; soil digests The volume of literature published on the flame spectrometric determination of selenium is small compared with that available for many other elements. The determination of selenium by flame spectrometry presents some problems ; for example, the selenium resonance lines lie in the far ultraviolet region of the spectrum below 200nm and this frequently leads to unfavourable signal to noise ratios resulting from atmospheric and back- ground absorption of these selenium lines. Rann and Hambly,l however, obtained a sensitivity (for 1% absorption) of 1.0 pg ~ m - ~ of selenium by atomic-absorption spectro- metry (AAS) using the 196.1-nm selenium resonance line and an air - acetylene flame.With the introduction of the argon - hydrogen air-entrained flame2 the problem of flame absorption was greatly reduced but more severe interference effects were observed in AAS for selenium in this cooler flame. In order to provide higher sensitivity and to overcome some of these interferences chemical separation procedures have been developed that are based on the evolution of hydrogen selenide into the The original method for the determination of selenium by AAS in this way employed reduction with metallic zinc, and a collection and storage device was used for the evolved hydrogen selenide prior to its introduction into the flame for AAS.6 Pollock and West' extended the technique to the determination of bismuth, antimony and tellurium by AAS using a magnesium metal- titanium( 111) chloride mixture as reductant.Schmidt and RoyeI.8 reported the determina- tion of selenium, arsenic, bismuth and antimony by AAS using the hydride generation technique in a procedure in which a solution of sodium tetrahydroborate( 111) was employed as the reductant; this reagent had previously been shown by other workersV~lO to provide rapid and efficient reduction. Sodium tetral~ydroborate(II1) has been employed in this work for the generation of hydrogen selenide prior to the determination of selenium by atomic-fluorescence spectroscopy (AFS) in an argon - hydrogen air-entrained diffusion flame using a simple non-dispersive atomic-fluorescence spectrometer.This paper reports the development of a reliable, simple method for the determination of selenium in aqueous solution by this technique and a study of the chemical interference effects encountered. The atomic-fluorescence spectrometric determination of selenium was first reported by Dagnall et aZ.ll using a dispersive spectrometeir equipped with an air - propane flame giving a detection limit of 0 . 2 5 p g ~ m - ~ of selenium on aspiration of aqueous solutions using a pneumatic nebuliser. Fluorescence from the 204.0-nm selenium resonance line was observed when the flame was irradiated by radiation from a selenium electrodeless discharge lamp, the optical axis of which was aligned at 90" to the optical axis of the monochromator. In this study a similar experimental arrangement has been employed using a non-dispersiveAZAD, KIRKBRIGHT AND SNOOK 233 spectrometer with which it was possible to observe fluorescence from the 196.1-, 214.3- and 204.0-nm lines simultaneously, thus enabling a detection limit of 10 ng cme3 to be observed using discrete sample introduction via the hydride generation technique. Two procedures have been investigated for the suppression or elimination of the well known interference of copperl29l3 on the determination of selenium by the hydride genera- tion technique.In the first procedure the copper and other interfering ions are removed when selenium is coprecipitated with lanthanum from alkaline medium and the precipitate containing selenium is taken for analysis by the procedure developed.In the second procedure the interference from copper is suppressed by utilising the addition of tellurium( IV) to the analyte solution; stable copper telluride is formed and the interference of copper on the hydrogen selenide generation step is suppressed. The chemical pre-treatment and atomic-fluorescence spectrometric procedures developed have been applied to the deter- mination of selenium in soil digests obtained after digestion with a perchloric acid - nitric acid mixture. Experimental Apparatus The instrumentation employed in this work was a purpose-built non-dispersive atomic- fluorescence spectrometer and a simple hydride generation apparatus. A schematic diagram of the equipment employed is shown in Fig.1 and the details of the components employed are listed in Table I. Radiation from a microwave-excited selenium electrodeless discharge lamp (EDL) was focused on to a rotating sector and then refocused into the argon - hydrogen flame. The atomic-fluorescence radiation stimulated from selenium atoms in the flame was then observed at 90" to the incident radiation by passage through a focusing lens to the solar-blind end-window photomultiplier. The output from the photomultiplier was taken to a lock-in amplifier whose reference signal was provided by the rotating sector in the incident radiation beam from the EDL source. The analytical atomic-fluorescence signals for selenium observed at the output from the lock-in amplifier were displayed at the pot en t iometric chart recorder.E. H.T I Baffle L_? I A Loc k- in amp1 if ier L' Reference P.S.D. Hydride generation cell Chart recorder Microwave generator I I Fig. 1. Schematic diagram of equipment employed. Reagents Selenium( IV) standard solutions were prepared by dissolving pure elemental selenium (Specpure grade, Johnson Matthey Ltd.) in a minimum volume of concentrated nitric acid and diluting to volume with 5 M hydrochloric acid. The sodium tetrahydroborate(II1)234 AZAD et al. : DETERMINATION OF SELENIUM IN SOIL Analyst, Vol. 104 reagent was used as a freshly prepared 5% (m/’V) solution in 1% sodium hydroxide solution. Analytical-reagent grade lanthanum nitrate, tellurium(1V) oxide, perchloric acid, hydro- chloric acid, nitric acid and concentrated ammonia solution were used in all experiments.TABLE I INSTRUMENTATION EMPLOYED Source .. .. .. Chopper. . . . . . Microwave generator . . Photomultiplier . , Lock-in amplifier . . Phase sensitive detector optics . . . . .. Chart recorder . . .. * . .. .. .. .. .. .. .. Selenium microwave electrodeless discharge lamp operated at 2450 MHz in a &wave resonant cavity. Radiation modulated with an eight-sector mechanical chopper Programmable Rofin, Model 7500, 3-800 Hz (Rofin Ltd., Egham, Surrey) Microtron 2001 (EMS Ltd. , Wantage, Berkshire) Solar blind, Type R431, Hamamatsu Co., Japan Brookdeal Electronics, Type 450s (Brookdeal Ltd., Bracknell, Berkshire) Brookdeal Electronics, Type 41 1 (Brookdeal Ltd.) Source focused as 1: 1 image on the flame using two 7.5-cm focal length fused silica convex lenses (L, and La).Flame focused as inverted 1: 1 image on PMT using 7.5-cm focal length lens (Id3) Servoscribe, Model RE 511.20 (Smiths Industries Ltd.) Procedure With the flame ignited and argon passing through the hydride generation cell, sufficient time (approximately 20s) was allowed for the_ replacement of any air in the apparatus. A 2-cm3 volume of sodium tetrahydroborate(II1) reagent solution was then transferred into the generation cell through the side-arm. Acidified selenium standard solution (or sample solution) (1 cm3) was then pipetted into the sodium tetrahydroborate(II1) solution using a syringe pipette whose tip was fitted with a rutbber sleeve to ensure a gas-tight fit with the side-arm of the cell during sample introduction.The hydrogen selenide generated was then swept into the argon-hydrogen flame by the argon supply to the flame. The selenium atomic-fluorescence signal was recorded at tlhe potentiometric chart recorder ; the signal duration observed was approximately 8 s for ii 5 pg ~ m - ~ selenium standard solution. The optimum operating conditions established for the procedure, with the particular instru- mental arrangement employed, are summariseti in Table 11. TABLE I1 OPTIMUM OPERATING CONDITIONS FOR DETERMINATION OF SELENIUM Microwave power to source . . .. .. .. .. .. Reflected power from cavity .. .. .. .. .. Applied voltage to PMT . . .. .. .. .. .. Hydrogen flow-rate . . .. .. .. .. .. * . Argon flow-rate . . .. .. .. .. .. .. Hydride generation cell volume .. .. .. .. .. Sodium tetrahydroborate(II1) reagent volume (5% m/ V ) Selenium sample solution volume . . .. .. .. .. .. 50 W 12 w 600 V 3.3 dma min-1 6.0 dm8 min-1 46 cma 2 cm3 1 cm3 Determination of Selenium in Soil Digests One-gram amounts of soil samples were weighed into a series of test-tubes, 3.5 cm3 of oncentrated nitric acid were added to each sample and the test-tubes were covered and llowed to stand overnight. A few glass boiling beads were added to each tube and thenMarch, 1979 DIGESTS BY NON-DISPERSIVE ATOMIC-FLUORESCENCE SPECTROMETRY 235 1.5 cm3 of concentrated perchloric acid (72% m/V) were added to each. The tubes were then transferred into a cold aluminium digestion block, the temperature of which was increased steadily to 100 "C over a period of 30 min.The block was maintained at this temperature for 30 min and then the temperature was increased to between 190 and 200 "C and main- tained at this temperature until digestion of the soil was complete. The final temperature of 200 "C should not be exceeded if charring and the loss of selenium by volatilisation are to be avoided. The test-tubes were then removed from the digestion block and allowed to cool. A 2-cm3 volume of potassium bromide solution (2% m/V) was added to each and the test-tubes were allowed to stand in boiling water for 15 min to ensure complete reduction of selenium(V1) to selenium(1V). The solutions were then centrifuged and the residues rejected. The supernatant solution was taken for analysis; either the lanthanum nitrate - ammonia or the tellurium(1V) addition procedure was applied in order to eliminate inter- ference from copper.The solutions were then made 5 M with respect to hydrochloric acid and analysed by the hydride generation technique using the atomic-fluorescence spectro- meter. Procedures for Suppression of Interferences Lanthanum nitrate coprecipitation procedure Lanthanum nitrate (0.5 cm3 of a 5% m/V solution) was added to each solution prepared for analysis using the digestion procedure described above, 2 cm3 of ammonia solution were then added and the solutions were mixed. After standing for 1 min the solutions were centrifuged and the liquid discarded. The precipitate was then dissolved in the appropriate amount of 5 M hydrochloric acid. Telluyiam(I V ) procedure Tellurium(1V) oxide (0.3 cm3 of a 0.1 M solution) was added to each solution prepared using the digestion procedure described above and then diluted to 5 cm3 with 5 M hydro- chloric acid.Results and Discussion Optimisation of Experimental Parameters Pure aqueous standard selenium(1V) solutions were used in order to optimise the experi- mental variables in the instrumental system, and to provide the best attainable sensitivity and precision in the determination of selenium by the atomic-fluorescence spectrometric technique, utilising the generation of hydrogen selenide for introduction of the analyte into the argon - hydrogen air-entrained flame. The operating power for the microwave-excited EDL source, argon and hydrogen gas flow-rates to the flame, photomultiplier operating voltage, hydride generation cell volume and sodium tetrahydroborate(II1) and selenium sample solution volumes used were each varied independently in order to establish optimum conditions for the determination of selenium.The optimum conditions established in this way are summarised in Table 11. E$ect of hydrochloric acid alzd sodium tetrahydroborate(III) concentrations The effect on the intensity of the atomic-fluorescence signal, observed for 0.5 pg of selenium introduced into the hydride generation cell in 1 cm3 of solution, of variation in the con- centration of hydrochloric acid in the sample solution was investigated. The sodium tetrahydroborate( 111) concentration was maintained constant at 5% (m/V) for this experi- ment. The results obtained are shown in Fig.2. Variation in the acid concentration present in the selenium sample solution has a pronounced effect on the efficiency of generation of hydrogen selenide only when the solution is less than 0.8 M with respect to hydrochloric acid; in all further work the hydrochloric acid concentration of solutions to be analysed was maintained at 5 M. Using 1 cm3 volumes of solution containing 0.5 pg of selenium, which were 1 M with respect to hydrochloric acid, the effect on the selenium atomic-fluorescence signal of vari- ation in the concentration of sodium tetrahydroborate(II1) solution in the generation cell236 AZAD et al. : DETERMINATION OF SELENIUM IN SOIL Analyst, Vol. 104 was investigated. The results obtained are shown in Fig. 3; little variation in hydride generation efficiency was observed over the concentration range 2 4 % (m/V) of sodium tetrahydroborate(II1). A concentration of 5y4, sodium tetrahydroborate( 111) was chosen for use in all further work.100 100 III- 2.0 4.0 6.0 ---i 0.5 1 .o Concentration of HWM Na BH4 concentration, % m/V O 0.1 Fig. 3, Effect of sodium tetra- hydroborate(II1) concentration on the centration on the determination of 0.5 determination of 0.5 pg cm-S of sele- pg cm-S of selenium. nium. Fig. 2. Effect of hydrochloric acid con- Calibration graph, limit of detection and $recisio;vt With the optimum instrumental operating conditions and reagent concentrations, analytical calibration graphs for selenium were found to be rectilinear for selenium solutions containing between 10 and 500 ng cmW3 of selenium in 1 cm3 sample volumes, i.e., 10500 ng of selenium (Fig.4). The relative standard deviation obtained in the repetitive determina- tion of selenium in a solution containing 100ng~m-~ was 2.5%. The detection limit for selenium, defined as that mass of selenium required to produce a signal to noise ratio of 2 for the atomic-fluorescence signal, was 10 ng of selenium under the conditions employed. A significant background blank signal was observed for selenium, equivalent to approxi- mately 16 ng cmV3 of selenium, caused by the presence of selenium as impurity in the sodium tetrahydroborate(II1) reagent ; this blank was corrected for by subtraction in all quantitative analytical work undertaken. Selenium concentration/pg cmv3 Fig.4. Aqueous calibration graph for the determinaMon of selenium by non-dispersive atomic-fluorescence spectrometry.March, 197'9 DIGESTS BY NON-DISPERSIVE ATOMIC-FLUORESCENCE SPECTROMETRY 237 Interference efects and their sufi$ression and elimination The determination of selenium by atomic-absorption spectrometry utilising the hydride generation technique is well known to be subject to interference from a number of heavy metal ions, and in particular copper(II), which depress the efficiency of the hydrogen selenide generation by sodium tetrahydroborate( 111). Similar interferences were expected in the atomic-fluorescence spectrometric procedure developed here and were confirmed in experi- ments in which the effects of metal ions on the atomic-fluorescence signal produced for 500ng of selenium were recorded.Table I11 illustrates typical results of the depressive effects of the presence of some metal ions on the analytical signals observed for 1 pg CM-~ selenium solutions. It is clear that the presence of copper(I1) as a concomitant element causes serious interference; in the presence of 1000 pg cm-3 no atomic-fluorescence signal was obtainable for selenium. As copper(I1) concentrations in soil digests were expected to be sufficiently high to interfere with the selenium determination, two procedures were investigated to minimise or eliminate interference from copper. TABLE I11 DEPRESSIVE EFFECT OF METAL IONS ON THE ANALYTICAL SIGNALS OBSERVED FOR 0.5 pg OF SELENIUM Concentration of interfering element 1 000 pg cm-8.Element Na(1) . . .. K(1) . . .. .. .. .. ::\::\ Ca(I1) . . Ba(I1) . . .. .. Hg(W Al(I1) . . .. Depression of signal, 0 2 0 0 2 0 0 0 % Element Fe(I1) . . .. Fe (I1 I) .. Pb(I1) . . .. Zn(I1) . . .. Co(I1) . . .. Cu(I1) . . .. A m * - .. Ni(I1) . . .. Depression of signal, 36 20 40 21 20 99 80 65 % The procedure reported by Bedard and Kerbyson,12 in which the interference of copper on the determination of selenium, by AAS via generation of hydrogen selenide, was elimi- nated by removal of the selenium from sample solutions by coprecipitation from an alkaline medium with lanthanum, was investigated. This procedure was observed to give good recovery of selenium using a double precipitation; the mean recovery for 10 replicate analyses was 99% with a relative standard deviation of 2.5%.The procedure was found to be most efficient when the lanthanum hydroxide precipitate was filtered off as soon as possible after precipitation. The effect on the recovery of selenium, monitored as the selenium atomic- fluorescence signal intensity, of elapsed time between precipitation and filtration is shown in Fig. 5. The effect of variation in pH of the solution on the selenium recovery by the coprecipitation procedure was found not to be critical provided that the pH was maintained above 9.0. The second procedure investigated for suppression of the interference from copper utilises the addition of tellurium( IV) to sample solutions immediately before the hydride generation procedure ; this procedure has been described elsewhere by Kirkbright and Taddia.13 In this procedure the interference from copper is suppressed by formation of copper telluride, which is more stable than copper selenide.The addition of excess of tellurium(1V) results in some suppression of the atomic-fluorescence signal observed for selenium ; Fig. 6 shows the effect of variation in the tellurium(1V) concentration added to 0.5 p g ~ m - ~ selenium sample solutions on the signal recorded. A constant suppression of approximately 30% is attained at tellurium(1V) concentrations between 0.06 and 0.08 M. The presence of 0.06 M tellurium(1V) enables relatively high concentrations of copper to be tolerated in the deter- mination of selenium; Fig. 7 shows the effect of increasing copper concentration on the atomic-fluorescence signal observed from 0.5 pg ~ m - ~ selenium sample solutions containing 0.06 M tellurium(1V) solution.Copper does not cause interference at levels up to 50 p g cmq3 although at higher concentrations the selenium recovery decreases. Addition of a con- centration of tellurium(1V). greater than 0.06 M to sample solutions would permit extension238 AZAD et al.: DETERMINATION OF SELENIUM IN SOIL Analyst, VoZ. 104 .- f c W W v) - J a - ' 10 20 30 I ' 0 a.04 0.06 0.08 Time lapsed before filtratiodmin Fig. 5. Effect of time elapsed Tellurium (IV) oxide concentration/M before filtration on the recovery of selenium observed when using the Fig. 6. Effect of tellurium(1V) con- lanthanum nitrate coprecipitation pro- centration, added to 0.5pgcm-S of cedure to remove interferences.selenium solutions, on the generation of selenium hydride. of the tolerance of the procedure to higher concentrations of copper(I1). Fig. 8 shows a comparison of the analytical calibration graphs obtained for aqueous selenium solutions in the presence and absence of copper utilising tellurium( IV) to suppress copper interference. These results confirm the restoration of the selenium signal to approximately 70% of its value in the absence of copper when tellurium(1V) is employed to suppress copper inter- f erence . I , I , 0 40 80 120 Copper concentration/pg cm-3 Fig. 7. Effect of copper on the deter- mination of 0.5 p g cm-s of selenium in the presence of 0.06 M tellurium (IV) . B C _- - C 6 Selenium concentration/pg cm-3 Fig. 8. Comparison of analytical calibrations in the presence and absence of 0.06 M tellurium(1V) oxide.A, Aqueous selenium; B, aqueous selenium + 0.06 M telluiium(1V) + 20 pg cm-S of copper; and C, aqueous selenium + 20 pg cm-8 of copper. Determination of Selenium in Soil Digests Soil samples were digested using a mixture of perchloric and nitric acid; care was taken not to char the sample during digestion at 200 "C and to avoid loss of selenium by volatilisa- tion. As the hydride generation procedure is only applicable to selenium in its oxidation state of four it was necessary to reduce any selenium(V1) produced in the strongly oxidising digestion mixture by the addition of potassium bromide solution after digestion. Sample digestion recoveries were evaluated by adding to l-g soil samples a known amount of selenium prior to their digestion.The recovery of the added selenium was then deter- mined. The results of these experiments are shown in Table IV.March, 1979 DIGESTS BY NON-DISPERSIVE ATOMIC-FLUORESCENCE SPECTROMETRY TABLE IV RECOVERY OF SELENIUM ADDED TO SOIL SAMPLE NO. 4 239 Selenium concentration Selenium added/ Selenium determined] Recovery, in sample/pg 8-1 Pg P8 % 0.7 f 0.014 0.1 0.83 104 0.7 f 0.014 0.2 0.87 97 0.7 f 0.014 0.3 0.94 94 0.7 f 0.014 0.4 1.18 107 Nine soil samples were digested using the procedure described. Each sample was then analysed by both the lanthanum coprecipitation and the tellurium(1V) methods of interference suppression. The results obtained for the selenium content of the soils analysed are shown in Table V for both methods of interference suppression. As can be seen from the table there is no appreciable quantitative difference between the results obtained by both methods.These results also show extremely good agreement with those obtained by the hydride generation technique and optical-emission spectrometry using an inductively coupled argon plasma source. TABLE V COMPARISON OF RESULTS FOR THE SELENIUM CONTENT OF SOIL DIGESTS Selenium found r L % soil sample 1 2 3 4 5 6 7 8 9 Lanthanum nitrate method Mean, SD, RSD, 0.37 0.017 4.5 0.36 0.016 4.4 0.24 0.010 4.1 0.70 0.014 2.0 18.7 0.64 3.4 111 2.93 2.6 0.29 0.014 4.8 0.31 0.020 6.4 0.30 0.023 7.6 A r > p.p.m. p.p.m. % Tellurium( IV) method Mean, p.p.m. 0.35 0.35 0.23 0.68 18.6 110 0.28 0.30 0.29 SDI p.p.m. 0.009 0.012 0.015 0.015 0.49 2.45 0.012 0.015 0.018 RSD,’ % 2.6 3.4 6.5 2.2 2.6 2.2 4.2 5.0 6.2 ICP* method, p.p.m.0.38 0.33 0.23 0.69 19.2 - 0.28 - * ICP = optical-emission spectrometry using an inductively coupled argon plasma source. Conclusion It has been demonstrated that selenium can be determined in soil digests using the technique of non-dispersive atomic-fluorescence spectrometry. The technique is both sensitive and precise. Although the hydride generation procedure is normally subject to severe interference from copper, this effect has been eliminated by employing chemical pre- treatment of the samples, using lanthanum hydroxide as a coprecipitant or the addition of tellurium(1V) to forrn stable copper telluride during reduction. Both methods have been applied successfully to the determination of selenium in soil digests.It is difficult to recommend which procedure is the most suitable for selenium determinations as each has its own advantages; the lanthanum hydroxide coprecipitation is well established and with care leads to excellent recoveries of selenium even in the presence of high concentrations of copper (approximately 2000 pg cm-3). In order to obtain good recoveries of selenium, however, re-precipitation must be employed. The tellurium( IV) procedure is a simple method but lowers the detection limit by 30% and at the concentrations employed in this study is only effective in removing interferences from copper up to a concentration of 5 0 ~ g c m - ~ of copper. This limitation may not be a serious disadvantage, however, as copper levels in soil digests should seldom exceed this figure.240 AZAD, KIRKBRIGHT AND SNOOK References Rann, C. S., and Hambly, A. W., Analytica Chim. Ada, 1965, 32, 346. Kahn, H. L., and Schallis, J. E., Atom. Absorption Newsl., 1968, 7, 5. Fernandez, F. J., and Manning, D. C., Atom. Absorption Newsl., 1971, 10, S6. Clinton, 0. E., Analyst, 1977, 102, 187. Siemer, D., and Hagemann, L., Analyt. Lett., 1!)75, 8, 323. Yamamoto, Y. Y., Kumamaru, Y., Hayashi, Y . , and Kanke, M., Analyt. Lett., 1972, 5, 717. Pollock, E. N., and West, S. J., Atom. Absorption Newsl., 1972, 11, 104. Schmidt, F. J., and Royer, J. L., Analyt. Lett., 1973, 6, 17. Pierce, F. D., Lamoreaux, T. C., Brown, H. R., and Fraser, R. S., Appl. Spectrosc., 1976, 38, 38. Fernandez, F. J., Atom. Absorption Newsl., 1973, 12, 93. Dagnall, R. M., Thompson, K. C., and West, T. S., Talanta, 1967, 14, 557. Bedard, M., and Kerbyson, J. D., Can. J. Speabosc., 1975, 21, 64. Kirkbright, G. F., and Taddia, M., Atom. Absovpion Newsl., in the press. Received September 7th, 1978 Accepted October 17th, 1978 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

ISSN:0003-2654    

DOI:10.1039/AN9790400232

出版商:RSC    

年代:1979

数据来源: RSC