ISSN:0003-2654    

DOI:10.1039/AN97499FX037

出版商:RSC    

年代:1974

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97499BX039

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

October, 19741 THE ANALYST iiiHUMPHRY DAVY'SBRILLIANT INVENTIONnIt is well known that Sir Humphry had quite a few laughs finding a use fornitrous oxide. There would have been no problem, however, finding a gooduse for 'Pronalys' analytical reagents because they are exceptionally purechemicals ideal for use in the most exacting analytical procedures.Comprehensive specifications are available for every product and these,with their constant composition, ensure consistency of results.But as 'Pronalys' wasn't around at the time, he invented a miner'ssafety-lamp instead.1aIIII~=IIIIIIIID~~IIIIIIIIm TO:May & Baker Ltd Dagenham Essex RM10 7XSPlease send me further information on 'Pronalys' highi purity analytical reagentsI Name ............................................................................................................mII Address ...................................................................................................... I......................................................................................................................... 1e11111111111111111111llllI'Pronalys' is a trade markRh6ne-hulenc Groupof Companieiv THE ANALYST [October, 1974THE ANALYSTEDITORIAL ADVISORY BOARDChairman: H.J. Cluley (Wembley)*L. S. Bark (Salford)R. Belcher (Birmingham)L. 1. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)'R. M. Dagnall (Huntingdon)E. A. M. F. Dahmen (The Netherlands)*J. B. Dawson (Leeds)A. C. Docherty (Billingham)D.Dyrssen (Sweden)*W. T. Elwell (Birmingham)*D. C. Garratt (London)J. Hoste (Belgium)D. N. Hume (U.S.A.)H. M. N. H. Irving (Leeds)M. T. Kelley (U.S.A.)*J. A. Hunter (Edinburgh)W. Kemula (Poland)*G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)*J. M. Ottaway (Glasgow)*G. E. Penketh (Billingham)S. A. Price (Tadworth)D. 1. Rees (London)E. B. Sandell (U.S.A.)*R. Sawyer (London)A. A. Smales, O.B.E. (Horwell)H. E. Stagg (Manchester)E. Stahl (Germany)A. Walsh (Australia)T. S. West (London)P. Zuman (U.S.A.)*A. Townshend (Birmingham)* Members of the Board serving on the Executive Committee.ANALYTICAL SCIENCES MONOGRAPHNo. IHigh- Precision Titri metryby C. Woodward and H.N. RedmanImperial Chemical Industries Limited (Agricultural Division)BRIEF CONTENTS:IntroductionVisual Titrations, with sections on Apparatus, Standard Substances and their preparation andInstrumental Methods, with sections on Photometric Titrations, Electrometric Titrations andassay, and Standard Solutions.Miscellaneous Techniques.References to the literature of high-precision titrimetry.Pp. viiif63Price f 2-50ISBN 0 85990 501 2Obtainable from :Society for Analytical Chemistry, Book Department,911 0 Savile Row, London, W1 X 1 AFMembers may buy personal copies at the special price of f2.0October, 19741 THE ANALYSTNOTICE TO SUBSCRIBERSJournals, 1975Subscriptions for The Analyst, Analytical Abstracts and Proceedings should besent to:The Chemical Society, Publications Sales Office, Blackhorse Road,Letchworth, Herts., SG6 1 HN.Subscription Rates for 1975 (January to December) (post free)The Analyst, Analytical Abstracts and Proceedings (including indexes) :(a) The Analyst, Analytical Abstracts, and Proceedings .. . . . . . . f48.00(b) The Analyst, AnalyticalAbstracts printed on one side of the paper, and Proceedings f 50-00The Analyst and Analytical Abstracts without Proceedings (including indexes) :( c ) The Analyst, and Analytical Abstracts . . . . . . . . . .(d) The Analyst, and Analytical Abstracts printed on one side of the paper . .(Subscriptions are NOT accepted for The Analyst and/or for Proceedings alone)Analytical Abstracts only (two volumes per year) :(e) Analytical Abstracts .. . . . . . . . .(f) Analytical Abstracts printed on one side of the paper . .Rates for single copies, including back numbers, post free ;The Analyst . . . . . . . . . . . . .The Analyst annual index. . . . . . . . . .Analytical Abstracts . . . . . . . . . .Analytical Abstracts index (whole-year volumes, 1-1 5)Analytical Abstracts index (half-year volumes, 16 onwards)Proceedings . . . . . . . . . . . .Proceedings annual index . . . . . . . .Rates for unbound volumes of the Journals, post free;Unbound volumes of The Analyst, including index (back issues). .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .Unbound volumes of Analytical Abstracts, including index (back issues)... .. .. .. .. .. .. .. .. .. .f 44.00f 46.00f 36.00f 38.00f 3-50f 3.50f 3.50f 8.00f 4.505OP5OPf 36.00. . f36.00vi SUMMARIES OF PAPERS IN THIS ISSUE [October, 1974Summaries of Papers in this IssueAtomic-fluorescence Spectrometry as an Analytical TechniqueA Critical ReviewSUMMARY OF CONTENTSIntroductionGeneralPresent status of atomic-fluorescence spectrometryTypes of fluorescence transitionsRadiance expressions in atomic-fluorescence spectrometryAnalytical curves in atomic-fluorescence spectrometryQuenching processesInterferences in atomic-fluorescence spectrometryGeneralOptical designMonochromator v c w m non-dispersive opticsMultiple-element systemsElectronic designSources for atomic-fluorescence spectrometryTheoryInstrument design for atomic-fluorescence spectrometryVapour discharge lampsContinuum radiation sourcesHollow-cath.ode lampsElectrodeless discharge lampsPulsed laser sourcesFlamesNon-flame cellsSolid sample atomisationReduction - aeration method for mercuryAtom cells for atomic-fluorescence spectrometryPractical applicationsAlloys and high-purity samplesEnvironmental samplesOilsMedicalemission flame analysisComparison between atomic fluorescence, atomic absorption and atomic-Future developments in atomic-fluorescence spectrometryREPRINTS of this Review paper can be obtained from The Society for AnalyticalChemistry, Book Department, 9/10 Savile Row, London, W1X IAF, at50p per copy (with a 25 per cent.discount for 4 or more copies), post free.A remittance for the correct amount, made out to The Society forAnalytical Chemistry, should accompany every order ; these reprints are notavailable through Trade Agents.R. F. BROWNERDepartment of Industry, Laboratory of the Government Chemist, Cornwall House,Stamford Street, London, SE1 9NQ.Analyst, 1974, 99, 617-644October, 19741 SUMMARIES OF PAPERS I N THIS ISSUESulphonated Alizarin Fluorine Blue : an Improved Reagent forthe Positive Absorptiometric Determination of the Fluoride IonAlizarin fluorine blue, [3-NN-di(carboxymethyl)aminomethyl]- 1,2-dihydroxyanthraquinone, has been modified by the introduction of a sulpho-nate group into the 5-position. The complex formed by lanthanuni(II1) withthis new compound demonstrates a reaction towards fluoride similar to thatshown by the complexes formed by lanthanum and cerium(II1) with theoriginal alizarin fluorine blue but with the added advantage of increasedsolubility.This allows a metal to reagent complex ratio of 2 : 1 to be used,which gives considerable increase in sensitivity towards fluoride.M. A. LEONARD and G. T. MURRAYviiDepartment of Analytical Chemistry, The Queen's University of Belfast, Belfast,BT9 5AG.Analyst, 1974, 99, 645-651.An Improved Procedure for Application of the Fujiwara Reactionin the Determination of Organic HalidesFujiwara discovered that when chloroform or other organic halides areheated with pyridine and sodium hydroxide a red colour is formed.Somehalides react very sensitively, others weakly or not a t all, depending on manyfactors.A standard procedure involving this reaction is recommended for use inanalytical toxicology and information on sensitivities of twenty-two organichalides when using this procedure is presented. This information permitsaccurate interpretation concerning the presence or absence of organic halides.Some commercial samples that gave a positive Fujiwara test reacted nega-tively after being purified by distillation or gas - liquid chromatography andthe literature on positive Fujiwara reactions should therefore be consulted.Methods for the determination of an organic halide on the basis of theFujiwara reaction should be used only when all other reactive halides areabsent.J.F. REITH, Miss W. C. van DITMARSCH and Th. de RUITERDepartment of Toxicology, State University, Catharijne Singe1 60, Utrecht, TheNetherlands.Analyst, 1974, 99, 652-656.Determination of Tryptophan and Indole Substances by aColorimetric Diazotisation MethodA colorimetric method for the determination of tryptophan in proteinhydrolysates by its conversion into nitrosamine with nitrous acid followed bydiazotisation with N-l-naphthylethylenediamine dihydrochloride has beenstudied. Selective nitrosation of tryptophan was best achieved a t 20 to35 "C using 1.0 to 1.2 M hydrochloric acid. Diazotisation was best achievedat 10 "C or below. Sodium chloride inhibited the nitrosation reaction to aconsiderable extent and, therefore, tryptophan standards should contain anamount of sodium chloride equal to the amount present in sample hydrolysates.Such standards are prepared from a portion of the sample hydrolysate pre-treated with activated charcoal.Twenty common protein amino-acids otherthan tryptophan did not interfere in the colour development, but the methodwas found to be applicable to indole and its derivatives in addition to tryptophan.Compounds such as phenols and aromatic amines were found to interfere.A. K. GOSWAMIDepartment of Soil Science and Chemistry, Himachal Pradesh University, Palampur176061, H,P., India.Analyst, 1974, 99, 657-660viii THE ANALYST [October, 1974We challenge you!Can you find another manufacturer who can supply AA systems of suchhigh quality at these prices? We don’t believe you can . . . . . . . . . . . . !We do believe we make:the best spectrophotometer for AA and flame emission with bothdirect and integrated read-out at less than $1,600,the best 4 lamp spectro-photometer for AA and flameemission with scanning and manyfeatures others provide as extrasat less than $2,000,the best system for AA, flameemission, flameless absorptionand solution photometry at less-a + than $2,500,the best range F of unique attach-ments for determ-ining Mercury,Arsenic, Selenium, etc.the best is available 7 Prices are ex works and exclude VATTel: Camberlw I02761 63401Industrial Division,Shandon Southern Instruments Ltd.,Frimley Road, Camberley, Surrey. GU16 5E

ISSN:0003-2654    

DOI:10.1039/AN97499FP111

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

October, 19741 THE ANALYST ixCLASSIFIED ADVERTISEMENTSThe ratefor classified advertisements is 35p a line (or spaceequivalent of a line) with a n extra chargz of xop for theuse of a Box Number. Sem-displayed classifiedadvertisements are Lr.60 per single columrc ccnlimctre(wain. 3 cm.)Copy for classi,fied adverliseinents required not latm thanthe 18th of the month precedingthe dale of fiublication whichi s on the 16th of each month. Advertisements should headdressed lo J . Arthur Cook, 0 Lloyd Square, London,WCrX 9BA. Tel.: 01-837 6315APPOINTMENTS VACANTWith an expanding research programme, the Paint Research Associ-ation has a vacancy for an analyst capable of applying moderninstrumental techniques to the solution of practical problems. Theappointment will be made a t Senior Research Officer level (presentsalary range f;2,280-L3,460).Applicants should have a good firstdegree and research or industrial experience with chromatographicand spectroscopic techniques or in polymer chemistry. Pleasc apply,giving details of qualifications arid previous experience to : The Headof Chemistry Division, The Paint Research Association, WaldegraveRoad, Teddington, Middlesex TWll 91-D.Please mention THE ANALYSTwhen replying to advertisementsSENIORASS I STANTANALYSTSalary f 2820-f 31 65 p.a. (pluscurrent threshold of f 146.1 6p.a. + pay award imminent).Applications are invited from persons withthe Associateship of the Royal Institute ofChemistry or equivalent qualification andexperience in the examination of Food,Drugs and Water.Application forms from the County Secre-tary (Room 210), County Hall, Glenfield,L eicester L E3 BRA.(Tel. L eicester 871 3 13,Ext. 7139).Andytical ChemistG. D. Searle are one of the world leaders in medical research.Our Analytical Development Department, based at High Wycombehas a number of new projects from many parts of the worldcurrently under investigation, owing to expansion into newmarkets. And this has caused a vacancy for an Analytical Chemist.You should hold a PhD or BSc degree specialising in organic orphysical chemistry and be experienced in classical andinstrumental methods of analysis including chromatographyand spectroscopic techniques.An attractive commencing salary will be offered and assistancewith re-location expenses will be given in appropriatecircumstances. Conditions of employment include 4 weeks holiday,pension scheme, RUPA and active social club.Initially, please write to H.W. Cooke, PersonnelManager, R & D Division, G. D. Searle & Co. Ltd.,Lane End Road, High Wycombe, Bucks.Please quote reference AC/106x SUMMARIES OF PAPERS I N THIS ISSUEThe Determination of Tannins with Cerium( IV) SutphateA method for the determination of tannin in tea samples is described.The technique is based on the oxidation of tannins with an excess of cerium(1V)sulphate and subsequent determination of the excess by back-titration withammonium iron(I1) sulphate solution. Although this method and theLowenthal procedure appear to determine the same tannins, the precision ofthe former is f 1.24 per cent., while that of the latter is only & 10.12 per cent.M.KAPEL and R. KARUNANITHY[October, 1974Procter Department of Food and Leather Science, University of Leeds, Leeds, LS29 TT.Analyst, 1974, 99, 661-665.The Detection and Determination of Residues of the HerbicideNitrofen in VegetablesA method for residue analysis, based on electron-capture gas chromato-graphy following clean-up on a sulphuric acid - Celite column, is presented.The limit of detection is 10 ng g-l. Recoveries have been investigated fordifferent kinds of vegetables. The extraction efficiency of acetonitrile veYsusdichloromethane was checked in a separate experiment. Spectroscopic dataand results for the herbicide are included.J.KVALVKGChemical Research Laboratory, Agricultural University of Norway, 1432 ~s-NLH,Norway.Analyst, 1974, 99, 666-669.The Simultaneous Collection of Data From Two AutomatedAmino-acid Analysers and Processing of the Data by ComputerAn account is given of the development of a logging system for acquiringdata on punched tape from two amino-acid analysers running at differentspeeds. The outputs of the analysers were sampled in such a way that a visualrecord of the progress of each separation was retained. A method of pro-gramming the computer to edit and process the data is outlined. Resultsobtained from the computer showed good agreement, both within analyticalsystems and between systems.Manual calculations from recorder charts,made by use of the time-consuming manual method, agreed well with thecomputer results from punched tape and cards. Problems encounteredduring the period of development are discussed.J. DAVIDSON, A. W. BOYNE, W. R. HEPBURN and N. L. MACKIERowett Research Institute, Bucksburn, Aberdeen, AB2 9SB.Analyst, 1974, 99, 670-682.The Electrochemical Determination of Vitamin APart I. Voltammetric Determination of Vitamin A inPharmaceutical PreparationsThe voltammetric oxidation of vitamin A a t a carbon paste electrode wasused to determine the vitamin A content of some pharmaceutical preparations.Chromatography is not required as a clean-up procedure, therefore time isgained and a potential source of errors is eliminated.Vitamin E (tocopherol),which, in large amounts, interferes,in the most widely used methods, is renderedelectrochemically inactive in the range of potentials that is of interest. This isachieved by acetylating all of the tocopherols that are present in the samplein alcohol form. The oxidation waves of vitamin A are linearly proportional tothe concentration range studied. The work described in this paper is con-cerned with the alcohol and acetate derivatives of vitamin A.SAMUEL S. ATUMA, JORGEN LINDQUIST and KENT LUNDSTROMDepartment of Analytical Chemistry, University of Uppsala, Box 531, S-751 21,Uppsala 1, Sweden.Analyst, 1974, 99, 683-689October, 19741 THE ANALYST xiRANK HILGERHOLLOWCATHODEL Ring 0843 24261 Ext.28for a 48 hourdelivery servicenates spurious discharges and ensuresexcellent short term stability.Reliability,Manufactured on an advanced produc-tion plant from materials carefullyselected for their high purity. Duringprocesring lamps are baked at a pres-sure of 1094 Torr to provide maxi-mum outgassing giving rise to lowbackground and clean line spectrum.UniversalDesigned t o fit not only Rank HilgerH1170and H1550 butalmorteveryAtomic Absorption Spectrophoto-meter , being suitable for 3 modeoperation: D.C., Pulsed or Modulated.ValueBeing at the lower end of the pricerange coupled with Long Life, HighEnergy and Reliability offer a remark-able value for such a high quality lamp.RANKHILGERWESTWOOD. MARGATEKENT-CT94JL.ENQLANDANALOlD compressed chemicalreagents offer a saving in the use oflaboratory chemicals.The range of over 50 chemicals in tabletform includes Oxidizing and ReducingAgents, Reagents for Colorimetric Analysisand Indicators for Complexometric titra-tions.Full details of all Analoid preparations freeon request from:RIDSDALE & 60. LTD,Newham Hall, Newby,Middlesbrough, Cleveland TS8 9EATelephone: Middlesbrough 3721 6SELECTED ANNUAL REVIEWSof theANALYTICAL SCIENCESVolume 2 - 1972CONTENTSThe Techniques and Theory of Thermal Analysis Applied to Studies on InorganicMaterials with Particular Reference to Dehydration and Single Oxide Systems -D. DollimoreDevelopments in Ion Exchange - F. VernonThermometric and Enthalpimetric Titrimetry - L.S. Bark, P. Bate and J. K. GrimeObtainable from-Pp. vi + 149 f5.00; U.S. $13.00 ISBN 0 85990 202 1The Society for Analytical Chemistry, Book Department,9/10 Savile Row, London W I X I A FMembers of The Chemical Society may buy personal copies at the special price of f3.00; U.S. $8.0xii SUMMARIES OF PAPERS I N THIS ISSUE [October, 1974An Ion-selective Electrode Method for the Determinationof Chloride in MilkA method for determining the chloride content of milk involving the useof a chloride-selective electrode is described. The method is simple, preciseand more rapid than the British Standard procedure, which involves pre-cipitation of the chloride as silver chloride and titration of unreacted silver.A. W. M.SWEETSURThe Hannah Research Institute, Ayr, KA6 5HL, Scotland.Analyst, 1974, 99, 690-692.The ‘Practising Chemists’A History of theSociety for Analytical Chemistry1874-1 974By R. C. CHIRNSIDE andJ. H. HAMENCEf 3-00; u.s.S8.00225 pages; 1 1 platesCS Members f2-50ISBN 0 85990 700 9Obtainable from The Society for Analytical Chemistry,(Book Department), 9/10 Savile Row, London W1 X 1 AOctober, 19741 THE ANALYST xiiiANALYTICAL CHEMISTRY'S editorial slant has beengoing in the direction of these new social concerns . . .but it is still the most dependable and accurate sourceof the basics. In its pages you'll find materials char-acterization and measurement information that ISvital in your functions as chemist metallurgist,biochemist, or engineer.Your subscription brings you 14 issues of ANALYTICALCHEMISTRY a year, making this the biggest bargainin chemical literature.With every subscription you get the currentlA6ORATORY GUlDE TO INSTRUMENTS,EQUIPMENT, AND CHEMKALS .. . PLUS THEApril ANNUAL REVIEWS.This year the ANNUAL REVIEW'S deals withAPPLICATIONS and is a most important sourceof valuable data for every chemist.At only $5.00 a year for members and only $7.00for nonmembers, you can't afford to be without it.Send in the order form today!1155 Sixteenth Street, N W.Washington, D C 20036Please send me ANALYTICAL CHEMISTRY at the following subscription rote IIINonmembers 05700 U$ll 00 US1900 0 5 2 0 0 0Note Subscriptions at ACS Member Rates are for personal ure only1IState/CountryNoture of Company sBusinessL.I am not an ACS member 0 I om an ACS member0 Bill me for $ ___Bill company [ I II L 1 Poyment enclosed in the omount of S chemistryxiv THE ANALYST [October, 1974Reprints of Review PapersREPRINTS of the following Review Papers published in The Analyst since 1963 are available fromthe Book Department, Society for Analytical Chemistry, 9/10 Savile Row, London, W1X 1AF(not through Trade Agents). A complete list of all reprints available from earlier years can beobtained on request.The price per reprint is 50p; orders for four or more reprints of the same or different Reviewsare subject to a discount of 25 per cent. Remittance with order, made out to “Society forAnalytical Chemistry, ’’ will prevent delays.“Classification of Methods for Determining Particle Size, ’’ by the Particle Size Analysis“Methods of Separation of Long-chain Unsaturated Fatty Acids,” by A.T. James (August,“Beer’s Law and its Use in Analysis,” by G. F. Lothian (September, 1963).“A Review of the Methods Available for the Detection and Determination of Small Amounts“Circular Dichroism,” by R. D. Gillard (November, 1963).“Information Retrieval in the Analytical Laboratory,” by D. R. Curry (November, 1963).“Thermogravimetric Analysis,” by A. W. Coats and J . P. Redfern (December, 1963).“Some Analytical Problems Involved in Determining the Structure of Proteins and Peptides,”“The Faraday Effect, Magnetic Rotatory Dispersion and Magnetic Circular Dichroism,” by“Electrophoresis in Stabilizing Media,” by D.Gross (July, 1965).“Recent Developments in the Measurement of Nucleic Acids in Biological Materials, ’’ by“Radioisotope X-ray Spectrometry,” by J. R. Rhodes (November, 1966).“The Determination of Iron(I1) Oxide in Silicate and Refractory Materials,” by H. N. S.“Activation Analysis,” by R. F. Coleman and T. B. Pierce (January, 1967).“Techniques in Gas Chromatography. Part I. Choice of Solid Supports,” by F. J. Palframan“Heterocyclic Azo Dyestuffs in Analytical Chemistry,” by R. G. Anderson and G. Nickless“Determination of Residues of Organophosphorus Pesticides in Food,” by D. C. Abbott and“Radioactive Tracer Methods in Inorganic Trace Analysis : Recent Advances, ” by J. W.“Gamma-activation Analysis, ” by C. A.Baker (October, 1967).“Precipitation from Homogeneous Solution,” by P. F. S. Cartwright, E. J. Newman and“Industrial Gas Analysis,” by (the late) H. N. Wilson and G. M. S. Duff (December, 1967).“The Application of Atomic-absorption Spectrophotometry to the Analysis of Iron and“Inorganic Ion Exchange in Organic and Aqueous - Organic Solvents,” by G. J. Moody and“Radiometric Methods for the Determination of Fluorine,” by J. K. Foreman (June, 1969).“Techniques in Gas Chromatography. Part 11. Developments in the van Deemter RateTheory of Column Performance,” by E. A. Walker and J . F. Palframan (August, 1969).“Techniques in Gas Chromatography. Part 111. Choice of Detectors,” by T. A. Gough andE. A. Walker (January, 1970).“Laser Raman Spectroscopy,” by P. J . Hendra and C. J. Vear (April, 1970).“Ion-selective Membrane Electrodes,” by Ern0 Pungor and KlAra T6th (July, 1970).“X-ray Fluorescence Analysis,” by K. G. Carr-Brion and K. W. Payne (December, 1970).“Mass Spectrometry for the Analysis of Organic Compounds,” by A. E. Williams and H. E.“The Application of Non-flame Atom Cells in Atomic-absorption and Atomic-fluorescence“Liquid Scintillation Counting as an Analytical Tool,” by J. A. B. Gibson and A. E. Lally“The Determination of Some 1,8-Benzodiazepines and Their Metabolites in Body Fluids,”“Atomic-fluoiescence Spectrometry as an Analytical Technique” by R. F. BrownerSub-committee of the Analytical Methods Committee (March, 1963).1963).of Cyanide,” by L. S. Bark and H. G. Higson (October, 1963).by Derek G. Smyth and D. F. Elliott (February, 1964).J. G. Dawber (December, 1964).H. N. Munro and A. Fleck (February, 1966).Schafer (December, 1966).and E. A. Walker (February, 1967).(April, 1967).H. Egan (August, 1967).McMillan (September, 1967).D. W. Wilson (November, 1967).Steel,” by P. H. Scholes (April, 1968).J. D. R. Thomas (September, 1968).Stagg (January, 1971).Spectroscopy,” by G. F. Kirkbright (September, 1971).(October, 1971).by J . M. Clifford and W. Franklin Smyth (May, 1974).(October, 1974)

ISSN:0003-2654    

DOI:10.1039/AN97499BP117

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

OCTOBER, 1974 Vol. 99, No. 1183 THE ANALYST Atomic-fluorescence Spectrometry as an Analytical Technique A Critical Review* By R. F. BROWNER (Department of Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SEX gNQ) SUMMARY OF CONTENTS Introduction General Present status of atomic-fluorescence spectrometry Types of fluorescence transitions Radiance expressions in atomic-fluorescence spectrometry Analytical curves in atomic-fluorescence spectrometry Quenching processes Interferences in atomic-fluorescence spectrometry General Optical design Monochromator vcvs'sus non-dispersive optics Multiple-element systems Electronic design Sources for atomic-fluorescence spectrometry Theory Instrument design for atomic-fluorescence spectrometry Vapour discharge lamps Continuum radiation sources Hollow-cathode lamps Electrodeless discharge lamps Pulsed laser sources Flames Non-flame cells Solid sample atomisation Reduction - aeration method for mercury Atom cells for atomic-fluorescence spectrometry Practical applications Alloys and high-purity samples Environmental samples Oils Medical emission flame analysis Comparison between atomic fluorescence, atomic absorption and atomic- Future developments in atomic-fluorescence spectrometry INTRODUCTION GENERAL- ATOMIC-FLUORESCENCE spectrometry has long been used by physicists investigating electronic perturbations in atoms.Early work included classic experiments by Wood in 1905,l Nichols and Howes in 19242 (the first flame experiments), Badger in 192g3 and Mankopff in 1933.* Much of the work was carried out in quartz vessels, and this aspect has been thoroughly reviewed by Mitchell and Zemansky5 and Pringsheim.6 The first suggestion of any analytical utility for atomic fluorescence was made by Alkemade in 1962.7 For details see summaries in advertisement pages.*Reprints of this paper will be available shortly. @ SAC; Crown Copyright Reserved. 617618 BROWNER: ATOMIC FLUORESCENCE AS AN [AnaZyst, Vol. 99 The practical analytical potential of atomic-fluorescence spectrometry was first demon- strated by Winefordner and Vickers8 and Winefordner and Staabg in 1964, and later by West10 and by Dagnall, West and Youngll in 1966. From 1964 to the present time, some 200 papers relating to the analytical use of atomic fluorescence have appeared in the world’s literature, compared with a total of about 3200 papers on all aspects of atomic-absorption and atomic-emission flame spectrometry.12 A number of excellent and comprehensive reviews have been p ~ b l i s h e d , l ~ - ~ ~ to which reference should be made for a fuller treatment of many aspects of the subject.It is the intention in this review to provide a sufficient description of the theoretical, instrumental and experimental background to atomic-fluorescence spectrometry to allow the reader to make a reasoned choice between this and other competitive techniques (such as at omic-absorp tion and flame-emission spectrometry) when confronted with a particular problem of analysis. Other atomic spectroscopic techniques (e.g., radiofrequency and micro- wave plasma emission, a.c.and d.c. arc emission and X-ray fluorescence) willnot be considered. PRESENT STATUS OF ATOMIC-FLUORESCENCE SPECTRONETR\T- At the present time, no commercial instrumentation is available for atomic-fluorescence spectrometry, following the demise of a six-channel multiple element instrument .28 Con- sequently, laboratory-constructed equipment is essential for practical work. Single-channel instruments can readily be assembled from selected optical and electronic components. However, the modification of many commercial atomic-absorption spectrometers is un- desirable29 because of poor optical efficiency. The construction of equipment with multiple element capability generally requires a higher degree of electro-mechanical expertise than is necessary for single-channel i n s t r u m e n t ~ ~ ~ y ~ ~ (with the exception of a simple rapid-scanning ~pectrometer~~).A more detailed discussion of instrumental design parameters is given in the section Instrument design for atomic-fluorescence spectrometry. Atomic-absorption spectrometry is so well established as an analytical technique that any competing technique must be able to offer significant advantages, by providing either lower limits of detection or greater speed of information retrieval, in order to become the method of choice. At the same time, simplicity of operation and (relative) freedom from interference must not be sacrificed. Atomic-fluorescence spectrometry offers considerable improvements in detection cap- ability for many elements, compared with conventional atomic absorption, particularly for metals with principal resonance lines that lie below about 320 nm.17 Conversely, there are certain instrumental drawbacks to atomic-fluorescence spectrometry, which are inherent in its nature and which have undoubtedly delayed its wider acceptance as a complementary technique to atomic-absorption spectrometry.The most important drawback is the problem of developing a source that combines the properties of very high radiant output at the relevant wavelength with stability, reproducibility and, above all, ease of operation. Thermostatically controlled electrodeless discharge l a r n p ~ , ~ ~ - ~ ~ , new designs of high-intensity hollow-cathode lamps 36937 and the p ~ l s e d ~ ~ s ~ * and intermittent-p~lsed~~ operation of conventional hollow- cathode lamps have all provided possible solutions to this problem (see the section Sources for atomic-fluorescence spectrometry).Secondary problems have arisen from (photon) shot noise induced by emission from either the flarne4Oy4l (or n~n-flarne)~~ cell, or from intensely emitting species introduced into the flame (or non-flame) cell (e.g., sodium) .29 These problems have led to some difficulties in practical analysis, but it is possible to minimise emission problems by good optical design and the use of low-background separated Scattering of primary radiation has led to erroneous signals, especially when continuum light sources have been used with low-temperature (e.g., air - hydrogen and argon - hydrogen - oxygen) turbulent but scattering is found to be low with high-temperature (e.g., air - acetylene and nitrous oxide - acetylene) pre-mixed laminar flames and narrow-line source^.^*^^^ Also, compensation techniques have been proposed for residual scattering encountered in flame45 and n ~ n - f l a m e ~ ~ analysis of samples with high salt contents. Recent instrumental developments have therefore led to a considerable improvement in the prospect that atomic-fluorescence spectrometry will be more widely used in the future than at present.October, 19741 AXALYTICAL TECHNIQUE. A CRITICAL REVIEW THEORY T Y P E S OF FLUORESCENCE TRANSITIONS- 619 There are fourteen possible discrete processes by which an atom can absorb radiation and re-emit it as fluorescence radiation.47 Of course, it is unusual for all of these possibilities to be open to most atoms under normal experimental conditions, and in practice five of these processes are likely to be most frequently encountered : resonance fluorescence, direct line fluorescence, excited state direct line fluorescence, stepwise line fluorescence and thermally assisted stepwise line fluorescence.These processes are illustrated diagrammatically in Fig. 1, using the consistent nomenclature recommended by Omenetto and Winef~rdner.~’ In 2 1 iki ( a ) 35 ( 6 ) 0 ( C ) =: -0 ( d ) II (el 0 Fig. 1. Major types of atomic-fluorescence transitions : ( a ) , resonance fluorescence ; ( b ) , direct line fluorescence; (c), stepwise line fluorescence; ( d ) , excited-state direct line fluorescence : and ( e ) , thermally assisted stepwise line fluorescence.Energy spacing is arbitrary. Broken lines represent non-radiational (collisional) processes resonance fluorescence [Fig. 1 ( a ) ] , radiational excitation and de-excitation are between the same upper and lower energy levels, and consequently absorbed and emitted radiation are of the same wavelength. In direct line fluorescence [Fig. 1 ( b ) ] , only the upper level is common to the radiational excitation and de-excitation processes, and when the excitation energy is greater than the fluorescence energy (hence Afluorescence > Aebsorption) the process is termed “Stokes.” The less common process, in which the fluorescence energy exceeds the radiational excitation energy, is termed “anti-Stokes” (Afluoreecence < Aabrorptfon).If different upper levels are involved in radiational excitation and de-excitation processes [Fig. 1 ( c ) ] , this phenomenon is termed “stepwise line fluorescence.” If absorption and fluorescence steps involve only excited states (as often occurs when low-lying energy levels of atoms are appreciably populated by thermal or chemi-excitational processes, or both), then the process is said to be “excited” [Fig. 1 (41. Finally, when thermal excitation follows the initial radiational excitation, prior to fluorescence emission [Fig. 1 (e)], the process is termed “therm- ally assisted.’’ Other possible combinations of these processes, together with two-photon excitation, are likely to be significant only with the high photon fluxes produced by laser ex~itation.~8-5~ Sensitised fluorescence, in which an excited species transfers excitational energy to another species, which subsequently emits this energy as atomic fluorescence, has yet to be observed in analytical flame or non-flame atomisers, mainly because of the generally low atomic particle densities (typically 10l2 ~ m - ~ ) .Ionic fluorescence has also been reported for rare-earth elements in the nitrous oxide - acetylene flame, with laser e ~ c i t a t i o n . ~ ~ A process involving triplet - singlet emission, following singlet - singlet excitation, has been observed for cadmium in non-flame cells, and termed atomic phosphore~cence.~3 RADL4NCE EXPRESSIONS I N ATOMIC-FLUORESCENCE SPECTROMETRY- The atomic-fluorescence radiance expressions for low atomic concentrations are relatively simple, and have been well c h a r a c t e r i ~ e d .~ , ~ ~ - ~ ~ ~ ~ ~ ~ ~ ~ The radiance expressions for high620 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, Vol. 99 atomic concentrations are not of analytical interest, as atomic fluorescence is essentially a trace technique. However, the shapes of the analytical curves at high concentration are important and are considered in a later section. There are slight differences between the expressions for narrow-line sources (e.g., hollow- cathode lamps, electrodeless discharge lamps, or vapour discharge lamps operated at low currents) and continuum sources (e.g., xenon arcs or quartz halogen lamps). These differ- ences result from the different source - atomic-absorption profiles.Considering the radiance of the atomic fluorescence, B,, (in erg s-l cm-2 sr-l) emitted at right-angles from a totally illuminated parallelepiped of gas, containing a uniform distribution of atoms, as in Fig. 2, the following simplified expressions can be derived : Line excitation source- Continuum excitation source- where K = atomic-absorption coefficient for Doppler broadening (nm cm2 s-l); n = number of atoms in lower level of absorption transition (cm-3); f = absorption oscillator strength (no units) ; 8 = factor to allow for finite width of line source compared with absorption line (no units); B , = radiance of line source (erg s-l cm-2 sr-1); Y' = fluorescence power yield [(ergs fluoresced per second) per (ergs absorbed per as = solid angle of source radiation reaching atom cell (sr); A , = surface area of fluorescence cell (cm2); Ah, = Doppler half-width of absorption line (nm); Bch = spectral radiance of continuum source (erg s-l cm-2 sr-1 nm-1).second)] ; L,l,L' = fluorescence cell dimensions (cm) ; As atomic-fluorescence radiation is isotropic, a further factor, QF, must be added in order to give the atomic-fluorescence signal observed by a detecting system. Here OF is the solid angle (sr) collected by the detecting optics. -L- Atom cell Fig. 2. Idealised atomic-fluorescence cell. B8, radiance of line source (erg s-I cm-2 sr-l) ; and BF, radiance of atomic fluorescence (erg s-l cm-2 sr-l). L , I and I' are fluorescence cell dimensions (cm) These equations are strictly applicable only to low-power excitation sources, With sources of very high irradiance (e.g., pulsed, tunable dye lasers), account must be taken of possible saturation effects, when the population of the excited state becomes comparable with that of the lower electronic state.Under these circumstances, the fluorescence radiance is no longer proportional to either the source radiance or the fluorescence power yield.56-58 Reference should be made to specialist publications for further details of laser-excited atomic fluorescence .48-52October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 62 1 Aspects of the radiance expressions that are important analytically are as follows. For both line and continuum sources (other than pulsed laser sources), the fluorescence radiance is linearly related to : (i) the number, n, of ground-state atoms; (ii) the radiance, Bs, of the narrow line source; (iii) the spectral radiance, Bch, of the continuum source; (iv) the solid angle of source radiation, a,, focused on to the atom cell; (vi) the power yield of fluorescence, Y’.The radiance expressions must be modified in order to allow for incomplete or non- homogeneous illumination of the atom cell, or for incomplete measurement of fluorescence from the cell. However, these situations should be avoided experimentally (or a reduction in the fluorescence signal may resultz2) and so do not require detailed treatment. (v) the solid angle of fluorescence radiation, QF, received by the detector; ANALYTICAL CURVES IN ATOMIC-FLUORESCENCE SPECTROMETRY- The shapes of analytical curves in atomic-fluorescence spectrometry have received considerable theoretical and have been verified experimentally for mag- n e s i ~ m .~ ~ Typical curves for line and continuum sources (Fig. 3) show the linear relationship that exists between the fluorescence signal and the number of ground-state atoms, at low atomic concentrations. At high atom concentrations, the curve becomes parallel to the concentration axis with continuum sources, and has a slope of -8 for line sources (on a log - log plot). A longer linear range (up to three decades) is usually found with atomic fluorescence than with atomic absorption (up to two decades). With sufficiently intense sources, the curves generally extend to lower concentrations than is possible with atomic absorption.With weak sources, no improvement in linear range may be obtained over atomic ab~orption.~g The potential “double value problem” arising from a convex analytical c ~ r v e ~ ~ ~ ~ can be avoided by either front-surface illumination of the atom ce1129959 or by sample dilution. At trace levels, duality of value in atomic fluorescence with a line source does not occur, and so presents no problem in practice. Both atomic fluorescence with a continuum source and conventional atomic absorption are subject to multiple value problems, but errors can be avoided with proper calibration procedures. (b 1 Slope = 0 Log n Fig. 3. (b) sources. which fluorescence transition arises Typical analytical curves for atomic fluorescence with line (a) and continuum BF, atomic-fluorescence radiance; and n, number of atoms in lower state from QUENCHING PROCESSES- The population of the excited state in atomic fluorescence is in excess of the equilibrium thermal population. Consequently, collisions between excited atoms and molecules present in the atom cell may result in radiationless deactivation of the excited state.This deactiva- tion will diminish the fluorescence radiance by reducing the fluorescence power yield (see the section Radiance expressions in atomic-fluorescence spectrometry). The efficiency of quench- ing, expressed as the quenching cross-section, 0, is unique for each combination of atomic excited state and quenching molecule. Measured values are relatively few,1*6$60-65 but from these values fluorescence power yields can be calculated.622 BROWNER: ATOMIC FLUORESCENCE AS AN Representing fluorescence emission by the process M* --+ M + hv rate A s-I and quenching by IT* + x -+ M + x rate K , TX1 s-1 [Analyst, VOl. 99 ’ - L A the fluorescence power yield is given by where the summations allow for all possible radiational and non-radiational deactivating pathways.24361 ,G6 The high values (Le., low quenching) for argon-diluted stoicheiometric hydrogen - oxygen flames are note- worthy (pure inert gases have no quenching properties) and account for the interest such flames have attracted in atomic-fluorescence spectrometry. Y’ = A/(CA + CK, [XI) Power yields for a few typical flames and atoms are given in Table I. TABLE I ATOMIC-FLUORESCENCE POWER YIELDS FOR SOME TYPICAL ATOMS AND FLAMES* Flame composition 2 H, - 0 2 - 4 N, 6 H, - 0 2 - 4 N, 0.4 C,H, - 0, - 4 N, H2 - 0, - 4 N2 2 H, - 0, - 10 Ar * All values from Jenkins.67 pre-mixed laminar flames.Fluorescence power yield f h I Temperature/K Na K Li T1 Pb 2100 0.066 0.047 0.021 0.070 0-079 1800 0.049 0.049 0-015 0.099 0.10 2200 0.042 0.028 0.017 0.042 0.067 1800 0.75 0-37 0.15 0.33 0.22 1600 0.044 0.03 - 0.051 0.069 Calculated from experimental results with flame-shielded Unfortunately, it is necessary to use air - acetylene and nitrous oxide - acetylene flames in order to obtain good atomisation efficiency and freedom from interference for most e1ements.29~37,40~41~68~6g Practical fluorescence yields are therefore usually much lower than those which can be obtained under ideal circumstances.However, it should be noted that when saturation of the upper state is approached, as with laser excitation, the power yield approaches unity, irrespective of the flame ~sed.~6-~8 INTERFERENCES IN ATOMIC-FLUORESCENCE SPECTROMETRY- Possible interferences in atomic-fluorescence spectrometry are summarised in Table 11. Chemical and physical interferences reduce the number of free atoms available to fluoresce, by interaction with matrix components, and are the major source of interference. However, these interferences are identical with those experienced in atomic absorption and flame atomic emission. Spectral interferences are most likely to occur with continuum sources,70-72 or with multi-element line sources and non-dispersive (see the section Optical design), Scattering interferences are most troublesome with continuum sources and high concen- trations of refractory elements present in the sample matrix.However, non-dispersive optics, without the use of filters to isolate the measurement wavelength, are also prone to this interference. All lines emitted from the source may be scattered and detected, while only resonance lines will give rise to atomic-fluorescence emission.37 In general, atomic fluorescence with a “spectrally pure” line source suffers from similar spectral, physical, chemical and scattering interferences to atomic absorption with a line source. Atomic fluorescence with a continuum source will suffer from more interference in each instance. Atomic flame emission is more prone to spectral interference than atomic fluorescence with a line source, but does not suffer from scattering interference.INSTRUMENT DESIGN FOR ATOMIC-FLUORESCENCE SPECTROMETRY GENERAL- Referring to the expressions for atomic-fluorescence signals given in the section Radiance expressions in atomic-fluorescence spectrometry, it can be seen that the measured signal inOctober, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 623 atomic-fluorescence spectrometry is proportional to : (i) the source radiance (or spectral radiance, for a continuum source); (ii) the solid angle of Source radiance incident on the atom cell; (iii) the solid angle of fluorescence radiance received by the detector; (iv) the power yield of the fluorescence transition. Other parameters, such as the number of ground-state atoms, the profile of the absorbing line and the oscillator strength, are not readily amenable to instrumental improvement and so need not be considered here.TABLE I1 INTERFERENCES I N ATOMIC-FLUORESCENCE SPECTROMETRY Occurrence r \ A Interference Effect on signal Atomic-fluorescence Atomic-fluorescence Spectral Measured signal too Occurs for elements Occurs only i f source element lines in atom cell. Potentially serious systems and “spectrally impure” sources continuum line high present in atom cell lines overlap matrix having lines within spectral band width of isolation device with non-dispersive Chemical Reduction in measured signal, due t o lowering of free metal concentration These effects are identical for both continuum and line sources, and occur as for atomic absorption and atomic emission Scattering of source Measured signal too Potentially serious Less serious than with radiation high scattering occurs continuum.Can occur over full spectral with non-dispersive band width of isolation device flames optics and turbulent Quenching from Reduction in signal, Has not been reported experimentally matrix elements due t o lowering of excited state population The lower working range of atomic-fluorescence spectrometry, as with all other instru- mental techniques, is dictated by the signal to noise ratio (SIN) measured at the read-out. Optimum SIN does not always result from improvements in the optics of the system. Multi- pass optics, for example, may not necessarily lower detection limits. Equally, careful design of source and detector electronics can produce large gains in SIN.As most atomic-fluores- cence spectrometers are laboratory constructed, a discussion of some important design parameters is given below. OPTICAL DESIGN- The signal to noise ratio at the amplifier readout can be represented by73 where k = constant; n = number of solid angles of value Qe gathered by suitable mirrors (no units); t = transmission factor of spectrometric system (no units); H = effective slit height of monochromator (cm); Re = solid angle of fluorescence viewed by monochromator or non-dispersive detector B, = atomic-fluorescence radiance (erg s-l cm-2 sr-1); (sr) ; Ai, = r.m.s. photoanodic noise current due to photon shot effect (A); = r.m.s. photoanodic noise current due to flicker in atomiser background emission (A) ;624 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, Vol. 99 A& = r.m.s.photoanodic noise current due to amplifier read-out (A); Scattering noise, optical aberrations, etc., have been neglected. An optimum SIN is essential in order to obtain low limits of detection in atomic-fluores- cence spectrometry, and can be obtained optically by maximising the following parameters : (i) n, the number of solid angles of fluorescence viewed by the detector, usually with a field mirror. (ii) t, the transmission of the optics, by using either a monochromator with efficient reflecting surfaces and a grating blazed close to the measurement wavelength, or by using a non-dispersive system with or without filters. For a good monochromator, t will be about 0.4.(iii) H , the effective slit height, which should be matched to the height of the primary radiation image in the atom cell so as to prevent excessive background emission and scattering . With mono- chromators, this angle is given approximately by the ratio of the illuminated area of the grating to the focal length of the monochromator. With non-dispersive systems, Q, is given by the area of the focusing lens (or mirror) to its focal length. (v) The slit width, W (of monochromators only), will have an optimum value.73 In practice, slit widths in excess of this optimum value produce little deterioration in SIN. Typical focusing systems are shown in Fig. 4. Systems with lenses, which should be quartz for good transmission of ultraviolet radiation, are relatively cheap and simple to align, but suffer from aberrations.Mirrors do not suffer from chromatic aberration. Undoubtedly, W = slit width (cm). For good interference filters at 250 nm, t will be about 0-2. ( i v ) Qe, the solid angle of fluorescence radiation viewed by the detector. +>source x A Secondary Detector mirror k\ Casse gra in mirrors lipse Scattering source \ I / Fluorescence source \ / - Fig. 4. Focusing optics for atomic fluorescence. Some commonly used systems are illustrated : (a), lens and mirror ~ y s t e r n ~ ~ 9 ~ ~ : ( b ) , elliptical mirror system76; (G), all-mirror ~ystem7~178; (d), inverse Cassegrain system28; and ( e ) , scattering correction system.45 t70 AC, atom cell; D, detectorOctober, 19741 ANALYTICAL TECHNIQUE.A CRITICAL REVIEW 625 the Cassegrainian system produces the greatest optical gain, but is expensive. Recent studies have shown that the gain to be obtained from each mirror is very depend- ent on the total system used, including the e l e ~ t r o n i c s . 7 ~ ~ ~ ~ ~ ~ ~ Generally, the secondary source mirror should approximately double the The field and detector mirrors collect background radiation from the atom cell, in addition to the fluores- cence radiation. If simple d.c. electronics are used, or a.c. electronics (especially tuned to frequencies less than 100 Hz, where l/f noise predominatess0), then flame flicker may become the limiting noise factor. If flame flicker is limiting, field and detector mirrors will produce no improvement in SIN.With synchronous, phase-sensitive (“lock-in”) amplifiers, the effect of high background flames (or non-flame cells) or strongly emitting additives will be to produce an SIN gain pro- portional only to the square root of the optical gain, due to shot-noise limitation. Non- dispersive systems, using an air - acetylene flame, have been reported to achieve little improve- ment in SIN with a simple spherical field mirror.37 An optical arrangement for the correction of background scatter (which may occur with solutions high in involatile components) has recently been d e ~ c r i b e d ~ ~ , ~ ~ [Fig. 4 (e)] . Radiation from an electrodeless discharge lamp and a xenon arc lamp is shone alternately on to a flame via a reflecting chopper. The signals are subtracted electronically with a lock-in amplifier, and higher accuracy is claimed than with an uncorrected system.(at low metal concentrations). MONOCHROMATOR aeYsus NON-DISPERSIVE OPTICS- Both have application to single and multiple-element analysis. Non-dispersive systems, owing to their large optical aperture and their ability to detect fluorescence at more than one wave- length simultaneously, give detection limits that compare well with those for medium-resolu- Some advantages and disadvantages of the two systems are set out in Table 111. TABLE I11 COMPARISON OF MONOCHROMATORS AND NON-DISPERSIVE SYSTEMS I N ATOMIC-FLUORESCENCE SPECTROMETRY System Advantages Monochromator Single-element analysas- Wide wavelength range. Good stray light rejection. Moderately fast optics.Easy selection of element by grating rotation. Ability to use sensitive, wide-range PMTs Multiple-element analysis- As for single-element analysis. Simultaneous analysis possible with Vidicon detector or Multiplex decoding. Rapid sequential analysis by automated grating rotation Non-dispersive optics Single-element analysis- Simple and inexpensive system Freedom from wavelength Good detection limits Multiple-element analysis- drift. As above. Can differentiate between elements with several lock-in amplifiers tuned to different frequencies. wheel can be used for mectral isolation Alternatively, rotating filter Disadvantages Kelatively high cost. Need for wavelength peaking. Possibility of wavelength drift. Transmission varies with wavelength. Smaller solid angle of acceptance than non-dispersive system As for single-element analysis.Complex and expensive equipment necessary for versatile multiple- el emen t analysis Needs solar blind PMT or UV- sensitive PMT and filters. Prone to scattering interference. Possibility of spectral interference As above. UV filters have low transmission626 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, Vol. 99 tion r n ~ n ~ c h r ~ m a t o r ~ . ~ ~ ~ ~ ~ ~ ~ ~ Typical improvements are shown in Table IV. Against this must be set the need to use either a solar blind photomultiplier tube (limiting the useful wavelength region to below 300 nm) or high-cost interference filters, in order to reduce the detected background emission to an acceptable level.25s37,40s82,83 TABLE IV COMPARISON OF DETECTION LIMITS BETWEEN NON-DISPERSIVE Hollow-cathode sources, air - acetylene flame* AND MONOCHROMATOR OPTICS Elemt tit Wavelength /nm Lampf Detection limit§/ng ml-1 r - Non-dispersive11 MonochromatorT/ -1 Separated** Unseparated Separated** Unseparated AS 193.7 I-ICL Au 242.8 HIL(A) Bi 223.1 ; 306.8 HIL(B) ; HCL Cd 228-8 HCL co 240.7 HIL(A) Fe 248.3 HIL(A) Ga 294.4; 41 7.2 HCL 253.7 HCL Hg In 303.9; 451.1 HCL Ir 254.4 HCL Mg 285.2 HIL(B) M n 279.5 HCL Ni 232.0 HIL(A) ; HCL Pb 283.3 HCL Pd 244.8; 340.5 HIL(R) Pt 265.9 HIL(A) Rh 343.5 HCL R u 287.5; 372.8 HCL Sb 217.6 HIL(A) Se 196.0 HCL Sn 284.0; 303.4 HCL Te 214.3 HCL T1 276.8; 377.6 HCL Zn 213.9 HIL(B) * All values taken from Larkins.37 6000 100 250 4 30 3 80 000 30 000 6000 4000 0.2 10 2 1000 2000 300 30 000 80 000 40 6000 3000 3000 10 000 0.3 150 000 900 2500 40 400 40 300 000 300 000 100 000 30 000 0.9 90 10 15 000 5000 - n.d.n.d. 400 90 000 40 000 40 000 100 000 3 n.d. 1000 5000 20 150 50 20 000 50 000 4000 17 000 0.15 30 300 3000 150 30 000 5000 600 50 000 70 000 20 000 2000 3 - n.d. 4000 33 000 60 1000 300 40 000 400 000 8000 600 000 0.6 200 600 20 000 500 70 000 15 000 1500 200 000 20 000 5 - - - t Where two wavelengths are quoted, the first refers to non-dispersive and the second to monochromator The second wavelength is generally more favourable for fluorescence radiance, but not suitable HCL = normal hollow-cathode lamp; HIL(A) = high-intensity lamp, Lowe type36; HIL(B) = high- Lower detection limits n.d. = not detectable a t any concen- detection. for use with solar blind PMT. intensity lamp, Sullivan and Walsh type.84 could be obtained with more intense sources for some elements.tration up to lo6 ng ml-l. 9 Detection limits for aqueous solutions and hollow-cathode sources only. 11 Simple optical system, built around solar-blind PMT (HTV, Type R166). 7 0-5-m Ebert monochromator, f / l O aperture (Techtron, Model M1-A). ** Argon or nitrogen flow of 15 1 min-1 around pre-mixed air - acetylene flame. Concentrated sodium solutions have been demonstrated to give rise to considerable noise with non-dispersive In contrast, a well designed monochromator should have sufficient stray light rejection to avoid this problem. Non-dispersive systems, based on solar blind photomultiplier tubes alone, also require spectrally pure sources, as emission from source impurities may give rise to spurious fluorescence if the impurity element is also present in the a n a l ~ t e .~ ' Before the choice of an experimental system is made, these factors (together with those discussed in Table 111) should be considered carefully. MULTIPLE-ELEMENT SYSTEMS- Atomic-fluorescence spectrometry is adaptable for multiple-element analysis rather more readily than atomic absorption, owing largely to the simpler optics required in the fluorescence technique. Approaches applicable to multiple-element atomic-absorption analysis include the continuum light s o ~ r c e , ~ ~ - ~ ~ which has been used for single-element analysis with rapid wavelength scanning by Fabry Perot dtalonsg or oscillating quartz plate.90October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 627 Both of these methods are readily applicable to multiple-element work, but the sensitivities and detection limits obtainable are inferior to those found with line sources.No practical applications have been reported for these methods. The use of multiple sourcesgI or multiple single sourcesg2 requires complex, expensive and inconvenient optics for spectral isolation. A Vidicon tube, placed at the exit slit of a low-resolution monochromator, has also been proposed for multiple-element atomic-absorption analysis.s3 An approach to multiple- element atomic absorption using resonance monochromators has been described.s4 Multiple-element flame-emission analysis requires either a polychromator of moderate resolution (working spectral band width less than 0.04 nm) and good wavelength stability, a rapid-scanning wavelength drive of great accuracy and reproducibility, or the use of a Vidicon detector.s5 However, background correction presents appreciable problems with rapid scanning or Vidicon detection.The cost and complexity of these systems are considerable, but elimination of any supplementary radiation source is a worthwhile gain over atomic- absorption or atomic-fluorescence multiple-element systems. Multiple-element atomic-fluorescence spectrometers can operate in either simultaneous or sequential modes. Again, fully automated systems for the determination of a large number of elements are expensive, whichever system is selected. An atomic-fluorescence spectrometer for the simultaneous determination of up to six elements has been described in detai1.30931 An improved version of this spectrometer was also marketed briefly, and its operating principles have been discussed.28 The system (Fig.5 ) is based upon the rapid sequential pulsing of six hollow-cathode lamps (100-ms pulses) a t peak currents of up to 190 mA. Coincidentally, a filter disc rotates into the optical path between the flame and detector, isolating the atomic-fluorescence signal from unwanted flame background radiation. The signal for each element is integrated before read-out. Fig, 5. Six-channel, multiple-element atomic-fluorescence spectrometer. (From Demcrs and Mitchell2*) Two other sequential atomic-fluorescence instruments have been described, both based on scanning monochromators.96-98 In the simpler device,98 two dual-element electrodeless discharge lamp sources irradiate a flame continuously, while the monochromator scans rapidly to produce a fluorescence spectrum (Fig.6). In a much more sophisticated, and instrument- ally more complex, approachS6pg7 (Fig. 7), twelve hollow-cathode lamps (arranged in a bank) are pulsed sequentially and intermittentlyss at high currents (up to 210 MA). While each lamp is on, the monochromator is peaked at the appropriate wavelength. A d.c. motor then slews the grating drive to another wavelength setting,loO the next lamp is pulsed, the atomic- fluorescence signal storedlol and the process repeated. The main limitations of multiple-element work are that operating conditions for the flame or non-flame atomiser will be a compromise and will not suit all elements equally, and that the concentrations and sensitivities of the elements detected may differ widely, requiring different gains for each measurement channel.This difference may also lead to fluorescence signals arising from non-linear portions of the analytical curve, which lead to calibration difficulties. However, these problems are less severe with atomic-fluorescence than with atomic-absorption systems.628 Program mab I e sample turntable BROWNER: ATOMIC FLUORESCENCE AS AN - Digital logic or mini - co m p u te r - Teletype [Analyst, VOl. 99 2-port divider n I Amplifier Recorder % X cylindrical microwave cavity IVlo no c h ro mat o r Rapid scan motor Fig. 6.Four-element scanning atomic-fluorescence spectrometer. (From Norris and Westg8) ELECTRONIC DESIGN- The method of electronic detection and signal processing must be appropriate to the system as a whole. Specialised electronics are necessary for some multiple-element spectro- m e t e r ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ in order to follow the pulse sequence of the sources, but in general the choice is between d.c., a.c., synchronous a.c. (lock-in), photon counting and gated detectors (with pulsed sources) .lo2,103 D.c. amplifiers (with continuously operated sources) make maximum use of the signal, but do not discriminate against changing background emission and are susceptible to background flicker n o i ~ e . ~ ~ , ~ ~ Simple narrow-band a.c. amplification, tuned to an a.c.source frequency well away from principal sources of system noise, can power supply Hollow -cathode lamps Burner and optics M Programmable monochromator Printer v Photo- I I I I Fig. 7. Twelve-element rapid sequential atomic-fluorescence spectrometer. (From Malmstadt and Cordos96 9')October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 629 improve SIN, but is inferior to the lock-in system,lo2 which, in addition, discriminates on the basis of phase. Photon counting provides digital information directly and is useful for following weak transient events, such as atomic-fluorescence signals produced with heated rod (or tube) non-flame furnaces. Photon counting also provides some slight improvement in SIN at low signal levels, in comparison with lock-in amplifi~ation,~~~-~O~ particularly if weak sources such as conventional hollow-cathode lamps are used.lo7 However, large signals may cause errors from “pulse pile-up.”lo6 With pulsed source - gated detector systems, either phase-sensitive amplifiers30 or gatedlol or boxcar i n t e g r a t o r ~ ~ ~ J ~ ~ can be used. Pulsed source - gated detector systems give large improvements in S / N , compared with continuous operation of the source at the same average power, provided that the atom cell background emission is high (e.g., with nitrous oxide - acetylene flames).The gain, G, in SIN is given by where ?, = average photodetector current due to atomic fluorescence from pulse source (A); i, = average photodetector current from continuously operated source (A) ; f = repetition rate of pulsed source and gated detector (Hz); t, = pulse width of fluorescence pulses due to pulsed source (s).Hence, for a typical duty factor, ftp, of Lock-in and pulsed source - gated detector systems seem to be most useful practically, with some potential for photon counting. Circuit descriptions for lock-in amplifiers, 7*y109--111 gated d e t e c t o r ~ ~ ~ , ~ ~ ~ ~ O ~ and photon counting equipment112-l14 have been published, and a wide range of equipment is available commercially. Automatic and cross-correlation techniques have been used for the atomic-fluorescence determination of rhodium115 and a 6-fold improvement over lock-in amplification claimed. However, the equipment is expensive. there will be an SIN gain for lo2. SOURCES FOR ATOMIC-FLUORESCENCE SPECTROMETRY- The requirements of sources for atomic-fluorescence spectrometry are : (i) high radiance (see the section Radiance expressions in atomic-fluorescence spectro- (ii) good short and long-term stability; (iiz) reproducible output ; (in) ease of operation; (v) availability for wide range of elements; metry) ; (ui) low cost; (vii) long shelf-life.Sources that have been used successfully to excite atomic fluorescence include vapour dis- charge lamps, electrodeless discharge lamps, hollow-cathode lamps (conventional and high intensity), xenon arc lamps and pulsed lasers. The main types are shown schematically in Fig. 8. Vapow discharge lamps-These lamps were used widely in early analytical work, but have received little attention since, because of the restricted range of elements for which they are available and their tendency to emit broad and self-reversed lines.Thermal stability is also troublesome. Nevertheless, good detection limits have been reported for cad- mium,9,11,21,116-118 zinc,9,75,116,118,119 mercury,9W gallium,7590 indium,75,120 thalli~m,67,75,118,120 sodium and potassium.67 Spectral overlaps have been used in order to excite the atomic fluorescence of iron, manganese, nickel, chromium, thallium, copper and magnesium with a high-pressure mercury lamp.75 Continuum radiation sources-Continuum sources also received considerable attention in early publications, because of the attractive prospect of using one source to determine many elements. Sources used have been 150-W,70,1197121 450-W12s124 and 500-W72 xenon arc lamps, and the elements determined were si1ver,72y119,121 gold,l21 bismuth,121 copper,72,119,121,123,124 calcium,72,122 c a d m i ~ m , 7 ~ 9 ~ ~ ~ ~ h r o m i u m , ~ ~ ~ ~ ~ iron,119,123,124 indium,72 magne- sium,72,119,121,122 manganese,72,114 lead,119,121,123,124 thallium,72,119,121 and zinc.72,119,121 Generally,630 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, VOl.99 detection limits have been poorer than with intense line sources (often less favourable, even, than determination by atomic emission in a premixed air - acetylene and scattering interferences have been Spectral line interferences have also been reported.72 In view of the poor detection limits found, the scattering problems arising from matrix components and the restricted range of elements determinable, continuum sources do not seem to have a wide range of application at present. Quartz window I Mica shield Ceramic insulator Cathode Anode Quartz window 8 C A B Evacuated jacket Heated f i larnen t Anode E I ect rod eless Quartz discharge lamp \ xenon 1 Cathode Quartz window C A B ( C i Heater coil “A” antenna (4 i e i ( f ) Fig.8. Sources for atomic-fluorescence spectrometry. Frequently used sources include : ( a ) , shielded hollow-cathode lamp (for pulsed operation) ; (b) , high-intensity hollow-cathode lamps4 (A = anode, B = booster anode and C = cathode); ( c ) , high-intensity hollow-cathode lamp36 (A = common anode, B = oxide-coated booster anode, C = cylindrical hollow cathode and D = discharge restricting baffles); (d), metal vapour discharge lamp; (e), xenon arc lamp; and (f), thermostatically controlled electrodeless discharge lamp HoZZow-cathode lamps-Conventional hollow-cathode lamps (including the newer shielded cathode design) do not have adequate radiance, when operated in either continuous or modu- lated modes, to produce intense atomic-fluorescence signals with either flame41p42s125 or non-flame126 atomisers.Repetitive pulsing, at peak currents up to 600 mA, has improved pulse - d.c. line ratio intensities by up to 800 x .38 Operation in an intermittent pulsed current mode (peak current typically 200 mA) raises the peak radiance by approximately 1 0 0 ~ compared with d.c. operation of the lamp at the same average current.99 Pulsed source operation is particularly suited to multiple-element work,28t30s31~99 where the pulse sequence may be readily synchron- ised with detecting electronics. High-intensity hollow-cathode lamps of the Sullivan and Walsh ~ a t t e r n * ~ $ l ~ ’ have been widely used in atomic-fluorescence work and provide good detection limits with flame37,68-70s 125,128--132 and non-flame ce11s.42,46,126,133-136 However, with flame cells it should be noted that only detection limits for gold,1s19132 cadmium125 and copper70 were appreciably (3 x or more) better than the best reported values obtained by atomic-absorption ~pectr0metry.l~~ A new style high-intensity lamp [Fig.8 (c)], designed specifically for atomic-fluorescence work,36 produces up to 12 times the radiance of normal high-intensity lamps.This source has been used successfully with non-dispersive atomic-fluorescence a p p a r a t ~ s . ~ ~ , ~ * Unfor- tunately, only conventional hollow-cathode lamps are readily available commercially.October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 631 EZectrodeZess discharge lamps-Electrodeless discharge lamps have a history of analytical use that is closely related to that of atomic-fluorescence spectrometry itself. Early workers found that electrodeless discharge lamps gave very high radiance, whether antenna138--140 or resonant cavity was used. However, problems associated with obtaining a stable, reproducible output (which required considerable operator ski1133y143) held back the widespread acceptance of these sources. A great deal of effort has been applied to improving the operation of electrodeless discharge lamps by modifying tuning stubs on the cavities,144J45 by careful control of the pressure of the lamp fill gas,140s146p147 by optimising the mass of fill material148 and by vacuum jacketing the 1amps.139J44J49 None of these approaches has proved wholly successful, as the vapour pressure of the element within the lamp still remains a very sensitive function of the absorbed microwave p0wer.~5 Recent work has ~ h o ~ n ~ ~ - ~ ~ ~ ~ ~ ~ ~ ~ that most problems involved in the operation of electrodeless discharge lamps are solved by separating the functions of heating and microwave excitation of the discharge, by the use of a flow of pre-heated air around the lamp.A simple device, built around an “A” antenna, has been described,34 which has been found to give better reproducibility and ease of operation than the equivalent thermostatically controlled ah ~ a v i t y .~ 5 A heater system for a $A cavity has also been described.lS2 The thermostatically controlled antenna system has been found to give very satisfactory results in atomic fluorescence. 35 s1509151 9153 9 154 Also, multiple-elemen t elect rodeless discharge lamps can be operated rationally and easily for a wide range of elements.150 Single and multiple electrodeless discharge lamps can be constructed in the laboratory for over fifty elements, and have been used to determine thirty eight of these elements by atomic-fluores- cence spectrometry. Preparative conditions for electrodeless discharge lamps have recently been revie~ed~6P~~~1~56 and reference should be made to these publications for full details.Electrodeless discharge lamps can readily be m ~ d u l a t e d ~ ~ ~ - l ~ ~ and circuits to inject a modulation waveform on to the magnetron output have been de~cribed.~~J5~-160 Pulsed operation of electrodeless discharge lamps has not yet been described, but would seem to have great potential for improving SIN in atomic fluorescence, as has been found with hollow- cathode lamps. Pulsed laser sources-The peak spectral power density available from pulsed tunable dye lasers can be in the range lo7 to lo9 W cm-2 nm-l, which is 1013 to 1015 times greater than the average power of an electrodeless discharge lamp operated continuously. Present-day dye lasers can provide 2 to 5-ns pulses, at a repetition rate of up to 25 Hz.By grating rotation and dye cell interchange, a wavelength region of 360 to 650 nm can be covered, with a spectral band width of between 0.1 and 1.0 nm. Within this wavelength region, good rather than outstanding detection limits have been obtained for a number of element~,~*--52 using either oscilloscope or boxcar integrator read-outs. Major limitations at present are the restricted low wavelength limit, which can be reduced to 265 nm by frequency doubling1619162 (but only at the expense of reduced radiant output), and high unit cost. Costs should decrease considerably in time, but even then it is likely that some major advances in laser technology will occur before these sources are suitable for routine work.A detailed review of pulsed source atomic-fluorescence spectrometry, including the use of dye lasers, has recently been p~blished.~’ ATOM CELLS FOR ATOMIC-FLUORESCENCE SPECTROMETRY- the following properties : In order to obtain full advantage from the fluorescence technique, atom cells should have (i) high atomisation efficiency; (ii) freedom from physical and chemical interferences ; (iii) low background emission at the measurement wavelength ; (iv) good stability (continuous atomisers) ; good reproducibility (single-shot atomisers) ; (v) low temperature, and low concentration of quenching species, in order to maximise Only requirement ( v ) is specific to atomic-fluorescence spectrometry, the others being held in common with atomic-absorption and atomic-emission spectrometry. fluorescence yield ; (vi) long residence time of atoms in the optical path.632 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, VOl.99 Flames-The requirements of good atomisation and little interference generally conflict directly with the need for low background emission, low temperature and minimum con- centrations of quenching species, particularly with flame atomisation. It is generally recognised that for practical analysis, pre-mixed laminar air - acetylene or nitrous oxide - acetylene flames must be used. In order to reduce background emission to a minimum, the secondary reaction zone of the flame is separated with a concentric stream of argon or nitrogen.43,163,164 Argon is preferred to nitrogen as it does not quench fluorescence.Much work has been carried out with hydrogen-based flames, using air or argon - oxygen mixtures as oxidant, but the severe interference^^^^^^^^^,^^^^^^^^^^^ and scattering problems noted44,165 restrict their use to volatile elements contained in essentially matrix-free solutions. Turbulent, or turbulent pre-mixed, flameP6 may produce greater SIN than laminar burners, particularly if observations are made above the luminous flame 20ne.l~~ However, turbulent flames magnify problems of interference and scattering found in the equivalent laminar fl ame and so are of little practical use.44~71~168~169 An excellent discussion of flame composition effects on atomic-fluorescence signals has been p ~ b l i s h e d , ~ ~ $ ~ ~ ~ but unfortunately most of the practical work included in this study refers to hydrogen-based flames.Burner geometry can have an important bearing upon the shapes of analytical curves,5@ but flame stability is the most important consideration. Theoretically, a rectangular geometry is preferred,59,170 but in practice few flames maintain this shape for a distance of more than 1 cm above the burner head. Circular burners, which are the simplest to construct, are therefore most widely used. Non--am cells-Theoretically non-flame cells are nearly ideal atomisers for atomic- fluorescence spectrometry, in that quenching can be reduced to a minimum by use of argon as the surround gas, while good atomisation efficiency is still maintained. However, non-flame cells suffer from particular problems of sampling reproducibility and physical interference, both in atomic-absorption and atomic-fluorescence spectrometry.(i) amplifier and photomultiplier saturation from the very high continuum emission (ii) scatter from matrices with a high inorganic content; (iii) noise from the short amplifier time constants that are necessary in order to follow Background emission and scatter can be minimised by careful optical design and the use of correction technique^.^^^^^ The measurement time constant may increase if a continuous non-flame cell is employed rather than a single-shot device. Continuous non-flame systems with platinuml'l or graphite tubeslo7 have been described, but the results showed no great improvement over discrete-sample devices. A re~iewl7~ of non-flame cells, applicable to atomic-fluorescence work, has been published that contains a full description of each type of cell.Although carbon-tube discrete systems173 and platinum-loop atomiser~l~~" have been used for atomic-fluorescence work, most of the published work refers to carbon-rod atomi~ers.~~?~~6,17~ Absolute detection limits obtained with carbon-rod atomisers are given in Table V, in com- parison with the best published figures for two commercially available carbon-tube atomic- absorption systems. In view of the introductory remarks to this section, it might seem surprising that atomic-fluorescence detection limits are superior (2 x or greater improvement) only for gold (2 x ), cadmium (1000 x ), copper (3 x ), mercury (14 x ), nickel (2 x ) and zinc (3 x ). Results are equivalent for bismuth, magnesium and tin, and inferior for silver (2 x ), cobalt (5 x ), gallium (2.5 x ), manganese (10 x ), lead (2 x ), antimony (30 x ) and thallium (7x).This apparent anomaly is emphasised by reference to detection limits in flames (Table VIII), where atomic fluorescence is seen to be superior to atomic absorption for fourteen elements and equivalent for six elements. This comparison includes four elements common to both tables (silver, cobalt, manganese and thallium), for which fluorescence is superior to atomic absorption in flames, but not with graphite-furnace atomisation. Firstly, the carbon-rod systems produce an atom cloud with a much shorter residence time in the optical path than that of the carbon-tube atomisers; secondly, the combination of short amplifier time constant and high gain necessary to measure small, rapid signals in fluorescence causes the SIN to Specific problems relating to atomic-fluorescence spectrometry are : from the cell; rapid events.The explanation is probably a combination of two factors.October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 633 deteriorate. Improvements, to allow the use of tube systems, or other systems with a longer atom residence time would probably be beneficial. Refinements in continuous non-flame systems may also lead to improvements in performance. Non-flame atomisers still have considerable potential as atom cells for atomic fluorescence, but more development work is necessary. Element Au Bi Cd co c u Ga Ag Hg Mg Mn Ni Pb Sb Sn T1 Zn TABLE V ABSOLUTE DETECTION LIMITS WITH GRAPHITE-FURNACE ATOMISERS I N ATOMIC-ABSORPTION AND ATOMIC-FLUORESCENCE SPECTROMETRY Limit of detection * /pg r A -I Atomic fluorescencef Atomic absorption (1) Atomic absorption (2)s 0.4 (153) 0.211 0.2511 411 (134) 10 8 0.1 0.1 1011 711 1011 (135) 0.311 (153) 7 50 (135) 2011 0-001511 (53) 20 (42) 611 ;I1 5 (135) yll 1 - - 100 0.0611 10 111 (135) 511 (42) 10 10 (135, 153) 511 611 2011 711 (153) - 1000 (42) 3011 20 (153) 311 14 0-06 10011 (153) 6011 0*02]] (135) 0.08 * Detection limits all refer to aqueous solutions and a signal to r.m.s.noise of 2. t Graphite-rod atomisers, 0.5 to 1-p1 samples. $ Varian tube furnace attachment to Model 63 graphite-rod a t ~ m i s e r , ~ ~ ~ 5-p1 sample. Reference numbers are given in parentheses. gignifies detection limit lower, by 2x or more, than other values. More than one symbol erkin-Elmer, Model HGA 72, tube f ~ r n a c e , ~ ' ~ 100-p1 sample.per row indicates essentially equivalent values. SoZid sample atomisation-Two systems directly applicable to solid samples are the arc pulse and the cathodic sputtering chamber.lg0 With the arc pulse atomiser, the sample is placed in the bowl of a wineglass-shaped graphite electrode and vaporised with a current pulse of up to 13 A. Cadmium and manganese in graphite powder and silver in geochemical materials have been determined in this way, with claimed improvements in detection limits over atomic-absorption measurements of between 10 and 100 x . In the sputtering chamber, an interchangeable water-cooled cathode specimen is held in a Pyrex glow-discharge lamp, provided with silica windows.Optimum operating con- ditions are found to be with argon as the fill-gas at 5 torr and an operating current of 35 mA. Sputtering, and hence analytical results, are from the sample surface only. This surface must therefore be thoroughly cleaned in order to remove any oxide film, first by abrasion and then by the application of a reversed-polarity d.c. discharge. Atomic fluorescence of the sputtered vapour, measured 20 mm below the cathode surface (in order to avoid the d.c. negative glow), is excited with high-intensity hollow-cathode lamps. The method has been applied to iron-based alloys, with estimated detection limits as shown in Table VI. A similar system, with simultaneous detection of iron, chromium and nickel in stainless steels, has also been described.180 The sources are modulated at different frequencies, and then three synchronous amplifiers are used to decode the atomic-fluorescence signal from each element. A single R106 photomultiplier with a glass UG5 filter is used as the detector. Reduction - aeration method for mercury-Two systems have been described,l81-184 which involve the use of an atomic-fluorescence measurement in place of the usual atomic-absorption finish in the reduction - aeration method for mercury.ls5Js6 In one of these systems,l81 a rectangular Pyrex cell with Vycor windows is irradiated with a mercury Penlite.A steady signal is obtained by re-circulation of the vapour, and an approximately five-fold improve- ment in detection limit obtained over the atomic-absorption method.A modification of the technique,ls3 with collection of the mercury on a silver wire, followed by heating release, has also been described.634 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, VOl. 99 The atomic-fluorescence signal is measured directly above an open tube in the second The argon stream, containing mercury, passes into the atmosphere and a A sensitivity similar to that obtained with the other atomic- single peak is recorded. fluorescence system is claimed, with a 30 detection limit of 2 x g. TABLE VI PERFORMANCE OF ATOMIC-FLUORESCENCE SPUTTERING CHAMBER Detection limits of several elements in iron matrix* Concentration Signal to noise ratio Estimated detection Element? in iron, per cent.(time constant = 1 s) limit, p.p.m.t Ni 0.12 70 40 Cr 0.08 26 60 Mn 0.15 32 10 Si 0.13 20 140 c u 0.08 0.7 800 * All values taken from Gough, Hannaford and Walsh.180 t High-intensity hollow-cathode lamps (of Lowe types6) used for all elements. Solar-blind PMT (HTV, Type R166) used as detector for Fe, Ni, Mn and Si. Wide-range PMT (HTV, Type R106) plus appropriate interference filters used as detector for Cr and Cu. Based on S/N of 2. Both atomic-fluorescence methods are essentially free from non-specific background absorptionls7 (owing to the nature of the fluorescence technique). Non-specific absorption presents a serious problem with the atomic-absorption method, and leads to high blank readings, especially in the presence of organic solvents (e.g., acetone and benzene).Also, the open design of the atomic-fluorescence cell completely avoids problems of cell fogging, which may arise from condensation of moisture in closed cells. Element Ag As Au Bi Ca Cd c o Cr TABLE VII ATOMIC-FLUORESCENCE APPLICATIONS (BY ELEMENT) Wavelength/ Matrix nm 328.1 338.3 328.1 193.7 242.8 223.1 306.8 422.7 228.8 240.7 357.9 Lubricating oil Lubricating oil Lubricating oil Base oil Ni alloy Mine water A1 alloy Steel Ni alloy Steel Soil extracts Bronze U and Th Graphite Serum Steel Sea water Steel Steel Zn alloy Bronze Sea water Atomisation Detection limit*t Air - H, 0.1 pg ml-l H, - Ar, turbulent 0.4 p.p.m. Carbon-rod atomiser 0.1 p.p.m. Carbon-rod atomiser 0.4 pg Air - C3H8, separated 1.4 p.p.m. (extract) Air - C,H,, separated 0.015 p.p.m.Air - C,H,, separated 30 p.p.m.1 Ar - H, - 0, 0.12 p.p.m. (extract) Air - C3H8, separated 0.8 p.p.m. (extract) Ar - H, - air 0.05 p.p.m.3 Air - C,H, 0.7 p.p.m. (extract) Ar - H, - air 2.5 p.p.m.1 Air - C,H, <0.06 p.p.m. (extract) Arc-pulse atomiser 3 x H, - N,O, pre-mixed, 4 pg; turbulent Air - C,H,, separated 0.5 p.p.m.1 Air - C,H,, separated 0.001 p.p.m. (extract) Ar - H, - air 25 p.p.m.3 Sputtering chamber 0.08% Ar - H, - air 25 p.p.m.$ Ar - H, - air 25 p.p.m.1 Air - C,H,, separated 0.003 p.p.m. (extract) Reference 189 190 191 153 192 131 193 194 192 97 195 97 196 177 197 198 199 97 180 97 97 199October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW TABLE VII-continued Wavelength1 Element nm Matrix Atomisation Detection limit*t c u 324.8 A1 alloy Steel Steel Bronze Zn alloy Soil extracts Blood serum Blood serum Sea water Lubricating oil Lubricating oil Lubricating oil Fuel oil Air - C,H,, separated Ar - H, - air Sputtering chamber Ar - H, - air Ar - H, - air Air - C,H, Air - H,, separated Air - C,H,, separated Air - C,H,, separated Air - H, H, - Ar, turbulent Carbon-rod atomiser 0, - N, - kerosene 30 p.p.m.0.5 p.p.m.1 0.15% 0.5 p.p.m.1 0-5 p.p.m.S 1 p.p.m. (extract) 25 pg 1-1 0.12 pg ml-l 0.0004 p.p.m. (extract) 0.9 p.p.m. 0.1 p.p.m. 0-004 p.p.m. 0.01 pg ml-1 Fe 248.3 Al alloy Air - C,H,, separated 100 p.p.m. Bronze Ar - H, - air 2.5 p.p.m.1 Zn alloy Ar - H, - air 2.5 p.p.m.3 Sea water Air - C,H,, separated 0.003 p.p.m. (extract) Lubricating oil Air - H, 1 pg ml-l Lubricating oil Ar - H, 2 p.p.m.Fuel oil 0, - N, - kerosene 0.04 p.p.m. Mn Ni Pb Se Si J <OV6 ng 0.003 mg 1-I 5 }3 ng Amalgamate on Ag + cold vapour Cold vapour Air - C2H2, separated 8 p.p.m. Ar - H, - air 0.05 p.p.m.3 Ar - H, - air 0-05 p.p.m.1 0.05 p.p.m.$ 0-3 pg ml-' 0-3 p.p.m. 0.4 p.p.m. (extract) Ar - H, - air Air - C2H2 253.7 Rock Sediments Sea water Urine Flour Air Water Rock Sediment 285.2 A1 alloy Zn alloy Urine Orchard leaves Lubricating oil Ar - H, Lubricating oil Ar - H, Soil extracts 279.5 A1 alloy A1 alloy Al alloy Bronze Orchard leaves Soil extracts Sea water Graphite Air - C,H,, separated Ar - H, - air Sputtering chamber Ar - H, - air Ar - H, - air Air - C,H, Air - C,H,, separated Arc-pulse atomiser 60 p.p.m. 1.5 p.p.m.1 o-13y0 1.5 p.p.m.1 1-5 p.p.m.$ 1 p.p.m. (extract) 0.001 p.p.m.(extract) 1 x 10-6% 232.0 A1 alloy Air - C,H,, separated 70 p.p.m. Steel Air - C,H,, separated 1 p.p.m.1 Steel Sputtering chamber 0.12 yo Lubricating oil Air - H, 2 pg ml-l Gas oil Air - C,H,, separated 0.04 p.p.m Blood serum H, - N,O, pre-mixed, 25 pg turbulent 405.8 Ni alloy Air - C,H,, separated 0.3 p.p.m. (extract) Lubricating oil Air - H, 5 pg ml-l Fuel oil 0, - N, - kerosene 0.06 p.p.m. Base oil Carbon-rod atomiser 15 pg Whole blood Carbon-rod atomiser 7-5 pg Urine Carbon-rod atomiser 7-5 pg 196.1 Ni alloy Air - C,H,, separated 0.5 p.p.m. (extract) 204.0 Steel Sputtering chamber 0.08% Steel N,O - C,H,, separated 70 p.p.m. 635 Reference 200 97 180 97 97 195 201 202 199 189 190 191 123, 124 200 97 97 199 189 190 123, 124 183 183 183 183 183 184 181 181 181 200 97 97 97 189 190 195 200 97 180 97 97 195 199 177 200 198 180 189 203 197 192 189 123, 124 153 46 46 192 180 204636 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, Vol.99 TABLE VII-continued Wavelength/ Element nm Matrix Atomisation Detection limit* 7 Reference Sn 303.4 Base oil Carbon-rod atomiser 100 pg 153 Te 214.3 Ni alloy Air - C,H,, separated 0.14 p.p.m. (extract) 192 Zn 213.9 A1 alloy Bronze High-purity Cu Soil extracts Sea water Boiler feed water Orchard leaves Orchard leaves Blood serum Air - C,H2, separated 8 p.p.m. Ar - H, - air 1 p.p.m.$ Air - C2H2 0.1 p.p.m. Air - C,H, 0.7 p.p.m. (extract) Air - C,H,, separated 0.0002 p.p.m. (extract) Air - C,H, 0.0004 pg ml-I Ar - H, - air 1 p.p.m.z - 0-1 pg g-1 Air - C,H, - 200 97 206 195 199 206 97 45 201 * Detection limits refer to the concentration of the element present in the original sample matrix, and Several values are estimates, (marked $), based on quoted Certain procedures involve solvent extraction prior to the final atomisation ; these are marked(extract) .are generally defined on the basis of SIN = 2. detection limits for aqueous solutions, and assuming a 2 per cent. solution of the original sample. PRACTICAL APPLICATIONS Atomic-fluorescence spectrometry is an area in which far more effort has been expended upon instrumental developments and improvements than upon the analysis of samples. The reverse situation applies to atomic-absorption spectrometry, where little genuine innovation has occurred in the last few years (as commented by Walshlss), but a vast number of samples are analysed each year by the technique.Nevertheless, atomic-fluorescence spectrometry has been applied to about twenty elements in a variety of sample matrices. A list of these applications, by element, is given in Table VII, together with the means of atomisation used and the detection limit achieved. Most reported applications are to the determination of minor components in alloys, or to trace impurities in high-purity metals. Approximately equal numbers of applications to environmental, medical and oil samples have been described. ALLOYS AND HIGH-PURITY SAMPLES- Simultaneous analysis, with a six-channel instrument, has been used to determine copper, iron, magnesium, manganese, nickel and zinc in aluminium alloys.200 Following dissolution in acid and removal of silicon as the volatile fluoride, only one dilution step was necessary.Analysis of seven British Chemical Standards and five independently analysed samples gave very rapid results for each element, with good agreement between experimental and published values. Rapid sequential analysis has also been applied to the determination of calcium, chrom- ium, copper and manganese in steels, zinc, cadmium, iron, manganese, copper, chromium and calcium in bronzes and cadmium, iron, magnesium, copper, chromium and calcium in zinc alloys.g7 The automated spectrometer gave a precision of 0.5 to 1 per cent. under optimum conditions. Major elements were determined (e.g., zinc at the 37 per cent. level) as well as trace components (e.g., magnesium at the 4 x per cent.level). Good agree- ment was obtained with materials analysed by the U.S. National Bureau of Standards (NBS). The cathodic sputtering chamberl8O has been used for both simultaneous and single- element determinations of British Chemical Standard (BCS) iron-base alloys. Iron, chromium and nickel were determined simultaneously, and nickel, chromium, copper, manganese and silicon singly, with a reproducibility of *1 per cent. The sample was analysed as a cleaned disc, the sputtered area being approximately 64 mm2. Eight standard aluminium alloys yielded good correlation between the atomic-fluorescence and spectro- photometric results.lg3 The problem of scattered radiation was avoided by using the direct line fluorescence at 302.5 nm.A relatively lengthy masking and extraction procedure was used for preparation of the steel samples,lg* followed by atomisation into a turbulent argon - hydrogen - oxygen flame. The atomic-fluorescence method was shown to be five times more Multiple-element techniques are particularly suited to alloy analysis. Bismuth has been determined in aluminium alloyslg3 and steels.lg4October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 637 sensitive than atomic-absorption measurements on the same extracted solutions. Solvent extraction was also used for sample preparation in the determination of lead in BCS steels, and arsenic, bismuth, lead, selenium and tellurium in nickel-base al10ys.l~~ Again, good correlation was obtained between analytical and reference results, when these were available.Dual-element electrodeless discharge lamps were used in the rapid sequential deter- mination of cobalt and nickel in a wide range of BCS steels. Standard deviations of 3.5 and 2.0 per cent., respectively, were found for a representative steel sample. Atomic fluorescence has also been applied to the determination of silicon in low-alloy steels with good accuracy and high sensitivity.204 The high sensitivity that can be obtained with atomic-fluorescence spectrometry has been used to good effect in the determination of very low levels of zinc in high-purity c0pper,~05 cadmium in uranium and thoriumlg6 and cadmium in graphite.177 ENVIRONMENTAL SAMPLES- Most work under this heading relates to either the simultaneous determination of several elements in soil, water and leaf samples, or the determination of mercury by various flameless atomic-fluorescence techniques.Calcium, copper, magnesium, manganese and zinc have been determined with a rapid, simultaneous spectrometer in a number of soil extracts.lg5 Again, the precision and accuracy obtained with this equipment are reported to be excellent. The same instrumentation has also been adapted for use in conjunction with an automated pre-concentration - solvent extraction procedure.lg9 Sea-water samples were analysed for cobalt, chromium, copper, iron, manganese and zinc. The improvement in detection limit resulting from solvent extraction varied from two (chromium) to ten fold (copper and zinc). An analysis rate of twenty-five samples per hour was possible.Rapid sequential multiple-element analysis has been used for the determination of zinc, iron, manganese, magnesium and calcium in NBS Standard Reference Material orchard leaves.97 Very good correlation was obtained over a wide concentration range in the original sample, from zinc at 28 mg g-l to calcium at 2.09 per cent. The importance of incorporating a highly accurate automatic pipette-diluter in the system, for the major elements, is emphasised. Flameless atomisation is particularly suitable for the atomic-fluorescence determination of very low levels of mercury in environmental samples (see the section Atom cells for atomic- fluorescence spectrometry). Mercury has been determined in rocks, sediments, sea water and flour using a reduction - aeration system for solution samples and a tube furnace for solid ~amples.1~~ After removal of volatile sulphides, the mercury was trapped in a silver amal- gamation tube, and final generation of mercury vapour as a “spike” resulted from electrical heating of the amalgam.Amounts as low as 0.6 ng of mercury could be quantitatively determined, with good agreement between results from the atomic-fluorescence technique and reference, or independently analysed, samples. A reduction - aeration method has also been used to determine mercury in air, with simple apparatus, to a limit of 0.003 mg m-3.184 Gold has been determined in mine waters131 and zinc in boiler feed waters.206 Direct analysis of mine water, with atomisation in a hydrogen - oxygen - argon flame, was possible at the 0.045 p.p.m.level. After solvent extraction, it was possible to determine down to 10-5 p.p.m. of gold. The atomic-fluorescence determination of zinc in feed waters to high- pressure boilers was preferred to atomic absorption, because the method was simpler and lower detection limits were possible. Atomic fluorescence has also been used to obtain reference analysis data for zinc in orchard leaves, using a scatter-corrected ~pectrometer.~~ OILS- involved the direct nebulisation of the oil into a turbulent hydrogen - argon - entrained air flame. This approach provided satisfactory results for silver, copper, iron, magnesium, nickel and lead, but not for aluminium, chromium, tin or titanium. Sample dilution with isobutyl methyl ketone and spraying into a pre-mixed air - hydrogen flame were not considered to provide any advantage over atomic-absorption methods.The direct atomic-fluorescence determination of copper, iron and lead in kerosene124 was The accuracy of both methods was similar. Early work on the determination of wear metals in lubricating638 BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, VOl. 99 considered to offer a marked advantage in analysis time, compared with the usual extraction and colorimetric procedures. A special burner allowed the kerosene to act as sole fuel, with an argon - oxygen oxidant mixture. A 450-W xenon arc lamp source was used for all three elements. Better detection limits were obtained for the kerosene solutions (copper, 0.004 p.p.m.; iron, 0.04 p.p.m.; lead, 0.06 p.p.m.) than for aqueous solutions (copper, 0.04 p.p.m.; iron, 0-16 p.p.m.; lead, 0-16 p.p.m.). A specially designed spectrometer, for the measurement of combined atomic fluorescence and flame emission from wear metals in lube oils, has been described.lgO When used for the determination of silver, copper, chromium, iron, magnesium, nickel and lead, satisfactory results were obtained for most iron, copper, silver and magnesium samples.The use of a collimated 150-W xenon arc lamp, turbulent flame and d.c. electrometer probably contributed to the difficulties experienced with lead, chromium and nickel. Nickel has been determined in gas oils and petroleum residues after dilutions with xylene of 4 to 10, and 10 to 500 fold, respectively.203 The method of additions was used and, with a separated air - acetylene flame, interference from scattering was found to be negligible.This result was in contrast with the situation in atomic-absorption determinations, in which scattering and non-specific molecular absorption were considered to present problems. Silver, lead and tin have been determined in base and silver and copper simultaneously in the same using a graphite-rod atomiser and thermostatically controlled multiple-element electrodeless discharge lamps. Good sensitivity and long linear analytical working ranges were obtained. A comparison was given between absolute and concentration detection limits for several elements in flame and non-flame atomisers. MEDICAL- Copper, zinc and magnesium have been determined simultaneously in a urine sample, using the rapid sequential spe~trometer.~~ As with previous applications of this apparatus, satisfactory correlation between experimental and reference results was obtained.Copper and zinc have also been determined directly in blood serum, after 25-fold dilution with water.201,202 Calibration graphs prepared from aqueous reference standards were adequate. Lead in whole blood and urine has been determined with a carbon-rod atomiser and hollow-cathode sources.46 In blood samples, the procedure involved 2.5-fold dilution with water, ashing at low temperature and measurement of the direct line fluorescence at 405.8 nm. A UG 5 filter was used with the lamp in order to reduce the source intensity at 405.8 nm to approximately 1 per cent. of its unfiltered value, and so effectively eliminate scattering inter- ference.With urine samples, the lower lead levels resulted in this scattering correction being inadequate, and it was necessary to alternate argon and nitrogen as surround gases. Sub- traction of the scatter signal obtained with nitrogen from the combined fluorescence and scatter signal obtained with argon provided good scatter correction. For both blood and urine samples, comparison with aqueous standards was adequate and standard additions were unnecessary. COMPARISON BETWEEN ATOMIC-FLUORESCENCE, ATOMIC-ABSORPTION AND ATOMIC-EMISSION A comparison between the three techniques, based upon the application of theoretical principles, has been published17 and should be consulted for a fuller treatment. In this section, discussion will be limited to the features of each technique that might lead to its selection for a particular analytical task.A few applications of atomic fluorescence to medical analysis have been reported. FLAME ANALYSIS The choice of technique will be based on the following considerations: (i) concentration of the element to be determined (in sample matrix); (ii) matrix composition, and concentration of matrix elements present in final solution, (iii) accuracy and precision required ; (iv) sample throughput required; ( v ) availability and cost of equipment; ( v i ) operator skill required, and availability. at the dilution necessary for analytical measurement ; All of these factors are interdependent to some degree, and the selection of the most appro- priate technique will depend to a considerable extent upon the weight given to each.C Ytober, 19741 ANALYTICAL TECHNIQUE.A CRITICAL REVIEW 639 In order to obtain good precision and accuracy, it is necessary to work with a final solutlm that has a concentration of the element to be determined at least ten times the detect1 3n limit shown in Table VIII. Therefore, if very low concentrations are to be measured for the 'ourteen elements of Group 1, atomic fluorescence would be the obvious choice. Similarly, atomic absorption would be chosen for the eight elements of Group 4 and atomic emission fol the eleven elements of Group 6. (The comparison includes only those elements for which atc\mic-fluorescence detection limits have been reported in flames.) For the six elements in Gi->up 2 and the two elements in Group 3, atomic fluorescence would be given equal weight wi,b atomic absorption and atomic emission, respectively.The choice 0; the most sensitive technique may also be made in order to enable greater dilutions to be used thus minimising chemical or physical interferences. Alternatively, this choice may also avoid the need for a pre-concentration step, with solvent extraction. TABLE VIII LILTITS OF DETECTION BY ATOMIC FLAME METHODS Limits of detection*/ng ml-l , Element Wavelength7 /nm Flame emission Atomic absorption Atomic fluorescence$§ (176) Group 1 : Atomic fluorescence superior to both atomic absorption and atomic emission- A€! 328.1 8 (207) 2 0.1 (139) Bi 223.1 2000 (207) 25 5 (213) Cd 326-1; 228-8; 228.8 800 (207) 2 0.001 (139) Ce 569.7; -; 394.2 10000 (207) n.d.5009 (51) co 345.4; 240.7; 240.7 30 (207) 10 5 (30, 214) c u 327.4; 324.7; 324.7 10 (207, 208) 1 1 (16, 167, 215) Ge 265.2 400 (207) 200 100 (51) 403.1; 279-5; 279.5 5 (208) 2 1 (216) Mn s c 402.0; 391.2; 402.2 30 (208) 20 10 (52) Te 238.3; 214.3; 214.3 20000 (207) 50 5 (213) T1 377-6; 276-3; 377.6 20 (207, 209) 30 8 (139) Zn 213.8 10 000 (207) 1 0.02 (139) Au 267.6; 242.8; 267.6 500 (208) 10 5 (92) 253.7 10 000 (207) 250 0.2 (81) Hg Group 2 : Atomic fluorescence = atomic absorption; both superior to atomic emission- As 235-0; 193.7; 193.7 10 000 (207) 100 100 (217) F e 372.0; 248-3; 248.3 30 (207) 5 8 (213) Ni 341-5; 232.0; 232.0 20 (207) 2 3 (69) Pb 405.8; 283.3; 405.8 200 (207) 10 10 (70, 74, 218) Sb 259.8; 217.5; 231.1 600 (207) 40 30 (219) Se 196.0 100 000 (207) 50 40 (213) Ga 417.2; 287.4; 417.2 10 (209) 50 10 (213) Sm 488.4; 429.7; 373.9 200 (207, 210) 2000 150s (51) Group 3 : Atomic fluovescence = atomic emission; both sufierior to atomic absorption- Group 4 : Atomic absorption superior to both atomic fluorescence and atomic emissiopa- Be 234.9 1000 (207) 2 10 (220, 221) Mg Hf 368.2; 307.3; 377.8 20 000 (207) 2000 100000 (51) Ho 410.4; 416.3; 410.4 20 (210) 10 150 (51) 285.2 5 (211) 0.1 1 (30, 136) 390.3; 313.3; 313-3 100 (208) 20 500 (222) Mo Pd 363.5; 274.6; 340-5 50 (207, 208) 20 40 (137) Si 251.6; 251.6; 204.0 3000 (207) 20 600 (222) Sn 284.0; 224.6; 303.4 100 (207) 10 50 (223) DY 421.2; 421.2; 418.7 50 (207) 50 500 151) Er 400.8 40 (210) 40 500 (51) Lu 331-2; 331.2; 513.5 1000 (207, 210) 700 3000 (5.) Rh 369.2; 343.5; 369.2 20 (212) 20 150 (61) Sm 488.4; 429.7; 429.7 200 (207, 210) 200 1505 (51) Group 5 : Atomic absorption = atomic emission; both superior to atomic fluor~scence- Group 6 : A tomic emission superior to both atomic absorption and atomic fluorescence- A1 396.2 5 (209) 20 100 (222) Ca 422.7 0.1 (199) 0.5 20 (139) E 11 459.4; 459.4; 462.7 0.5 (207) 20 20 (51)640 BROWNER: ATOMIC FLUORESCENCE AS AN TABLE VIII-continued [Analyst, VOl.99 Limits of detection*/ng ml-l I A \ Element Wavelengtht/nm Flame emission Atomic absorption Atomic fluorescencef 9 (176) Gd In Nd Pr Ru Sr V Yb 434-6; 368.4; 368.4 451.1; 303.9; 451.1 488.4; 463-4; 489.7 494.0; 495.1 ; 495-1 372.8; 349.9; 350.0 460.7 437.9; 318.4; 318.4 398.8; 398.8; 346.4 600 (207) 2 (209) 2 (208) 20 (162) 10 (208) 2 (208) 200 (210) 0.1 (211) 1200 800 (51) 20 100 (139) 1000 2000 (51) 5000 3000 (51) 70 500 (130) 2 30 (139) 4 70 (222) 5 10 (51) Group 7 : Atomic fluorescence = atomic absorption = atomic emission- Cr 425.4; 357.9; 357.9 4 (207) 3 5 (224) Nb 405.9; 405.9; 408.0 1000 (207, 208) 1000 1500 (51) Tb 432.7; 432.7; 433-8 400 (210) 600 500 (51) * All detection limits are for the direct determination of elements in aqueous solution, using mono- chromators and flame atomisers. Detection limits generally defined on a basis of SIN (r.m.s.) = 2.Reference numbers are given in parentheses. ?Where more than one wavelength is given, the values refer in order to flame emission, atomic absorption and atomic fluorescence.$ References 51 and 52 describe the use of a dye laser as source; all other references describe non-laser sources. = Ionic fluorescence. It is possible to obtain lower concentration detection limits, and very much lower absolute detection limits for many elements, by using a non-flame graphite or metal atomiser in place of a flame. By this means, non-flame atomic-absorption limits may equal or surpass flame atomic-fluorescence detection limits, and may make it possible to achieve the desired analysis without the need to use atomic fluorescence. However, problems of interference are inevitably more severe with non-flame atomisation in all but the simplest matrices, and the flame will certainly remain the most useful atomisation cell for some time to come. Of course, atomic fluorescence itself is also very useful with non-flame atomisers (see the section Atom cells for atomic-fluorescence spectrometry).The number of samples to be analysed, and the number of elements to be determined in each, will indicate whether a single-element manual technique, an automated single-element technique or a multiple-element technique is most appropriate. Atomic fluorescence is best suited to the last of these techniques, Practical experience indicates that atomic emission is more demanding of operator skill than atomic absorption or atomic fluorescence, and it is probably easier to obtain an erroneous result with a complex matrix in emission than with the other techniques. Emission tends to be used largely for the alkali metals, the alkaline earths and the rare-earth metals, although atomic absorption is probably used as widely as emission for the determination of the alkali and alkaline earth metals.At the present time, no equipment is available commercially which has been designed specifically for atomic fluorescence use. Although it is possible to modify single-beam atomic-absorption spectrometers for atomic fluorescence use, the results are less favourable than those obtained with more carefully chosen optical and electronic components (as dis- cussed in the section Instrument design for atomic-fluorescence spectrometry). Hence, for single-element atomic-fluorescence work, the spectrometer must be assembled from individual components, and for simultaneous multiple-element work a major constructional project is necessary.Atomic-absorption spectrometry is the method of choice for a wide variety of trace- element analyses. It possesses the capability for determination of up to sixty-five elements with good precision (1 per cent. or better) at levels down to (and often well below) 1 p.p.m. Both atomic fluorescence and atomic emission occupy a position complementary to atomic absorption. Atomic fluorescence is especially valuable for the determination of very low levels of certain elements and for multiple-element determinations. It has the further advantage (in common with atomic emission) of possessing a long linear working range (generally ten times longer than for atomic absorption).October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 641 FUTURE DEVELOPMENTS IN ATOMIC-FLUORESCENCE SPECTROMETRY A number of the atomic-fluorescence systems described in this review will undoubtedly undergo further development and improvement in the future.For example, the graphite and tantalum heated atomisers that have been used for most non-flame work will probably undergo modification in order to ensure a longer atom residence time in the optical path. Work on continuous non-flame atomisers could also improve detection limits in atomic fluorescence, by allowing longer measurement time constants to be used. Modifications of the atomic-fluorescence sputtering chamber could permit its use for solid samples other than iron alloys. It is also likely that further systems for multiple-element analysis will be developed.However, the areas that are most likely to see major developments are in the further development of pulsed, high-radiance sources and in the application of novel signal processing techniques. The work with pulsed, tunable dye lasers has shown the great potential that these sources possess for single or multiple-element analyses. The present limitations of low output below 360 nm and high cost must, however, be overcome before these sources are used more widely. Electrodeless discharge lamps, operated in a thermostatically controlled mode for high stability, should also be investigated as pulsed sources. The advantage to be obtained from pulsed-source operation, used in conjunction with gated detectors, has been discussed in the section Electronic design. As yet, relatively little attention has been paid to signal processing in atomic-fluorescence spectrometry.Generally, lock-in amplification has been used, although other techniques, such as photon counting, have been examined more recently. However, great potential exists for the application of multiplex techniques to the decoding of analytical information, particularly for multiple-element analysis. The 1. 2. 3. 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. author thanks the Government Chemist for permission to publish this paper. REFERENCES Wood, R. W., Phil. Mag., 1905, 10, 513. Nichols, E. L.. and Howes, H. L., Phys. Rev., 1924, 23, 472. Badger, R. M., 2. Plays., 1929, 55, 56. Mankopff, R., Vevh. dt. phys.Ges., 1933, 14, 16. Mitchell, A. C. G., and Zemansky, M. W., “Resonance Radiation and Excited Atoms,” University Pringsheim, P., “Fluorescence and Phosphorescence,” Interscience Publishers, New York, 1949. Alkemade, C. Th. J., in Lippincott, E. R., and Margoshes, M., Editors, “Proceedings of the Xth Winefordner, J . D., and Vickers, T. J., AnaEyt. Chem., 1964, 36, 161. Winefordner, J . D., and Staab, R. A., Ibid., 1964, 36, 165. West, T. S., Analyst, 1966, 91, 69. Dagnall, R. M., West, T. S., and Young, P., Talanta, 1966, 13, 803. Slavin, S., Atom. Absorption Newsl., 1973, 12, 77. Winefordner, J. D., and Mansfield, J. M., in Guilbault, G. G., Editor, “Fluorescence,” Marcel Ellis, D. W., and Demers, D. R., Adv. Chem. Ser., 1968, 73, 326. Price, W. J., in Browning, D.R., Editor, “Spectroscopy,” McGraw-Hill, New York, 1969. Champy, J., Metk. Phys. Autalysis, 1969, 5, 311. Winefordner, J. D., Svoboda, V., and Cline, L. J., CRC Crit. Rev. Analyt. Chem., 1970, 1, 233. Smith, R., in Winefordner, J. D., Editor, “Spectrochemical Methods of Analysis,” John Wiley Cresser, M. S., and Keliher, P. N., Amer. Lab., 1970, August, 8. Winefordner, J . D., and Smith, R., in Mavrodineanu, R., Editor, “Analytical Flame Spectro- metry, Selected Topics,” Macmillan, London, 1970. West, T. S., Minerals Sci. Eng., 1970, 2, 31. Alkemade, C. Th. J., Pure Appl. Chem., 1970, 23, 73. Winefordner, J . D., and Elser, R. C., Analyt. Chem., 1971, 43 (2)) 24A. Winefordner, J. D., Schulman, S. G., and O’Haver, T. C., “Luminescence Spectrometry in Analytical Chemistry,” Interscience Publishers, New York, 1972.Kirkbright, G. F., and West, T. S., Chewy. Brit., 1972, 8, 428. West, T. S., and Cresser, M. S., Appl. Spectrosc. Rev., 1973, 7, 79. Omenetto, N., Fraser, L. M., and Winefordner, J. D., Ibid., 1973, 7, 147. Demers, D. R., and Mitchell, D. G., in “Advances in Automated Analysis,” Halos and Associates, Barnett, W. B., and Kahn, H. L., Analyt. Chem., 1972, 44, 935. Mitchell, D. G., and Johansson, A., Spectrochim. Ada, 1970, 25B, 175. Press, Cambridge, 1972. Colloquium Spectroscopicum Internationale,” Spartan Books, Washington, D.C., 1963. Dekker, New York; Arnold, London, 1967. & Sons, New York, 1970. Miami, Fla., 1970.BROWNER: ATOMIC FLUORESCENCE AS AN [Analyst, VOl. 99 -- , Ibid., 1971, 26B, 677. West,’T.S., Pure Appl. Chem., 1971, 26, 47. Browner, R. F., Rietta, M. E., and Winefordner, J. D., Pittsb. Conf. Analyt. Chem. Appl. Browner, R. F., Patel, B. M., Glenn, T. H., Rietta, M. E., and Winefordner, J. D., Spectrosc. Browner, R. F., and Winefordner, J . D., Spectrochim. Acta, 1973, 28B, 263. Lowe, R. M., Ibid., 1971, 26B, 201. Larkins, P. L., Ibid., 1971, 26B, 477. Dawson, J . B. and Ellis, D. J.. Ibid., 1967, 23A, 565. Cordos, E., and Malmstadt, H. V., Analyt. Chem., 1973, 45, 27. Larkins, P. L., and Willis, J . B., Spectrochim. Acta, 1971, 26B, 491. Browner, R. F., and Manning, D. C., Analyt. Chem., 1972, 44, 843. Alger, D., Anderson, R. G., Maines, I. S., and West, T. S., Analytzca Chim. Acta, 1971, 57, 271. Kirkbright, G. F., and West, T. S., Appl. Opt., 1968, 7, 1305.Omenetto, N., Hart, L. P., and Winefordner, J. D., Appl. Spectrosc., 1972, 26, 612. Rains, T. C., Epstein, M. S., and Menis, O., Proc. Colloquium Spectroscopicurn Internationale Amos, J . D., Bennett, P. A., Brodie, K. G., Lung, P. W. Y . , and MatouSek, J. P., Analyt. Chem., Omenetto, N., and Winefordner, J . D., Appl. Spectrosc., 1972, 26, 555. Denton, &I. B., and Malmstadt, H. V., Apfil. Phys. Lett., 1971, 18, 486. Fraser, L. M., and Winefordner, J. D., Analyt. Chem., 1971, 43, 1693. ~- , Ibid., 1972, 44, 1444. Orne&tto, N., Hatch, N. N., Fraser, L. M., and Winefordner, J . D., Ibid., 1973, 45, 195. ---- , Spectrochim. Acta, 1973, 28B, 65. Aide;, J . F.,’and West, T. S., Analytica Cham. Acta, 1970, 51, 365. Hooymayers, H. P., Spectrochim. Acia, 1968, 23B, 567.Zeegers, P. J. Th., and Winefordner, J. D., Ibid., 1971, 26B, 161. Picpmeier, E. H., Ibid., 1972, 27B 431. -, Ibid., 1972, 27B, 445. Omenetto, N., Benetti, P., Hart, L. P., Winefordner. J, D., and Alkemade, C. Th. J., Ibid., 1973, Svoboda, V., Browner, R. F., and Winefordner, J. D., Appl. Spectrosc., 1972, 26, 505. Hooymayers, H. P., and Rlkemade, C. Th. J . , J . Quant. Spectrosc. Radiative Transfer, 1964,6, 501. Jenkins, D. R., PYOC. R. SOC., 1966, A293, 493. -, Ibid., 1968, A303, 453. ---, Ibid., 1968, A303, 467. -, Ibid., 1968, A306, 413. Hooymayers, H. P., and Lijnse, P. L., J . Quant. Spectrosc. Radiative Transfer, 1969, 9, 995. McCarthy, \V. J., Parsons, M. L., and Winefordner, J . D., Spectvochim. Acta, 1967, 23B, 25. Jenkins, D.R., Ibid., 1970, 25B, 47. Armentrout, D. W., Analyt. Chem., 1966, 30, 1235. MatouSek, J., and Sychra, V., Ibid., 1969, 41, 618. Manning, D. C., and Heneage, P., Atom. Absorption Newsl., 1967, 6, 124. Demers, D. R., and Ellis, D. W., Analyt. Chem., 1968, 40, 860. Cresser, M. S., and West, T. S., Spectrochim. Acta, 1970, 25B, 61. Winefordner, J. D., Parsons, M. L., Mansfield, J . M., and McCarthy, W. J., Analyt. Chem., 1967, Browner, R. F., Dagnall, R. M., and West, T. S., Analytica Chirn. Acta, 1970, 50, 375. Omenetto, N., and Rossi, G., Ibid., 1968, 40, 195. Shull, M., and Winefordner, J. D., Analyt. Chem., 1971, 43, 799. Benetti, P., Omenetto, N., and Rossi, G., AppZ. Spectvosc., 1971, 25, 57. Hubbard, D. P., and Michel, R. G., Awalytica Chim. Acta, 1973, 67, 55.Rains, T. C., Epstein, M. S., and Menis, O., Analyt. Chem., 1974, 46, 207. Alkemade, C. Th. J., Hooymayers, H. P., Lijnse, P. L., and Vierbergen, T. J. M. J., Spectrochim. Warr, P. D., Talanta, 1970, 17, 543. Elser, R. C., and Winefordner, J. D., A$@. Spectrosc., 1971, 25, 345. Vickers, T. J., and Vaught, R. M., ABaZyt. Chem., 1969, 41, 1476. Sullivan, J. V., and Walsh, A., Spectvochim. Acla, 1965, 21, 721. Fassel, V. A,, Mossotti, V. G., Grossman, W. E. L., and Kniseley, R. N., Ibid., 1966, 22, 347. Gibson, J . H., Grossman, W. E. L., and Cooke, W. D., Appl. Spectvosc., 1962, 16, 47. De Galan, L., McGee, W. W., and Winefordner, J. D., Analytica Chim. Acta, 1967, 37, 436. McGee, W. W., and Winefordner, J . D., Ibid., 1967, 37, 429. Nitis, G. J., Svoboda, V,, and Winefordner, J. D., Spectrochim.Acta, 1972, 27B, 345. Elser, R. C., and Winefordner, J. D., Analyt. Chem., 1972, 44, 698. Mavrodineanu, R., and Hughes, R. C., A$$. Opt., 1968, 7, 1281. Silvester, RI. D., Koop, D. J., and Barringer, A. R., 4th I n t . Conf. Atom. Spectrosc., Toronto, Aldous, I<. M., Mitchell, D. G., and Jackson, K. W., 4th I n t . Conf. Atom. Spectrosc., Toronto, Spectrosc., Cleveland, Ohio, 1972, Paper No. 136. Lett., 1972, 5, 311. X V I I , Florence, 1973, 3, 101. 1971, 43, 211. 28B, 289. 39, 436. Acta, 1972, 27B, 149. 1973. Paper TO31. 1973. Paper WC6. 31. 32. 33. 31. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 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.October, 19741 ANALYTICAL TECHNIQUE. A CRITICAL REVIEW 643 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. Sullivan, J. V., and Walsh, A., Appl. Opt., 1968, 7, 1271. Knapp, D. O., Omenetto, N., Plankey, F., Hart, L., and Winefordner, J. D., Analytica Chim. Malmstadt, H. V., and Cordos, E., Pittsb. Conf. Analyt. Chcm. Appl. Spectrosc., Cleveland, -- , Amer. Lab., 1972, 4 (8), 35. Nor& J. D., and West, T. S., Analyt. Chcm., 1973, 45, 226.Cordos, E., and Malmstadt, H. V., Ibid., 1972, 44, 2407. _ _ _ - , Ibid., 1973, 45, 425. - _ _ , Ibid., 2972, 44, 2277. Hieftje, G. M., Ibad., 1972, 44 (6), 81A. -, Ibid., 1972, 44 (7), 69A. Rolfe, J., and Moore, S. E., Appl. Opt., 1970, 9, 63. Robben, F., Ibid., 1971, 10, 776. Dawson, J . B., Meth. Phys. Analysis, 1971, 7, 10. Murphy, M. K., Clyburn, S. A., and Veillon, C., Analyt. Chcm., 1973, 45, 1468. Weide, J. O., and Parsons, M. L., Analyt. Lett., 1972, 5, 363. O’Haver, T. C., J . Chcm. Educ., 1972, 49 (3), A131. -, Ibid., 1972, 49 (41, A211. Caplan, L. C., and Stern, R., Rev. Scient. Instvum., 1971, 42, 689. Eather, R. H., and Reasoner, D. L., A9pZ. Opt., 1969, 8, 227. Alger, D., Dagnall, R. M., Sharp, B. L., and West, T. S., Analytica Chim. Acta, 1971, 57, 1.Dagnall, R. M., Sharp, B. L., aad West, T. S., Nature Phys. Sci., 1972, 235, 65. Hieftje, G. M., Bystroff, R. I., and Lim, R., Analyt. Chem., 1973, 45, 253. Mansfield, J. M., Winefordner, J. D., and Veillon, C., Ibzd., 1965, 37, 1049. Bratzel, M. P., Mansfield, J . M., and Winefordner, J . D., Analytica Chim. Acta, 1967, 39, 394. Winefordner, J . D., and Staab, R., Analyt. Chem., 1964, 36, 1367. Dagnall, 13. M., Thompson, K. C., and West, T. S., Analytica Chim. Acta, 1966, 36, 269. Omenetto, N., and Kossi, G., Spectrochirn. Acta, 1969, 24B, 95. Veillon, C., Parsons, M. L., Mansfield, J. M., and Winefordner, J . D., Analyt. Chem., 1966, 38, 204. Ellis, D. W., and Demers, D. Ii., Ibid., 1968, 40, 860. Cotton, D. H., and Jenkins. D. R., I n t . Conf. Atom. Absorption Spectrosc., Sheffield, 1969.Paper Acta, 1974, 69, 455. Ohio, 1972. Paper 132. No. U3. -~ , , Spectrochinz. Acta, 1970, 25B, 283. Bennett, P. A., Resonance Lines, 1969, 1, 1. Aggett, J., and West, T. S., Analytica Chim. Acta, 1971, 57, 15. Cartwright, J . S., Sebens, C., and Slavin, W., Atom. Absorption Newsl., 1966, 5, 22. Larkins, P. L., Lowe, R. M., Sullivan, J . V., and Walsh, A., Spectrochim. Acta, 1969, 24B, 187. West, T. S., and Williams, X. K., Analyt. Chem., 1968, 40, 335. Manning, D. C., and Heneage, P., Atom. Absorption Newsl., 1968, 7, 80. Matou;ek, J., and Sychra, V., Analytica Chim. Acta, 1970, 49, 175. __ -__ I n t . Conf. Atom. Absorption Spectrosc., Sheffield, 1969. Paper No. C5. West: T. S., and Williams, X. K., Analytica Chim. Acta, 1968, 42, 29.Aggett, J., and West, T. S., Ibid., 1971, 55. 349. Anderson, R. G., Maines, I. S., and West, T. S.. Ibid., 1970, 51, 355. West, T. S., and Williams, X. K., Ibid., 1969, 45, 27. Sychra, V., Slevin, P. J., Matougek, J., and Bek, F., Ibid., 1970, 52, 259. Bratzel. M. P., and Winefordner, J. D., Analyt. Lett., 1967, 1, 43. Zacha, K. E., Bratzel, M. P., Winefordner, J . D., and Mansfield, J. M., Analyt. Chem., 1968, 40, Mansfield, J . M., Bratzel, M. P., Norgordon, H. O., Knapp, D. O., Zacha, K. E., and Winefordner, Dagnall, R. M., Thompson, K. C., and West, T. S., Talanta. 1967, 14, 551. Dagnall, R. M., and West, T. S.. Appl. Opt., 1968, 7, 1287. Hoare, H. C., Mostyn, R. A., and ru’ewland, B. T. N., I n t . Conf. Atom. Absorption Sfiectvosc., Aldous, K. M., Alger, D , Dagna.11, R.M., and West, T. S., Lab. Pract., 1970, 19, 587. Cooke, D. O., Dagnall, R. M., and West, T. S., Analytica Chim. Acta, 1971, 56, 17. Silvester, M. D., and McCarthy, W. J., Analyt. Lett., 1969, 2, 305. -- , Spectrochim. Acta, 1970, 25B, 229. Cook;, D. O., Dagnall. R. M.. and West, T. S., Analytzca Chim. Acta, 1971, 54, 381. Cresser, M. S., and West, T. S., Ibid., 1970, 50, 517. Patel, B. M., Browner, R. F., and Winefordner, J. D., Analyt. Chem., 1972, 44, 2272. Rains, T. C., and Rush, T. A,, Tech. Note Nat. Bur. Stand., No. 504, 1969. Ball, J . J., Rev. Scient. lnstrum., 1973, 44, 1141. Patel, B. M., Reeves, R. D., Browner, R. F., Molnar, C. J., and Winefordner, J . D., AppZ. Browner, R. F., Patel, B. M., and Winefordner, J . D., Proc. Colloquium Spectroscopicurn Inter- Haarsma, J .P. J., de Jong, G. J., and Agterdenbos, J . , Spectrochim. Acta, 1974, 29B, 1. Beckwith, P. M., Browner, R. F., and Winefordner, J . D., “High Frequency Excited Electrode- 1733. J . D., Spectrochim. Acta, 1968, 23B, 389. Sheffield, 1969. Paper No. D6. Spectrosc., 1973, 27, 171. nationale X V I l , Florence, 1973, 1, 227. less Sources in Analytical Chemistry,” Marcel Dekker, New York, in the press.BROWNER Browner, R. F., Dagnall, R. M., and West, T. S., Analytica Chim. Acta, 1969, 45, 163. Thompson, I<. C., and Wildy, P. C., Analyst, 1970, 95, 562. hlger, D., Dagnall, R. M., Silvester, M. D., and West, T. S., Analyt. Chem., 1972, 44, 2255. Smith, K. L., Rev. Scient. Instrum., 1973, 44, 1108. Johnson, F. M., and Swagel, M.W., Appl. Opt., 1971, 10, 1624. Kuhl, J., and Spitschan, H., Opt., Commun., 1973, 7, 256. Hobbs, R. S., Kirkbright, G. F., Sargent, M., and West, T. S., Talanta, 1968, 15, 997. Aldous, K. M., Browner, R. F., Dagnall, R. M., and West, T. S., Analyt. Chem., 1970, 42, 939. Dinnin, J . I., Ibid., 1968, 40, 1825. Bratzel, M. P., Dagnall, R. M., and Winefordner, J. D., Ibid., 1969, 41, 713. Smith, R., Elser, R. C., and Winefordner, J . D., Analytica Chim. Acta, 1969, 48, 35. O'Haver, T. C., and Winefordner. J. D., J . Chem. Edzkc., 1969, 46, 435. Mansfield, J . M., Parsons, M. L., Veillon, C., and Winefordner, J . D., Analyt. Chem., 1966,38, 204. Jenkins, D. R., Spectrochim. Acta, 1967, 23B, 167. Black, M. S., Glenn, T. H., Bratzel, M. P., and Winefordner, J . D., Analyt. Chem., 1971,43, 1769. Kirkbright, G. F., Analyst, 1971, 96, 609. Massmann, H., Spectrochim. Acta, 1968, 23B, 215. Bratzel, M. P., Dagnall, R. M., and Winefordner, J . D., Analytica Chim. Acta, 1969, 48, 197. Molnar, C. J . , Reeves, R. D., Winefordner, J . D., Glenn, M. T., Ahlstrom, J . R., and Savory, J , , Detection Limits for Model 63 Carbon Tube Atomiser, Varian Techtron Ltd., 1972. Slavin, S., Barnett, W. B., and Kahn, H. L., Atom. Absorption Newsl., 1972, 11, 37. Belyaev, Y. I., Karyakin, A. V., and Pchelintsev, A. M., J . Analyt. Chem. USSR, 1970,25,735. Belyaev, Y. I., and Pchelintsev, A. M., Ibid., 1970, 25, 1799. ~- , Ibid., 1970, 25, 1922. Gough, D. S., Hannaford, P., and Walsh, A., Spectrochim. Acta, 1973, 28B, 197. Muscat, V. I., and Vickers, T. J., Analytica Chim. Acta, 1971, 57, 23. Thompson, K. C., and Reynolds, G. D., Analyst, 1971, 96, 771. Muscat, V. I., Vickers, T. J., and Andren, A., Analyt. Chem., 1972, 44, 218. Thompson, K. C., Lab. Pract., 1972, 21, 645. Hatch, W. R., and Ott, W. L., Analyt. Chem., 1968, 40, 2085. Manning, D. C., Atom. Absorption Newsl., 1970, 9, 109. West, C. D., Analyt. Chem., 1974, 46, 797. Walsh, A., Pure Appl. Chem., 1973, 34, 145. Smith, R., Stafford, C. M., and Winefordner, J . D., Can. Spectrosc., 1969, 14, 2. Miller, R. L., Fraser, L. M., and Winefordner, J. D., Appl. Spectrosc., 1971, 25, 477. Patel, B. M., and Winefordner, J . D., Analytica Chim. Acta, 1973, 64, 135. Dagnall, R. M., Taylor, M. R. G., and West, T. S., Lab. Pract., 1971, 20, 209. Hobbs, R. S., Kirkbright, G. F., and West, T. S., Talanta, 1971, 18, 859. Hofton, M. E., and Hubbard, D. P., Analytica Chim. Acta, 1972, 62, 311. Dagnall, R. M., Kirkbright, G. F., West, T. S., and Wood, R., Analyt. Chem., 1971, 43, 1765. Murugaiyan, P., Natarajan, S., and Venkatswarlu, Ch., Analytica Chim. Acta, 1973, 64, 132. Sarbeck, J . R., St. John P. A., and Winefordner, J. D., Mikrochim. Acta, 1972, 55. Norris, J . D., and West, T. S., Analytica Chim. Acta, 1971, 55, 359. Jones, M., Kirkbright, G. F., Ransom, L., and West, T. S., Ibid., 1973, 63, 210. Dagnall, R. M., Kirkbright, G. F., West, T. S., and Wood, R., Analyst, 1972, 97, 245. KolihovA, D., and Sychra, V., Chemicke' Listy, 1972, 66, 93. Sychra, V., and MatouSek, J . , Ibid., 1970, 52, 376. Kirkbright. G. F., Rao, A. P., and West, T. S., Analyt. Lett., 1969, 2, 465. Warr, P. D., Talanta, 1971, 18, 234. Marshall, G. B., and Smith, A. C., Analyst, 1972, 97, 447. Christian, G. D., and Feldman, F. J., Appl. Spectrosc., 1971, 25, 660. Pickett, E. E., and Koirtyohann, S. R., Spectrochim. Acta, 1968, 23B, 235. _ _ - , Ibid., 1969, 24B, 325. KnisLley, R. N., Butler, C. C., and Fassel, V. A., Analyt. Chem., 1969, 41, 1494. Pickett, E. E., and Koirtyohann, S. R., Spectrochim. Acta, 1968, 23B, 673. -- , Analyt. Chem., 1969, 41 (14), 28A. Hell, 'A., and Ricchio, S., Pittsb. CoNf. Analyt. Chem. Appl. Spectrosc., Cleveland, Ohio, 1970. Fleet B., Liberty, K. V., and West, T. S., Analytica Chim. Actci, 1969, 45, 205. Dinnin, J . I., Analyt. Chem., 1967, 39, 1491. Ebdon, L., Kirkbright, G. F., and West, T. S., Talanta, 1970, 17, 965. Dagnall, R. M., Taylor, M. R. G., and West, T. S., Spectrosc. Lett., 1968, 1, 397. Sychra, V., and Matougek, J., Talanta, 1970, 17, 363. KolihovA, D., and Sychra, V., Analytica Chim. Acta, 1972, 59, 477. Hingle, D. N., Kirkbright, G. F., and West, T. S., Analyst, 1968, 93, 522. Bratzel, M. P., Dagnall, R. M., and Winefordner, J . D., Analyt. Chem., 1969, 41, 1527. Dagnall, R. M., Kirkbright, G. F., West, T. S., and Wood, R., Analytica Chim. Acta, 1969, 47, 407. Browner, R. F., Dagnall, R. M., and West, T. S., Ibid., 1969, 46, 207. Norris, J . D., and West, T. S., Ibid., 1972, 59, 355. Appl. Spectrosc., 1972, 26, 606. -- , , Analytica Chim. Acta, 1973, 63. 479. Received April 8th, 1974 Accepted June 13th, 1974 644 157. 168. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 173a. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224.

ISSN:0003-2654    

DOI:10.1039/AN9749900617

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, October, 1974, Vol. 99, $9. 645-651 645 Sulphonated Alizarin Fluorine Blue : an Improved Reagent for the Positive Absorptiometric Determination of the Fluoride Ion BY M. A. LEONARD AND G. T. MURRAY (Department of Analytical Chemistry, The Queen’s University of Belfast, Belfast, B T9 5AG) Alizarin fluorine blue, [3-NN-di(carboxymethyl)aminomethyl]- 1,2- dihydroxyanthraquinone, has been modified by the introduction of a sulpho- nate group into the 5-position. The complex formed by lanthanum(II1) with this new compound demonstrates a reaction towards fluoride similar to that shown by the complexes formed by lanthanum and cerium(II1) with the original alizarin fluorine blue but with the added advantage of increased solubility. This allows a metal to reagent complex ratio of 2 : 1 to be used, which gives considerable increase in sensitivity towards fluoride.ALIZARIN fluorine blue { [3-NhT-di (carboxymethyl) aminomet hyl] -1, Z-dihydroxyanthraqui- none; AFB} has been used for the determination of fluoride ions in a great variety of matrices since 1959.1-3 The complex formed by lanthanum with this reagent (A,,,, 500 nm) reacts with fluoride to give a ternary complex (A,,,. 567 nm) rather than undergo decomposition to give alizarin fluorine blue and lanthanum fluoride. While the reagent is highly effective at the final concentration level of 5 x M, the insolubility of the lanthanum complex in water at the operating pH of 4.5 limits the range of fluoride concentrations that can be examined. This problem can be largely overcome by operating with mixed aqueous organic solvents or at a higher pH, or both, but undesirable spectral shifts then occur, which reduce the difference between the absorption spectra given by the alizarin fluorine blue - lanthanum and alizarin fluorine blue - lanthanum - fluoride complexes.Additionally, even at 5 x 10-5 M concentration, the presence of a buffer system that possesses mildly complexing anions such as acetate or succinate appears to be necessary in order to hold the complexes in solution and produce results of satisfactory reproducibility. Such buffers compete with alizarin fluorine blue for the metal ion and reduce sensitivity, although this effect is not serious. Numerous thorough studies of the influence of solvent, pH and reagent to metal ratio on the efficacy of the reaction have been r e p ~ r t e d .~ - l ~ For many years improved versions of alizarin fluorine blue have been sought but without success. 14-16 Recently, however, we have succeeded in synthesising the alizarin fluorine blue analogue of alizarin-5-sulphonic acid, the structural formula for the potassium salt of which {potassium sulphonate; AFBS) 3-[NN-di(carboxymethyl) aminomethyl] - 1,Z-dihydroxqant hraquinone-5- is shown below. ,CH,COOH ‘CH,COOH CH,-N This compound demonstrates the useful chemical and physical features of the original alizarin fluorine blue in improved measure and possesses the desired advantage of increased solubility. However, in the preparation of the pure compound, this last attribute leads to problems in its isolation.EXPERIMENTAL SYNTHESIS OF THE REAGENT- Recrystallise sodium alizarin-5-sulphonate from water, including an efficient filtration step, then dissolve 1.0 g of the recrystallised material in 20 ml of water containing 0.32 g (2 mol) of potassium hydroxide. Dissolve 4-8 g (10 mol) of iminodiacetic acid (disodium salt) in 50 ml of water, add 4.0 ml of 36 per cent. formaldehyde solution and transfer the mixture into a 250-ml three-necked, round-bottomed flask fitted with a reflux condenser, stirrer and dropping funnel containing the alizarin-5-sulphonate Dilute the solution to 100 ml. @ SAC and the authors.646 LEONARD AND MURRAY: SULPHONATED ALIZARIN BLUE: A REAGENT [Analyst, Vol. 99 solution and a further stopper if necessary. Raise the temperature of the flask to 75 "C, add 30 ml of the alizarin-5-sulphonate solution and stir the mixture for 1 hour.When the tem- perature is steady lightly stopper the reflux condenser so as to retain formaldehyde vapour. Then add about a further 30ml of alizarin sulphonate solution and 1.0ml of 36 per cent. formaldehyde solution. Finally, after a further 1 hour, add the remaining 40 ml of alizarin- 5-sulphonate solution together with 1.0 ml of the formaldehyde solution. Cool the mixture to room tem- perature, transfer it into a 400-ml beaker and lower the pH to 3 by addition of dilute hydro- chloric acid. Allow the mixture to stand overnight, then filter off the orange solid, which was shown by elemental analysis to consist of the potassium salt of unreacted alizarin-5-sulphonic acid.Gradually concentrate the filtrate, in 25-ml steps, at about 75°C by rotary film vacuum evaporation or vacuum distillation. Collect the material precipitated at each stage and examine it for alizarin-5-sulphonate or iminodi- acetic acid by the electrophoretic or thin-layer chromatographic methods given at the end of this section. Early fractions may be contaminated with potassium alizarin-5-sulphonate; if so, recrystallise them from a small amount of water, with careful filtration, which should remove the less soluble starting material. Late fractions will be contaminated with iminodi- acetic acid; recrystallise these fractions from a small amount of water. Addition of acetone (up to about 20 per cent.) as the mother liquor cools aids the production of good material.Bulk all good samples of potassium alizarin fluorine blue 5-sulphonate and perform a final recrystallisation from water containing 20 per cent. of acetone. Wash the crystals with small volumes of water containing increasing proportions of acetone, then with diethyl ether. Dry the product at 70 "C under vacuum in the presence of phosphorus(V) oxide (see Note). Maintain the mixture at 75 "C for a total of 3 hours. After each volume reduction, set the solution aside to cool to room temperature. NOTE- Alizarin fluorine blue S was prepared by a Mannich condensation reaction involving sodium alizarin 5-sulphonate, formaldehyde and iminodiacetic acid, which is similar to, but not identical with, that used originally to prepare alizarin fluorine blue. Potassium hydroxide was used in this procedure in order to form the less soluble potassium salt of AFBS.Free iminodiacetic acid can be used in the synthesis if sufficient potassium hy- droxide is introduced simultaneously so as to form the dipotassium salt. Commercially available sodium alizarin-5-sulphonate (Bayer, Leverkusen) was examined for purity by extraction of the solid with diethyl ether and electrophoretic examination of the residue (see below). The extract was subjected to thin-layer chromatography on Kodak pre-coated silica gel sheets [developing solvent n-butanol - pyridine - water - acetic acid (15 + 10 + 12 + 3)] and gave only one spot corresponding to alizarin. The residue from extraction showed on electrophoresis only one line of mobility 0.66 x 10-4 cm2 V-1 s-1 (the alizarin-3-sulphonate anion shows a mobility of 0-33 x 10-4 cm2 V-1 s-1 under the same conditions).ELECTROPHORETIC EXAMINATION OF ALIZARIN-5-SULPHONATE AND PRODUCT FRACTIONS- Planar electrophoresis was carried out on Cellogel or thick filter-paper sheets, using as supporting electrolyte a 0.04 M sodium dihydrogen orthophosphate solution adjusted to pH 5.5. Anthraquinones were made visible with ammonia vapour and iminodiacetic acid was located separately with a 0.2 per cent. ninhydrin in acetone spray followed by heating to 100 "C. Several applications of ninhydrin may be required. Paper electrophoresis was more effective for the detection of iminodiacetic acid. THIK-LAYER CHROMATOGRAPHY- Product fractions can be examined by thin-layer chromatography on cellulosc layers by using the typical amino-acid developing solvent n-butanol - acetic acid - water (12 + 3 + 5) ; they are rendered visible by using ammonia vapour or ninhydrin separately as described above.Potassium alizarin sulphonatc is more soluble than AFBS in ethanol but extraction of product fractions with ethanol appears to be of no use for the removal of this starting material from them. It is very difficult to obtain a product free from potassium chloride. For practical purposes this diffi- culty is no great drawback ; stock solutions can be prepared from amounts of reagent adjusted according to the reagent purityas assessed from the results of microanalysis or a determination of chloride. Starting materials are readily removed by careful recrystallisation.When AFBS is dried under fairly vigorous conditions a distinct dehydration reaction takes place accompanied by a colour change from orange - yellow to dark brown. However, tests show that no degradation takes place; indeed the change is reversed by exposure to water vapour.October, 19741 FOR POSITIVE ARSORPTIOMETRIC DETERMINATION OF FLUORIDE ION 647 Results for a good product by elemental analysis, per cent., were: carbon 45-48; hydrogen 2.68; nitrogen 2.50; sulphur 6.13; and potassium 7.60. Calculated, per cent., for AFBS (Cl9H,,0,,NSK) : carbon 45-32; hydrogen 2.80; nitrogen 2.78; sulphur 6.37; and potassium 7.70. Yield 86.5 per cent. SOLUTIONS- Alizarin fluorine blue solution, 5 x lo-' M-Suspend 0.1927 g of alizarin fluorine blue (Hopkin & Williams; purified by the method of Hall13) in 100 ml of water, then add 2-0 ml (4 equivalents) of 1 N sodium hydroxide solution.Agitate the mixture vigorously so as to dissolve the solid, then lower the pH to approximately 6 by addition of dilute hydrochloric acid and dilute the solution to 1 litre. Alizarin fluorine blue S solution, 5 X lo-, M-Dissolve 0-2517 g (see above) of freshly dried alizarin fluorine blue S (this compound is hygroscopic) in 500 ml of water and dilute the solution to 1 litre. Lanthanum nitrate solution, M-Dissolve 4-3303 g of lanthanum nitrate hexahydrate in water and dilute the solution to 1 litre. The solution can be standardised against EDTA at pH 5.5, using xylenol orange as indicator, although in our experience this is unnecessary. Lanthanum nitrate solution, 5 x M-Dilute 50.0 ml of the above solution to 1 litre. Sodium$uoride solution, 5 x M-Dissolve 0.0210 g of sodium fluoride (Koch-Light, electronic grade, more than 99.99 per cent.pure) in water and dilute the solution to 1 litre. Store the solution in a polythene bottle. Potassium nitrate solution (for adjustment of ionic strength), 1 bf-Dissolve 101.11 g of AnalaR potassium nitrate in water and dilute the solution to 1 litre. Acetate bufer, pH 4.63-Dissolve 41.02 g of AnalaR anhydrous sodium acetate in 500 ml of water, add 28-71 ml of glacial acetic acid and dilute the mixture to 1 litre. APPARATUS- Absorption spectra were recorded over the range 350 to 750 nm by using a Perkin-Elmer 402 ultraviolet - visible spectrophotometer. A Unicam SP600 spectrophotometer was used for single absorbance measurements, while pH values were taken with an EIL Vibret pH meter, Model 46A, with a shielded glass - calomel electrode pair; 2-cm glass cells were used for absorbance measurements.INVESTIGATION OF THE REAGENT ABSORPTION SPECTRA OF THE ALIZARIN FLUORINE BLUE AND ALIZARIN FLUORINE BLUE S In Fig. 1 spectra for the species of interest are compared. Initial absorbance - time studies showed that maximum absorbance was achieved after 1 minute for the AFBS - lanthanum complex and after 15 minutes for the AFBS - lanthanum - fluoride complex. The corresponding times for the AFB species were 1 minute and 45 minutes, respectively. Fig. 1 shows clearly the,bathochromic wavelength shift of complexes of AFBS in com- parison with those of AFB, a summary of which is given below.COMPLEXES- AFBS - AFB - AFB - AFBS - lanthanum lanthanum AFB AFBS lanthanum lanthanum -fluoride -fluoride A max./nm . . .. 423 430 500 543 570 583 The AFBS - lanthanum - fluoride absorbance maximum at 583 nm gives rise to a pure blue solution. IONISATION OF ALIZARIN FLUORINE BLUE S- The presence of three isosbestic points (457, 498 and 528 nm) is indicative of ionisation transitions between four ionised forms of the reagent, these forms having A,,,. values of 430, 525, 543 and 590 nm. Although the spectrum of the reagent in the higher pH range is composed of more than one peak, the shift of the main peak to 590nm and the fact that the other peaks, as well as having Amax. values close to Relevant spectra are shown in Fig.2.648 LEONARD AND MURRAY: SULPHONATED ALIZARIN BLUE: A REAGENT [Analyst, Vol. 99 W avel engthh rn Fig. 1. Absorption spectra for AFB and AFBS: a, AFB; b, AFB - lanthanum (1 + 1) ; c, AFB - lanthanum - fluoride (1 + 1 + 2) ; d, AFBS; e, AFBS - lanthanum (1 + 1) ; f, AFBS - lanthanum - fluoride (1 + 1 + 2) ; and g, AFBS - lanthanum - fluoride (1 + 2 + 2). Ca = 5 x M ; pH, 4-63 (0.1 ml of buffer per 50 ml); cell path length, 2.0 cm; p = 0.1 (KNO,) 590 nm, are of lower intensity compared with the main peak (cf., alizarin fluorine blue) cause the solution to have a blue coloration rather than purple as occurs with alizarin fluorine blue. Ionisation data for the system were obtained by using the techniques of pH - absorbance curves and pH titration.We would ascribe the following, as yet tentative, pK values to the four ionisable hydrogen atoms: 2.7, 6-1, 10.2 and 12.5. 1-01 a o n.c o.8t c " " -2 a 2 0.4 0.2 700 0 350 400 450 500 5 50 600 650 Wavelengthhm Fig. 2. Variation of the absorption spectrum for alizarin fluorine blue S with pH. pH: 1, 2.18; 2, 5.60; 3, 6.20; 4, 7.38; 5, 8.96; 6, 11-27; 7, 12.21; 8, 12.83; and 9, 13.58. CR = 5 x M ; cell path length, 2.0 cm; p = 0.1 (KNO,) STUDY OF THE BINARY AND TERNARY COMPLEXES- The techniques described above were applied to the analysis of the binary (reagent - lanthanum) and ternary (reagent - lanthanum - fluoride) complexes. The variation of the absorption spectra with pH for these complexes can be seen in Figs. 3 and 4. Again, well defined isosbestic points indicate that simple equilibria are involved.Figs. 5 and 6 show the respective pH - absorbance curves for the binary and ternary complexes. Mathematical analysis of these curves17 showed the release of two protons and three protons, respectively, from the undissociated form of the reagent during complexation. Relevant details are givenOctober, 19741 FOR POSITIVE ABSORPTIOMETRIC DETERMINATION OF FLUORIDE ION 649 with the appropriate diagrams. We would draw attention to the remarkable colour change with pH of the ternary complex, with which the transition from a 90 per cent. yellow to 90 per cent. blue colour takes place over 0.66 pH unit. Oa8 t 0.6 0.4 0.2 0 350 400 450 500 550 600 650 700 W avelengthh m Fig. 3. Variation of the absorption spectrum for the AFBS - lantha- num complex.pH: 1, 3.64; 2, 3.79; 3, 3.92; 4, 3-96; 5, 4.05; 6, 4.27; 7, 4-70; and 8, pH 6.00. CR = 5 X 1 0 - 6 M ; c~ = 6 X 1 0 - 6 M ; 2.0-cm cells; p = 0.1 (KNO,) SENSITIVITY TOWARDS FLUORIDE IONS- As a conclusion to this preliminary study of alizarin fluorine blue S, a fluoride sensitivity comparison with alizarin fluorine blue was undertaken, which consisted in the construction of graphs of Beer's law for the addition of fluoride ions to the binary complexes, using the 0 I 350 400 450 500 550 600 650 700 Wavelength/nm Fig. 4. Variation of the absorption spectrum for the AFBS - lanthanum - fluoride complex. pH: 1, 3-60; 2, 3-73; 3, 3.82; 4, 3.93; 5, 3.99; and 6, 4.66. CR = 5 X M ; cy = 1 X M; CF = 1 x 1 0 - 4 ~ ; 2-0-cm cells; p = 0.1 (KNO,)650 LEONARD AND MURRAY: SULPHONATED ALIZARIN BLUE: A REAGENT [Analyst, Vol.99 " 1 2 3 4 5 6 7 8 9 PH Fig. 5 . Absorbance - pH curves for AFBS and the AFBS - lanthanum complex. A, CM/CR = 0; and B, c ~ / c ~ = 100. CR = 5 X W 5 M ; wavelength 520nm; 2-0-cm cells; p = 0.1 (KNO,) respective wavelengths of maximum separation between the spectra of the binary and ternary species. These optimum wavelengths were obtained by recording the absorption spectra of the ternary fluoride complex systems relative to the binary systems as blank, the maximum peak on the resulting trace indicating the most suitable wavelength. The resulting graphs are shown in Fig. 7. As can be seen, the performance of AFBS compares well with that of AFB especially if an AFBS to lanthanum ratio of 1 : 2 is used.Attempts to employ an AFB to lanthanum ratio of 1:2 result in precipitation. Beer's law is obeyed over the range 0 to 12 pg of fluoride per 50 ml. With the ratio of 1:2 for the AFBS - lanthanum complex the fluoride sensitivity AA/AF for 100 ml and a 1-cm cell is 4.67 x 10-3 absorbance unit per microgram of fluoride. The linearity of the graph at low fluoride concentrations with AFBS appears to be superior to that obtained with AFB. 0 1 2 3 4 5 6 PH Fig. 6. Absorbance - pH curve for the AFBS - lantha- num - fluoride system. CR = 5 x 1 0 - 5 ~ ; c R : c M : c ~ = 1:2:2; wavelength 520 nm ; 2.0-cm cells ; p = 0.1 (KNO,)October, 19741 FOR POSITIVE ABSORPTIOMETRIC DETERMINATION OF FLUORIDE ION 651 0-35 0.30 d) 0-25 0 x 0.20 i 2 0.15 0.1 0 0.05 0 5 x M Fluoride/ml Fig.7. Beer’s law graphs for the AFBS - lanthanum - fluoride complex. A, C A F B ~ : CLa: Cp = 1 : 1 : X ; B, Cap~g: CLa : CF = 1:2:X; and C, CApB : CLa : Cp = 1:l:X. CR = 5 X lop5 M ; pH = 4.63 (0.1 ml of acetate buffer per 50 ml) ; p = 0-1 (KNO,) ; X A F B ~ = 635 nm; XAFB = 610 nm; 2-0-cm cells CONCLUSIONS On the basis of the results obtained from the preliminary study of alizarin fluorine blue S, it would appear that the reagent, while retaining the basic complexing features of alizarin fluorine blue , possesses the desired characteristics of increased solubility in aqueous medium and enhanced sensitivity towards the fluoride ion, as well as a more rapid attainment of maximum absorbance for the ternary complex. Interferences have not been studied at this stage but in all probability would prove to be similar to those which afflict alizarin fluorine blue.A more intensive study of the reagent and its binary and ternary complexes is in progress at present and the results of this work will follow in a later paper. We are grateful to Messrs. Bayer A.G. (Leverkusen, Germany) for the gift of the com- mercial sample of sodium alizarin-5-sulphonate and to the staff of our analytical services laboratory for many elemental analyses. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 1 1 . 12. 13. 14. 15. 16. 17. REFERENCES Belcher, R., Leonard, M. A., and West, T. S., J . Chem. SOG., 1959, 3577. Leonard, M. A., and West, T. S., Ibid., 1960, 4477. Leonard, M. A., in Johnson, W. C., Editor, “Organic Reagents for Metals and for Certain Radicals,” Greenhalgh, R., and Riley, J. I?., Analytica Chim. Acta, 1961, 25, 179. Belcher, R., and West, T. S., Talanta, 1961, 8, 853 and 863. Yamamura, S. S., Wade, M. A., and Sikes, J. H., Analyt. Chem., 1962, 34, 1308. Hanocq, M., and Molle, L., Analytica Chim. A d a , 1968, 40, 13. Langmyhr, F. J., Klausen, K. S., and Nouri-Nekoui, M. H., Ibid., 1971, 57, 341. Weinstein, L. H., Mandl, R. H., McCune, D. C., Jacobson, J. S., and Hitchcock, A. E., Contr. Boyce Fernandopulle, M. E., and Macdonald, A. M. G., Microchem. J., 1966, 11, 41. Analytical Methods Committee, Analyst, 1971, 96, 384. Buck, M., 2. analyt. Chem., 1963, 193, 101. Hall, R. J., Analyst, 1963, 88, 76. Belcher, R., Leonard, M. A., and West, T. S., J . Chem. SOC., 1958, 2390. Vaughan, A., “Some Derivatives of Hydroxyanthraquinones as Reagents for the Fluoride Ion,” Al-Ani, K., “The Use of Strongly Chelating Derivatives of Anthraquinones as Analytical Reagents,” Reilley, C. N., and Sawyer, D. T., “Experiments for Instrumental Methods,’’ McGraw-Hill, New Received January 14th, 1974 Accepted May 6th, 1974 Volume 2, Hopkin & Williams, Chadwell Heath, Essex, 1964. Thomson Inst. PI. Res., 1963, 22, 207. hl.Sc. Thesis, University of Birmingham, 1967. Ph.D. Thesis, Queen’s University of Belfast, 1971. York, 1961.

ISSN:0003-2654    

DOI:10.1039/AN9749900645

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

652 Analyst, October, 1974, Vol. 99, $9. 652-656 An Improved Procedure for Application of the Fuj iwara Reaction in the Determination of Organic Halides BY J. F. REITH, MISS W. C. VAN DITMARSCH AND TH. DE RUITER (Department of Toxicology, State University, Catharijne Singe1 60, Utrecht, The Netherlands) Fuj iwara discovered that when chloroform or other organic halides are heated with pyridine and sodium hydroxide a red colour is formed. Some halides react very sensitively, others weakly or not at all, depending on many factors. A standard procedure involving this reaction is recommended for use in analytical toxicology and information on sensitivities of twenty-two organic halides when using this procedure is presented. This information permits accurate interpretation concerning the presence or absence of organic halides.Some commercial samples that gave a positive Fujiwara test reacted nega- tively after being purified by distillation or gas - liquid chromatography and the literature on positive Fujiwara reactions should therefore be consulted. Methods for the determination of an organic halide on the basis of the Fuj iwara reaction should be used only when all other reactive halides are absent. IN 1914, Fujiwaral discovered that solutions containing chloroform or other organic halides added to a mixture of aqueous sodium hydroxide solution and pyridine produce a deep red colour when heated. The “Fuj iwara reaction” soon found application in analytical toxicology. In his procedure, Fujiwara transferred 3 ml of 10 per cent. aqueous sodium hydroxide solution and 2 ml of pyridine into a test-tube, heated the mixture to boiling tem- perature, added to it 1 ml of the solution to be tested and mixed the contents of the tube by shaking it.The formation of a purple - red colour in the pyridine layer within a few seconds indicates the presence of chloroform, trichloroacetaldehyde, bromoform, iodoform, 1,Z- dichloroethane or trichloroacetic acid. Saturated iodine solution also gives this colour. This procedure has been modified by many author^^-^ and has been variously described in handbooks. Several authors have stressed that the colour intensity is controlled by many factors, e.g., the presence of other organic solvents (ketones, alcohols and pyridine), the volume and concentration of the sodium hydroxide solution, temperature during the reaction, and time elapsing between mixing and observing of the colour.These and other factors produce different effects with several organic halides. The stability of the colour also depends on the reacting organic halide and the factors mentioned above. Sunlight and oxygen are also considered to have a harmful effect on colour stability. Some ~ ~ r k e r ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ have recommended the use of amounts of pyridine, water and sodium hydroxide in such proportions as to produce a homogeneous phase. In their procedures the effects of the factors mentioned above may be different. Lists of substances that give positive and negative reactions with Fujiwara’s test have been published, 2 s l 2 ~ l4 based on publications by workers who used different procedures; it is evident that such lists are misleading.As so many factors have an influence on whether or not an organic halide is “Fujiwara positive,” it is necessary to adopt a standard procedure in order that reliable information on the reactivity of organic halides may be compiled. For this purpose, we have chosen the procedure recommended by Lugg7 suitably modified so as to permit its reproducible application. EXPERIMENTAL Sodium hydroxide solution-Dissolve 430 g of analytical-reagent grade sodium hydroxide Pyridine-Purify by distillation if a blank test gives a colour. @ SAC and the authors. REAGENTS- in water to give 1 litre of solution.REITH, VAN DITMARSCH AND DE RUITER 653 APPARATUS FOR SAMPLE PURIFICATION- flame-ionisation detect or. 20 per cent.Fractonitril I11 on silanised, acid-washed, 80 to 100-mesh Chromosorb W. column temperature, 40 to 150 "C, depending on the boiling-point of the halide. Gas chromatograph-Becker (Delft), Type 1964, equipped with a 1:l beam splitter and a CoZumns-Aluminium, 3 or 4 m x 4 or 6 mm i.d., packed with 20 per cent. DC 11 or The conditions used were: carrier gas, nitrogen at the rate of 30 to 45 ml min-l; and COLORIMETRIC PROCEDURE- Transfer an adequate volume of a solution of the sample in pyridine into a calibrated 10-ml test-tube that has a glass stopper; make the volume up to 5 ml with pyridine and add 5 ml of 10.75 M sodium hydroxide solution. Stopper the tube and shake it vigorously for several seconds. Loosen the stopper and heat the tube in boiling water for 5 minutes.Note the colour of the pyridine layer during and immediately after heating. For quantitative purposes, immediately after heating place the test-tube in water at room temperature for 2 minutes. Separate the layers and immediately determine the extinction of the pyridine layer a t 368 nm. QUALITATIVE RE su LTS- Several commercial samples of organic halides showed many peaks in the gas - liquid chromatograms that were caused by impurities, Some samples gave a positive reaction, but a negative reaction after being purified by distillation or by gas - liquid chromatography. Results obtained with the above standard procedure are shown in Table I. TABLE I BEHAVIOUR OF TWENTY-TWO ORGANIC HALIDES IN THE FU JIWARA REACTION Compound Method of purification Formula of sample Positive reactions- Trichloroacetic acid CCl3.COOH None Chloroform CHC13 None Bromoform CHBr3 None 2-Trichloromethvlpropan-2-01 C(CC13) (CH3) 20H CHCl2. CHCl2 GLC CBr2 = CHBr GLC Cryst. t 1,1,2,2-Tetrachl&oet6ane Tribromoethylene cis- 1 ,2-Dichloroethylene Dibromomethane Carbon tetrachloride 1,l-Dichloroethylene a-Trichlorotoluene Trichloroethylene Dichloroacetic acid 1, 1, 1-Trichloroethane 1,1,2,2-Tetrabromoethane 1,l-Dibromoethane Hexachloroethane a-Dichlorotoluene Dichloromethane 1,l-Dichloroethane 1,1,8-Trichloroethane Tetrachloroethylene Negative reactions- CHCl= CHCl CH,Br2 cc14 CCl2 = CH2 C&5. cc13 CCl2 = CHCl CHCl2. COOH CC13. CH3 CHBrz . CHBr2 CHBr2. CH3 CCl3. CC13 C&5. CHC12 CH2C12 CHCl2. CH3 CHC12.CH2Cl CCl2 = CCl2 GLC GLC GLC GLC GLC None GLC GLC GLC GLC Cryst.7 GLC GLC GLC GLC GLC Minimum amount Colour* detectable Pink 0.0025 mg Purple - red 0.0012 p1 Purple - red 0.005 mg Pink - red 0.001 p1 Orange: 0.001 pl Brownish 0.008 p1 Pink 0*008 p1 Pinks 10 p1 orange: Purple - red 0.01 p1 Red 0.07 p1 Purple - red 1 pl Orange - yellow: 0.01 pl Orange - red 6 p1 Purple - red 2 pl Orange 7 0.003 p1 Brown - None - None - None - None - None - Orange 10 p1 El moll- 1 om (368 nm) 28 500 25 100 16 500 13 800 12 500 8000 4600 770 330 50 13 700 100 7 11 200 30 12 110 50 <2 <2 < 2 120-680 * Visual observation of colour after 5 minutes' heating, if not indicated otherwise. t Crystallisation. : First purple, changing into orange. 3 First pink, then yellow, but after shaking, pink again.7 Red only after heating for more than 5 minutes. It is of interest to note that observations on the behaviour of carbon tetrachloride in Truhaut5 indicated that the colour obtained with this Fujiwara tests have differed widely.654 [Analyst, Vol. 99 substance was “a little less” than that with chloroform. Burke and Southernl8 observed a colour intensity of about one-third of that given by chloroform. L u g 7 however, placed carbon tetrachloride as No. 7 in his list of sensitivities. Such differences may have been caused by impurities in the samples investigated. Indeed, Mecke and Oswaldlg warned that most commercial samples of carbon tetrachloride contain chloroform and trichloroethylene. JenSovsky20 and JenGovksy and Bardodgj 21 distilled carbon tetrachloride without obtaining a fraction that was Fujiwara negative.They believed that trace amounts of chloro- form were present in the distillates. It has been t h o ~ g h t ~ 9 ~ ~ that chloroform is formed in the carbon tetrachloride by the combined action of moisture and light and experiments carried out by Webb, Kay and Nicho14 support this supposition. We tried to purify carbon tetrachloride by subjecting it to gas - liquid chromatography in different ways, but the fractions obtained reacted positively. Moreover, we heated 25 ml of pyridine containing 5 pl of purified carbon tetrachloride with 50 ml of 10.75 M sodium hydroxide solution in the hope that trace amounts of chloroform, if present, might react with the pyridine and thus be removed.The pyridine layer was then separated, dried with an- hydrous sodium sulphate and distilled. The distillate was Fuj iwara positive, although the reaction was less distinct than usual. The above experiment was repeated in a room that was almost dark; the colour obtained with the Fujiwara test was appreciably less than when the whole procedure was performed in daylight. The conclusion was reached that under normal conditions carbon tetrachloride is slightly Fujiwara positive. QUANTITATIVE RESULTS- In Lugg’s procedure,7 the pyridine layer shows different absorption maxima, depending on the heating time and the substance tested. The maximum at 535nm shows highest intensity after heating for 5 minutes when chloroform is tested. The absorption maximum at 368 nm increases with the heating time and is more intense than the maximum at 535nm after heating for 5 minutes.The use of this maximum, corresponding to the maximum absorption of glutaconic dialdehyde, 23 was first advised by Friedman and Cooper.24 The results of our absorption measurements at 368 nm are summarised in Table I. REITH et al. : IMPROVED PROCEDURE FOR APPLICATION OF THE DISCUSSION AND CONCLUSIONS The mechanism of the Fujiwara reaction is based on the remarkable ease with which the pyridine ring can be split under special conditions. Numerous examples of this ring breakage, preceded by the formation of unstable pyridinium salts, have been r e c ~ r d e d . ~ ~ - ~ ~ Two examples are important in explaining the mechanism of the Fujiwara reaction, namely, the reactions suggested by Vongerichten and Konig.In 1899, Vongerichten28 and, independently, two other investigator~~~~3~ observed that an intense red coloration is produced when 2,4-dinitrochlorobenzene reacts with pyridine in an alkaline medium. Later, Reitzenstein31 isolated and identified a pyridinium salt (I) as a reaction product. Thereupon Z i n ~ k e ~ ~ - ~ 6 isolated and analysed the reaction products and concluded that a pyridinium salt (I) is formed, followed by breakage of the ring to give compound (11), which hydrolyses to glutaconic dialdehyde (111) or its tautomeric enolic form (IV): H H H H HC/C+CH HC/‘\CH H~C/‘+CH HCHC‘CH HC I‘ CH HC, CHO OHC CHO OHC CHOH 1 - 1 1 1 - 1 l o r I II ‘Nf NU I R R’ Hal I I I I l l IV In Vongerichten’s reaction the halogen was chlorine and R a dinitrophenyl radical.The reaction products were red because of the presence of conjugated double bonds. Zincke did not identify glutaconic dialdehyde itself, but its dianilide. D i e ~ k m a n n ~ ~ , ~ * also isolated a derivative and Ba~rngarten~~~~O identified the sodium salt of compound (IV), formed in another instance of ring breakage.October, 19741 FU JIWARA REACTION IN DETERMINATION OF ORGANIC HALIDES 655 Konig41-43 observed an intense red coloration when cyanohalogen reacts with pyridine in alkaline medium in the presence of certain amines. He explained the reaction mechanism, based on Zincke’s report, by the formation of compounds I, I1 and I11 in which R is a CN radical. The aldehyde groups react with the amine to form coloured final products.The Fujiwara reaction is generally u n d e r ~ t o o d ~ ~ ~ ~ ~ ~ ~ to have the same mechanism. In the formulae I and 11, R is -CHCl, for chloroform. Among products formed in the reaction between chloroform, pyridine and alkali, Treibs 27 identified /3-vinylacrylic acid. He discussed the possibility of its formation from compound I11 by isomerisation. JenSovsky and Bardodgj 21 demonstrated glutaconic dialdehyde as a reaction product. Some authors make a distinction between “true” Fujiwara positive compounds and other compounds that produce red substances under the same conditions. JenSovksy and Bardodgj indicate the following to be true Fujiwara positive compounds. (1) Chloroform, cc-trichlorotoluene and 2-trichloromethylpropan-1-01. (2) (3) Four “true” Fujiwara-positive compounds have been the subject of Moss and Rylance’s investigation^.^^ They observed that the solutions (after the normal hot conditions of the reaction) had absorption maxima at 365 and 530 nm, and thin-layer chromatography of the solutions showed a red spot at R, 0.54 in all four instances.However, when the reaction took place at 0 “C, a maximum was also found at 390 to 400 nm. Moss and Rylance believe that the latter absorption maximum relates to compound I, which is so unstable that it is not found when the reaction is carried out at a higher temperature. The maximum at 530 nm corresponds to compound 11; it decomposes slowly into the final product 111, which has a maximum at 365 nm.24 The same authors stress the point that other organic halides show different behaviour in several ways, for example, the red reaction product formed by trichloroethylene is not an acid - base indicator as are the true Fujiwara products and its colour rapidly changes to orange.20$21 The solution shows a peak at 440 nm24 and at 545 nm.46 Moss and R y l a n ~ e * ~ obtained a coloured product with an RF value of 0-75 instead of 0.54 in their thin-layer chromatography. The present study shows that the formation of red coloured products in the Fujiwara test by any organic halide cannot be predicted on the basis of existing knowledge and that different mechanisms may be involved.Nevertheless, this test remains useful for screening purposes when the procedure is standardised and sufficient evidence is gained on organic halides that react positively. A test with 1 pl of an unknown sample by the standard pro- cedure and reference to Table I enables a conclusion to be drawn whether one or more members of the first group are present or absent.The differences between the results obtained by former investigators and ours may partly be caused by differences in procedures, but it is more likely that their samples were contami- nated. We observed that commercial samples of 1,1,2-trichloroethane and of 1,1,2,2-tetra- chloroethylene gave a positive Fujiwara test, but after purification by gas - liquid chromato- graphy the reaction was negative. Webb et aL4 indicated that both substances gave a positive reaction and Lugg inserted the latter as No. 14 in his list of reacting halides.The Fujiwara reaction has been the basis of methods for the determination of numerous organic halides3$5-7,9,16,17~1*, 24,47 including carbon tetrachloride and even tetrachloroethylene, the latter of which was found by us to be Fujiwara negative. In our opinion the Fujiwara reaction should not be applied for quantitative purposes, because so many factors influence the colour formation and because it is non-specific. Contamination of a sample with trace amounts of chloroform or other strongly reacting halides seriously interferes in the determina- tion of more weakly reacting substances. Organic halides that yield chloroform during the reaction, e.g., trichloroacetic acid and t richloroacet aldehyde. Bromo- and iodo-analogues of (1) and (2). REFERENCES 1.2. 3. Fujiwara, K., Sber. Abh. naturf. Ges. Rostock, 1914-1915, 6, 33. Briining, A., and Schnetka, M., Arch. Gewerbepath. Gewerbehyg., 1933, 4, 740. Daroga, R. P., and Pollard, A. G., J . SOC. Chem. Ind., 1941, 60, 218.656 REITH, VAN DITMARSCH AND DE RUITER 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. Webb, F. J., Kay, I<. K., and Nichol, W. E., J . I n d . Hyg. Toxicol., 1941, 27, 249. Truhaut, R., Bull. Fe‘d. I n t . Pharm., 1949, 23, 432. - , Annls. Pharm. Fr., 1951, 9, 175. Lugg, G. A., Analyt. Chem., 1966, 38, 1532. Moss, M. S., and Kenyon, M. W., Analyst, 1964, 89, 802. Leibmann, K. C., and Hindman, J. D., Analyt. Chem., 1964, 36, 348.Pesez, M., and Pokier, P., “MCthodes et RCactions de l’Analyse Organique,” Masson & Cie., Paris, Feigl, F., “Spot Tests in Organic Analysis,” Sixth Edition, Elsevier, London, 1960, p. 327. Maehly, A. C., “Volatile Toxic Compounds,” in Stolman, A., Editor, “Progress in Chemical Toxi- Preuss, Fr. R., “Gadamers Lehrbuch der chemischen Toxikologie,” Volume I/1, Verlag Vanden- Curry, A., “Poison Detection in Human Organs,” Charles C. Thomas, Springfield, Ill., Second Sunshine, I., Editor, and Maes, R. A., Consulting editor, “Handbook of Analytical Toxicology,” Rogers, G. W., and Kay, K. K., J . Ind. Hyg. Toxicol., 1947, 29, 229. Seto, T. A., and Schultze, M. O., Analyt. Chem., 1956, 28, 1625. Burke, T. E., and Southern, H. K., Analyst, 1958, 83, 316.Mecke, R., and Oswald, F., Spectrochim. Acta, 1951, 4, 348. JenBovsky, L., Chemicke’ Listy, 1954, 48, 1419. Jenbovsky, L., and BardodGj, Z., Pracovni LLk., 1954, 6, 301. Pestemer, M., Angew. Chem., 1951, 63, 118. Schwarzenbach, G., Lutz, K., and Felder, E., Helv. Chim. Acla, 1944, 27, 576. Friedman, P. J., and Cooper, J. F., Analyt. Chem., 1958, 30, 1674. Reitzenstein, F., and Breuning, W., J . prakt. Chem., 1911, 83, 97. Konig, W., and Bayer, R., Ibid., 191 1, 83, 325. Treibs, A., Justus Liebigs A n n l n Chem., 1943, 497, 297. Vongerichten, E., Ber. dt. chem. Ges., 1899, 32, 2571. Gail, G., Inaug. Diss., Marburg, 1899. Spiegel, L., Ber. dt. chem. Ges., 1899, 32, 2834. Reitzenstein, F., J . Prakt. Chem., 1903, 68, 251. Zincke, T., Justus Liebigs A n n l n Chem., 1904, 330, 361. - , Ibid., 1905, 338, 107. - , Ibid., 1905, 339, 193. Zincke, T., Heuser, G., and Miiller, W., Ibid., 1904, 333, 296. Zincke, T., and Wurker, W., Ibid., 1905, 341, 365. Dieckmsnn, W., Rer. dt. chem. Ges., 1902, 35, 3201. -, Ibid., 1905, 38, 1650. Baumgarten, P., Ibid., 1924, 57, 1622. - , Ibid., 1926, 59, 1166. Konig, W., J. prakt. Chem., 1904, 69, 1 and 105. - , Ibid., 1904, 70, 19. -, Ibid., 1911, 83, 406. Gautier, J. A,, Annls Pharm. Fr., 1948, 6, 180. Moss, M. S., and Rylance, H. J., Nature, Lond., 1966, 210, 945. Bonnichsen, R., and Maehly, A. C . , J . Forens. Sci., 1966, 11, 414. Kondos, A. C., and McClymont, G. L., Analyst, 1959, 84, 67. 1954, Volume 111, pp. 111 and 120. cology,” Volume 3, Academic Press, New York and London, 1967, pp. 63-82. hoeck & Ruprecht, Gottingen, 1969, pp. 34-37. Edition, 1969, pp. 49-52. Chemical Rubber Co., 18901 Granwood Parkway, Cleveland, Ohio, 44128, 1969, p. 401. Received January 7th, 1974 Accepted APril 29th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900652

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, October, 1974, Vol. 99, pp. 657-660 657 Determination of Tryptophan and Indole Substances by a Colorimetric Diazotisation Method BY A. K. GOSWAMI (Department of Soil Science and Chemistry, Himachal Pradesh University, Palampur 176061, H. P., India) A colorimetric method for the determination of tryptophan in protein hydrolysates by its conversion into nitrosamine with nitrous acid followed by diazotisation with N-1-naphthylethylenediamine dihydrochloride has been studied. Selective nitrosation of tryptophan was best achieved at 20 to 35 "C using 1.0 to 1.2 M hydrochloric acid. Diazotisation was best achieved a t 10 "C or below. Sodium chloride inhibited the nitrosation reaction to a considerable extent and, therefore, tryptophan standards should contain an amount of sodium chloride equal to the amount present in sample hydrolysates.Such standards are prepared from a portion of the sample hydrolysate pre- treated with activated charcoal. Twenty common protein amino-acids other than tryptophan did not interfere in the colour development, but the method was found to be applicable to indole and its derivatives in addition to tryptophan. Compounds such as phenols and aromatic amines were found to interfere. THE preparation of a chromogen from tryptophan by its conversion into nitrosamine with nitrous acid followed by diazotisation with an amine was first used as a method for the colorimetric determination of tryptophan by Eckert.l The method was further modified2 to eliminate the effect of colouring materials in the alkaline hydrolysates of plant samples.Although these methods were very sensitive and followed Beer's law over a wide range of concentrations, the recovery from alkaline hydrolysates of added tryptophan rarely exceeded 80 per cent. The slope of the standard curve depended on the room temperature. The present study was, therefore, undertaken to investigate the reasons behind these anomalies and find solutions for them. EXPERIMENTAL Sodium nitrite solution, 1 per cent.-This solution should be prepared just before use. Ammonium sulphamate solution, 8 per cent.-This solution can be preserved. N- 1-Naphthylethylerediamine dihydrochloride solution, 0.5 per cent.-This solution should be kept in a dark bottle, placed in a refrigerator when not in use and should be discarded when it becomes coloured.REAGENTS- PREPARATION OF HYDROLYSATE- Hydrolyse 100 mg of sample with 10 ml of 5 M sodium hydroxide solution a t 110 "C for 18 hours in a sealed tube. To the hydrolysate add 15 ml of 5 M hydrochloric acid; the hydro- lysate will now be 1.0 M in hydrochloric acid and have a sodium chloride concentration of 2 M. When the acidified hydrolysate is turbid, centrifuge it at a low speed for 30 minutes and use the top, clear layer. When the colorimetric reading is high, dilute the hydrolysate with 1 M hydrochloric acid that is 2 M in sodium chloride. PREPARATION OF STANDARD TRYPTOPHAN SOLUTION- Prepare a standard solution of tryptophan in 1 M hydrochloric acid that is 2 M in sodium chloride so that the concentration is in the range 5 to 25 ,ug ml-l.If a sample contains sodium chloride or the sodium chloride content of a hydrolysate is not known, take a portion of the hydrolysate, treat it with activated charcoal, filter it and prepare the standard tryptophan solution from this purified hydrolysate. 0 SAC and the author.658 GOSWAMI : DETERMINATION OF TRYPTOPHAN AND INDOLES [Analyst, VOl. 99 PROCEDURE- To 5 ml of hydrolysate or standard tryptophan solution add 1 ml of 1 per cent. sodium nitrite solution, mix well and leave the mixture at room temperature for 1 hour. Add 1 ml of 8 per cent. ammonium sulpliamate solution with a l-ml pipette, making sure that the liquid on the wall of the test-tube is completely washed in while pouring. Shake the tube intermittently for 5 minutes and place it in a B.O.D.incubator set at 10 "C (or on the lower shelf of a refrigerator at 6 "C). After about 15 minutes, add 1 ml of 0.5 per cent. N-l-naphthylethylenediarnine dihydrochloride solution and leave the mixture for 1 hour at 10 or 6 "C. Bring the tube to room temperature by placing it under running tap water and immediately read the absorbance at 530 nm against a blank. Prepare the sample blank with the hydrolysate in a similar manner except that the 1 per cent. sodium nitrite solution is replaced with 1 ml of water. The recovery is 97 to 102 per cent. RESULTS AND DISCUSSION The results presented in the tables are absorbances measured at 530 nm on a Spectronic 20 instrument (Bausch and Lomb). The method for the determination of tryptophan described under Procedure was used in all instances and any deviations from this method are mentioned in the text.EFFECT OF TEMPERATURE ON COLOUR DEVELOPMENT- The effect of temperature was studied by varying the temperatures at which the nitrosation and diazotisation reactions were carried out. I t can be seen from Table I that at both 10 "C and 30 "C the extent of colour development was low, and the best results were obtained at 20 "C. However, when nitrosation was conducted at 20 to 35 "C and diazotisation at 10 "C and below, a consistent and maximum amount of colour was obtained. Hence the use of lower temperatures reduced the extent of nitrosation and higher temperatures reduced the extent of diazotisation. In the presence of a reagent that reacts rapidly with nitrous acid, the aromatic nitrosamine is slowly converted back into the aromatic secondary amine.The present method does not show this behaviour when ammonium sulphamate reagent, which rapidly reacts with nitrous acid, is added and left at room temperature for 1 hour. Therefore, a Fischer - Hepp transformation of the aromatic nitrosamine into a nitroso-secondary amine in a solution of hydrochloric acid can account for the beneficial effect of higher temperatures. The diazotisation reaction is exothermic and many diazonium compounds are stable only at low temperatures,4 which explain the optimum temperature of 10 or 6 "C in the diazotisation reaction. TABLE I EFFECT OF TEMPERATURE ON NITROSATION AND DIAZOTISATION REACTIONS OF TRYPTOPHAN Nitrosation was conducted with an initial hydrochloric acid concentration of 1.0 M.Results are absorbances at 530 nm Nitrosation Diazotisation temperature/ temperature/ "C "C 10 10 20 20 30 30 35 35 20 10 30 10 35 10 20 6 30 6 35 6 - 100 0.230 0-245 0.150 0.125 0.325 0.325 0.335 0.325 0.330 0.325 20 0.045 0.060 0.040 0.035 0-060 0.065 0.065 0.060 0.070 0.065 40 0.085 0.1 10 0.080 0.060 0.130 0.130 0.140 0.130 0.135 0.140 60 0.135 0-155 0.115 0-085 0.190 0.190 0*200 0.195 0.200 0-195 80 0.170 0.195 0.125 0.1 15 0.260 0.270 0.260 0-270 0.265 0.260 EFFECT OF SALT AND PROTEIN HYDROLYSATE ON COLOUR DEVELOPMENT- It was noted that acidified alkaline hydrolysates always contained a considerable amount of sodium chloride and this salt considerably depressed the colour development. Table I1 gives the results of a systematic study on this effect.An acid hydrolysate of protein at the 1 per cent. level in the absence of sodium chloride had virtually no effect on colour developmentOctober, 19741 BY A COLORIMETRIC DIAZOTISATION METHOD 659 and that at the 2-5 per cent. level also had a very small effect in comparison with the effect of 2 M sodium chloride solution. When a 1 per cent. solution of a mixture of twenty synthetic protein amino-acids in equal amounts was used for the study of tryptophan recovery there was no depression of colour development and the recovery ranged from 96 to 102 per cent. However, the colour developed in the presence and absence of protein hydrolysates using 1 M hydrochloric acid that was 2 M in sodium chloride was the same and the recovery was close to 100 per cent.TABLE I1 EFFECT OF SODIUM CHLORIDE AND PROTEIN HYDROLYSATE ON THE DETERMINATION OF Nitrosation was conducted at room temperature with an initial hydrochloric acid concentration TRYPTOPHAN was conducted at 10 "C. Results are absorbances at 530 nm of 1.0 M and diazotisation Initial concentration of P protein hydro1 ysate, NaCl/M per cent. 0 0 2 0 2 2.5 2 1.0 0 2.5 0 1-0 0 1*0* f 20 0.065 0.045 0.035 0.045 0.055 0.060 0.065 40 0.128 0.075 0-065 0.065 0.1 15 0.120 0-130 60 0.185 0.1 15 0.100 0.105 0.170 0.180 0.180 80 0.250 0.160 0-140 0.145 0.220 0.240 0.255 7 100 0.325 0.210 0.170 0.195 0-270 0.300 0.330 * 20 common protein amino-acids (with the exception of tryptophan) were mixed in equal amounts and Efforts were made to establish the relative extents of the depression by sodium chloride of the nitrosation and diazotisation reactions separately by adding sodium chloride before and after nitrosation, prior to the diazotisation reaction.From Table 111, it is evident that most of the depressing effect of sodium chloride was on the nitrosation reaction. Table I11 also indicates that by increasing or decreasing the acidity of the hydrochloric acid from 1.0 M, it was not possible to overcome the depressing effect of sodium chloride. Although at the lower acidity the extent of colour development by tryptophan was the same as that when 1.0 M hydro- chloric acid was used, it was found that a hydrochloric acid concentration ranging from 1.0 to 1.2 M was the most suitable. In this range of acidity interference by amino-acids such as tyrosine, histidine and proline in the reactions of protein hydrolysates was completely eliminated (cf., Table 111).The use of hydrochloric acid of concentration greater than 1.2 M also eliminated the interference but reduced the sensitivity of the method. a solution of this mixture was prepared. TABLE I11 EFFECT OF ACIDITY, SODIUM CHLORIDE AND PROTEIN HYDROLYSATE ON THE COLOUR Nitrosation was conducted at room temperature and diazotisation at 10 "C for 100 pg of DEVELOPMENT OFTRYPTOPHAN tryptophan. Results are absorbances at 530 nm Concentration of r--Ap- 7 protein HC1 acidity/M hydrolysate,, h 1 NaCl per cent. 0-5 0.7 0.9 1.0 1.2 1.4 1.8 2.5 0 0 0.340 0.335 0.340 0-325 0.325 0.300 0.260 0.185 2 M , added a t the 0 0.265 0.270 0.260 0.260 0.260 0.210 0.160 0.120 initial stage nitrosation 2 M, added after 0 - - - 0.310 0.315 0.295 0.250 0.160 0 1 0.380 0-400 0.335 0.330 0.325 0.295 0.255 0.190 2 M, added a t the 1 0.280 0.280 0.265 0.260 0-255 0.215 0.165 0.120 initial stage660 GOSWAMI SPECIFICITY OF THE METHOD- Twenty common amino-acids other than tryptophan, including proline and histidine, which are secondary amines, did not give colours by this method.However, the method was found to be applicable to indole and its derivatives in addition t o tryptophan. Thus indole, skatole and indoleacetic acid were found to give colours, mostly ranging from pink to deep red. Surprisingly, phenol gave a colour while tyrosine, the only phenolic amino-acid, gave no colour at concentrations up to 100 pg ml-l. Cresols and aromatic amines also gave colours. REFERENCES 1. 2. 3. 4, Eckert, H. W., J . Biol. Chem., 1943, 148, 208. Smith, A. M., and Agiza, A. H., J . Sci. Fd Agric., 1951, 2, 508. Sidgwick, N. V., “The Organic Chemistry of Nitrogen,” Third Edition by Miller, I. T., and Springal, Peircey, M. R., and Ward, E. R., J . Chew. Soc., 1962, 3841. H. D., Clarendon Press, Oxford, 1966, p. 593. Received December 31st, 1973 Amended April23rd, 1974 Accepted April 29th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900657

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, October, 1974, Vol. 99, pp. 661-665 661 The Determination of Tannins with Cerium( IV) Sulphate By M. KAPEL AND R. KARUNANITHY* (Procter Department of Food and Leather Science, University of Leeds, Leeds, LS2 9 JT) A method for the determination of tannin in tea samples is described. The technique is based on the oxidation of tannins with an excess of cerium(1V) sulphate and subsequent determination of the excess by back-titration with ammonium iron( 11) sulphate solution. Although this method and the Lowenthal procedure appear to determine the same tannins, the precision of the former is f 1-24 per cent., while that of the latter is only -f 10.12 per cent. A CONVENIENT method for the determination of tannins in foodstuffs has long been required. Although numerous methods appear in the literat~re,l-~ Procter’s modification‘ of that described by Lowenthal is generally favoured and has remained in use for almost a century. This technique is, however, beset with serious difficulties, particularly those deriving from the indistinct and unreliable end-point of the titration. This end-point is difficult to determine, not only because of the nature of the colours involved, but also, as the present investigation has confirmed, because of a degree of time dependence.Thus, the actual position of the end-point can be altered by variation of the rate of titration. Essentially, the Lowenthal method consists in titration of an aqueous extract of the plant in question with potassium permanganate solution. As this reagent oxidises many other substances in addition to polyphenols, the titration is repeated on a separate aliquot from which tannins have been removed with gelatin and kaolin, the polyphenol content then being calculated by difference.The number of other substances oxidised by the potassium per- manganate is further controlled by the addition of measured amounts of indigo carmine to the two titrations. This material, which is oxidised fairly slowly by permanganate, acts as an indicator and also ensures that only those substances which react more rapidly with the titrant influence the result. Nevertheless, it is clear that the rate of titration will affect the volume of permanganate solution required, a conclusion that is amply corroborated by the experiment a1 results. Inouye and Okamura5 sought to overcome this difficulty by the use of excess of cerium(1V) sulphate, followed by an iodimetric back-titration.Their interest was in certain barks containing polyphenols that are used for leather tanning, so that they were not concerned with the conversion factors required for the application of the method to foodstuffs. Back- titration is effected by means of ammonium iron(I1) sulphate solution and a conversion factor suitable for the determination of tannins in tea has been worked out. The precision of the method has been thoroughly investigated and found to compare very favourably with the Lowenthal procedure. The method has also been investigated for the analysis of a com- mercial sample of tannic acid. The present investigation is also based upon a cerium(1V) sulphate method.EXPERIMENTAL REAGENTS- All reagents were of analytical-reagent grade. Gelatin soZzdion-Soak 25 g of gelatin in saturated sodium chloride solution for 1 hour. Then, when it is cool, make the volume up to Acidic sodium chloride solution-Add 25 ml of concentrated sulphuric acid to 975 ml of Kaolin-This substance conformed with the specifications laid down by the Society of Warm the mixture until the solid has dissolved. 1 litre with saturated sodium chloride solution. saturated sodium chloride solution. Leather Trades’ Chemists.6 Present address: School of Pharmacy, University of Singapore, Sepoy Line, Singapore 3. @ SAC and the authors.662 KAPEL AND KARUNANITHY: THE DETERMINATION [Analyst, Vol. 99 Cerium(1V) sulphate solution, 0.1 N-Prepare the solution from purified ammonium cerium(1V) nitrate.Eliminate nitrate ions by precipitation of cerium(1V) hydroxide with ammonia and subsequent dissolution in 5 N sulphuric acid. Standardise the final solution with arsenic(II1) oxide. Ammonium iron(II) sulphate solution, 0.1 N. Sulphuric acid, 5 N. 1,lO-Phenanthroline - iron(II) su1Phate complex, 0-025 M-Dissolve 1.485 g of 1,lO- phenanthroline monohydrate in 100 ml of 0.025 M iron(I1) sulphate solution. PREPARATION OF THE EXTRACT- Five grams of tea were boiled with 400 ml of water for 1 hour, the mixture then being filtered through a Buchner funnel containing a Whatman No. 1 filter-paper. When cool, the filtrate was transferred quantitatively into a 500-ml calibrated flask and made up to the mark with water.REMOVAL OF TANNINS- A 100-ml volume of extract A was transferred by pipette into a 250-ml calibrated flask; 50 ml of gelatin solution were added and the mixture was made up to the mark with acidic sodium chloride solution. The mixture was shaken well and poured into a conical flask containing 20 g of kaolin; the flask was shaken well for 15 minutes and the contents were filtered. The resulting liquid is referred to below as extract A. The filtrate is referred to below as solution B. 25 ml of solution B = 10 ml of solution A = 0.1 g of sample. CERIUM@) SULPHATE METHOD- A 10-ml volume of extract A was transferred by use of a pipette into a 1-litre beaker and diluted to about 600 ml with water; 10 ml of 5 N sulphuric acid and 25 ml of cerium(1V) sulphate solution were then added.After 1 hour, 3 drops of 0.025 M 1,lO-phenanthroline - iron(I1) sulphate complex were added and the mixture was titrated with 0.1 N ammonium iron(I1) sulphate solution until the first sharp colour change to orange - red occurred. The process was repeated with 25 ml of solution B. The determination on each of the samples was carried out in duplicate. MODIFIED LOWENTHAL METHOD- followed. The Lowenthal method as modified by the Indian Tea Research Station in Assam7 was EFFECT OF REACTION TIME IN THE CERIUM(IV) SULPHATE METHOD- A sample of tea extract was analysed by the cerium(1V) sulphate method. The sample solution was left in contact with cerium(1V) sulphate for varying periods of time before back- titration with ammonium iron(I1) sulphate solution, and it was found that a minimum reaction time of 45 minutes was necessary to complete the oxidation.A reaction time as long as 2 hours did not cause any further change in the result. Accordingly, a reaction time of 1 hour is recommended for the method. COMPARISON OF THE CERIUM@) SULPHATE AND MODIFIED LOWENTHAL METHODS- The results obtained from seven different commercial samples of black tea and seven different samples of tannic acid solution prepared from BDH Chemicals Ltd. tannic acid are given in Tables I and 11, respectively. The correlation coefficient for tea samples was 0.9952 and that for tannic acid solution was 0.9861. These are excellent correlation values, a fact which may also be noted from a study of the corresponding regression lines in Figs.1 and 2. An analysis of variance on these results yielded an F ratio of 1.6478 for the tea samples and 1.5360 for tannic acid solution. In both instances, the number of degrees of freedom was 1 for the greater variance estimate and 12 for the lesser. From the table for the variance ratio with these degrees of freedom, the 20 per cent. level of F is found to be 1.84. Clearly, on the hypothesis that the two methods are essentially equivalent, one could expect chance variations to produce the observed variance ratios of 1.6478 and 1.5360 in more thanOctober, 19741 OF TANNINS WITH CERIUM(IV) SULPHATE 663 20 per cent of the instances. Accordingly, the hypothesis of equivalence can be considered justified for the range investigated, The question of concentrations below this level is discussed later.TABLE I COMPARISON OF MODIFIED LOWENTHAL AND CERIUM(IV) SULPHATE METHODS FOR TEA SAMPLES Sample number 1 2 3 4 5 6 7 Modified Lowenthal method 0.04 N KMnO,/ Tannin, ml per cent. 3.30 5.49 3.95 6.57 4.35 7.24 5.15 8-57 5.50 9.15 6.65 11.07 7.35 12.23 r Cerium(1V) sulphate method f- 0.1 N Ce(SO,),/ ml 4.35 5.05 5.15 5.95 6.65 7-80 8.40 7 Tannin, per cent. 6.05 7.02 7-16 8.27 9.24 10.84 11-68 TABLE I1 COMPARISON OF MODIFIED LOWENTHAL AND CERIUM(IV) SULPHATE METHODS FOR TANNIC ACID SOLUTIONS Modified Lowenthal method Cerium(1V) sulphate method W r R 1 Sample 0.04 N KMnO,/ Tannin, 0.1 N Ce(SO,),/ Tannin, number ml per cent. ml per cent. 1 5.15 8.57 6.15 8-92 2 3.48 5.79 4.37 6.34 3 5-73 9.53 6.60 9.57 4 6-40 10.65 7.25 10.51 5 6.95 11-56 8.00 11-60 6 6.80 11.32 7-50 10.88 7 7.35 12.23 8.10 11.75 On the basis of the arguments explained above, it was concluded that the results of the two methods show a satisfactory degree of correlation. Accordingly, the factor to be used in the calculation of tannin content from the results of the cerium(1V) sulphate procedure is derived from that of the Lowenthal method on a proportional basis.The factors thus calculated are as follows. For tea samples : For tannic acid: 1 ml of 0-1 N cerium(1V) sulphate = 0.001 39 g of tannin (gallotannic acid). 1 ml of 0.1 N cerium(1V) sulphate = 0.001 45 g of tannin (gallotannic acid). By means of these factors, the tannin contents of the tea samples and the tannic acid solutions were calculated. They are given in Tables I and 11, respectively.Clearly, any other type of sample could be subjected to this analytical procedure only after determination of the appropriate factor. The precisions of the two methods were compared by means of a series of ten deter- minations on a sample of tea, with both procedures. The standard deviation and the coefficient of variation of the titration before removal of tannins in the Lowenthal method are 0.534 and 3.315 per cent., respectively, while the corresponding figures for the titration after removal of tannins are 0-473 and 5.209 per cent. Similar calculations for the cerium(1V) sulphate method give results of 0.079 and 0-424 per cent. in the first titration and 0-070 and 0.689 per cent. in the second. It can be seen that, in each method, both sources of variation contribute to the error to about the same extent, the second titration possibly having a slightly greater influence than the first.From the mean values and variances of both titrations, the over-all coefficient of variation of each method has been calculated and found to be 10.12 per cent. for the modified Lowenthal procedure, but only 1.24 per cent. for the cerium(1V) sulphate technique. This criterion clearly establishes the superiority of the cerium( IV) sulphate method.664 KAPEL AND KARUNANITHY: THE DETERMINATION [A%a&St, VOl. 99 0 1 2 3 4 5 6 7 8 0.04 N KMn04/ml Modified Lowenthal method Fig. 1. Regression line of cerium(1V) sul- phate method on modified Lowenthal method for tea samples. Value of 0.1 N Ce(SO,), solution: x , predicted; and 0, observed Owing to the large coefficient of variation of the Lowenthal method, its usefulness for determining very low concentrations of tannin is doubtful.For this reason the cerium(1V) sulphate procedure could not be compared with the Lowenthal method for low tannin contents. In fact, Hermann and Enge8 have pointed out that it is not possible to determine tannin contents of 1 per cent. or less with reasonable accuracy by any method. Because of the uncertainty over the position of the end-point, it is recommended that, for materials known to be low in tannin, a larger aliquot of extract A be used for the analysis. The regression lines for tea samples and tannic acid (Figs. 1 and 2) show small intercepts, the presence of which require an explanation. The cause of the phenomenon may possibly I I I I I I I ’ 1 2 3 4 5 6 7 8 0.04 N KMn04/ml Modified Lowenthal method Fig.2. Regression line of cerium(1V) sul- phate method on modified Lowenthal method for tannic acid solution. Value of 0.1 N Ce(SO,), solution: x , predicted; and 0, observedOctober, 19741 OF TANNINS WITH CERIUM(IV) SULPHATE 665 be that the results of the Lowenthal method are achieved by direct titration with potassium permanganate, whereas in the cerium(1V) sulphate procedure the oxidising agent remains in contact with the sample for some time before the back-titration is complete. In view of the known time dependence of the oxidation in both instances, it is feasible to assume that the treatment with cerium(1V) sulphate involves some components of the sample which remain immune to the action of the potassium permanganate under the conditions of the experiment. 1. 2. 3. 4. 5. 6. 7. 8. REFERENCES Schanderl, S. H., in Joslyn, M. A., Editor, “Methods in Food Analysis,” Second edition, Academic Swain, T., and Goldstein, J. L., in Pridham, J. B., Editor, “Methods in Polyphenol Chemistry,” King, H. G. C., and White, T., in “The Chemistry of Vegetable Tannins, A Symposium,” Society Procter, H. R., J . Soc. Cham. Ind., 1884, 3, 82. Inouye, Y., and Okamura, H., Nappon Hikaku Gijutsu Kyokaishi, 1957, 3, 12. “Official Methods of Analysis of the Society of Leather Trades’ Chemists,’’ Society of Leather Pearson, D., “The Chemical Analysis of Foods,” Churchill, London, 1970, p. 280. Hermann, K., and Enge, W., Dt. ApothZtg, 1959, 99, 325. Press, London, 1970, p. 701. Pergamon Press, Oxford, 1964, p. 131. of Leather Trades’ Chemists, Croydon, 1956, p. 31. Trades’ Chemists, Croydon, 1965, SLT 2/1. Received March 25th, 1974 Accepted May 30th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900661

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

666 Analyst, October, 1974, Vol. 99, @. 666-669 The Detection and Determination of Residues of the Herbicide Nitrofen in Vegetables BY J. KVALVAG (Chemical Research Laboratory, Agricultuval University of Norway, 1432 A*s-NLH, Norway) A method for residue analysis, based on electron-capture gas chromato- graphy following clean-up on a sulphuric acid - Celite column, is presented. The limit of detection is 10 ngg-l. Recoveries have been investigated for different kinds of vegetables. The extraction efficiency of acetonitrile uersm dichloromethane was checked in a separate experiment. Spectroscopic data and results for the herbicide are included. THE herbicide nitrofen (2,4-dichlorophenyl 4’-nitrophenyl ether) has been thoroughly tested as an agent for the chemical control of weeds.The method for residue analysis described by the manufacturer, Rohm and Haas, is based on extraction with dichloromethane, clean-up on a Florisil column and, finally, determination by means of gas chromatography.1 For crops containing much oil and crops that are rich in carotene, additional clean-up steps have to be carried out. The U.S. Food and Drug Administration2 lists nitrofen among the pesticides that pass through the fuming sulphuric acid - Celite column that was first applied by Davidow3 to the determination of DDT in fatty materials, and as a clean-up method that is applicable to all kinds of material is desirable, an investigation of the modified Davidow column described by Jensen and co-workers4 was undertaken. This investigation resulted in a residue analysis method, the procedure fur which is given in this paper.SPECTROSCOPIC INVESTIGATION OF THE STANDARD SUBSTANCE- A standard sample of nitrofen was kindly supplied through the manufacturer’s European representative, Minoc S.A., Pans, France. To obtain confirmation of the structure for this substance a number of spectrochemical methods were applied. Infrared spectroscopy and ultraviolet spectroscopy were carried out : the principal infrared absorption bands were found at frequencies of 1250, 1262 and 1340 cm-l; the ultraviolet spectrum had a single absorbance maximum at 292 nm, as has been reported by Dahlgard and Brewster. Nuclear magnetic resonance spectra were observed on a Varian HA-100 nuclear magnetic resonance spectrometer under the following conditions: 100 mg of the standard sample were dissolved in 0.7 ml of tetrachloromethane while an equal amount was dissolved in deutero- benzene-d,.The spectra were recorded at a frequency of 100 MHz with reference to tetramethylsilane as standard. The sweep width was 1000 Hz, and the field of interest was re-scanned at a sweep width of 250 Hz. The results are given in Table I. TABLE I RESULTS OF NUCLEAR MAGNETIC RESONANCE SPECTROSCOPIC INVESTIGATION OF NITROFEN Solvent Position/Hz cc1, 691 709 730 749 813 637 639 679 714 779 Form Doublet Doublet Quartet Doublet Doublet Doublet Doublet Quartet Doublet Doublet Coupling constant/Hz 9 8.5 2 - 5 ; 8.5 2.5 9 9 8.5 2 - 5 ; 8.5 2.5 9 Intensity 2 1 1 1 2 2 1 1 1 2 The spectra obtained can be interpreted as resulting from 2,4-dichlorophenyl 4’-nitro- phenyl ether, even though it is somewhat curious that para-coupling between the protons in the 3- and 6-positions of the chlorine substituted ring was not observed.@ SAC and the author.KVALVAG 667 RESIDUE ANALYSIS METHOD APPARATUS- Varian-Aerograph gas chromatograph, Model 20GThis instrument was equipped with a hydrogen-3 electron-capture detector and a glass column, 5 foot x Q inch id., packed with 3 per cent. QF-1 fluorosilicone on 100 to 120-mesh Gas-Chrom Q. Chromatographic columns of 2 cm id. REAGENTS- sulphuric acid with 30 g of Johns-Manville Celite 545 in a mortar. Sulphuric acid - Celite mixture-Mix 20 ml of analytical-reagent grade concentrated Acetonitrile and light petroleum (boiling range 40 to 60 "C), redistilled.PROCEDURE- Extract the pesticide by using the following method, which is based on procedures of Getz,G Mills, Onley and Gaither' and Nelson,* with modifications. The solvent used in the extraction is acet oni t rile. Place 100 g of homogenised plant material into the container of a high-speed mixer and mix it with 200 ml of acetonitrile and 10 g of Celite for 5 minutes. Filter the mixture through a 7-cm glass Riichner funnel with the aid of suction. Transfer the filtrate to a 1-litre separating funnel. Wash the residue on the filter with 50 ml of acetonitrile and combine the washings with the extract in the funnel. Add 100 ml of light petroleum and shake the mixture for 1 minute. Next add 600 ml of water and 10 ml of saturated sodium chloride solution and shake the mixture for 2 minutes.Allow the two phases to separate, discarding the aqueous phase. Then, dry the light petroleum phase by passing it through a column of sodium sulphate and into a 100-ml calibrated flask. Wash the separating funnel with 10 ml of light petroleum,g add the washings to the calibrated flask through the drying column and adjust the contents of the calibrated flask to the mark by adding more light petroleum via the same route. Pack a chromatographic column to a depth of 10 cm with sulphuric acid - Celite mixture by adding it in small portions and tamping with a glass rod to consolidate the mixture in the column. Add an aliquot of the sample solution to the column and elute the sample with light petroleum to give a total volume of eluate of 100 ml.Concentrate the eluate to 10 ml in a Kuderna-Danish evaporation apparatus. Then, inject a 5-pl volume into a gas chromato- graph operated at the following temperatures: column, 195 "C; injector, 210 "C; and detector, 205 "C. As tailing, with the consequent loss of linearity, is especially troublesome on old gas- chromatographic columns when used for the analysis of nitrofen, it is preferable to use newly conditioned columns and to change the column packing frequently. With a freshly packed column, injections of standard of 250 pg or less give a satisfactory response. Take out 5 ml of solution from the flask for clean-up. Compare the results with those given by the nitrofen standard. RECOVERIES AND PRECISION OF THE METHOD- Solutions of nitrofen containing 5 or 10 pg of the herbicide were added to different kinds of vegetables in the container of the mixer, there being 100 g of plant material.As interest was focused on the use of the sulphuric acid - Celite column and the precision obtainable with use of this clean-up stage, six aliquots of the light petroleum phase from each extraction were carried through the clean-up and determinative stages of the procedure. The results are given in Table 11. Because no residues above the detection limit of 10 ng g-l were discovered in any of the crops investigated, the following experiment was carried out. Nitrofen was applied to Swedish turnips (Brassica napobrassica L.) at the cotyledon stage at a rate of 2 kg ha-l.* The plants were harvested 1 week after spraying, then homogenised, and divided to give four samples.Two samples were extracted with dichloromethane and the nitrofen determined according to the manufacturer's method, while two samples were extracted with acetonitrile and the determination was carried out as described here. The content of herbicide residue was about 3 pg g-1 and no significant difference was observed between the extraction methods. *1 ha=104 m2.668 KVALVAG : DETECTION AND DETERMINATION OF THE TABLE I1 [Analyst, Vol. 99 RECOVERIES OF NITROFEN ADDED TO PLANT MATERIAL AND RELATIVE STANDARD DEVIATION OF THE COMBINED CLEAN-UP AND DETERMINATION* Amount of nitrofen Recovery, Relative standard Crop addedlpg per lOOg per cent. deviation, per cent. Cabbage , . .. 10 5 Carrots . . . . 5 Cauliflower . . 5 Leek .. . . 10 5 Onion . . . . 5 Swedish turnips 5 *Based on six determinations on each crop. 93 92 113 87 90 92 99 92 3.3 2.8 1.8 4-6 2.7 3.2 8.5 4-1 DISCUSSION The manufacturer of nitrofen has found that many pesticides do not interfere in its determination. Moreover, as the sulphuric acid - Celite column is selective, interference from most other herbicides is ruled out by the present method anyway. However, toxaphene (containing polychlorinated camphenes) is not removed by the clean-up procedure and can interfere on the gas chromatogram. On the other hand, the determination of toxaphene is not appreciably affected by the presence of nitrofen as toxaphene is a mixture that produces a characteristic chromatographic pattern. Further, from a toxicity point of view, it is scarcely of interest to know whether a sample contains nitrofen when toxaphene has already been detected.Of greater environmental significance are the PCBs (polychlorinated biphenyls). On chromatograms resulting from this procedure, the nitrofen peak appears between the last peak (PCB-15) and the penultimate peak (PCB-14) of Chlophen A-50.1° The nitrofen peak overlaps slightly with that of PCB-14. The PCBs mentioned are, however, less abundant in samples presented for residue analysis than those with shorter retention times. Confirmation of the results can be obtained by injection of the sample on to another column, but the use of an independent method, for instance colorimetry, provides more convincing confirmatory results. A colorimetric procedure is described by the manufacturer, and nitrofen was included by Friestadll in a study of automated colorimetry.Both methods are based on reduction of the nitro group to an amino group, followed by colorimetry of the product resulting from diazotisation and coupling. The resistance to attack by sulphuric acid has been regarded as a suitable reaction for testing the persistency of pollutants.12 It is therefore interesting to note that nitrofen was not found in any of the crops investigated except Swedish turnips at the cotyledon stage, when harvested 1 week after spraying. The vapour pressure of nitrofen is relatively high (8 x 10-6 mm of mercury at 40 OC),13 but as pesticides with even higher vapour pressure are known to remain as residues, evaporation is unlikely to be the full reason for the general absence of residues.However, this absence becomes less remarkable when the photochemical lability found by Crosby and Nakagawa14 is taken into account. The forces binding pesticide residues to biological material differ in character and vary in polarity. As extraction procedures with a polar and a non-polar solvent gave the same result, the binding of the extracted residues is likely to have been weak and possibly the herbicide had not entered the plant tissue to any significant extent. The spectrochemical part of this work was carried out by the Central Institute for Industrial Research under the supervision of Jan Bder. Growing and spraying of the Swedish turnips was carried out by Tor J. Fiveland, Norwegian Plant Protection Institute.REFERENCES 1. 2. 3. 4. “Pesticide Analytical Manual,” Volume 11, U.S. Food and Drug Administration, Washington, D.C., Op. cit., Volume I, 1971, Table 201-B. Davidow, R., J . Ass. 08. Agric. Chem., 1950, 33, 130. Jensen, J . A., Cueto, C., Dale, W. E., Rohte, C. F,, Pearce, G. W., and Mattson, A. M., J . Agric. 197 1. Fd Chem., 1957, 5, 919.October, 19741 HERBICIDE NITROFEN IN VEGETABLES 669 5. 6. 7. 8. 9. 10. 1 1 . Dahlgard, M., and Brewster, R. Q., J . Amer. Chem. SOC., 1958, 80, 5861. Getz, M. E., J . Ass. OH. Agric. Chem., 1962, 45,393. Mills, P. A., Onley, J. H., and Gaither, R. A., Ibid., 1963, 46, 186. Nelson, R. C., Ibid., 1964, 47, 289. Munks, B., “Pesticide Residue Manual,” Florida Department of Agriculture, Tallahassee, 1965, p. 850. Jensen, S., Ambio, 1972, 1, 123. Friestad, H. O., “Methods in Residue Analysis,” in Tahori, A. S., Editor, “Proceedings of the Second International IUPAC Congress of Pesticide Chemistry,” Volume IV, Gordon and Breach Science Publishers, New York, 1971, p. 299. 12. 13. 14. Lunde, G., Gether, J., and Josefsson, B., in preparation. Perkow, W., “Wirksubstanzen der Pflanzenschutz- und Schadlingsbekampfungsmittel,” Verlag Crosby, D. G., and Nakagawa, M., Abstr. 162nd Meet. Amer. Chem. Soc., Washington, D.C., 1971. Paul Parey, Hamburg, 1971. Received March 5th, 1974 Accepted May Sth, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900666

出版商:RSC    

年代:1974

数据来源: RSC