摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRYADVISORY BOARD"Chairman: J. M. Ottaway (Glasgow, U.K.)"L. S. Bark (Salford, U.K.)E. Bishop (Exeter, U.K.)W. L. Budde ( U S A . )D. T. Burns (Belfast, U.K.)L. R. P. Butler (South Africa)H. J. Cluley (Wembley, U.K.)E. A. M. F. Dahmen (The Netherlands)L. de Galan (The Netherlands)A. C. Docherty (Billingham, U.K.)D. Dyrssen (Sweden)G. Ghersini (ltaly)J. Hoste (Belgium)A. Hulanicki (Poland)"G. W. Kirby (Glasgow, U.K.)W. S. Lyon (U.S.A.)H. V. Malmstadt (U.S.A.)G. W. C. Milner (Harwell, U.K.)"A. C. Moffat (Aldermaston, U.K.)E. J. Newman (Poole, U.K.)H. W. Nurnberg (West Germany)"T. B. Pierce (Harwell, U.K.)E. Pungor (Hungary)P. H. Scholes (Middlesbrough, U.K.)D. Simpson ( Thorpe-le-Soken, U.K.)"J.M. Skinner (Billingham, U.K.)"J. D. R. Thomas (Cardiff, U.K.)K. C. Thompson (Sheffield, U.K.)"A. M. Ure (Aberdeen, U.K.)A. Walsh, K.B. (Australia)G. Werner (German Democratic Republic)T. S. West (Aberdeen. U.K.)'P. C. Weston (London, U.K.)"J. Whitehead (Stockton-on-Tees, U.K.)J. D. Winefordner (U.S.A.)P. Zuman (U.S.A.)"G. J. Dickes (Bristol, U.K.)"Members of the Board serving on the Analytical Editorial BoardEditor: P. C. WestonSenior Assistant Editor: R. A. YoungAssistant Editors: Mrs. J. Brew, Miss D. ChevinREGIONAL ADVISORY EDITORSDr. J. Aggett, Department qf Chemistry, University of Auckland, Private Bag, Auckland, NEW ZEALAND.Professor L. Gierst, Universit6 Libre de Bruxelles, Facult6 des Sciences, Avenue F.- D.Roosevelt 50,Professor H. M. N. H. Irving, Department of Theoretical Chemistry, University of Cape Town, Ronde-Professor W. A. E. McBryde, Faculty of Science, University of Waterloo, Waterloo, Ontario, CANADA.Dr. 0. Osibanjo, Department of Chemistry, University of Ibadan, Ibadan, NIGERIA.Dr. G. Rossi, Chemistry Division, Spectroscopy Sector, CEC Joint Research Centre, EURATOM, lspraDr. 1. Rubeska, Geological Survey of Czechoslovakia, Malostranske 19, 11 8 21 Prague 1, CZECHO-Professor J . R&ikka, Chemistry Department A, Technical University of Denmark, 2800 Lyngby,Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN.Professor L. E. Smythe, Department of Chemistry, University of New South Wales, P.O. Box 1,Professor P.C. Uden, Department of Chemistry, University of Massachusetts, Amherst, MA 01 003,Editorial: Editor, The Analyst, The Royal Society of Chemistry, Burlington House,Piccadilly, London, WIV OBN. Telephone 01 -734 9864. Telex No. 268001Advertisements: Advertisement Department, The Royal Society of Chemistry, Burlington House,Piccadilly, London, W1 V OBN. Telephone 01 -734 9864. Telex No. 268001The Analyst (ISSN 0003-2654) is published monthly by The Royal Society of Chemistry, BurlingtonHouse, London W I V OBN, England. All orders accompanied with payment should be sent directly toThe Royal Society of Chemistry, The Distribution Centre, Blackhorse Road, Letchworth, Herts. SG6 1 HN,England. 1983 Annual subscription rate UK f93.50, Rest of World €99.00, USA $201 .OO. Purchased withAnalytical Abstracts UK f226.50, Rest of World f238.50, USA $487.00. Purchased with AnalyticalAbstracts plus Analytical Proceedings U K f 251 .OO, Rest of World f 265.00, USA $539.00. Purchasedwith Analytical Proceedings UK fl17.50, Rest of World f 124.50, USA $253.00. Air freight and mailingin the USA by Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11003.USA Postmaster: Send address changes to: The Analyst, Publications Expediting Inc., 200 MeacharnAvenue, Elmont, NY 11003. Second class postage paid at Jamaica, NY 11431. All otherdespatches outside the UK by Bulk Airmail within Europe, Accelerated Surface Post outside Europe.PRINTED IN THE UK.Volume 108 No 1286 0 The Royal Society of Chemistry 1983 May 1983Bruxelles, BELGIUM.bosch 7700, SOUTH AFRICA.Establishment, 21 020 lspra (Varese), ITALY.SLOVAK I A.DENMARK.Kensington, N.S.W. 2033, AUSTRALIA.U.S.A

ISSN:0003-2654    

DOI:10.1039/AN98308FX017

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

ANALAO 108 (1 286) 553-648 (1 983) May 1983THE ANALYST55357358159159760360861 5621626633636639641645648THE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRYCONTENTSPlatform Atomisation in Carbon Furnace Atomic-emission Spectrometry-Laszlo Bezur, John Marshall, John M. Ottaway and Riad Fakhrul-AldeenIdentification and Determination o f Microgram Amounts of Phenol and Chloro-phenols with Methylbenzothiatolinone Hydrazone Using the Ring-ovenTechnique-Nelly G. Buckman, John 0. Hill and Robert J. MageeCalcium lon-selective Electrode Studies: Covalent Bonding o f Organic Phos-phates and Phosphonates t o Polymer Matrices-P. C. Hobby, G. J. Moody andJ. D. R. ThomasDetermination o f the Moisture Content o f Starch Using Near Infrared Photo-acoustic Spectroscopy-Peter s.Belton and Steven F. TannerSubstitution in Cellulose Ethers. Determination o f the Distribution o fAlkoxyl Substituents on the Glucose Units Using High-performance LiquidChromatography-Konrad Sachse, Klaus Metzner and Thomas WelschDevelopment and Evaluation of a Radioimmunoassay f o r t h e Detection ofAmphetamine and Related Compounds in Biological Fluids-Peter Alan Mason,Tara Singh Bal, Brian Law and Anthony Christopher MoffatSpectrophotometric Determination o f Trace Amounts o f Quaternary AmmoniumSalts in Drugs by lon-pair Extraction with Bromophenol Blue and Quinine-Tadao SakaiSimple Method f o r the Determination o f Gypsum, with Some Observations onthe Solubilities o f Gypsum, Anhydrite, Calcite and Dolomite-Henry A.Foner and Sarah EhrlichEffect of On-line Complex Formation Kinetics on the Flow Injection AnalysisSignal : the Spectrophotometric Determination o f Chromium(V1)-JoaoCarlos de Andrade, Julio Cesar Rocha, Celio Pasquini and Nivaldi BaccanTitrimetric Determination o f Some N-Substituted Phenothiazine Derivatives-Mohamed lbrahim Walash, Mohamed Rizk, Abdel-Malik Abou-Ouf and Fathalla BelalPart II.SHORT PAPERSLigand-exchange Reactions in the Analysis o f Cobalt Complexes by Atomic-absorption Spectrophotometry-W. James Swindall, Duncan Thorburn Burnsand Elmugdad A.AliDifferential Determination o f Cationic and Anionic Surfactants in Mixtures byTwo-phase Titration-Masahiro Tsubouchi and John H. MalloryAdsorption o f Metals on Polypropylene During Cold Vapour Atomic-absorptionDetermination o f Mercury-Hart B.MacPherson and Shier S. BermanSpectrophotometric Study o f the Reaction o f Titanium(lV) with Chloro-phosphonazo I-Qiu Xing-chu and Zhu Ying-quanBOOK REVIEWSERRATUMSummaries o f Papers in this issue-Pages iv, v, vi, vii, viii, ix, xPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class at New York. USA, Post OfficNEWSTAINLESS STEELCERTIFIED REFERENCEMATERIALSavailable fromBUREAU OF ANALYSEDSAMPLES LTDFor full details of these, and for copiesof new catalogue of all CRMs suppliedby BAS, write, telephone or telex to:BAS Ltd., Newham Hall, Newby,Middlesbrough, Cleveland, TS8 9EATelephone: Middlesbrough 31 721 6Telex: 587765 BASRIDTHE QUEEN'SUNIVERSITYOF BELFASTMSc COURSE inI ANALYTICAL CHEMISTRYApplications are invited for admission to thisestablished 12 month full-time MSc coursewhich provides a comprehensive training inthe theory and practice of modern chemicaland instrumental methods of analysis.Applicants should normally possess anhonours degree (or equivalent) in chemistryor cognate subjects.Part-time courses areavailable.The Science and Engineering ResearchCouncil has recognised the course for tenureof its Advanced Course Studentships.A description booklet and application formscan be obtained from Professor D. ThorburnBurns, Dept. of Chemistry, Queen's Universityof Belfast, Belfast BT7 1 NN, Northern Ireland.A202 for further information.See page x A203 for further information. See page xSpecial Publication No. 44Pollution: Causes, Effectsand ControlEdited by Roy M. HarrisonThe chapters in this book are in the main derivedfrom the course notes provided by lecturers at anR.S.C. Residential School on this topic held atLancaster University in September, 1982. Thesehave been supplemented by a few additionalcontributions aimed at improving the overallcoverage of this very broad subject area. TheResidential School had a teaching function andthe chapters are pitched at a level appropriate tothis objective.This book will therefore be of value not only toteachers and students but also to scientists andtechnologists working in the field of pollution.Softcover 330pp 0 85186 875 4 Price f12.00CONTENTSThe Control of Industrial Pollution; Water Quality and Health;Aspects of the Chemistry and Analysis of Substances ofConcern in the Water Cycle; The Role of WastewaterTreatment Processes in the Removal of Toxic Pollutants;Sewage and Sewage Sludge Treatment; The chemistry ofMetal Pollutants in Water; Effects of Pollutants in the AquaticEnvironment; Important Air Pollutants and Their ChemicalAnalysis; Pollutant Pathways in the Atmosphere;Atmospheric Dispersal of Pollutants and Modelling of AirPollution; Legislation and the Control of Air Pollution;Catalyst Systems for Emission Control from Motor Vehicles;Evaluating Pollution Effects on Plant Productivity: ACautionary Tale; Epidemics of Non-infectious Disease;Systems Methods in the Evaluation of EnvironmentalPollution Problems; Organometallic Compounds in theEnvironment.($21.00) RSC Members f8.00~ ~~ ~ , RSC members should send their orders to: The Royal Society of Chemistry, The Membership Officer,l 30 Russell Square, London WClB 5DT.Non-RSC members should send their orders to: The RoyalThe Royal Society of ChemistryBurlington HouseLondon W1V OBNI istry, Distribution Centre, Blackhorse Road, Letchworth, Herts SG6 1 HN.

ISSN:0003-2654    

DOI:10.1039/AN98308BX019

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

iv SUMMARIES OF PAPERS I N THIS ISSUE May, 1983Summaries of Papers in this IssuePlatform Atomisation in Carbon Furnace Atomic-emission SpectrometryThe design and operational characteristics of platforms used in carbon furnaceatomic-emission spectrometry are described. Platforms made of graphite andcoated with pyrolytic graphite give significant improvements in the sensitivityand detection limits achieved for 23 out of 24 elements investigated. Thedelay in atomisation allows the production of the atomic vapour a t a time whenthe vapour phase inside the furnace atomiser has reached higher temperatures.Improvements in signal to noise ratio have also been achieved through use of ahigh-resolution bchelle spectrometer, a novel square-wave wavelength motlula-tion system and the better separation of the analyte signals from blank signals.Keywords : Carbon furnace atomic-emission spectrometry ; Platfornz atomisationLASZLO BEZUR, JOHN MARSHALL, JOHN M.OTTAWAY and RIADDepartment of Pure and Applied Chemistry, University of Strathclydc, CathedralStreet, Glasgow, G1 1XL.FAKHRUL- ALDEENAnalyst, 1983, 108, 553-572.Identification and Determination of Microgram Amounts of Phenoland Chlorophenols with Methylbenzothiazolinone HydrazoneUsing the Ring-oven TechniqueThe ring oven, in conjunction with thc reagent metl~ylbenzothiazolinonehydrazone, has been used for the determination of cl~loroplienols in water.Four different methods were employed : aqueous in sitzi solvent extractionusing a single and a mixed organic solvent, in situ total cxtraction of a mixtureof ten phenols and in situ separation of binary plicnol mixtures.Thesemethods appear to be suitable for the determination of phenols in water with adetection limit of 2 x 10-5 mg 1-1 for solvent extraction and lW4 mg 1-1 for thein situ methods, the detection limit being defined as the concentration forwhich no visible ring is formed for use a t a lower limit on the standard scale.Keywords : Water analysis ; ring-oven technique ; metkylbenzothiazolinonehydrazone ; phenol and chlorophenol determinationNELLY G. BUCKMAN, JOHN 0. HILL and ROBERT J. MAGEEDepartment of Inorganic and Analytical Chemistry, La Trobe University, Bundoora,Melbourne, Australia.Analyst, 1983, 108, 573-580May, 1983 SUMMARIES OF PAPERS I N THIS ISSUECalcium Ion-selective Electrode Studies : Covalent Bonding ofOrganic Phosphates and Phosphonates to Polymer MatricesThe covalent linking of organophosphates and phosphonates to the copolymerVAGH (a partially liyclrolysecl copolymer of vinyl chloride and vinyl acetate)is describcrl along with attempts to phosphonate polystyrene by Friedel -Crafts and free-radical processes.The various products obtained have beenevaluated for use in calcium ion-selective electrode membranes. The bestmembranes were all based on VAGH, a product from each type being selectedfor further study for their calcium ion-sensing qualities, namely, VAGH PI,VAGH PIT and VAGH PIII.VAGH PI is based on the covalent bonding of monodecylphosphate toVAGH, while VAGH PI1 resulted from the binding of mono-[4-( 1 , 1 3 3 -tetramethylbutyl)plienyl]pIiosphate. Viable electrode membranes requiredadditional poly (vinyl chloride) support and dioctyl phenylpliosphonatesolvent mediator.VAGH PI11 was obtained by a synthesis amounting tocovalently binding monooctyl phenylphosphonate to VAGH and was usedwith some additional dioctyl phenylphosphonate solvent mediator and calciumbis{di-[4-( 1,1,3,3-tetramethylbutyl)phenyl]pliosphate) sensor in electrodemembranes.The VAGH PI electrode matched those based on equivalent ungrafteddidecylphospliate sensors, except that calcium ion selectivity was poor in ahigh sodium ion background. The VAGH PI1 electrode, on the other hand,did not fulfil the performance of electrodes based on free calcium bis{di-[4-(1,1,3,3-tetramethylbutyl)phenylJphospliate} as ion sensor, and there wasless selectivity for calcium over sodium and magnesium. Some loss ofselectivity was also characteristic of the VAGH PI11 electrode, but itsslightly lengthened lifetime was insufficient to warrant synthesis of thephosphonated polymer matrix.Keywords : Calcium ion-selective membrane electrode ; covalent bonding of ionsensors ; grafted ion sensorsP. C.HOBBY, G. J. MOODY and J. D. R. THOMASApplied Chemistry Department, Redwood Building, UWIST, Cardiff, CF1 3NU.Analyst, 1983, 108, 581-590.Determination of the Moisture Content of Starch UsingNear Infrared Photoacoustic SpectroscopyNear infrared photoacoustic spectroscopy has been used to measure the watercontent of wet starch samples (over the range 0-30% of water).The relation-ship between moisture content and observed intensity of the photoacousticsignal is analysetl in terms of the theory of the photoacoustic effect. It isshown that the non-linear dependence of signal intensity on moisture contentis consistent witli tlwory and that this non-linearity is to be generally expectedin high water content systems. A method of linearising the data is described.Keywords : Neav infrared ; photoacoustic spectroscopy ; moisture determination ;starchPETER S. BELTON and STEVEN F. TANNERAgricultural Research Council, Food Research Institute, Colney Lane, Norwich,NK4 7UA.VAnalyst, 1983, 108, 591-596vi SUMMARIES OF PAPERS I N THIS ISSUEDetermination of the Distribution of Alkoxyl Substituents on theGlucose Units Using High-performance Liquid ChromatographyMay, 1983Substitution in Cellulose Ethers.Part 11.A procedure is described for the quantitative characterisation of the distribu-tion of alkoxyl groups in methyl- and ethylcelluloses. The total hydrolysate isderivatised by reaction with benzoyl chloride in sodium hydroxide solution.Several favourable implications of pre-column derivatisation for the subsequentseparation by high-performance liquid chromatography are cliscussed.Isocratic separations are carried out on LiChrosorb Si100, 10 ,urn, usingchloroform with 0.1 and 0.2% V / V ethanol as the eluent. All eight individualglucose ethers present in the hydrolysates, including the positional isomers ofalkyl- and dialkylglucoses, can be determined quantitatively from the chroma-tograms.Thus, the procedure provides data on the distribution of substituentsaccording to their number and position in the anhydroglucose units.Keywords : Alkyl cellulose ; substituent distribution ; position of substituents ;$re-column derivatisation; high-pevformance liquid chromatograpliyKONRAD SACHSE and KLAUS METZNERKombinat VEB Chemische Werke Buna, DDR-42 12 Schkopau, German DemocraticRepublic.and THOMAS WELSCHKarl-Marx-Universitat Leipzig, Sektion Chemie, Analytisches Zentrum, Liebigstr.18, DDR-7010 Leipzig, German Democratic Republic.Analyst, 1983, 108, 597-602.Development and Evaluation of a Radioimmunoassay for theDetection of Amphetamine and Related Compoundsin Biological FluidsA radioimniunoassay has been devclopetl for the tlctcction of amplietarnine andits analogues in blood and urine without any prc-treatnicwt of the saniples.Itis based on a commercially available antiserum and a I lioclinated derivativeof amplietarnine.in \‘crysmall samples ( 5 0 ~ 1 ) of blood and urine. I t is c l i c ~ ~ p ( 3 pcnce p i - test), rapid,simple to perform and is specific for compounds clost~l!. rclatul to ~~nipl~c.taniiiie.A high, positive correlation was obtainctl ( r := 0.93) \vlic.n rcwilts o f theanalyses of urine samples from voluntecn \vlio Imtl ingestc(1 ~~rnl)lic.t~~~iiiiic wcrecompared with those produced by gas clironiatogi-~Il)li!~ - inass spcctrometry.The assay has proved very useful for tlic tlctcction of ~tniI’hi~t~Lliiiiic.~ andclosely related compounds in biological fluids.The assay can detect low levels of amplictamiiic~ (lws t l i n i i 10 ngKeywords Amfilietaiizine defection ; blood; w i n e ; i.nrlioiiiziiiunonssa.\lPETER ALAN MASON, TARA SINGH BAL, BRIAN LAW and ANTHONYCHRISTOPHER MOFFATCentral liesearch Establishmcnt, Honic Otlicc Forensic Scicncc Service, .\ldcrmaston,Reading, Berkshire RG7 41”.Analyst, 1983, 108, G03-607

ISSN:0003-2654    

DOI:10.1039/AN98308FP049

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

May, 1983 SUMMARIES OF PAPERS I N THIS ISSUESpectrophotometric Determination of Trace Amounts ofQuaternary Ammonium Salts in Drugs by Ion-pairExtraction with Bromophenol Blue and QuinineThe sensitive and selective method reported here is based on the reaction ofquinine cations with bromoplienol blue (HPJ3) to form a 1 : 1 complex anion,which, with a quaternary ammonium salt a t pH 0.7, is extractable intochloroform as an ion pair. Berberine and benzethonium were extracted inlarge amounts into chloroform with UPR only when quinine, in a pH 6.7buffered aqueous solution, co-existed with them. In adtlition, when tliisshowed a strong ion associate ability, then bcrbcrine antl benzcthonium couldreact in the organic phase with 131’13 to give a blue product. Tlie absorbanceof the ion associate in chloroform was measured a t 610 nm. Calibrationgraphs were linear in tlie range 0.5 x 10-6-3.0 x M for berbcrine andbenzetlionium.Molar absorptivities were 3.75 x lo4 1 mol-l cm-l for theberberine associate and 3.14 x lo4 1 mol-l cm-l for the benzethoniumassociate. The blue associate can be used for selective and sensitive spectro-photometric determination of berberine and benzethonium in multi-com-ponent drugs.Keywords : Rerberine nssay ; benzethoniunz assay ; co-extraction spectrophoto-metry ; bronzopkenol blue - quinine associateTADAO SAKAIDepartment of Chemistry, Gifu College of Dentistry, 185 1 Takano, Hozumi-cho,Gifu 501-02, Japan.Analyst, 1983, 108, 608-614.Simple Method for the Determination of Gypsum, withSome Observations on the Solubilities of Gypsum, Anhydrite,Calcite and DolomiteA rapid method is presented for the determination of gypsum and relatedminerals in natural antl synthetic materials, based on tlic extraction of calciumfrom the matrix material using slightly amnioniacal kvatcr as a solvent withsubsequent complcxometric titration of the calcium using 131)TA. Tlie resultscompare reasonably well with tliosc obtaincd by tlie conventional gravimetricprocedure.Tlic interferences due to other sparingly soluble calcium com-pounds, sucli as calcite antl dolomite, commonly found in conjunction withgypsum were examined, as were tlic solubilities of anhydrite, calcite anddolomite in both water and sodium chloride solution.‘The method is particu-larly suitable for industrial control and deposit assessment purposes.Keywords : Gypsum determination ; anhydrite ; calcite ; dolomite ; solubilityviiHENRY A. FONER and SARAH EHRLICHGeochemistry Department, Geological Survey of Israel, 30 NIalchei Israel Strect,Jerusalem, 95 501, Israel.Analyst, 1983, 108, 615-620...Vlll SUMMARIES OF PAPERS I N THIS ISSUEEffect of On-line Complex Formation Kinetics on theFlow Injection Analysis Signal : the SpectrophotometricDetermination of Chromium(V1)May, 1983The sensitivity of the flow injection analysis (FIA) method for determinationof chromium(V1) using lI5-dipheny1carbazide (DT’C) clepentls on the concentra-tion of the acid used in confluence with the I ) K solution.At low acid con-centrations slower on-line reaction kinetics are observed for Cr - DI’C complexformation. Hence the chemical contribution to the over-all dispersion valuecannot be ignored. This indicates that the best experimental conditions forthe static procedure may not always translate directly to the dynamic contli-tions of FIA. In the present instance the maximum signal is obtained withacid concentrations a t or above 0.80 M. Although sulpliuric acid may be used,as in the conventional procedure, the best working conditions are acliievedusing nitric acid.Keywords : Flow injection analysis ; on-line kinetics ; clwomiuiii ( V I ) determina-tionJOAO CARLOS DE ANDRADE, JULIO CESAR ROCHA, CELIO PASQUINIand NIVALDI BACCANUniversidade Estadual de Campinas, Instituto de Quimica,,C.P.6154,13100-Campinas,S%o Paulo, Brazil.Analyst, 1983, 108, 621-625.Titrimetric Determination of Some N- SubstitutedPhenothiazine DerivativesAn indirect titrimetric metliocl is described for the clctcwiiination of someN-substituted plienothiazine derivatives. The mcthotl involvcbs the use of1,3-dibro1iio-5,5-ti1~ietli~lliy~lantoi1i or lY-~~romosuccinimitlc as the titrant.A known excess o f citlier reagent is adtled and, after ;I specifid time, theresidual reagent is deterniinecl ioclinictrically. The proposctl mctliotl wasapplied to the analysis of pliarmaceutical preparations containing thc drugs,and the results obtained compared favourably with tliose obtained bypharmacopoeia1 methods.Keywords : N-Substituted Phenothiazine deteviiiination ; titrimetry ; 1,3-dibrovno-5,5-divnethyl?iydantoin ; N-bvovnosucciniiizideMOHAMED IBRAHIM WALASH, MOHAMED RIZK, ABDEL-MALIKABOU-OUF and FATHALLA BELALDepartment of Analytical Chemistry, Faculty of Pharmacy, Mansoura University,Mansoura, Egypt.Analyst, 1983, 108, 626-632.Ligand-exchange Reactions in the Analysis of Cobalt Complexesby Atomic-absorption SpectrophotometryShort PaPevKeywords : Ligand-exchange yeactions ; cobalt complexes ; ditliizone ; atomic-absorption spectrophotovnetryW.JAMES SWINDALL, DUNCAN THORBURN BURNS and ELMUGDADA. ALIDepartment of Analytical Chemistry, Thc Queens University of 13clfast, Iklfast,BT9 5AG.Analyst, 1983, 108, 633-636May, 1983 SUMMARIES OF PAPERS I N THIS ISSUEDifferential Determination of Cationic and Anionic Surfactantsin Mixtures by Two-Phase TitrationShort PaperKeyurovds : Two-filtrrse tityntion ; crrtionic and anionic surfactant deteriizination ;sodium tetvn~lte~~~lOornte titrant; tetrabroi?zo~Jienol~~~t~talein ethyl esterindicatovMASAHIRO TSUBOUCHIT,al,oratory o f Chcniistry, liochi Mcclical School, Oko, Nankoku, Kochi 781-51, Japan,and JOHN H. MALLORYPurex Corporation, 24600 South Main Strcct, Carson, CA 90749, USA.Analyst, 1983, 108, 636-639.Adsorption of Metals on Polypropylene During Cold VapourAtomic-absorption Determination of MercuryShort PaperKeywords : Mercury determination ; atomic-absorption spectrophotometry ;metal adsorption ; polypropyleneHART B.MAcPHERSON and SHIER S.BERMANDivision of Chemistry, National Research Council of Canada, Ottawa, Ontario,Canada, 1KA OR9.Analyst, 1983, 108, 639-641.Spectrophotometric Study of the Reaction of Titanium(1V) withChlorophosphonazo IShort Pa9erKeywords ; Titanium determination ; chlorophosphonazo I ; spectroplzoto-metryQIU XING-CHUAgricultural Science Research Institute of Ganzhou Prefecture, Jiangxi, China.and ZHU YING-QUANPO Box 82, Chengdu, China.ixAnalyst, 1983, 108, 641-644THE ANALYST May, 1983READER ENQUIRY SERVICEFor further information about any of the products featured in the advertise- fments in this issue, please write the appropriate A number in one of the 3 iPostage paid if posted in the British Isles but overseas readers must affixboxes below. g ia stamp.(Please use BLOCK CAPITALS)NAM E .........................................................................................................................................................................................OCCUPATION .................................................................................................................................................................ADDRESS .............................................................................................................................................................................SECOND FOLDPostageDo not affix Postage Stamps if posted inGt. Britain, Channel Islands or N. Ireland I will bePaid byLicenseeI REPLY SERVICELicence No. W.D. 106Reader Enquiry ServiceThe AnalystThe Royal Society of ChemistryBurlington HousePiccadilly London W1 E 6WFENGLANDTHIRD FOLDFLAP

ISSN:0003-2654    

DOI:10.1039/AN98308BP053

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

May 1983 The Analyst Vol. 108 No. 1286 Platform Atomisation in Carbon Furnace Atomic-em ission Spectrometry Laszlo Bezur," John Marshall John M. Ottawayt and Riad Fakhrul-Aldeen Department of Pure and Applied Chemistry University of Stvathclyde Cathedral Street Glasgow G1 1XL The design and operational characteristics of platforms used in carbon furnace atomic-emission spectrometry are described. Platforms made of graphite and coated with pyrolytic graphite give significant improvements in the sensitivity and detection limits achieved for 23 out of 24 elements investigated. The delay in atomisation allows the production of the atomic vapour at a time when the vapour phase inside the furnace atomiser has reached higher temperatures. Improvements in signal to noise ratio have also been achieved through use of a high-resolution 6chelle spectrometer a novel square-wave wavelength modula-tion system and the better separation of the analyte signals from blank signals.Keywords Carbon furnace atomic-emission spectvometvy ; platform atomisation A number of publications in recent years have demonstrated the utility of the carbon furnace as a source for the generation of atomic-emission signals and a state of the art review has recently been pub1ished.l Much of the previous work has employed atomic-absorption instrumentation imposing severe limitations on the technique as neither the design of the furnaces nor the optical arrangement of these systems has been optimised for the measurement of carbon furnace atomic-emission spectrometric (CFAES) signals.In the measurement of carbon furnace atomic-emission signals a continuum background signal from the tube wall is always present .2 Even under optimum focusing a residual background signal persists which has been shown to be the result of Rayleigh scattering of black-body radiation from the tube wall into the monochromator.3 Correction for such continuum background radiation has been successfully achieved by wavelength modulation generated by an oscillating refractor plate.4 A novel method of square-wave wavelength modulation was employed in the instrument used for the work described in this paper and its construction and operation have been de-scribed Using this system the background signal from the furnace can be auto-matically subtracted obviating the need to obtain a separate background measurement for each analytical signal.Previous work has been based on the optimisation of the signal to background ratio in the measurement of CFAES signals. I t follows that the amount of scattered light entering the monochromator has a considerable influence on the over-all detection limit of the technique. For such a system it is important that the design of the furnace should not contribute an increase in the scatter signal if real improvements in sensitivity are to be made. However with a system incorporating an automatic background correction device the noise rather than the background is the limiting factor on detection power. Selection of the optimum operating temperature will therefore be based on signal to noise ratios rather than signal to background ratios as previously described.8 The temperature experienced by the analyte atomic vapour greatly affects the sensitivity obtained by CFAES.2y9 Improvements in sensitivity have been achieved by modification of the graphite tube itself and hence its thermal properties.Such modifications include the reduction of the thickness of graphite in various sections of the tube2S1* and the addition of a sample cup to the base of the tube.ll These design modifications are essentially permanent as the heating characteristics of the tubes are irreversibly altered. In addition although effective the modifications described require the precision machining of * Present address Institute for General and Analytical Chemistry Technical University of Budapest, Gellert ter.4 1521 Budapest Hungary. t To whom correspondence should be addressed. 55 554 BEZUR et a,!. PLATFORM ATOMISATION Analyst Vol. 108 tubes and tend to be time consuming. In some instances tube designs limit the range of ele-ments that may be determined,g and this is undesirable from the point of view of obtaining compromise conditions for simultaneous multi-element analysis by CFAES. A tube design is required that has specific advantages for furnace emission measurement which at the same time does not result in a degradation in sensitivity for any class of element. The over-riding consideration in the design of a graphite tube for the measurement of sensitive atomic-emission signals is the temperature attained in the tube and which is experienced by the analyte atoms.9 With an exponential temperature dependence of emission intensity at a constant free atom population the desired solution is one that results in the atomisation of the sample after the maximum furnace vapour temperature has been attained.As the vapour temperature depends on the wall temperature it is necessary to delay the sample atomisation until the vapour-phase temperature is maximised. This has been achieved to some extent by the use of the volatile-element tubelo and the cup-tube,ll where selective heating of parts of the atomiser tube heats the vapour prior to the formation of atoms. However it is also important that the maximum temperature that the tube can reach is not reduced as a result of the design. With the cup-tube the cup temperature does not reach the appearance temperature of some elements and cleaning the tube of an involatile matrix is difficult.The effect of diffusion of the atomic vapour from the furnace must also be considered, especially with respect to the temperature obtained in the vapour phase at the time of atomisa-tion as this will affect the sensitivity. Consequently the heating rate of the tube and the efficiency of atomisation are important parameters in evaluating a tube design with respect to the effective excitation temperature and rate of atom loss due to diffusion. Ideally a tube is required in which the sample is atomised almost instantaneously into a maximum temperature furnace atmosphere. L’vov12 has discussed the advantages of atomisation of the analyte into an atmosphere that has reached isothermal conditions for atomic absorption.Platform atomisation has been suggested as a method of obtaining such conditions and the technique has been investigated by a number of workers.13-17 The platform is a small pyrolytic graphite plate that is placed in the centre of the furnace beneath the injection port and the sample is deposited on it. Because the platform is not part of the tube itself thermal contact with the tube is minimal resulting in a delay in the heating of the platform with respect to the tube. As a result the sample is atomised from the platform at some time later than it would be from the tube wall and is introduced into a vapour that has reached or is close to isothermal conditions. This is precisely the situation required for improvement in the sensitivity of CFAES.Indeed increases in sensitivity would be expected to be much greater with atomic emission as opposed to atomic absorption by virtue of the strong temperature dependence of the emission signal. It is also expected that advantages derived from the use of the platform in atomic absorption would be equally applicable to CFAES. In an earlier paper,5 the detection limits for a range of elements achieved by the use of a platform in an HGA 72 furnace used with an dchelle spectrometer system were reported. In this paper more extensive information is given that includes the optimisation of platform design for CFAES measurement and the analytical parameters used in the operation of the furnace - platform for CFAES.Experimental Instrumentation The instrument system used was described briefly in a previous paper.5 The spectrometer is a Spectrametrics SMI I1 I khelle monochromator and a square-wave wavelength-modulation system was incorporated for automatic background correction of carbon furnace atomic-emission signals. Wavelength modulation was achieved by a rotating quartz chopper with four quadrants of different thickness and was analogous to a system recently described in detail for use in continuum source atomic-fluorescence spectrometry.18 In contrast to the earlier work with the dchelle ~ y s t e m ~ in which the wavelength modulation chopper was mounted near the exit slit in this instance it was positioned near the entrance slit. Operation in this mode is clearly preferable if the potential use of the dchelle system.for simultaneous multi-element analysis is to be realised. The other components of the instrument the Perkin-Elmer HGA 7 May 1983 I N CARBON FURNACE AES 555 carbon furnace atomiser and Brookdeal precision lock-in-amplifier were as described previously5 and are listed in Table I. I t should be noted that the figures for dispersion were given incorrectly in the previous paper.5 hlonochromator . . Wavelength modulation Lens . . . . . . Atomiser . . Detector . . . . . . . . Power and reference unit . . Signal processing . . TABLE I INSTRUMENTATION Spectrametrics SMI I11 Cchelle monochromator Focal length 0.75 m F number 13 Optical speed f / l O Reciprocal linear dispersion-a t 200 nm 0.062 nm mm-l a t 400 nm 0.124 nm mm-l a t 600 nm 0.180 nm mm-l Slit widths 0.2 mm Slit heights 0.5 mm Rand pass a t 400 nm 0.025 nm Laboratory constructed18 Chopper rotation speed 20 rev s-l Modulation frequency 40 Hz Modulation interval see text f = 130 mm silica Perkin-Elmer HGA 72 carbon furnace Laboratory-prepared platforms Purge gas argon 1.5 1 min-' Hammamatsu R292 photomultiplier Laboratory constructed 28 V d.c.supply to the EHT d.c. - d.c. converter f 15 V d.c. power supply and 6 V d.c. for the chopper motor Ortec Brookdeal 5002 current pre-amplifier Ortec Brookdeal 9503D precision lock-in-amplifier The square-wave wavelength-modulation system was developed to give a maximum signal to noise ratiolg and has been used to correct both for residual tube-wall background emission and variable background emission caused by various sample matrices.The modulation system consists of a chopper disc made of four quartz segments of different thickness a d.c. motor and a reference signal generated using an infrared light-emitting source and photodiode used to trigger the lock-in-amplifier. The quartz quadrants were each made of radius 30 mm and thickness 2.5 1.0 2.5 and 4.0 mm as described previously.18 A Perspex rim was attached around the chopper disc with a balancing weight opposite the 4.0-mm quadrant. The lock-in-amplifier was used to achieve automatic continuous background correction by subtracting the background signal obtained during observation through the 1 .O- and 4.0-mm quadrants from the analyte plus background signals observed through the 2.5-mm quadrants.The background-corrected signal was recorded using a Servoscribe R 541.20 potentiometric recorder but a digital transient recorder Type DL 901 (Electroplan Ltd. Royston Hertford-shire) was also used to study rapid signals. To measure the uncorrected signal the d.c. output of the current pre-amplifier was directly connected to the chart recorder. A l-m optical bench was fixed to the base of the spectrometer and the focusing lens and HGA 72 furnace head were mounted on the rail using screw adjustable holders. Wavelength adjustment was normally achieved for convenience using a hollow-cathode lamp of the appropriate element. Where necessary this was also achieved by injection of a suitable con-centration of the element required.In this instance care must be taken to avoid memory effects particularly for involatile elements. Problems due to residual atom signals were often particularly troublesome with this instrument owing to the high sensitivity achieved for many elements and it was found preferable to use an old graphite tube for wavelength adjustment when using a high-concentration analyte solution. In preliminary measurements a silica lens of focal length 50 mm and diameter 40 mm was used to produce a 1 1 image of the centre of the HGA 72 tube at the entrance slit of the mono-chromator. Subsequently and for most of the results reported a silica lens of focal lengt 556 BEZUR et al. PLATFORM ATOMISATION Analyst Vol. 108 130 mm and diameter 40 mm was substituted in order to reduce background emission reaching the detector.The furnace operation programme was optimised for each element separately as will be described, but in all instances the gas-stop mode was used during the atomisation stage. The HGA 72 graphite tube has dimensions of length 53 mm and i.d. 8.6 mm and is significantly larger and wider than most current commercial furnace systems. The absence of windows also helps to reduce problems related to scattering of tube-wall radiation from external surfaces into the monochromator. The gas flow and dispersion of atom populations formed in the tube pro-ceeds from the centre to the ends increasing the atom residence time and particularly under gas-stop conditions increasing the temperature experienced by the atom population.2 Standard Perkin-Elmer HGA 72 graphite tubes were used but were coated with pyrolytic graphite in sits prior to use.This was achieved by the introduction of a flow of 50 ml min-1 of methane into the argon purge gas stream for 3-6 min whilst the furnace was set to operate at 2100 “C. Platforms were prepared in the laboratory and were made of different designs as shown in Fig. 1. It was desirable to minimise the contact between the platform and the tube to reduce the rate of heat transfer. Ideally the platform should be wide enough to make contact with the sides of the tube without touching the bottom. Sections of graphite cut from standard HGA 72 tubes were found to be very suitable as their curvature matched that of the tube wall and gave the desired degree of contact when mounted in the centre of the tube.A Perkin-Elmer HGA 72 graphite furnace was used for atomisation and excitation. 1.5 t 4 L1 Fig. 1. Graphite platforms used A curved “rectangular” platform; and B as A but with sections removed from the contact edges. The platforms used were of the two basic designs shown in Fig. 1. The simplest was a curved “rectangular” piece cut from the wall of the HGA 72 tube of dimensions 8 x 11 nim. The mass of this piece was approximately 0.12 g. The second type had similar dimensions but sections were removed from the two contact edges to reduce further the tube to platform contact to that at the four corners of the platform. Platforms of different dimensions were prepared to arrive at this optimum the thickness in each instance being maintained at the thickness of the HGA 72 tube wall.Platforms were coated with pyrolytic graphite under the same conditions as those used for coating the tubes. Platforms were placed in the centre of the HGA 72 tube and were introduced easily via the open ends of the atomiser. Both types could accommodate up to 50 p1 of sample solution, which was transferred through the normal injection port of the atomiser using an Oxford micro pipette. Atomic-absorption measurements were also made with this system using the appropriate hollow-cathode lamp positioned on the optical bench and focused through tlie HGA 72 atomiser and on to the entrance slit of the monochromator. A mechanical chopper was used to achieve intensity modulation of the hollow-cathode lamp radiation and a similar reference signal wa May 1983 IN CARBON FURNACE AES 557 generated to operate the lock-in amplifier.The Wavelength modulation unit was disconnected during absorption measurements. The mechanical chopper was mounted in front of the hollow-cathode lamp and between it and the HGA 72 furnace. Temperature measurements of the graphite tube and platform were made using an Ircon Series 1100 recording optical pyrometer using the appropriate current shunts and the R 541.20 potentiometric recorder. The pyrometer was focused as required on the inside tube wall or platform through the injection port of the HGA 72 atomiser. Temperature measurements were made assuming an emissivity of unity. Reagents All reagents were of the highest available purity and high-purity distilled water was used for the preparation of all solutions.Stock solutions (1 000 pg ml-l) of each element were pre-pared from the metal or appropriate sulphate or nitrate salts and acidified to a final concentra-tion of 10-2 M sulphuric or nitric acid. Working solutions of the appropriate concentrations were prepared when required. Detection Limits signal equal to the noise of the blank signal. The detection limits were calculated as the concentration of analyte element that produced a Results and Discussion Optimisation of Signal to Noise Ratio using commercial atomic-absorption systems to measure furnace atomic-emission signals detection limits were defined in terms of signal to background ratio. Using wavelength modulation for continuous automatic background correction detection limits be-come dependent on the signal to noise ratio.4 Assuming thermal equilibrium in the furnace tube2* and consequently a Boltzmann distribution of energy the measured emission signal or spectral line intensity at any instant will be given by2Ov2l In previous e x p ( g ) .. . . . . * - (1) KhcNtAijgiL 47~B ( T ) hij I = where I is the intensity over the total line width in W m-2 sr-l K is a combined unit conver-sion factor and instrument constant N t m-3 is the total instantaneous concentration of atoms in all states in the source L m is the length of the source in the direction of viewing Aij s-l is the Einstein transition probability for spontaneous emission gi is the statistical weight of the upper energy level B(T) is the partition function over all states Ei J is the energy of the upper state i T K is the temperature of the vapour phase in the furnace hij m is the wavelength of the spectral line and h c and R have their usual meanings and values.The dominating noise component of carbon furnace atomic-emission signals using continuous background correction is in most instances the shot noise associated with the residual black-body radiation from the tube all.^,^ To improve detection limits of the carbon furnace atomic-emission technique it is therefore required to increase I without significantly increasing the residual black-body radiation. Because in optimising the conditions for CFAES measurement the most sensitive spectral line will be chosen and fixed and the optical aperture parameters included in the value of K will be made as large as possible within instrumental constraints the only factors that can be used to increase the value of I are Nt L and T .In electrothermal atomisers an excessive increase in L is generally unacceptable owing to the increased possibility of self-absorption and increased temperature gradients of the furnace leading to condensation etc. at the cooler parts of the tube. Although some modified graphite tubesg?l1 have been designed to vary the heating characteristics with respect to length Nt and T remain the most important parameters. In a graphite furnace operated under gas-stop conditions the vapour-phase temperature T , corresponds closely to the tube-wall temperature.20 The maximum tube-wall temperature is limited by the rapidly diminishing lifetimes of graphite tubes above 3300 K and the power available in most commercial systems is restricted to levels that prevent this temperature being exceeded.The instantaneous total atom concentration Nt is a result of a series of comple 558 BEZUR et at. PLATFORM ATOMISATION Analyst Vol. 108 solid melt and gas phase reactions that depend on the chemical and physical properties of the analyte and matrix as well as instrumental factors such as the heating rate the final equilib-rium temperature set and the surface material of the atomiser. According to L'vov's model of the atomisation process,12 the maximum number of analyte atoms in the furnace Nmax. is related to the total number of analyte atoms introduced into the furnace No and is given by Nmax.= 2N0 2 - - 1 + e-Ti/T,> . . 71 '(" 7 2 where r1 is the atomisation time and r2 is the residence time. The loss of analyte atoms in graphite tube furnaces is expressed by r2 and is due partly to the internal gas flow and partly to diffusion. Under interrupted gas-flow or gas-stop atomisation conditions diffusion and convection caused by the expanding gases are the important factors and will be dependent on the internal dimensions of the tube. Under diffusion-controlled conditions r2 is a function of temperature through the tempera-ture dependence of the diffusion coefficient where - - (3) 7 2 Gcc-" and m varies between 1.5 and 2. The rate of loss therefore increases with increasing tempera-ture and atomic-absorption results often show that Nmax.and hence the absorption maximum ( A max,) occur at an optimum temperature that may be considerably less than the maximum or equilibrium temperature. When atomic-emission measurements are made the maximum signal occurs after Nmax. (or Amax.)2s22 and at a stage when the number of atoms available for excitation is diminishing. This is because IA is also dependent on T which for most elements is still increasing at the time of Nmax The time of peak intensity of the emission signal there-fore depends on the two changing parameters N t and T. The fact that under normal operation the temperature is changing during atomisation is not ideal for atomic-absorption measurements. An improvement in sensitivity and reduction or elimination of matrix interference effects was achieved by increasing the heating rate of the furnace and also by delaying the evaporation or atomisation of the sample.15 In this instance Nmax.occurs at a time when the furnace is closer to the equilibrium or maximum temperature, and molecules that may be formed as a result of interfering reactions are more completely dissociated. The use of higher heating rates and the delay of atomisation to times at which higher temperatures are available is even more advantageous in emission measurement than absorption as the possibility then exists of simultaneously increasing Nma,. and T . Thus maximum power heating using the HGA 2200 carbon furnace gave a significant improvement in the atomic-emission detection limits for involatile elements,23 owing to increases in Nma,.. In this work platform atomisation was used to delay atomisation to give an increase in T during the lifetime of the atom population. Detection limits in CFAES are limited by the noise generated by the intense emission from the graphite tube wall. In a properly aligned and baffled optical system it should not be possible for tube wall radiation to enter the spectrometer slit directly but a residual signal is still ~bserved.~ The main sources of this residual radiation which has the characteristics of black-body radiation may be the following : (a) tube-wall radiation scattered by molecules and particles in the hot part of the furnace (Rayleigh scatter) ; (b) tube-wall radiation scattered by particles formed by condensation at the cooler ends of the tube (Mie scatter); (c) radiation scattered by incandescent carbon particles from the decomposition of the graph-ite tube ; (d) radiation reflected into the spectrometer from furnace windows and other optical com-ponents; and (e) aberrations or dirt particles in the optical system focusing lens furnace windows etc., which might allow tube-wall radiation to enter the spectrometer directly despite correct alignment.The intensity of the emission signal given by any of the above sources is proportional to the intensity of the radiation from the tube wall which is given by Planck's Law which for May 1983 I N CARBON FURNACE AES 559 narrow band pass and in the wavelength region and temperatures of interest in carbon furnace atomic emission has the form3 .. . . ' ' (4) CI E -- A5 exp(C,/AT) where E is the integrated intensity of the black-body radiation C and C2 are constants A is the wavelength and T the absolute temperature. Thus the black-body radiation and the fraction of it scattered are exponentially dependent on the temperature in the same way as the analyte atomic-emission intensity given by equation (1). Thus an increase in temperature (at the same atom concentrations) increases the signal and background in proportion and gives no advantage in signal to background ratio.* The noise component of the background-corrected signal however is proportional to the square root of the background intensity in the shot noise-limited examples and will therefore change less rapidly with temperature.In CFAES with wavelength modulation background correction detection limits can be improved by maximising the temperature as far as possible and also by minimising all sources of residual background emission. The angular distribution of the scattered light intensity I, in the Rayleigh region (A> diameter of particles d) is expressed by T12 where I is the impingement intensity 8 is the angle between the incident and scattered light beams n is the refractive index of the particle V is the volume of the particle h is the wave-length and Y is the distance of the particle from the point of observation. Thus the intensity of Rayleigh scattered light is proportional to the incident intensity and to the sixth power of the particle diameter and inversely proportional to the fourth power of the ~ a v e l e n g t h .~ ~ ~ The intensity is largest at angles near 0 and 180". The intensity of scattered light caused by larger particles (d > A) Mie scatter is not dependent on wavelength but is dependent on the angle 8 giving maximum intensity at small angles and a minimum at 180". The intensity of Mie scatter is also proportional to the number of scattering particles and will thus increase in the presence of a sample matrix that forms condensed vapours at the cooler ends of the tube.24 This is particularly relevant in the open-ended HGA 72 atomiser used in this work but will be minimised in other atomiser designs. Mie scattering will also reduce the analyte atom intensity reaching the monochromator and will thus degrade both the signal to background and signal to noise ratios by simultaneously changing both factors.Modifications to the tube or gas flow could be used to reduce the effect of Mie scattering. Scattering of black-body radiation by carbon particles will increase with the age of the graphite tube. A fine loose carbon dust is gradually formed in most types of graphite tubes, but the effect is reduced by coating with pyrolytic graphite. The latter was used throughout this study. Optical Design Both were used to form a 1 1 image of the central section of the tube at the entrance slit of the monochromator. The apertures of the two lenses were identical producing a signal from a central cylindrical section of the graphite tube of 2 mm diameter. The 50-mm lens gave a more deformed image of the 53-mm long HGA 72 tube and even in the optimum position a proportion of tube-wall radiation entered the monochromator directly.The 130-mm lens gave a much improved image with a distinct dark area in the middle of the slit indicating that no tube radiation entered the monochromator directly. The detection limit for manganese was determined for each system giving 0.14 pg 1-1 for the 50-mm lens and 0.029 pg 1-1 for the 130-mm lens an improvement of 4.8 in signal to noise ratio. In CFAES correct lateral alignment of the furnace and optical components is essential. The emission intensity of the tube-wall radiation (background) and the total emission for 50 pl of 1 pg 1-1 manganese solution were measured at the correct position and with the furnace tilted out of the exact horizontal axis.Several positions were used and the results are shown in Fig. 2 as a function of the angle. The signal to background ratio changed from 4.3 to 0.6 when the angle was changed from 0 to 2". Two projector lenses of focal lengths 50 and 130 mm were compared 560 BEZUR et al. PLATFORM ATOMISATION Analyst Vol. 108 1 .o 0.8 0.2 \.C !I 4 2 h I I 0 0 1 2 a, Fig. 2. Effect of the lateral alignment of the graphite furnace on (A) the emission signal from 50 p1 of 1 pg 1-l manganese (B) the background signal a t 403.08 nm and (C) the signal to back-ground ratio (A/B). Wavelength Modulation The resolution of the SMI bchelle monochromator is about an order of magnitude better than that usually available in commercial atomic-absorption instruments and results in an im-proved signal to background ratio.The square-wave wavelength-modulation system not only provided automatic correction for the residual background signal but also correction for the less reproducible background emission from various matrices. The chopper consisting of four quartz quadrants was mounted at the entrance slit of the monochromator at an incident angle of 24" (at a slit width of 0.2 mm) to the optical axis. The extent to which the light is displaced by refraction on passing through the chopper is dependent on the thickness of the quartz,18 and it was arranged that light at the analyte atom wavelength passed through the two 2.5-mm quadrants. The 1 .O- and 4.0-mm quadrants then allow the background signal to be measured on either side of the atomic line.The chopper rotation speed was 20 s-l which produced a square wave modulated signal of frequency 40 Hz. The incident angle of 24" was chosen to give efficient modulation over the range from 200 to 800 nm at tlie 0.2-mm slit width. The total modulation distance was 0.52 mm at 200 mi 0.50 mm at 400 nm and 0.46 mm at 800 nm and gave modulation intervals of 0.032 0.062 and 0.115 nm at the respective wavelengths. The minimum acceptable angle of incidence was determined by calculation and was also measured using a chromium hollow-cathode lamp as the line source at 425.4 nm. The anode current of the photomultiplier tube was measured on a 1-MR resistance using an oscilloscope in the d.c. mode. If the angle of incidence is large enough the minimum current is equal to the dark current.The measured minimum angle 23" was larger than that calculated 21" owing to the aperture of tlie spectrometer and the slight de-focusing of the slit image caused by the chopper disc. To ensure correct modulation an incident angle of 24" was chosen. Dimensions and Design of Platforms Previous studies of platform atomisation have concerned their use in atomic absorption in which the most important requirement is to minimise the reduction of intensity of the hollow-cathode lamp. Increased background emission reaching tlie detector is rarely important but assumes much greater significance in atomic-emission measurements not only because of the generally higher atomisation temperatures used but also because of the direct relat ionsliip between detection limits and the residual background emission signal.In addition flat or plane platforms appeared to be unsuitable in the HGA 72 as the sample tended to spread to tlie edges of the platform and then wet the tube wall. To minimise both problems curved platforms were designed with tlie shape and dimensions shown in Fig. 1. These curved platforms can be prepared from a tube with dimension May 1983 IN CARBON FURNACE AES 561 identical with those of the tube in which they are to be used. Only the lower edge of the plat-form touches the tube wall satisfactorily reducing heat conduction. The upper edge of the platform is some distance from the tube wall reducing the tendency of the sample to wet the wall and ensuring that samples are more accurately located even compared with injection directly on the tube wall.Normal injection of samples through the injection hole of the HGA 72 tube is straightforward and 5O-pl aliquots are conveniently atomised using platforms of optimum dimensions. Platforms with sections removed from the contact edges were prepared which further reduced the rate of heating and cooling. The platforms indicated in Fig. 1B were found to be significantly better for volatile elements such as gallium and indium. The effect of the different platform configurations in the light beam is indicated by the avail-able angular apertures at the centre of the tube and the resulting free tube diameter shown in Table 11. In the system used the aperture is limited by the optical system but the reduction in free tube diameter is significant as the radiating surface approaches more closely to the optical observation zone resulting in a higher scattered light intensity.TABLE I1 EFFECT OF PLATFORM ON AVAILABLE ANGULAR APERTURE AND FREE INNER DIAMETER OF AN HGA 72 TUBE Available angular Free inner Tube aperture O diameter/mm HGA 72 . . . . . . 18.8 8.5 HGA 72 with single curved platform . . . . 18.8 7.0 HGA 72 with double curved platform . . 18.8 6.0 Temperature Characteristics of Platforms heat to the p1atf0rm.l~~~~ the change of platform temperature with time is therefore given by16 I t is clearly established that the reduced tube - platform contact results in a slow transfer of Heat is transferred to the platform almost entirely by radiation and where E C Q S and V correspond to the platform and E is the emissivity C is the specific heat, Q is the density S is the surface area and V is the volume of the platform 8 is the Boltzmann constant and T and T are the temperatures of the tube wall and platform respectively.To increase the temperature lag of the platform only a few of the parameters in equation (6) are significant. The choice of material is restricted to the different forms of graphite tantalum and tungsten alternative materials being subject to chemical and thermal decomposition. With any specific material the only means of increasing the temperature lag is to increase the thickness of the platform. The effect of platform dimensions and mass is shown in Fig. 3. The use of a longer platform of the same thickness is unsatisfactory as the surface area in-creases almost in proportion to the mass or volume and the temperature lag is not greatly increased (compare curves B and D in Fig.3). Longer platforms also suffer from the dis-advantage of the greater spreading of the sample in the tube. It may then be subject to variable heating rates owing to the temperature gradient existing in the tube. They will also reduce the available angular aperture. Almost the same increased delay was achieved by using the platform with reduced contact edges shown in Fig. 1B (see curve C Fig. 3). This also increased the time required for cooling the platform between injections. Some practical advantage was found in using two platforms mounted one on top of the other.The tempera-ture lag is about the same as achieved by doubling the mass of a single platform. The platforms only touch at the edges and are heated equally by the tube wall on either side. The temperature of the platforms is then the same and there is no heat transfer between them. Although no tube or platform emission was allowed to enter the monochromator directly, background emission signals increased owing to increased scattering of radiation. Fig. 4 illustrates the increase in background from the use of a single platform ( x 1.28) and a double platform ( x 1.83) these signals being taken at 500 nm directly from the pre-amplifier i.e., without background correction 562 BEZUR et al. PLATFORM ATOMISATION Analyst Vol. 108 3000 s !?? t? 2000 tl n I- : 1000 2 4 6 8 10 Time/s I I I I 1 1 I 0 8 6 4 2 0 Ti me/s Fig.3. Temperature veYsus time graphs for an HGA 72 furnace operated a t 2600 K. (A) Standard tube; (B) standard tube plus platform Fig. 4. Background signals taken shown in Fig. 1A; (C) standard tube plus platform from the pre-amplifier a t 500 nm. HGA shown in Fig. 1B; and (D) standard tube plus 72 operated a t maximum power 999 platform of dimensions 8 x 20 mm. units for 10 s. (A) Standard tube; (B) standard tube plus platform; and (C) standard tube plus two platforms placed one on top of the other. Platform Atomisation The effect of the platform on atomisation was investigated for chromium and gallium, representing elements of different volatilities. Both atomic-absorption and atomic-emission signals were observed under identical conditions using a standard HGA 72 tube and a standard tube with platform.Background correction was used in the emission mode but not in absorp-tion where it was unnecessary for the pure analyte solutions used. Tube and platform temperatures were also observed under the same conditions and all signals were synchronised in time from the start of the atomisation stage of the furnace. Platform temperature measure-ments made using an optical pyrometer are likely to be in error owing to the reflection of light from the tube wall scattered off the platform. In this work no account has been taken of this source of error. As indicated platform temperatures will only be too high it will be noted that this error does not affect any conclusions drawn which clearly demonstrates the characteristic properties of platform atomisation.For chromium 50-pl aliquots of a 50 pg 1-1 solution were used for both absorption and emission and the results are illustrated in Fig. 5. At this concentration of chromium the start of the atomic-absorption signal in the standard tube (curve C Fig. 5) occurs at 1920 K (curve A). Atomisation from the platform appears to start at a platform temperature of 1950 K (compare curves E and B). These atomisation temperatures are in good agreement but are higher than the appearance temperature of chromium at 1 740 K,f2 possibly owing to the effect of c~ncentration.~~ When atomisation from the platform starts the tube temperature has reached 2350 K giving a lag of 400 K.I t is clear from curve E that at the time of maximum atom population 5.5 s the temperature of the tube has not reached its maximum. A further delay of about 2 s giving 7.5 s in total would be required to achieve coincidence of Nmax. and the equilibrium temperature Tmax The peak absorbance for chromium was about the same with and without the platform (compare curves C and E) but the emission intensity increased by a factor of 1.6 using the platform (compare curves D and 1;). As at the temperatures used in electrothermal atomisers most atoms remain in the ground energy state the total concentration of atoms in all states at any time N t is proportional to the absorbance at that time A t . Substitution in equation (1) and rearrangement give May 1983 6 -C d -I $ 5 -4 -IN CARBON FURNACE AES 563 I I I 3000 2000 g 2 2 4-n E 1000 + 0 2 4 6 8 10 Ti me/s Fig.5. Absorption and emission curves for 50 p1 of 50 pg 1-' chrom-ium atomised at an HGA 72 furnace setting of 2800 K (A) change of temperature of the graphite tube; (B) change of temperature of the plat-form of type A (Fig. 1) ; (C) chromium atomic absorption following tube-wall atomisation ; (D) chromium atomic emission following tube-wall atomisation ; (E) chromium atomic absorption with platform atomisation ; and (F) chromium atomic emission with platform atomisation. where K' and E are constants for a specific transition and K' includes all the non-exponential terms in equation ( l ) plus the conversion constant relating N t to At.Hence a plot of the In IA/At veysus T-l should be a straight line of slope - E i / k . For the chromium wavelength of 425.43 nm this slope will be -3.38 x 104 K-1. Vapour temperatures were not measured in this study. Therefore tube-wall and where relevant measured platform temperatures were used to examine this relationship. In Fig. 6 In IA/A t is plotted for chromium atomisation in a standard HGA 72 tube without platform against the reciprocal tube-wall temperature, T;:. A linear relationship is obtained but with a greater negative slope than the theoretical value suggesting that the average vapour temperature experienced by the analyte atoms a t any time is slightly less than the tube-wall temperature at that time. In Fig. 7 values 3 ' I I I 3 3.5 4 4.5 W K - ~ x 10-4 Fig.6. Plot of InTA/A versus T-l for 50 pl of 50 pg I-' chromium atomised from the tube wall a t a furnace setting of 2800 K. Broken line indicates theoretical slope -E,/k, of -3.38 x lo4 I<-l for chromium a t 425.43 nm. 7 6 55 -I 4 C B ' A I I I 3 3.5 4 4.5 T V K - ~ x 1 0 - ~ Fig. 7. Plot of 1nI)./At versus T-l for 50 p1 of 50 pg 1-' chromium atomised from the platform a t a furnace setting of 2800 I<. A, T T,; 13 ,T = Tev; and C T = T,. Ihoken line indicates theoretical slope as for Fig. 6 564 BEZUR et al. PLATFORM ATOMISATION Analyst Vol. 108 obtained for chromium atomisation from a platform are plotted against T -4 and against the measured temperature of the platform T-t as well as the average of the two temperatures, T;:.All give linear relationships and the theoretical slope (broken line) falls between T; and T-;. Similar measurements were carried out for gallium using 50 p1 of a 50 pg 1-1 standard solution. Absorbance and emission measurements were made at 403.29 nm and are illustrated in Fig. 8. Both the absorption signals from the tube wall and the platform start at 1400 K, which is close to the average literature value for the appearance temperature of 1350 K.l2 7 - 0.7 6 - 0.6 .- 5 5 - 0.5 m a .- E 4 -; 0.4 S E a $ U 0 2 4 6 8 10 Absorption and emission curves for 50 pl of 50 pg 1-1 gallium at an HGA 72 furnace setting of 2950 I<. (A) Change of temperature of the graphite tube; (B) change of temperature a t the platform of type A (Fig. 1) ; (C) gallium atomic absorption following atomisation froin the tube wall ; (U) gallium atomic emission for tube-wall atomisation; (E) gallium atomic absorption for platform atomisation ; and (F) gallium atomic emission for platform atomisation.Time/s Fig. 8. The tube-wall temperature at the start of the absorption signal from the platform is 2 180K, giving a platform delay in temperature of 780 I< at this time. For gallium an increase in peak absorbance by a factor of 3 is obtained on using the platform (curves C and E) suggesting more complete dissociation of molecular species in the higher vapour-phase temperature available following platform atomisation. The enhancement in the emission signal on the use of the platform is a factor of 5 (curves D and F).The maximum absorption signal for gallium with 3 3.5 4 4.5 5 T ’ / K - ’ x Plot of lnl~,/A vevszfs T-l for 60 pl of 50 p g 1-1 gallium atomised from the platform a t a furnace setting of 2950 I<. A T = T,; B,T = Tav; and C T = T,. Hroken line indicates theoretical slope -E,/k’ of 3.56 x lo4 1i-l for gallium at 403.29 nm. Fig. 9 May 1983 I N CARBON FURNACE AES 565 the platform still occurs about 3 4 s before the tube reaches its maximum or equilibrium temperature. All three are linear but in contrast to chromium the slope of the Ti? line is closer to the theoretical slope of 3.56 x lo4 K-l. The relationship examined in Figs. 7 and 9 provides an interesting method of testing the application of the Boltzmann equation to the excitation of atoms during electrothermal atomisation and of estimating the average vapour-phase temperature experienced by the transient atom population.The differences between chromium and gallium cannot be explained at present. Clearly the platform temperature lags behind the tube-wall tempera-ture during the atomisation of most elements and the vapour-phase temperature may not follow the tube wall exactly either. At least the part of the tube that contains the platform is at a temperature lower than the tube wall and this may help to reduce the over-all vapour-phase temperature. It is clear that the above relationship may be used to predict the emission curve and the time of maximum emission intensity for an element if the absorbance curve of the element and the temperature characteristics of the tube are known.The expected increases in emission intensity as a function of temperature at constant atom concentrations can also be easily calculated from equation (1). Typical values for chromium, gallium and lead are shown in Table I11 and are similar to those values observed experimentally (shown in Table VII and discussed later). The actual increases in peak emission intensity will depend on both the temperature and atom concentrations at the time of each peak and on the energy of the upper state of the transition. The In (IA/At) relationships with T;’ TF’ and T;;;1 are shown in Fig. 9. TABLE I11 THEORETICAL INCREASES IN EMISSION INTENSITY WITH TEMPERATURE (Nt CONSTANT) Element and wavelengthlnm Ei/eV T J K Cr 425.43 . . . . . . . . 2.91 2 800 2 700 2 700 Ga 403.29 .. . . 3.07 2 800 2 700 2 700 Pb 405.78. . . . 4.38 2 800 2 700 2 700 TZIK 2 900 2 900 3 000 2 900 2 900 3 000 2 900 2 900 3 000 I2lIl 1.52 2.38 3.51 1.5 2.5 3.7 1.87 3.66 6.62 Platform Drying Temperatures The reduced heat transfer to the platform means that much longer drying times are required if the tube temperature is set as usual close to the boiling-point of the solvent. Using 373 K for water and a 50-pl aliquot of sample drying times increased to 180-200 s from the normal 30-40 s. Increasing the tube temperature to 473-573 K reduced the drying time to 3040 s without sample losses by spluttering. Heat transfer to the platform at low temperatures is by conduction and energy losses arising from the evaporation of the water balances the plat-form temperature allowing a smooth drying rate even at higher tube temperatures.When the solvent has evaporated the temperature of the platform rises to that of the tube wall, which might then cause losses of very volatile elements. In practice this has only been observed for indium when a lower drying temperature of <450 K must be used. Drying temperatures are best chosen by visual observation of solvent evaporation. Effect of Platform on Emission Peak Shape In our experience the use of platform atomisation results in improvements in signal to background and signal to noise ratios above that due to the increased vapour-phase temperature. This arises from improved peak characterisation and shape and can be illus-trated for a number of elements.Background-corrected signals for a blank sample and 50 pl of appropriate concentrations for silver indium and lead both with and without a platform are shown in Figs. 10-12. Without a platform large and variable blank signals are obtained for silver and lead which appear to b 566 BEZUR et al. PLATFORM ATOMISATION Analyst Vol. 108 -\ \ \ \ \ \ I 1.1 \ A I D +- Time Fig. 11. Emission peak shapes for indium at 4- Time 410.f7 nm. (A) Blank-firing of standard tube; (B) 50 pl of 100 pg 1-1 indium from tube wall; (C) blank firing with platform; and (D) 50 pl of 0.5 pg 1-1 32Ft 1;''. " ~ ~ i ~ p e a ' ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ indium from platform. Temperature settlng, 2900 K. tube; (B) 50 p1 of 50 pg 1-' silver from tube wall; (C) blank firing with platform; and (D) 50 pl of 5 pg 1-l silver from platform.Tem-perature setting of the HGA 72 furnace, 2900 K. due to a memory effect caused by evaporation of analyte from the ends of the tube and contact cones. The analyte peak after the sample injection is ill-defined and superimposed on the variable blank signal. Using the platform a substantial increase in sensitivity is observed, but in addition the peak is more clearly separated from the blank signal and much more sharply defined. At the measurement of peak emission with the platform both sensitivity and signal to noise ratio are improved. With indium (Fig. 11) the blank signal is similar for the tube-wall and platform atomisation but both sensitivity and peak shape are improved I A d Fig.12. Emission peak shapes for lead at 405.78 nm. (A) Blank firing of standard tube; (B) 50 pi of 1000 pg 1r1 lead from tube wall; (C) blank firing with platform; and (D) 50 pl of 100 pg 1-1 lead from plat-form. Temperature setting 2 900 K May 1983 I N CARBON FURNACE AES 567 when using the platform. Several reasons can be given for the improved atomisation obtained with the platform : (a) the sample is localised on the platform and is therefore prevented from spreading away from the centre of the tube. The sample is therefore evaporated in the hottest part of the tube and in a shorter time; (b) evaporation takes place into a much hotter gas which results in faster evaporation and dissociation of gas molecules; and (c) during platform atomisation the temperature at the ends of the tube is higher when the sample is evaporated and thus condensation is reduced.Blank signals were found to be significant for a number of elements and arise in part from the very high sensitivities achieved using this technique. Several causes may be considered. The blank signal may be an analyte signal derived from contamination present in the tube material from slow release of refractory species formed on the graphite surface or from con-densed analyte on the ends of the tube or contact cones. I t may also be due to a failure in the wavelength modulation system due to the presence of a highly structured background signal, or to the partial intensity modulation given by the wavelength modulation system used.26 Characteristic blank signals are shown in Fig.13 together with appropriate analyte signals using concentrations near the detection limits. For manganese gallium lead indium and silver there is no problem from the blank signal which is either negligible or well separated from the analyte peak. With copper and chromium a slowly rising blank signal is obtained, which was shown to be due to analyte atoms present in the form of a memory effect. Thus the magnitude increased after the injection of higher concentrations of analyte and decreased when gas-flow atomisation was used. In many instances the memory effects shown here are com-parable to the "double peaks" observed in the atomic-absorption determination of volatile elements. +Time Emission peak shapes for several elements together with blank signals [EL) at the same wavelength.Platform atomisation. Mn 1 pgl-1 manganese; Cu 1 pg 1-' copper; Cr 1 pg 1-1 chromium; and Ga 1 pg 1-' gallium. Wavelengths as in Table VI; 50-p1 aliquots; temperature setting 2 900 K. Fig. 13. Owing to the sensitivity achieved by furnace atomic emission with platform atomisation it is essential to remove the analyte completely from the tube after each injection. Although gas-stop atomisation is required for maximum sensitivity it tends to increase the condensation effects and hence blank signals. Maximum sensitivity with minimal memory effects is achieved by gas-stop atomisation maintained only until the peak emission intensity has been observed, followed by a clean stage at maximum temperature under gas-flow conditions until the signal returns to zero.Such criteria will also be relevant in the use of platform atomisation in atomic absorption although the effects for each element will vary with the sensitivity given by each technique 568 BEZUR et al. PLATFORM ATOMISATION Analyst Vol. 108 Temperature Programme Mention was made above of the choice of drying temperature which is critical if analysis times are not to be unnecessarily long. The platform also takes longer to cool to ambient temperatures and an extra 20-30 s delay must be allowed at this stage. The selection of the maximum ashing temperature is very similar to that for the standard tube without platform and recommended values were normally used without difficulty. The choice of atomisation temperature however is different from that reported earlier for a system without wavelength modulation.8 Above a specific temperature the signals remained constant with increase in temperature ie.a plateau region was reached. Atomisation temper-atures given in Table IV are either the maximum available (2873 K) or close to it and as TABLE IV OPTIMUM ATOMISATION TEMPERATURES FOR CFAES MEASUREMENT Element . . . . . . . . 3 :: Be . . Cd . . Cr . . c u . . . . . . Fe . . . . Ga . . K . . . . Mg . . Mn . . . . Na . . . . Ba Co Dy Eu Ho In Ni, Pb Sc Sr Yb . . T / K -7 STD tube Platform >2 773 -~ 2 8 2 3 >2 823 - >2 823 >2773 2 873 >2 723 ->2673 >2773 ~ 2 6 2 3 > 2 673 >2773 ->2823 2 873 >2 823 2 873 ~ 2 4 7 3 >2673 >2673 >2 823 2 873 2 873 predicted,8 are higher than for a system without wavelength modulation.The optimum temperatures given are the minimum required to give maximum intensity and little change is observed at higher temperatures. The differences between the temperature found to be optimum with and without the platform are very small and the major difference in conditions is caused by the introduction of wavelength modulation background correction. This is a function of the change in dependence of detective power from signal to background to signal to noise ratio as originally reported by Epstein et aZ.4 The use of the platform and wavelength modulation will clearly facilitate the selection of compromise conditions for a potential simul-taneous multi-element analysis system based on CFAES without significant sacrifice in sensitivity for any element.Improvement factors for the improvement in sensitivity with the use of the gas-stop mode are shown in Table V for both platform and tube atomisation. The improvement with the TABLE V EFFECT OF GAS FLOW ON EMISSION SIGNAL SENSITIVITY MEASURED USING 50-pl ALIQUOTS OF ANALYTE AT 1 pg ml-I Element Mg . . In . . Cr . . Mn . . Fe . . cu . , co . . Ni Na Sensitivity gas stop Sensitivity gas flow 7-Platform Standard tube 6.5 9 6 90 3 4.5 10 50 6 30 4 G 5.8 17 4.4 4.3 19 4 May 1983 IN CARBON FURNACE AES 569 platform is in all instances smaller than with the tube alone. The main advantage of the gas-stop mode in atomic emission is to reduce the rate of loss of analyte and allow a greater vapour-phase temperature to be obtained at gas flow.The delay caused by the platform acts in an analogous manner and hence reduces the effect of the gas-stop mode. However the improve-ment factors with the platform are still significant and the gas-stop mode should be used for the highest sensitivity. Effect of Platform on Reproducibility When large sample volumes are injected directly on to the tube wall of a furnace the sample spreads out either instantaneously or sometimes during the drying stage to an extent that depends on the surface tension of the solvent. With many sample types reproducibility is impaired through variable spreading effects. The platform counteracts this by preventing the sample from extending over its perimeter.The surface tension of the droplet ensures that it does not fall off the top surface of the platform. The reproducibility of sample placement is therefore improved and the reproducibility of analysis is consequently also improved. The results in Table VI indicate this to be so particularly for the more volatile elements. Precision is also improved because of the faster and more complete evaporation and dissociation of the analyte and because of the reduction in condensation effects at the tube ends which results in the sharper peak shapes illustrated in Figs. 10-12. The platforms used in this study could accommodate sample volumes of up to 50 pl. TABLE VI REPRODUCIBILITY OF CARBON FURNACE ATOMIC-EMISSION MEASUREMENT OF SELECTED ELEMENTS Wavelengths given in Table VII ; optimuni conditions as given in Table I V ; 5O-pl injections.Relative standard deviation yo * Concentration/ Element pg 1-1 Cr . . 50 c u . . 50 Fe . . 100 In . . 500 K 1000 Mn . . 10 Na . . 1000 A f \ Standard tube Standard tube with platform 2.6 1.9 4.1 2.3 4.3 3.1 5.5 3.2 2.4 1.9 2.6 2.3 2.6 2.0 * 10 replicates. Effect of Platform Atomisation on Sensitivity and Detection Limits As reported previ~usly,~ the detection limits achieved with this system are extremely low for many elements. The number of elements investigated to date has been considerably extended and the detection limits obtained are shown in Table VII and compared with values obtained using other relevant spectroscopic systems.Table VII includes the wavelengths used for each element and the corresponding energies of the upper level of each transition and also the appearance temperatures of most elements.12 Also included are the improvement factors obtained by use of the platform in terms of both detection limit and sensitivity. Column 5 in Table VII lists the best detection limits achieved without background correction but using the optimum temperature and tube type referred to. Column 6 gives the values obtained with a 0.75-m Ebert monochromator combined with an HGA 2200 furnace and the oscillating refractor plate form of wavelength modulation. Comparison of columns 5 and 6 shows the advantages of the use of wavelength modulation and a higher resolution .monochromator as this system gave significant improvements for all elements except lead.Further significant improvements have been achieved with the present system both with and without platform atomisation. Beryllium and barium are the only elements for which detec-tion limits are not lower using the present system barium undoubtedly owing to the poo 570 BEZUR et aE. PLATFORM ATOMISATION TABLE VII COMPARISON OF CFAES DETECTION LIMITS Detection limit (DL)/yg I-’ Element 2 * * * * Au Ba ,. Be Cd Cr Go c u Fe Ga . . Ho In K Na ._ Ni Pb s c Sr Yb : 2 a . * * - I nm 328.07 396.15 267.59 553.55 234.86 326.11 425.43 345.35 324.75 404.89 459.40 371.99 403.29 405.39 410.17 404.41 285.21 403.08 330.23 341.48 405.78 402.04 460.73 398.79 eii 3.78 3.14 4.63 2.24 5.28 3.80 2.91 4.02 3.78 3.06 2.70 3.33 3.08 3.06 3.02 3.06 4.34 3.0R 3.75 3.11 4.38 3.08 2.69 3.11 ‘ K 1120 2070 1370 2 200 730 1740 1640 1460 (-) (-) 1540 1350 (-4 lkb 1580 1530 1510 1250 1675 1060 2 450 2100 (-) 3 Appearance Without With Present work Wavelenethl Ed temueraturel*/ wavelength wavelength Present work STD tube + modulation mod~lation4.~~ STD tube platform 0.44t 4.17 12s 160t 4601 52s 35a “?7,, 1.45 l.3** 0.66t 2 y 505 217 (-1 38 1.511 (-4 23% 27t 147 1.27 0.367 1.4 0.9 0.4 0.044 2.4 0.1 13.0 1.0 0.3.5 0.1x 2.1 2.6 0.9 0.06 5.7 2.1 0.92 0.31 0.026 360 .5 0 2 8on 26 370 0.051 0.030 0.26 5.0 0.023 1.6 0 .In1 9.1 0.73 0.1 8 0.039 17.4 0.025 2.0 0.3 0.029 3.6 2.1 3.4 0.35 0.25 0.016 25 5 0 Analyst Vol. 108 Improvement in detection Improvement limits in sensitivity, DL(STD) S(PL)* ~ -DL(PL) 27 30 14 10 56 1.5 1.9 1.5 2.9 1.4 I .4 1.4 1.6 1.5 H4 1.3 7 2.6 1.6 1.0 2.6 1.2 1.6 108 I , S (STD) 10 20 3.3 1.3 4.5 2.0 1.4 2.3 I .3 1.5 2 5.8 1.4 1.3 3.5 2.9 1.6 1.0 5 3.3 1.2 1.6 22 89 * Ratio of sensitivities measured under same conditions. t With cup tube.ll $ (-) Not measured.5 Standard tube. 7 H-igh temperature tube.* 11 With tapered tube.D ** With volatile-elements tube.D sensitivity of the photomultiplier in the wavelength region of the barium line. Comparison of columns 7 and 8 in Table VII indicates the improvements obtained by use of a singlecurved platform. Calculated improvement factors for both the detection limits and sensitivities are given in columns 9 and 10. The effectiveness of platform atomisation in CFAES is shown by the improvement in detection limits for all elements varying between 1.5 and 108 and in sensitivity between 2 and 89. For elements such as lead and silver the improvement is partly due to the reduction in noise and for others such as gallium it is almost certainly partly due to improved atomisation efficiency.The main contribution however derives from the increased emission intensity provided by the higher temperatures available during the lifetime of the atom population. The increases in sensitivity are consistent with the calculated values shown in Table 111. The linear range of the method is considerably extended as a result of the improvement in detection limits. Some typical growth curves using platform atomisation are shown in Fig. 14. Curvature begins to recur at roughly the same analyte concentration as with tube-1 I 1 I I I -4 -3 -2 - 1 0 1 2 3 Log c Fig. 14. Analytical curves of growth using plat-form atomisation for chromium lithium manganese, iron and nickel log (relative intensity IR) veYszts log (concentration pg ml-l). Wavelengths as in Table VI; conditions as in Table IV; 5O-pl aliquots May 1983 IN CARBON FURNACE AES 57 1 wall atomisation,2s but the linear part of the graph extends to lower concentrations owing to the improvement in detection limit offered by the platform.The linear dynamic range is about three orders of magnitude for most elements which is adequate for most analytical purposes, but not as extensive as the linear range obtained in inductively coupled plasma emission spectrometry. For many elements alternative wavelengths are available for the higher con-centration regions. I t should be noted that the detection limits for several elements are approaching the ultimate measurement levels where the greatest problems do not involve the sensitivity of the technique but those derived from contamination.The sensitivity of the method using the main lines for sodium and potassium is so great that for most practical purposes weaker lines must be used, as indicated in Table VII. Precise comparison of the two spectrometer wavelength modulation systems shown in columns 6 and 7 of Table VII is difficult. The square-wave wavelength modulation system used here should give an over-all improvement of 1 .8,19 but variations in performance are bound to be contributed by the different spectrometers and photomultiplier tubes used and particu-larly by the different carbon furnaces. The HGA 2200 and HGA 72 atomisers have different tube sizes heating rates and diffusional loss mechanisms and consequently both the atomisa-tion efficiencies and also the average vapoiir temperatures experienced by the atom population will be different.Direct comparisons of the alternative methods of wavelength modulation and alternative furnace types using the same spectrometer system have yet to be carried out and remain of interest. The characteristics of a monochromator optimum for CFAES will be a balance between aperture and band pass. In such a system furnace design will also be an important parameter With automatic background correction better results may possibly be obtained with a spectrometer of wider aperture and band pass than that used in this work. The results presented in this paper suggest that CFAES is one of the most sensitive emission techniques for a wide range of elements and that excellent results may be achieved with at least two different optical systems.Platform atomisation offers significant advantages in terms of sensitivity and combined with the reduction in chemical interferences,15~17 represents the most convenient and efficient means of atomisation so far described. Platforms can be used with any commercial electrothermal atomiser require slight adjustments to the atomisa-tion programme and are simple to prepare. The increase in tube lifetimes which also results, is an attractive by-product in view of the current price of graphite tubes for most instruments. The possibility of simultaneous multi-element analysis by CFAES is made more feasible by the use of platform atomisation as compromise atomisation conditions can be used without signifi-cant reductions in sensitivity for any element.The very low detection limits reported in this paper suggest the application of the technique to many real analytical problems. A specific interest in this laboratory is the analysis of trace elements such as manganese chromium and lead in biological materials.6 The detection limits for these and other elements are well below the currently accepted normal levels in blood serum and urine and indicate that results of good accuracy and precision could be obtained with little or no sample pre-treatment. The wave-length modulation used will also correct for increased background signals from light scattered by matrix particles. Detailed results of these studies will be presented in the near future. This work was made possible by the award of grants (to J.M.O.) by the Royal Society for the purchase of the HGA 72 carbon furnace atomiser and by the SEKC for the purchase of the khelle spectrometer.The award of a visiting fellowship (to L.B.) as part of the cultural agreement between the Ministry of Education of Hungary and the British Council is also gratefully acknowledged as is the leave of absence granted to L.B. from the Technical Uni-versity of Budapest. References 1. 2. 3. 4. 5 . 6. Ottaway J . M. Hutton 13. C. Littlejohn D. and Shaw,.F. Wzss. 2. Karl-Alnrx-Univ. Lezpzzg 1979, Ottaway J. M. and Shaw F. AppZ. Spectrosc. 1977 31 12. Littlejohn D. and Ottaway J. M. Analyst 1977 102 553. Epstein $1. S. Itains T. C. and O’Haver T. C. APPZ. Spectrosc. 1976 30 324. Ottaway J . A I . 13czur L. antl Marshall J. ‘4nalyst 1980 105 1130. Ottaway J . A I . I3czur Id. Fakhrul .4ltlecn I t . Frech W. antl Marshall J . “‘l‘racc 1’:lemcnt Analytical 28 357. Chemistry in Jlcdiciiic and 13iology,” Walter dc Gruytcr Munich 1980 p. 575 572 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. BEZUR MARSHALL OTTAWAY AND FAKHRUL-ALDEEN Marshall J. Bezur L. Fakhrul Aldeen R. and Ottaway J. M. Anal. Proc. 1981 18 10. Littlejohn D. and Ottaway J. M. Anal. Chim. Acta 1979 107 139. Littlejohn D. and Ottaway J. M. Analyst 1979 104 1138. Ottaway J. M. and Hutton R. C . Analyst 1976 101 683. Littlejohn D. and Ottaway J.M. Analyst 1978 103 662. L’vov B. V. Spectrochim. Acta Part B 1978 33 153. Gregoire D. C. and Chakrabarti Anal. Chem. 1977 49 2018. L’vov B. V. Pelieva L. A. and Sharnopolskii A. I. Zh. Prikl. Spektrosk. 1977 27 395. Slavin W. and Manning D. C. Anal. Chem. 1979 51 261. Katskov D. A. and Grinstein I . L. Zh. Prikl. Spektrosk. 1978 28 968. Slavin W. and Manning D. C . Spectrochim. Acta Part B 1980 35 701. Michel R. G. Sneddon J. Hunter J. K. Ottaway J. M. and Fell G. S. Analyst 1981 106 288. O’Haver T. C. Epstein M. S. and Zander A. T. Anal. Chem. 1977 49 458. Littlejohn D. and Ottaway J. M. Analyst 1979 104 208. Littlejohn D. and Ottaway J. M. Analyst 1978 103 595. Ottaway J. M. and Shaw F. Analyst 1975 100 438. Littlejohn D. and Ottaway J . M. Anal. Chim. Acta 1978 98 279. L’vov B. V. “Atomic Absorption Spectrochemical Analysis,” Hilger London 1970. Rowston W. B. and Ottaway J . M. Analyst 1979 104 645. Bezur L. Marshall J. and Ottaway J . M. in the press. Epstein M. S. Rains T. C. Brady T. J. Moody J . R. and Barnes I. L. Anal. Chenz. 1978 50 874. Littlejohn D. and Ottaway J. M. Can. J. Spectrosc. 1979 24 154. Received September 30th 1981 Accepted December 20th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800553

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst May 1983 Vol. 108 pp. 573-580 573 Identification and Determination of Microgram Amounts of Phenol and Chlorophenols with Methylbenzothiazolinone Hydrazone Using the Ring -oven Technique Nelly G. Buckman John 0. Hill and Robert J. Magee Department of Inorganic and Analytical Chemistry La Trobe University Bundoora Melbourne Australia The ring oven in conjunction with the reagent methylbenzothiazolinone hydrazone has been used for the determination of chlorophenols in water. Four different methods were employed aqueous in situ solvent extraction using a single and a mixed organic solvent in situ total extraction of a mixture of ten phenols and in situ separation of binary phenol mixtures. These methods appear to be suitable for the determination of phenols in water with a detection limit of 2 x mg 1-1 for the in situ methods the detection limit being defined as the concentration for which no visible ring is formed for use a t a lower limit on the standard scale.mg 1-1 for solvent extraction and Keywords Water analysis ; ring-oven technique ; methylbenzothiazolinone hydrazone ; phenol and chlorop?ienol determination The use of 3-methyl-2-benzothiazolinone hydrazone (MBTH) for the colorimetric determ' ina-tion of phenols is well documented. MBTH was first synthesised by Besthornl but it was not until nearly 50 years later that Hunig and Fritsch2 reported that new azo dyes could be obtained by the oxidative coupling of MRTH with phenols. The reagent was initially used by Kamata3 as a spot test for phenol derivatives and for the determination of phenol by Umeda.* I t formed the basis of an automated procedure for the determination of phenols developed by Friestad et aL5 Gasparic et aL6 investigated the colour reaction of a selected series of phenols with MBTH reagent using spectrophotometric and chromatographic methods.According to Hiinig and Fritsch,2 phenol reacts in the para-position to the hydroxyl group, so that the reaction may be written as follows : CH3 CH3 However in all of the reported work on the reaction of MBTH with phenols little or no atten-tion has been given to reaction with the chloro-substituted phenols which comprise a large group of environmentally significant compounds. Further no work has been reported as far as we are aware on the use of the ring oven for the detection and determination of phenols using this reagent.In an earlier paper,' an investigation into the applicability of 4-aminoantipyrine (4-AAP) as a spot reaction for chloro-substituted phenols using the ring-oven technique was described. This paper reports a comparable investigation with MBTH and its application to the ring-oven technique. Experimental Reagents Freshly prepared daily and diluted with Milli-Q water to the required concentration. Obtained in the form of the hydrochlor-ide monohydrate from Aldrich Chemical Co. A working solution of 0.125 g in 250 ml of Milli-Q water was used. This was stable for 1 week if kept in a refrigerator. Phenol stock aqueous solations. 3-Methyl-2-benzothiaxolinone hydrazone (MBTH) . Obtained from Aldrich Chemical Co 574 BUCKMAN et al.IDENTIFICATION OF PHENOL AND Analyst Vol. 108 Obtained from Mallinckrodt Inc. A 2-g amount was dissolved in 150 ml of Milli-Q water and 1.5 ml of sulphuric acid were added the volume being made up to 200 ml. Bufer solution. Prepared by dissolving 8 g of sodium hydroxide 2 g of the disodium salt of EDTA and 8 g of boric acid in this order in 200 ml of Milli-Q water and diluting to 250 ml. Working solutions were made by mixing equal volumes of this solution with ethanol. For aqueous determinations successful washing to the ring zone was carried out using a solution that consisted of 10 ml of Milli-Q water containing 4 drops of concentrated sulphuric acid. Ammonium cerium(1V) sulphate. Washing solution. Apparatus The ring oven used was supplied by ROFA Vienna Austria.Results and Discussion Phenol and 12 chlorophenols as listed in Table I were studied. Preparation of Standards A 0.1 mg 1-1 standard aqueous solution of each phenol was prepared from the corresponding stock solution. With the ring oven operating at 100 "C a series of aqueous standard rings were prepared from 1- 2- 4- 6- 8- and 10-pl volumes of the appropriate standard. The range investigated was 0.000 1-0.001 pg of each phenol. For example 1 p1 of phenol standard was added to the centre of the filter-paper and allowed to dry in a flow of cool air. A 10-p1 volume of MBTH was added with subsequent washing to the ring zone. The resultant faint ring was allowed to come almost to dryness and the filter-paper was then sprayed with ammonium cerium(1V) sulphate solution.This was allowed to reach incipient dryness and sprayed with buffer solution. The results obtained are shown in Table I. Four washings each of 20 pl sufficed. On final drying it was covered with paraffin wax. TABLE I RING COLOURS AQUEOUS EXTRACTION Phenol Phenol 2-Chlorophenol . . 4-Chlorophenol . . 2,4-Dichlorophenol . . 3,4-Dichlorophenol . . 2,3,5-Trichlorophenol . . 2,3,6-Trichlorophenol . . 2,4,5-Trichlorophenol . . 2,4,6-Trichlorophenol . . 3,4,5-Trichlorophenol . . . . 2,3,4,5-Tetrachlorophenol . . . . 2,3,5,6-Tetrachlorophenol . . . Pentachlorophenol . . Ring colour Red - brown Cobalt blue Maroon Vermilion Orange Bright green Prussian blue Crimson - pink Scarlet Magenta Blue - green Orange Yellow - orange To test the accuracy of the standards for the determination of concentration test solutions were prepared containing 0.300 0.500 and 0.700 mg 1-l.Standards and test solutions were treated in an identical manner. Three rings of each test solution were prepared and using the method of Weisz,B concentrations were calculated. It can be seen that identification and determination of phenols of the order of 10-4-10-3 pg is possible by development of rings in the aqueous phase using the ring oven with an error of &0.27&. The results are shown in Table XI. Extraction of the Complexes Formed with MBTH 0.2 mm for the extraction of phenols by organic solvents was undertaken. A preliminary thin-layer chromatographic study employing DC-Alufolien Kieselguhr F254, Chlorofor May 1983 CHLOROPHENOLS WITH MBTH USING RING-OVEN TECHNIQUE TABLE I1 EVALUATION OF in-situ AQUEOUS EXTRACTION PROCEDURE Phenol Phenol .. * . 2-Chlorophenol . . 4-Chlorophenol . . 2,4-Dichlorophenol 3,4-Dichlorophenol 2,3,5-Trichlorophenol 2,3,6-Trichlorophenol 2,4,5-Trichlorophenol 2,4,6-Trichlorophenol 3,4,5-Trichlorophenol . . . . . . 2,3,4,5-Tetrachlorophenol 2,3,5,6-Tetrachlorophenol Pentachlorophenol . . . . . . . . . . . . . . Amount added 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 3.00 5.00 7.00 x 1o-4ipg Amount found x 10-41~g Day 1 3.00 5.00 7.01 2.99 5.00 7.00 3.00 5.00 6.99 3.00 5.00 7.02 2.99 4.98 7.00 2.98 4.99 6.99 3.00 5.01 7.00 3.01 5.01 7.00 3.00 5.00 7.00 2.99 5.00 6.99 2.99 5.00 7.01 2.98 5.02 7.01 3.06 5.02 7.03 Day 14 2.99 5.00 7.00 2.99 4.99 6.99 3.00 5.00 6.98 2.98 4.97 7.00 2.99 4.98 6.98 2.97 4.98 6.97 3.00 5.00 7.00 3.00 5.00 6.97 3.01 5.01 7.03 3.00 5.02 7.01 3.01 5.01 7.03 2.97 4.99 6.96 3.00 4.97 6.97 Recovery % Day 1 100.0 100.0 100.1 99.7 100.0 100.0 100.0 100.0 99.9 100.0 100.0 100.3 99.7 99.6 100.0 99.3 99.8 99.9 100.0 100.2 100.0 100.3 100.2 100.0 100.0 100.0 1oo.i) 99.7 100.0 99.9 99.7 100.0 100.1 99.3 100.4 100.1 102.0 100.4 100.4 Day 14 99.7 100.0 100.0 99.7 98.8 99.9 100.0 100.0 99.7 99.3 99.4 100.0 99.7 99.6 99.7 99.0 99.6 99.6 100.0 100.0 100.0 100.0 100.0 99.6 100.3 100.2 100.4 100.0 100.4 100.1 100.3 100.2 100.4 99.0 99.8 99.4 100.0 99.4 99.6 575 extracted all complexes rapidly but non-selectively.Carbon tetrachloride extracted all com-plexes slowly but selectively. An optimisation study of carbon tetrachloride - chloroform (9 + 1) provided good selectivity and rapid elution of all complexes.Each phenol (2 x mg 1-l) was added followed by 2 ml of MBTH and after 5 min 1 ml of ammonium cerium(1V) sulphate solution. After a further 5 min 2 ml of buffer solution were added and the volume was made up to 25 ml. The two sets of flasks were allowed to stand for 15 min and the solutions in one set were extracted with 2.5 ml of chloroform and the other with 2.5 ml of 9 + 1 carbon tetrachloride - chloroform. After shaking vigorously for 5 min and allowing the layers to settle the organic solvent layers were collected after passing through phase-separation paper. Rings containing 10-6-10-5 pg were prepared by placing 5,- lo- 15 20- 30- 40- and 50-p1 volumes of extract on the centre of the filter-paper in the ring oven. Rings were formed by washing the coloured extract with solvent (chloroform or solvent mixture) to the ring zone.The ring oven was operated at 100 "C. The results obtained are shown in Table 111. Using the rings as standards tests were carried out to compare the recovery of each complex using chloroform alone as extractant and a mixture of carbon tetrachloride - chloroform (9 + 1). I t can be seen that chloroform effects an almost complete extraction of each complex from the aqueous phase. Hence for ring-oven analyses two series of 25-ml calibrated flasks were employed. The results are shown in Table IV 576 BUCKMAN et at?. IDENTIFICATION OF PHENOL AND TABLE I11 Analyst VoZ. 108 RING COLOURS ORGANIC SOLVENT EXTRACTION Phenol Phenol . . 2-Chlorophenol . . 4-Chlorophenol . . 2,4-Dichlorophenol .. 3,4-Dichlorophenol . . 2,3,5-Trichlorophenol . . 2,3,6-Trichlorophenol . . . . 2,4,5-Trichlorophenol . . . . 2,4,6-Trichlorophenol . . 3,4,5-Trichlorophenol . . 2,3,4,5-Tetrachlorophenol . . 2,3,5,6-Tetrachlorophenol . . Pentachlorophenol . . . . . . Blank . . Colour of extract in solvent Blood red Cobalt blue Maroon Vermilion Orange Bright green Prussian blue Crimson - pink Scarlet Magenta Blue - green Orange Orange - yellow Pale yellow Colour of ring obtained Grass green Orange - yellow Blue - green Vermilion Orange - chrome Deep purple Crimson - pink Deep orange Scarlet Dark green Magenta Spectrum blue Cobalt blue Cream All calculations were made using the method of Weisz8 The paraffin wax-covered rings To test the reliability of the procedure rings containing 3.00,5.00 and 7.00 x loAs pg of each The results indicate that the use of the mixed solvent system slightly decreases the over-all Using chloroform the error was & O .l % and with the mixed solvent it was &0.4%. were still reliable after storage for 6 weeks. phenol were prepared and compared with standards. accuracy. The results are shown in Table V. TABLE IV EXTRACTION EFFICIENCY ORGANIC SOLVENTS Phenol Phenol . . 4-Chlorophenol . . 2,4-Dichlorophenol . . 2,3,5-Trichlorophenol 2,3,6-Trichlorophenol 3,4,5-Trichlorophenol 2,3,4,5-Tetrachlorophenol 2,3,5,6-Tetrachlorophenol Pentachlorophenol . . Amount added x 10-6/pg . . 1.00 2.00 3.00 .. 1.00 2.00 3.00 . . 1.00 2.00 3.00 . . 1.00 2.00 3.00 . . 1.00 2.00 3.00 . . 1.00 2.00 3.00 . . 1.00 2.00 3.00 . . 1.00 2.00 3.00 . . 1.00 2.00 3.00 Amount found by chloroform extraction x 10-6/pg 1 .oo 2.00 3.00 1 .oo 2.01 3.00 1.00 1.99 3.00 1 .oo 2.00 3.00 1.00 2.00 3.01 1 .oo 2.00 3.00 1.00 2.00 2.99 1.00 2.00 3.00 1.00 2.01 3.00 Amount found in carbon tetrachloride -chloroform (9 + 1) eluent mixture x 10-61pg & Week 1 Week 6 0.89 0.89 1.78 1.77 2.69 2.69 0.79 0.79 1.58 1.57 2.36 2.35 0.92 0.93 1.86 1.85 2.78 2.77 0.99 1 .oo 1.99 1.99 2.99 2.98 0.41 0.40 0.81 0.79 1.24 1.24 0.87 0.87 1.75 1.75 2.62 2.61 0.50 0.48 0.99 0.97 1.51 1.49 0.98 0.98 1.99 1.98 2.97 2.96 0.76 0.76 1.53 1.52 2.20 2.29 Recovery by eluent mixture % -7 Week 1 Week 6 89.0 89.0 89.0 88.5 89.7 89.7 79.0 79.0 78.6 78.5 78.7 78.3 92.0 93.0 93.5 92.5 92.7 92.3 99.0 100.0 99.5 99.5 99.7 99.3 41.0 40.0 41 .O 39.5 41.1 41.1 87.0 87.0 87.5 87.5 87.3 87.0 50.0 48.0 49.5 48.5 50.5 49.7 98.0 98.0 99.5 99.0 99.0 98.7 76.0 76.0 76.1 76.0 76.3 76.May 1983 CHLOROPHENOLS WITH MBTH USING RING-OVEN TECHNIQUE 577 In-situ Extraction For the in sit24 extraction and separation of mixtures of phenols two procedures were investigated. In the first the following mixtures of the phenols were made up from stock solutions using Milli-Q water containing 0.001 mg 1-1 of each phenol (1) phenol + 2,3,6-trichlorophenol ; (2) 2,3,6- + 2,3,5-trichlorophenol; (3) 2,3,6-trichlorophenol + 2,3,5,6-tetrachlorophenol ; and (4) 2,3,5-trichlorophenol + 2,3,4,5-tetrachlorophenol.The two solvents used were (A) chloroform and (B) carbon tetrachloride - chloroform (9 + 1). TABLE V EVALUATION OF ORGANIC SOLVENT EXTRACTION PROCEDURE Phenol Phenol . . 2-Chlorophenol . . . . 4-Chlorophenol . . . . 2,4-DichIorophenol . . . . 3,4-Dichlorophmol . . . . 2,3,5-TrichlorophenoI . . 2,3,6-TrichIorophenol . . 2,4,5-Trichlorophenol . . 2,4,6-TrichlorophenoI . . 3,4,5-TrichlorophenoI . . 2,3,4,5-Tetrachlorophenol . . 2,3,5.6-Tetrachlorophenol . . Pentachloropheiiol . . . . Amount added x 10-6/vg Chloro- Carbon tetrachloridr -form chloroform (9 + 1)* .. 3.00 2.70 5.00 4.51 7.on 6.30 . . 3.nn 2.41 5.00 4.01 7.00 2.80 . . 3.00 2.36 5.00 3.94 5.00 5 3 1 . . 3.00 2.78 5.00 4.63 7.00 6.49 . . 3.00 2.73 5.00 4.55 7.00 6.37 . . 3.00 3.00 5.00 5.00 7.00 6.99 . . 3.00 1.24 5.00 2.06 7.00 2.88 . . 8.00 2.20 5.00 3.66 7.00 5.12 . . 5.00 2.05 5.00 3.42 7.00 4.59 . . 3.00 2.61 5.00 4.35 7.00 6.09 . . 3.00 1.51 5.00 2.50 7.00 3.50 . . 3.00 2.97 5.00 4.95 7.n0 6.93 . . 3.00 2.29 5.00 3.82 5.00 5.35 A 7- 7 Amount found x 1W6/vg Chloro- Carbon tetrachloride -form chloroform (9 + 1) 7 h--3.00 2.69 5.02 4.50 5.02 6.30 3.00 2.40 5.01 4.00 7.02 2.79 2.99 2.35 4.99 3.93 6.98 5.50 2.99 5.00 .LOO 4.99 6.95 4.98 7.00 3.01 5.01 7.02 3.00 5.00 7.00 8.00 4.99 7.01 2.99 5.00 7.00 3.00 4.99 7.01 3.01) 5.00 3.00 5.02 7.02 7.01 3.m 5.00 2.78 4.65 6.50 2.72 4.54 6.35 3.01 5.01 7.00 1.26 2.04 2.85 2.20 3.65 5.15 2.04 3.41 4.79 2.60 4.37 6.10 1.50 2.49 3.47 2.96 4.91 6.90 2.30 3.86 5.37 r-Chloro-form 100.0 100.4 100.1 100.0 700.2 100.3 99.7 99.X 99.7 99.7 100.0 100.1 100.0 99.8 99.6 100.0 99.6 100.0 100.3 100.2 100.3 100.0 100.0 100.0 100.0 99.8 100.1 99.7 100.0 100.0 100.0 99.x 100.0 100.0 100.0 100.0 100.4 100.3 i o n s Recovery yo Carbon tetrachloride -chloroform (9 + 1) 99.6 99.8 100.0 99.6 99.8 99.6 99.6 99.7 100.4 100.0 100.4 100.2 99.6 99.9 99.7 100.5 100.1 99.9 101.6 99.0 98.9 100.0 99.7 100.6 99.5 99.7 100.0 99.6 100.5 100.1 99.3 99.6 99.1 99.7 99.2 99.6 100.4 101.0 100.4 7 -Ap-* Theoretical amount added is bawd on thc mean percentage recovrry data (Table IV prnultimatp column) relating t o the mixed solvent system.Rings were prepared containing 5 x 10-6 pg of each phenol. Two 5-pl volumes of each mixture were added to the centre of the filter-paper with a micropipette and allowed to dry. Next 10 p1 of MBTH were added and allowed to dry followed by 3 pl of ammonium cerium(1V) sulphate solution. After drying 4 pl of buffer solution were added and the spot was allowed to dry. For each determination rings were formed by washing with solvent A and solvent B.With the former 80 pl were used 10 p1 at a time to form a single sharp ring zone. With solvent B sufficient solvent mixture was used to form two ring zones one more diffuse than the other. The temperature of the ring oven for this operation was maintained between 80 and 100 "C. The results are shown in Table VI and give an indication of the phenols present the colour of the ring correlating with those of the standards in Table 111. To test the in situ procedure for a series of phenol mixtures ten phenols (phenol 2-chloro-phenol 2,4-dichlorophenol 2,4,5- 2,4,6- 2,3,6- and 3,4,5-trichlorophenol 2,3,4,5- and 2,3,5,6-tetrachlorophenol and pentachlorophenol) were diluted in a 100-ml calibrated flask so that the solution contained From 1 2 4 6 8 and 10 p1 of this mixture (mixture 1) standards were produced using the procedure outlined previously using chloro-form as eluting solvent.mg of each phenol 578 BUCKMAN et al. IDENTIFICATION OF PHENOL AND TABLE VI Analyst VoZ. 108 RING COLOURS in-SitU CHLOROFORM EXTRACTION Ring colour I A 1 Solvent B I A > Inner ring Mixture Solvent A Outer ring Inner ring diameterlmm 1 Brown Deep green Red - brown 9.0 2 Magenta Purple Crimson 12.0 3 Purple Cobalt blue Violet - blue 11.0 4 Deep purple Purple Red - purple 10.0 For mixture 2 standard rings were prepared by the same procedure consisting of 2-chloro-phenol 2,4-dichlorophenol 2,4,5- 2,4,6- 2,3,6- and 3,4,5-trichlorophenol (1.5 x mg), and phenol 2,3,4,5- and 2,3,4,6-tetrachlorophenol and pentachlorophenol (2.5 x mg).This mixture contains a large proportion of “red” complexes; the total phenolic species con-centration is 10-5mg per 100ml. For mixture 3 standard rings were also prepared the mixture consisting of 2-chlorophenol 2,4-dichlorophenol 2,4,5- 2,4,6- 2,3,6- and 3,4,5-tri-chlorophenol (2.0 x mg) ; and phenol 2,3,4,5- and 2,3,5,6-tetrachlorophenol and penta-chlorophenol(2.2 x 10-6 mg). This mixture contains a large proportion of “blue” complexes; the total phenolic species concentration is 10-5 nig per 100 ml. In all instances the inner ring diameter was found to be inherent to the phenol binary mixture provided that the pro-cedures and solvent volumes were identical for all mixtures. pg of total phenol were prepared and compared with standards.The results are shown in Table VII. The rings exhibited stability for at least 6 weeks. Recoveries for mixtures 2 and 3 in com-parison with 1 are of the order of -0.4 and + 1.4% respectively. To test the validity of the procedure rings containing 3.00 5.00 and 7.00 x TABLE VII EVALUATION OF in-situ SOLVENT EXTRACTION PROCEDURE Mixture Amount of total & phenol added No. Colour x 1o-7lpg 1 Maroon 3.000 5.000 7.000 2 Burgundy 3.000 5.000 7.000 3 Purple 3.000 5.000 7.000 Amount of total phenol found & Day 1 Week 6 3.000 2.998 4.995 4.998 7.007 7.006 2.991 2.990 4.978 4.976 6.971 6.971 3.057 3.057 5.064 5.063 7.066 7.064 x 1o-7lpg Recovery % & Day 1 Week 6 100.0 99.9 99.9 100.0 100.1 100.1 99.7 99.7 99.6 99.5 99.6 99.6 101.9 101.9 101.3 101.3 100.9 100.9 A second procedure was also investigated for the analysis of total phenol mixtures.The following mixtures were prepared containing 0.01 mg 1-1 of each phenol (1) 2,3,5-trichloro-phenol + 2,3,6-trichlorophenol and (2) 2,3,5,6-tetrachlorophenol + 2,3,6-trichlorophenol. To prepare the rings for each mixture 2 p1 of solution were added to the centre of the filter-paper by means of a micropipette and allowed to dry. The phenols in each mixture were then washed outwards in the following way. For mixture 1 20 p1 of the solvent carbon tetra-chloride - chloroform (9 + 1) were followed by 10 pl when the initial amount was almost dry. For mixture 2,20 pl of the same solvent mixture were followed by 10 p1 while still moist and a further 10 p1 when the second addition was dry.The paper was then sprayed with MBTI-I reagent and while still moist with ammonium cerium(1V) sulphate solution. When dry the papers were sprayed with buffer solution and after drying were coated with paraffin wax. The results obtained are shown in Table VIII May 1983 CHLOROPHENOLS WITH MBTH USING RING-OVEN TECHNIQUE TABLE VIII RING COLOURS in-situ ORGANIC SOLVENT EXTRACTION BINARY MIXTURES Colour of ring’ formed 579 r 1 Inner ring Mixture Outer ring Inner ring diameterlmm 1 Purple Crimson 13.0 2 Cobalt blue Violet - blue 12.0 Standard rings were prepared for mixtures 1 and 2 of different concentrations using 1,2,4,6, 8 and 10 pl of a solution containing 0.01 mg 1-1 of each phenol in the mixture.The centre of the ring was then cut out and placed on another filter-paper on the ring oven and the complex washed quickly to the ring zone with 60 pl of chloroform; the operating temperature of the ring oven was 100 “C. The ring colours were compared with those shown in Table VIII for the respective colours. To evaluate the method rings were prepared containing 1.5 3.0 and 5.0 pl of mixtures 1 and 2 and the results calculated for the “unknowns” using the Weisz method. The results are shown in Table IX. TABLE IX EVALUATION OF in-situ SOLVENT EXTRACTION Phenol added to ring Mixture Phenol x 1o-5ipg 1 2,3,5-Trichlorophenol . . . . 1.500 3.000 5.000 2,3,6-Trichlorophenol . . . . 1.500 3.000 5.000 2 2,3,5,6-Tetrachlorophenol .. 1.500 3.000 5.000 2,3,6-Trichlorophenol . . . . 1.500 3.000 5.000 PROCEDURE BINARY MIXTURES Phenol found in ring x 10-5/~g Recovery % r Apparent * 1.500 3.000 5.006 1.499 3.001 5.002 1.500 3.002 5.000 1.499 3.001 5.000 1 Actual? 1.496 2.990 4.985 1.316 2.628 4.385 1.480 2.958 4.930 1.217 2.433 4.060 kpparent* Actual! 100.0 99.7 100.0 99.7 100.1 99.7 99.9 87.7 100.0 87.6 100.0 87.7 100.0 98.7 100.1 98.6 100.0 98.6 100.0 81.1 100.0 81.1 100.0 81.2 * Apparent values were determined by comparison of rings with standards for the mixtures 1 and 2 t On basis of Table 10. containing 1 2 4 6 8 and 10 pl. It should be noted however that the colour of the 2,3,5-trichlorophenol and the 2,3,5,6-tetrachlorophenol rings was due to the presence of some 2,3,6-trichloropheno1.It was necessary to correct determinations by carrying out a phenol distribution study that permitted a better determination of individual phenols in the mixture as shown in Table IX. To carry out the phenol distribution study it was necessary to work at higher concentrations, as the 2,3,6-trichlorophenol could not be detected at the lower concentrations in the outer ring. This study has shown that the components of an aqueous mixture of phenols can be identi-fied and that the total phenol content can be determined using the ring oven. Also some binary mixtures of phenols can be separated into the components by in sit% mixed solvent extraction on the ring oven. Hence the method is useful for the determination of phenols in pure water. Investigations on sea water obtained from the Port Phillip Ray area south of Melbourne Australia which is known to contain phenols is in progress in order to assess the application of this method to the analysis of environmental samples. The results are shown in Table X 580 BUCKMAN HILL AND MAGEE TABLE X Phenol 2,3,5-Trichlorophenol . . 2,3,5,6-Tetrachlorophenol 2,3,6-Trichlorophenol . . 1. 2. 3. 4. 5. 6. 7. 8. PHENOL DISTRIBUTION STUDY Phenol added x 10-5/~g . . 2.000 4.000 6.000 . . 2.000 4.000 6.000 . . 2.000 4.000 6.000 2.000 4.000 6.000 Amount of eluent Phenol found 20 1.994 20 3.980 20 5.980 20 + 10 1.970 20 + 10 3.944 20 + 10 5.910 20 0.25 20 0.49 20 0.74 20 + 10 0.37 20 + 10 0.13 20 + 10 1.10 added/pl x 10-5jpg References Recovery yo 99.7 99.5 99.7 98.5 98.6 98.5 12.5 12.3 12.3 18.5 18.3 18.3 Besthorn E. Bar. Dtsch. Chem. Ges. 1910 43 1519. Hunig S. and Fritsch K. H. Am. Chem. 1957 609 143. Kamata E. Nippon Kagaku Zasshi 1966 87 380; Chem. Abstv. 1966 65 16052a. Umeda M. Yakugaku Zasshi 1963 83 951; Chem. Abstr. 1964 60 11384b. Friestad H. O. Oii,D.E. and Gunther F. A. Anal. Chem. 1969 41 1750. Gasparit J. Svobodovh D. and PospiSilovA M. Mikrochim. Acta 1977 1 241. Buckman N. Hill J. O. and Magee R. J. Microchem. J. in the press. Weisz H. “Microanalysis by the Ring-Oven Technique,” Pergamon Press Oxford 1970. Received August 2nd 1982 Accepted December 7th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800573

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst May 1983 Vol. 108 pp. 581-590 581 Calcium Ion-selective Electrode Studies Covalent Bonding of Organic Phosphates and Phosphonates to Polymer Matrices P. C. Hobby G. J. Moodyand J. D. R. Thomas Applied Chemistry Department Redwood Building U WIST Car&# CF1 3NU The covalent linking of organophosphates and phosphonates to the copolymer VAGH (a partially hydrolysed copolymer of vinyl chloride and vinyl acetate) is described along with attempts to phosphonate polystyrene by Friedel -Crafts and free-radical processes. The various products obtainecl have been evaluated for use in calcium ion-selective electrode membranes. The best membranes were all based on VAGH a product from each type being selected for further study for their calcium ion-sensing qualities namely VAGH PI, VAGH PI1 and VAGH PIII.VAGH PI is based on the covalent bonding of monodecylphosphate to VAGH while VAGH PI1 resulted from the binding of mono-[4-( 1,133-tetramethylbutyl)phenyl]phosphate. Viable electrode membranes required additional poly(viny1 chloride) support and dioctyl phenylphosphonate solvent mediator. VAGH PI11 was obtained by a synthesis amounting to covalently binding monooctyl phenylphosphonate to VAGH and was used with some additional dioctyl phenylphosphonate solvent mediator and calcium bistdi-[4-( 1,1,3,3-tetramethylbutyl)phenyl]phosphate} sensor in electrode membranes. The VAGH PI electrode matched those based on equivalent ungrafted didecylphosphate sensors except that calcium ion selectivity was poor in a high sodium ion background.The VAGH PI1 electrode on the other hand, did not fulfil the performance of electrodes based on free calcium bis{di-[4-(1,1,3,3-tetramethylbutyl)phenyl]phosphate} as ion sensor and there was less selectivity for calcium over sodium and magnesium. Some loss of selectivity was also characteristic of the VAGH PI11 electrode but its slightly lengthened lifetime was insufficient to warrant synthesis of the phosphonated polymer matrix. Keywords Calcium ion-selective membrane electrode ; covalent bonding of ion sensors ; grafted ion sensors Functional lifetimes of ion-selective electrodes based on polymer matrix membranes with trapped liquid ion-exchanger sensing components are normally considerably shorter than for solid-state membrane electrodes. Among the reasons for this is the leaching of active com-ponents from the membrane especially at the membrane - solution interface.Various attempts have been made to lessen this detrimental characteristic by the covalent bonding immobilisation of active components to the polymer matrix membrane support The deterioration of surfactant-sensitive membrane electrodes and their inability to dis-tinguish between anionic and cationic surfactants can be overcome by using membranes with fixed charges chemically bound to a polymer matrix.l Thus for anionic surfactant-sensitive electrodes a tertiary amine was bound to the ends of poly(viny1 chloride) (PVC) chains and converted into a positively charged quaternary ammonium site by reaction with an alkyl br0mide.l Electrode membranes cast from tetrahydrofuran solution were conditioned in aqueous solutions of sodium dodecylsulphate in order to exchange the bromide ions with dodec ylsulphate.Grafted cationic surfactant-sensitive electrode membranes were made by the low-tempera-ture polymerisation of vinyl chloride using the SOs- radical anion. This gave a polymer of relative molecular mass (approximately 77000) in which about one third of the polymer chains terminated with sulphonate end gr0ups.l Membranes were conditioned by exchanging the associated hydrogen ions for the desired surfactant cation. Both the anionic and cationic surfactant electrodes could be used to determine surfactant ion activities a t above and below the critical micellar c0ncentration. 582 Analyst Vo,?. 208 Among the approaches to immobilising the ion-exchange site of calcium ion-selective elec-trodes is the binding of the styrene - b-butadiene - b-styrene (SBS) triblock ela~tomer.~-~ Thus, membranes were produced by firstly cross-linking SBS with triallyl phosphate followed by alkaline hydrolysis of the resulting covalently bound trialkylphosphate grouping to yield a pen-dant dialkylphosphate.Conditioning of the membrane then yielded a covalently bound calcium dialkylphosphate capable of acting as a calcium ion sensor.2 Good calcium ion-selective electrodes of lifetimes in an excess of 6 months were obtained although of limited selectivity towards calcium over important ions such as sodium and magnesium.2 Use of triundec-10-enyl phosphate and of diallyl phenylphosphonate instead of triallylphosphate gave robust cross-linked membranes but without significant improvements in ~electivity.~ While the lack of selectivity precludes the use of these electrodes in biomedical applications the durability of the triallyl phosphate-based electrode was demonstrated for monitoring coke-oven effluentss An alternative approach for grafting the organophosphate sensor to the polymer matrices is to condense monodecyl dihydrogen phosphate with the hydroxyl groups of a partially hydro-lysed vinyl chloride - vinyl acetate copolymer (copolymer VAGH) .6 The resultant product, when fabricated into membranes incorporating dioctyl phenylphosphonate yielded good calcium ion-selective electrodes of fast and steady response with near normal selectivity but without the advantage of an extended lifetime over the PVC matrix membrane electrodes with the physically trapped sensor.6 Among possible reasons for the relatively poor selectivity of the triallyl phosphate - SBS electrode is the lack of dioctyl phenylphosphonate solvent mediator which has been shown to be a necessary component in PVC membrane-based electrodes for promoting good calcium ion selectivity.'-9 The relatively short lifetimes of the PVC matrix membrane electrodes with grafted alkyl phosphate can on the other hand be due to leaching of the solvent mediator from the membrane.Therefore the earlier studies on grafting of phosphate to copolymer VAGH has been re-examined and extended to grafting of improved alkyl phenylphosphate calcium ion sensor and to grafting of the phosphonate solvent mediator.HOBBY et al. CALCIUM ISE STUDIES COVALENT BONDING Experimental Phosphate Ester Grafting to Copolymer VAGH The procedure previously employed6 has been modified by having an alternative method for removing pyridine and by varying the method of bringing the product into solution. The typical procedure for the grafting of two phosphate esters is described below and the variations in reactant amounts for reaction parameter studies are summarised in Table I. A 10-g mass of VAGH (relative molecular mass approximately 23000) [Union Carbide (UK) Ltd.] was purified by dissolving it in 100 cm3 of acetone and stirring the solution into 500 cms of methanol. The resulting fibrous white precipitate of VAGH was filtered and dried at 60 "C. Higher temperatures caused discoloration and degradation of the polymer.Mono-decyl dihydrogen phosphatelo (MDDP) and mono- [4-( 1,1,3,3-tetramethylbutyl)phenyl] dihydrogen phosphatell (MTMBPDP) were grafted to the VAGH by a modification of the procedure of Blackburn et aZ.12 for insoluble polymers. Thus 1.30 g of MDDP (0.00055 mol) or 1.57 g of MTMBPDP (0.0055 mol) previously dried in a vacuum desiccator over silica gel were added to a solution of VAGH (4.00 g i.e. 0.0055 mol of hydroxyl) in 30 cm3 of anhydrous pyridine which was freshly distilled from potassium hydroxide. To this was added a solution of dicyclohexylcarbodiimide DCC (2.25 g 0.01 1 mol), also in anhydrous pyridine (20 cm3). The mixture was shaken at room temperature (it was necessary to run at 35 "C for MTMBPDP) for between 24 and 72 h (some of the trial runs reported in Table I used different times).The product was precipitated by pouring the reaction mixture into 500 c1n3 of hot water and filtering. In this way most of the pyridine was removed from the fibrous mass after four successive washings with hot water. The product was then Soxhlet extracted with methanol to remove dicyclohexylurea unreacted MDDP (or MTMBPDP) and final trace amounts of pyridine. The crude polymer was insoluble in various common solvents although it was soluble in pyridine under reflux and swelled in tetrahydrofuran. The insoluble polymer (approximately 4 g) was brought into solution by refluxing with 150 cm3 of methanol for 2-3 h after which the polymer was dissolved in tetrahydrofuran.It was purified by precipitation into cold wel TABLE I PREPARATIVE DETAILS AND ELEMENTAL ANALYSES FOR PHOSPHORYLATED VAGH Reaction conditions Reactants and molar ratios normalised to VAGH A Elemental analysis % (s.d.) I Tempera-Time/d ture/"C Batch No. P (0.10 to.10 t0.10 1.20 (0.05) 1.31 (0.07) 1.66 (0.19) 1.88 (0.08) 1.55 (0.08) 1.64 (0.09) <0.10 C 40.3 --65.3 47.4 41.9 47.5 45.2 41.9 ----48.2 41.5 46.2 45.2 H 5.0 10.0 6.7 5.6 6.7 6.3 5.8 - ---- -6.4 5.2 5.6 5.8 N VAGH DCC MDDP 0 1 0 3 1 1 1 1 1 0 0 0 0 0 0 0 0 MTMBPDP 0 0 0 0 0 0 0 0 0 1 1 2 d 1 0 1 0 1 1 1 30 1 2 0 10.2 2.4 0 2.2 1.5 0 - - 3 20 3 20 1 20 0.33 20 0.83 20 1 20 1.5 20 2 20 3 20 4 5 6 7 8 9 10 1 2 1 2 1 2 1 2 1 0 11 12 13 14 Calcium salt 15 Calcium salt <o.m <0.10 <0.10 2.06 (0.04) 1.82 (0.03) 2.15 (0.08) 2.12 (0.01) 1 a 1 2 1 4 1 4 1 4 1 4 1 4 3 20 8 35 4 20 ---1.6 1.2 8.30 0 3.98 0.5 3 35 3 35 3 35 3 35 TABLE I1 PREPARATIVE DETAILS AND ELEMENTAL ANALYSES FOR PHOSPHONATED Reactants and molar ratios normalised to Reaction conditions Elemental A - f Prepwa- polymer tion Time/h remp./ C P Polymer - - - - - - - - Polystyrene (PS) .. Phosphonated PS (Friedel - Crafts) . . . . . . PS AICla PCl POCl, 1 1.00 1.013 4.17 0.063 0.5 70 4.79 Phosphonated PS (free radical) .. . . . PS Phosphite Benzoyl peroxide 1.00 1.00 0.1 8 100 0.43 3 1.00 1.00 0.1 17 100-160 0.50 4 1.00 1.00 0.5 16 100 1.04 VAGH Phenylphosphoric dichloride 5 1.00 1.45 plus octanolysis 2 + 1 60-105 2.20 7 1.00 1.45 2 60 2.14 6 1.00 1.45 plus methanolysis 2 + 1 60-105 2.09 Phosphonated VAGH . 584 Analyst VoZ. 108 stirred methanol. The purification procedures did not always completely remove the nitrogen and intermediates (Table I) but this was not a factor in electrode performance (Table 111). Calcium salts were prepared by adding a solution of the grafted product (1 g) in 50 cm3 of tetrahydrofuran to a stirred saturated calcium hydroxide solution (500 cm3) blanketed with nitrogen. The white granular calcium salt of the polymer was filtered with Whatman No.50 filter-paper. The hydrophobic product was wetted by 30% methanol in 300 cm3 of water and emulsified by ultrasonic bath treatment for 20 min. The emulsion was "cracked" with 10 g of sodium chloride and the resultant very fine white precipitate was filtered washed and dried. HOBBY et at?. CALCIUM ISE STUDIES COVALENT BONDING The product was filtered and dried at 60 "C. Friedel - Crafts Phosphonation of Polystyrene Phosphonation of polystyrene was based on the Friedel - Crafts approach a procedure for synthesis of 4-vinyldialkylphenyl phosphate13J4 but using polystyrene instead of ethylbenzene as the starting material. The polystyrene used was purified by re-precipitation into cold methanol from tetrahydrofuran solution. For the Friedel - Crafts preparation anhydrous aluminium chloride (7 g or 0.052 mol) was added to 100 cm3 of re-distilled chlorobenzene in a resin kettle with an electrically driven stirrer and protected by a calcium chloride drying tube.The solvent was stirred and heated on an oil-bath to 70 "C to dissolve the aluminium chloride. A solution of purified polystyrene (5 g or 0.048 mol of aromatic substituent) in 30 cm3 of chlorobenzene was added to the reaction mixture followed by phosphorus(II1) chloride (27.5 g or 0.20 mol). The mixture was stirred and heated to 70 "C for 30 min when a brown precipitate had collected on the flask walls. The solid was filtered and washed with 100 cm3 of hot chlorobenzene and the crude product (7.0 g) was stored in a desiccator to prevent the formation of a colourless odourless oil on the solid particles.For displacing the aluminium chloride from the aromatic ring the crude product was placed in toluene with phosphoryl chloride (0.46 g or 0.031 mol) and allowed to stand overnight. The product was filtered washed with nitrobenzene in an ultrasonic bath, filtered and the precipitate washed with tetrahydrofuran under reflux. The dried polymer product (3.9 g) was substantially insoluble but swellable in organic solvents. The synthesis was not completed because of this insolubility and the probable consequent unsuitability for use in ion-selective electrode membranes. Free Radical Phosphonation of Polystyrene This procedure is based on the synthesis of flame-retarding hydrocarbon polymers15 by reacting a hydrocarbon polymer (substantially free of aliphatic substitution but containing aromatic substituents such as polystyrene) with a phosphite [HP(O)(OR),] in the presence of a free-radical source.Purified polystyrene was ground to a particle size of <400 pm with a high-speed blender (Iona Model No. GPE 46D) and the fine particles filtered washed with water and methanol and vacuum dried at 60 "C. Benzoyl peroxide (BDH Chemicals) was purified by re-precipitation into methanol from acetone solution and vacuum drying the precipitate with calcium chloride desiccant. Long storage was with approximately 30% added water in a refrigerator freezer compartment. The purified dry polystyrene (10.4 g or 0.1 mol of aromatic substituent) was stirred with di-(1,1,3,3-tetramethylbutyl) phosphite (30.6 g or 0.1 mol) in a three-necked flask and heated on an oil-bath to approximately 80 "C to form a clear viscous solution.After cooling to room temperature dry benzoyl peroxide (2.42 g or 0.01 mol) was added to the flask and the mixture was slowly heated to approximately 40 "C when stirring was commenced. Further heating was applied and the appropriate reaction temperature maintained for up to 17 h (Table 11). The product was then cooled to approximately 40 "C 50 cm3 of acetone were added and the solution was poured slowly into 500 cm3 of methanol and cooled in an ice-bath. After filtering, the phosphonated polymer (yield 10.1 g) was purified by precipitation in 500 cm3 of methanol from tetrahydrofuran solution filtered washed with 200 cm3 of methanol ground to a fine particle size washed with water and dried at 60 "C.The material was used to fabricate electrode membranes. Other preparations with different ratios of polystyrene phosphite and benzoyl peroxide were made (see Table 11) May 1983 OF ORGANIC PHOSPHATES AND PHOSPHONATES TO POLYMERS 585 Grafting of Phosphonate to Copolymer VAGH The approach used was that used for the synthesis of dialkyl phenylphosphonates from phenylphosphonic dichloride and the appropriate alcohol.16 A solution of dry purified VAGH (10 g or 0.013 8 mol of hydroxyl) in anhydrous pyridine (70 cm3) (freshly distilled from potas-sium hydroxide) was treated whilst stirring in a resin kettle with phenylphosphonic dichloride (4.0 g or 0.02 mol) solution in anhydrous pyridine which was added dropwise over a period of 30 min.Near the end of the addition a gel-like mass was formed. After a further 30 min the reaction mixture was heated with stirring in an oil-bath to 60 "C for 1 h. Octan-1-01 (2.6 g or 0.02 mol) or methanol (1 g or 0.03 mol) was added to the viscous polymer solution and refluxed at 105 "C for 1 h. The product was precipitated by addition to 200 cm3 of methanol and filtered. This was purified by precipitation into 500 cm3 of methanol from tetrahydrofuran solution, filtered washed with 200 cm3 of methanol and dried at 60 "C. The 10.1 g of purified product were used for fabricating electrode membranes. Further batches were prepared with reactant ratios according to Table 11 and with octanolysis being either omitted or replaced by methan-alysis.Membrane Preparation Control membranes were prepared by the standard procedure1' of incorporating 0.04 g of the appropriate phosphate sensor and 0.36 g of dioctyl phenylphosphonate solvent mediator in a solution of 0.17 g of PVC in 6 cm3 of tetrahydrofuran and allowing the solvent to evaporate under controlled conditions. Membranes of soluble acid forms of grafted phosphate prepara-tions VAGH PI and VAGH PI1 (see Table I) were prepared by the same general procedure, that is by taking the preparation with PVC dioctyl phenylphosphonate or other solvent mediators in 6 cm3 of tetrahydrofuran with the modification of casting on to blocks of poly-(tetrafluoroethene) in place of the normal glass plates. Membranes of the calcium salt of VAGH PI were cast in the same way but the insoluble calcium salt of VAGH PI1 was dis-persed in PVC dioctyl phenylphosphonate - tetrahydrofuran as fine swollen particles obtained by treatment for 10 min in an ultrasonic bath.The various component loadings of membranes incorporating grafted phosphate preparations are shown in Tables I11 and IV. TABLE I11 CHARACTERISTICS OF ELECTRODES BASED ON VAGH PI - PVC MATRIX MEMBRANES Preparation I A Electrode Mem- slope at Limit of Life-brane Batch VAGH PI/ PVC/ DOPP/ 25 "C/mV detection/ time/ kgt&a ~~x~ No. No. g g g decade-' Y weeks (5 x 50-8) (a x 10 ) . . 9 0.01 0.17 0.36 27.0 2.0 x lo-' 2-3 2.8 8.0 x lo-' 'IA ' * } 0.05 0.17 0.36 27.6 1.2 x lo-' -2 2.3 7.0 X lo-' IIB :::$ : \ } 0.05 0.12 0.36 26.7-32.7 4.8 x lo-' 2-6 1.2 4.8 x lo-* :;$ : ) 0.07 0.10 0.36 28.2 2.0 x 2 1.8 5.6 x lo-' vA * * } 0.09 0.08 0.36 27.2 1.5 x lo-* 2 3.0 7.8 x lo-* VB VIB VII .. 9 0.20 0.10 0.20 23.7 1.7 x 2 Drifting noisyresponse " } 0.0 0.17 0.31 30.0 10-6 2-5 1.9 4.3 X lo-' VIII 9 0.40 0.10 0.36 28-30 lo-' -2 2.2 5.3 X lo-' IXA * * } 0.60 - 0.36 - - - - - IXB x 5 0.53 - - WouM not calibrate - - -* Numbers in parentheses refer to levels (M) of interferent. Condition of membrane Clear dry soft rubbery homo-geneous Yellow tinge slightly opaque some inhomogeneity lieh hi tackiness As membrane 11 with increased tackiness good electrode mern-brane As membrane 111 As membrane IV but increased tackiness Yellowish dry tough good elec-trode membrane As membrane VI but increased inhomogeneity Dry tough inhomogeneous Taeky membrane with insufficient strength for electrode use Stiff bright yellow transparent, homogeneous The various phosphonated materials prepared were with few exceptions unsuitable for membrane preparation and even for the exceptions no really satisfactory electrodes were obtained.The sensor membranes were fabricated by mixing the Friedel - Crafts phosphonated polystyrene with the phosphonate sensor and pressed into a disc at 10 ton in-2. The phos 586 Mem- Prepara-brane tion batch No. No. I 14 I1 14 I11 14 IV 14 V 15 VI 1s VII 15 VIII 16 HOBBY et al. CALCIUM ISE STUDIES COVALENT BONDING Analyst Vol. 108 TABLE IV CHARACTERISTICS OF ELECTRODES BASED ON VAGH PI1 - PVC MATRIX MEMVRANES VAGH PI1 VAGH PI1 (Ca*+)/g (H+)/g - 0.05 - 0.10 - 0.15 - 0.25 0.04 -- 0.05 - 0.05 - 0.05 PVC/ g 0.12 0.17 0.17 -0.17 0.12 0.12 0.12 Solvent mediator/ g 0.36t 0.31t 0.36t 0.25t 0.36t 0.36$ 0.368 0.348 Electrode Life-slope at Limit of time 35 OC/mV detection/ at 35 "C/ decade-I M weeks 32 8 x lo-' 5-6 28-31.7 lo-' -3 32 2 x -3 30.0 -5 31.3 2 x 10-a -5 31.7 -lo-' -3 (over 2 decades) Would not calibrate -k@ot* Ca Na (5 x 10-8) 40 34 44 -46 14 52 -kE?Na (5 x 10-4) 7.4 x 10-9 5.5 x lo-' 2.0 x lo-' -4.0 x lo-' 6.5 x lo-* 8.1 x lo-' -Condition of membrane Yellowish slight tackiness clear, rubbery some inhomogeneity.Best electrode membranes Yellowish slight tackiness rub-Ta%; and of insufficient me-chanical stre-ngth for elect-rode membrane Clear tough and rubbery Yellowish tough with slight in-homogeneity * Numbers in parentheses refer to levels (M) of interferent.1 Tripentyl phosphate. 9 Trioctyl phosphate. 7 Diphenyl 2-ethylhexylphosphate. Dioctyl phenylphosphonate. phonated polymers were normally mixed with additional dioctyl phenylphosphonate and calcium bis {di-[4-( 1,1,3,3-tetramethylbuty1)phenyl]phosphate) sensor for membrane evalu-ation (Table V). The best membrane examined was membrane VIII (Table V) and the phosphonated polymer of this membrane is called VAGH PIII. TABLE V CHARACTERISTICS OF PHOSPHONATED POLYMER MEMBRANES Mem- Phosphonated Amount of brane polymer polymer/g CaX,*/g DOPP/g Condition of membrane Electrode response I Polystyrene (Friedel - Crafts) I1 Polystyrene (free radical) I11 Three membranes prepared from each of three batches synthesised IV V VI VI I 5 6 7t [ :t IX X 0.0815 0.02 0.53 0.04 0.153 0.04 0.50 0.04 0.50 0.04 0.40 0.04 0.35 0.04 0.30 0.04 0.255 0.04 0.20 0.04 - Light brown solid mechanically - Heterogeneous containing un-fragile dissolved sensor opaque, brittle yellow film membrane 11 little mechanical strength Similar to membrane 111.Labile DOPP extrudes to membrane surface Stiff but deforms under pressure Soft slightly more flexible than 0.055 Slightly less brittle than 0.10 - Stiff yellow clear film 0.102 5 0.15 membrane VI 0.20 Soft fairly flexible 0.25 Soft tacky just sufficiently strong to form a membrane 0.30 Sticky film with little mechanical strength Anionic response of 30 mV decade-' over 10-1-10-' M calcium chloride solution No calibration No calibration No calibration No calibration No calibration Sub-Nernstian slope of -18 mV decade-' Working electrode (VAGH PIII) Working electrode but large drifts in standard cell potential-Insufficient membrane strength CaX = calcium bis{di-[4-(l,l,3,3-tetramethylbuty1)phenyl]phosphate).t Preparation number from Table 11. Membrane Evaluation Membrane circles of sufficient mechanical strength were taken for assembling electrodes with internal solutions of 10-1 M calcium chloride in the usual way. E.m.f. responses of the elec-trodes versus an EIL ceramic-junction calomel electrode (Cat.No. R J23) were measured with a Corning-EE1 Model 112 digital millivoltmeter/pH meter used in conjunction with a Servo-scribe Model RE 4541 potentiometric chart recorder. The various test solutions were main-tained at 25 * 0.1 "C. While making electrode calibrations with unbuffered and pH buffered 10-6-10-1 M calcium chloride solutions static response times1* were noted to within IfIl mV. Dynamic respons May 1983 OF ORGANIC PHOSPHATES AND PHOSPHONATES TO POLYMERS 587 times18 were also recorded to &1 mV the activity level of calcium with which the electrodes were in contact being changed by the dumping technique using a piston syringe and fast stirring. Cation interferences were assessed by the mixed-solution method based on a constant inter-ferent level except that pH interference-free ranges were assessed by maintaining a constant calcium level For pH-buffered calibrations the pH 7.5was set with Tris [tris(hydroxymethyl)methylamine] solution composed of 50 cm3 of 0.1 M Tris (12.114 g 1-l) and 40.3 cm3 of 0.1 M hydrochloric acid diluted to 4 1 with de-ionised water.Between measurements fhe various electrodes were stored in lob3 M calcium chloride solu-tions. Results Phosphate Ester Grafted VAGH Copolymer Table I summarises data for obtaining various products by the described synthetic procedure. The useful products for grafting MDDP are coded VAGH PI (synthetic runs 5-9 in Table I), while those for grafting MTMBPDP are coded VAGH PI1 (synthetic runs 14 and 15 in Table I).The characteristics of membranes and their electrodes prepared from VAGH PI and VAGH PI1 are summarised in Tables I11 and IV respectively. Table VI compares the specifications of the best VAGH PI electrode from Table I11 (mem-brane IIIB) with a control PVC matrix membrane with physically trapped calcium bis-(didecylphosphate) sensor and dioctyl phenylphosphonate (DOPP) solvent mediator and also the specifications of the best VAGH PI1 electrodes from Table IV (membranes I1 and V) with a control of physically trapped calcium bis{di- [a-( 1,1,3,3-tetramethylbutyl)phenyl]phosphate) and dioctyl phenylphosphonate. Details of the phosphorylated membrane of Keil et aLs are included for comparison. Phosphonated Polymer Membranes Table V summarises the main features of products of phosphonation of VAGH and of polystyrene by the Friedel - Crafts and free-radical reactions set up as membranes for electrodes.Table VI includes specifications of the best electrode of Table V namely membranes VIII of Table V (VAGH PIII) set alongside normal PVC matrix membrane controls with physically trapped sensor and dioctyl phenylphosphonate components and electrodes of VAGH PI and VAGH PII. Discussion I t is because liquid ion exchangers are endowed with greater selectivity that they have attracted more attention as ion-selective electrode sensors than the classical resinous ion exchangers. However cation-selective electrodes based on poly(crown ethers) in PVClg and on Dupont’s Nafion polymers20 raise the prospect of selective ion sensors based on solid exchangers and encourages studies for the bonding of liquid ion exchanger type of selective ion sensors to polymer matrices in order to increase electrode lifetimes.Nevertheless the Nafion polymer-based electrodes for potassium caesium and tetrabutylammonium ions are limited to a Nernstian response over just 1.4 orders of magnitude although there is high selectivity for tetrabutylammonium and other large organic cations for example = 8680 (DoTA = dodecyltrimethylammonium). Lifetimes of longer than 6 months have been quoted for the Nafion polymer electrodes.20 As mentioned above the extended lifetime of an electrode based on a dialkylphosphate group covalently bound to a thermoplastic poly(styrene - b-butadiene) triblock ela~tomer~-~ has been foiled by relatively poor selectivity for calcium over other ions of physiological importance.This supports the essential role of an appropriate solvent mediator for promoting selectivity towards calcium ions. Electrodes based on phosphate esters grafted to copolymer VAGH do not exhibit extended lifetimes (Tables 111 I V and VI) possibly because the essential dioctyl phenylphosphonate (or other appropriate solvent mediator) is leached from the membrane. Indeed leaching of the solvent mediator from the acid form of VAGH PI1 phosphorylated membrane into solution ha TABLE VI CALCIUM ION-SELECTIVE ELECTRODE PROPERTIES FOR BEST MEMBRANES OF TABLES 111-V Phosphor ylated VAGH (data from ref. 6) VAGH PI* (free acid) 4.8 x lo-' 4.0 x lo-# 30 (25 "C) 5 5.5-8 1-5 1 2-5 VAGH PIIS (free acid) 1.0 x 10-6 8.0 x 32 (35 "C) 5 No plateau 1-5 1 i) VAGH PIIT (Ca2+ form) 1.0 x 10-30 (35 " 5 No plateau 1-5 1 2-3 -Specification Detection limit in CaCl /M Detection limit in Tris buffered ipH 7:5) CaCl; M : Slope/mV decade-' .. . . . . . . . . Long-term drift/mV d-* . . . . . . . . pH range ( [Ca2+] = 1 0 - l / ~ ) . . . . . . . . Static response times (dilute to concentrated)/min . . Dynamic response time/s (1 cma 0.1 M CaCI, added to lo cm3 0.VOl M CaCl,) . . . . . . . . Opera t ioiial life time/wee ks Selectivity coefficients (mixed sbiution' methdd): . kKtxa ( 5 x 10-3)fS . . . . . . . . . . Standard? 6.6 x lo-' 2.0 x lo-' 32 (25 "C) 1 5-8.5 1-6 1 2-2 Standard$ 3.0 x 10-6 2.0 x 10-8 32 (35 "C) 1 4.5-9 1-5 1.2 x 10-6 .. 4.5 x 10-6 . . 2! (25 "C) i) . . 5.5-8 1-2 I 2 1 i-9 . . 1.0 M Na+) -. . 0.05 M Mg") -1.2 4 0.x4 0.02 14 2 Zero 5.0 x 10-4 46 6 0.048 1.1 0.017 0.72 0.033 28 0.012 Zero 0.040 38 kfl::= ( 5 x 10-s)SS . . . . . . . . . . * Membrane IIIB (Table 111). 1. Membrane of calcium bis(didecy1phosphate) (0.04 g) dioctyl phenylphosphonate (0.36 g) and PVC (0.17 g). $ Membrane I1 (Table IV). f Membrane of calcium bis{di-[4-(1,1,~,3-tetramethylbutyl)phenyl[phosphate} (0.04 g ) dioctyl phenylphosphonate (0.36 g) and PVC (0.17 g). 7 Membrane V (Table IV). 11 Membrane VIIL (Table V preparation 5 of Table 11). ** Membrane VIII (Table V preparation 6 of Table 11).tt Membrane VIII (Table V preparation 7 of Table 11). $$ Figures in parentheses are levels of interferent (B) ion in k!: B May 1983 OF ORGANIC PHOSPHATES AND PHOSPHONATES TO POLYMERS 589 been demonstrated as part of the present work by thin-layer chromatography. In this respect it is interesting to note previous observations22 relating the lifetime of PVC membrane electrodes based on neutral ligands to the rate of loss of plasticiser and/or ionophore from the polymeric phase into the sample solution. Similarly potassium ions are rapidly leached from the sensor disc of PVC potassium ion-selective electrodes while the resistance of the sensor disc eventually approaches that of PVC alone.23 Phosphate Grafted VAGH Copolymer With regard to attempts to fabricate electrode membranes from just the phosphate grafted VAGH with solvent mediator it is noted (Tables I11 and IV) that membrane mechanical strength considerations demand the support of added PVC (M approximately 70000).However in order to obtain functional electrodes from membranes as little as 0.01 g of phos-phate grafted VAGH product is sufficient (Table 111) although the best membranes (IIIB of Table I11 and I1 and V of Table IV) contained 0.05 g of VAGH P I and 0.04 g or 0.10 g of VAGH PII respectively with dioctyl phenylphosphonate and PVC (Table 111 IV and VI). Except for selectivity in the presence of high sodium levels the VAGH P I membranes gave electrodes of similar selectivity to the non-grafted PVC matrix membrane control (Tables I11 and VI) and slightly better response features at lower calcium ion levels but the VAGH PI1 membranes gave electrodes that responded reasonably well in solutions containing calcium ions only but had poor selectivity especially in the presence of sodium and potassium (Tables IV and VI).Also the pH range was poor and no advantage was gained on using either the calcium salt or free acid of VAGH PI1 as sensor (Table IV) or alternative solvent mediators. Therefore the covalent bonding of organophosphate esters to the polymer matrix has not given any further advantages in calcium ion-selective electrode design. On the contrary the electrodes with grafted MTMBPDP are very inferior in general performance when compared with the superior qualities of ordinary PVC matrix membrane control electrodes of the same sensor type (Table VI).Phosphonated Polymer Membrane Electrodes In pursuing this approach it was considered in the light of studies on membranes with covalently bonded phosphates and of leaching experiments that loss of solvent mediator may be a more significant factor in electrode ageing. Moreover the use of a water-soluble di-(4-nitropheny1)phosphoric acid as a calcium ion sensor in electrodes of lifetimes of greater than 2 weeks suggests that loss of sensor is not the limiting factor concerning operational lifetimes of PVC ion-selective electrodes. The insoluble product of the Friedel - Crafts phosphonation of polystyrene did not material-ise as envisaged and when set up in the membrane of an electrode pressed as a disc with cal-cium bis (di- [4-( l l ,3,3-tetramethylbutylphenyl)]phosphate} ion sensor the system was fragile and the response anionic in slope (Table V).The free-radical phosphonation of polystyrene gave products that did not plasticise with free dioctyl phenylphosphonate. However regardless of whether this product was set up as membranes with calcium ion sensor with or without free dioctyl phenylphosphonate no worth-while response was obtained for electrodes incorporating the membranes (Table V). On the matter of poor plasticisation of these polystyrene systems it is observed that a plasticised polystyrene film containing quaternary ammonium ion and sulphonic acid groups has been used successfully as a chloride ion-selective electrode for blood serum analysis.25 Clearly in this part of the study the phosphonated VAGH offers the best approach to this aspect of covalent grafting of solvent mediator (Table V) although only certain compositions provide working electrodes.Membrane VIII (Table V) that is VAGH PI11 provides respectable calcium ion-selective electrodes [VAGH PI11 (5-7) Table VI]. They at least match the original calcium ion-selective electrodes based on dialkylphosphate sensors but fall short of the good calcium ion-selectivity exhibited by the control electrode of the newer calcium bis(di- [4-( 1,1,3,3-tetramethylbutyl)phenyl]phosphate 1 sensor (Table VI). However, the PI11 electrodes offer a slight increase in operational lifetime but at 8-10 weeks this is only a marginal gain relative to the effort involved in synthesising the essential phosphonated membrane ingredient 590 HOBBY MOODY AND THOMAS Conclusion A partially hydrolysed vinyl chloride - vinyl acetate copolymer (VAGH) provides sites for grafting of organophosphate sensors and phosphonate solvent mediators.The former gives functional calcium ion-selective electrodes of good response with grafted decylphosphate. However there is a considerable loss of calcium ion selectivity compared with free sensor, when 4-( 1,1,3,3-tetramethylbutyl)pho~phate is grafted to VAGH copolymer. For VAGH copolymer linked to phenylphosphonate with some added dioctyl phenylphos-phonate for functional membranes and used in membranes in place of PVC and free dioctyl phenylphosphonate the calcium bis (di- [4-( 1 ,I ,3,3-tetramethylbutyl)phenyl]phosphate) sensor does not display its full calcium ion selectivity.Such membranes have however, slightly longer lifetimes than is normally so with PVC matrix membrane calcium electrodes based on organophosphates. This still leaves the easy to fabricate membranes of physical mixtures of sensor plus solvent mediator in PVC as the best viable alternative. The authors thank the Science and Engineering Research Council for a studentship (to P.C.H.) under the CASE scheme in association with the Central Electricity Research Labora-tories Leatherhead Surrey. 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. References Cutler S. G. and Meares P. J . Electroanal. Chem. 1977 85 145. Ebdon L. Ellis A.T. and Corfield G. C. Analyst 1979 104 730. Ellis A. T. Corfield G. C. and Ebdon L. Anal. Pvoc. 1980 17 48. Corfield G. C. Ebdon L. and Ellis A. T. Anal. Proc. 1981 18 112. Ebdon L. Ellis A T. and Corfield G. C. Analyst 1982 107 288. Keil L. Moody G. J. and Thomas J. D. R. Analyst 1977 102 274. Craggs A. Keil L. Moody G. J. and Thomas J . D. R. Talanta 1975 22 907. Moody G. J. Nassory N. S. and Thomas J . D. I<. Analyst 1978 103 68. Moody G. J. and Thomas J. D. R. Ion-Sel. Electrode Rev. 1979 1 3. Nelson A. K. and Toy A. D. F. Inorg. Chem. 1963 2 775. Craggs A. Delduca P. G. Keil L. Key S. J. Moody C . J. and Thomas J. D. R. J . Inorg. Nucl. Blackburn G. M. Brown M. J. Harris M. Ti. and Shire D. J . Chem. Soc. C 1969 676. Davankov A. B. Kabachnik M. I. Korshak V. V. Leikin Yu. A. Okhovetsker R. F. and Leikin Yu. A. Davankov A. B. Korshak V. V. Tsvetkov E. N. and Kabachnik M. I. Vysokomol. Rolich R. J. Fields E. K. U S Pat. 3 220 989 1965. Craggs A. Delduca P. G. Keil L. Moody G. J. and Thomas J. D. K. J . Inorg. Nucl. Chem., Craggs A. Moody G. J. and Thomas J D. R. J . Chem. Educ. 1974 51 541. Moody G. J. and Thomas J. D. R. Lab. Pract. 1974 23 475. Kimura K. Maeda T. Tamura H. and Shono T. J. J . Electroanal. Chem. 1979 95 91. Martin C. R. and Freiser H. Anal. Chem. 1981 53 902. Hobby P. C. Moody G. J. and Thomas J. D. R. Anal. Proc. 1982 19 316. Oesch U. and Simon W. Anal. Chem. 1980 53 692. Davies J. E. W. Moody G. J . Price W. M. and Thomas J. D. R. Lab. Pract. 1973 22 20. Keil L. Moody G. J. and Thomas J. D. R. Anal. Chim. Acta 1978 96 171. Oka S. Sibazaki Y. and Tahra S. Anal. Chem. 1981 53 588. Chem. 1978 40 1483. Tsvetkov E. N. Zh. Obshch. Khim. 1967 37 1605. Soedin. 1969 11 564. 1978 40 1943. Received November 25th 1982 Accepted January 12th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800581

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Amalyst May 1983 Vol. 108 $9. 591-596 591 Determination of the Moisture Content of Starch Using Near Infrared Photoacoustic Spectroscopy Peter S. Belton and Steven F. Tanner Agricultural Research Council Food Research Institute Colney Lane Norwich NR4 7 UA Near infrared photoacoustic spectroscopy has been used t o measure the water content of wet starch samples (over the range 0-30% of water). The relation-ship between moisture content and observed intensity of the photoacoustic signal is analysed in terms of the theory of the photoacoustic effect. It is shown that the non-linear dependence of signal intensity on moisture content is consistent with theory and that this non-linearity is to be generally expected in high water content systems. A method of linearising the data is described.Keywords Near infrared ; photoacoustic spectroscopy ; moisture deternaination ; starch Photoacoustic spectroscopy (PAS) in its simplest form involves the irradiation of a sample using a pulsed light source. Energy is absorbed by the sample and some of it is subsequently converted into heat that is the sample warms and cools at a frequency identical with that at which the light is chopped. The thermal wave generated in this way travels to the surface of the sample and causes a periodic heating and cooling of the surrounding gas layer. The consequent expansion and contraction of the gas give rise to an acoustic wave which is detected using a microphone. A spectrum may be obtained by measuring the intensity of the acoustic signal as a function of the wavelength of the exciting light.The photoacoustic technique is being used increasingly to study the optical properties of opaque samples.lP2 Spectra have been obtained not only from pure solid compounds but also from such diverse materials as plant leaves3 and compounds absorbed on thin-layer chromato-graphic (TLC) plate^.^ PAS is not limited to the measurement of optical absorption however. Adams et al. have used PAS to determine such parameters as sample thickness thermal conductivity5 and the quantum yield of radiative de-excitation6 Because the technique can be used to study opaque and inhomogeneous samples it has considerable potential as an analytical tool in the food industry. A particularly interesting application of PAS is its use to measure water content of food samples using the water absorption peak in the near infrared region of the spectrum.We have examined the near infrared photoacoustic spectrum of the starch - water system not only because this is of considerable interest in itself but also because it serves as a model for many foodstuffs. Starch is often a major food constituent and has a sigmoidal water sorption isotherm typical of many foodstuffs. In addition the range of water contents that can be examined (0-30% by mass) is large and the sorption occurs in a repro-ducible way. There have been a number of studies on quantitative photoacoustic spectros~opy~-~ includ-ing one study on the measurement of moisture ~ o n t e n t . ~ Experimental The starch samples (BDH Chemicals Ltd. potato starch) were pre-dried by placing 0.5 g of the sample in weighing bottles and heating for 3 h at 100 "C.Subsequently they were dried over phosphorus(V) oxide for 1 week. The samples were then allowed to equilibrate with saturated salt solutions for 4 d in 1-1 closed containers. At the end of this time the amount of water sorbed was determined gravimetrically. The photoacoustic spectra of the samples in the 1-2.5 pm region of the near infrared were measured using a commercial (EDT Model OAS 400) photoacoustic spectrometer. This instrument consists of a 300-W high-pressure xenon-arc lamp monochromator mechanical rotating sector PAS cell and a detection system consisting of two lock-in amplifiers. Varia-tions in lamp intensity in this single beam spectrometer are corrected by ratioing the signal to that from a pyroelectric detector.Most of the bands occurring in the near infrared region involve overtone and combinatio 692 BELTON AND TANNER DETERMINATION OF MOISTURE IN AIzaZyst Vd. 108 bands of the fundamental vibrations of functional groups containing hydrogen. The funda-mental oxygen-hydrogen stretching vibration in water occurs at roughly 2.8 pm and its first overtone is found a t 1.4 pm.l0 The combination band at 1.9 pm is useful in detecting water hydroxy groups and was used in this study to monitor water uptake. The range avail-able on our instrument was 10-240 Hz. The 10-Hz frequency was chosen to ensure firstly, that the signal to noise ratio was maximised and secondly that the depth from which the thermal wave escaped was as great as possible (>5 x mm).' As foodstuffs are often heterogeneous in character it is desirable to sample a representative portion of the material under study and thus avoid possible surface anomalies.A further advantage gained in using a chopping frequency of 10 Hz was that any direct elastic coupling between sample and gas was minimised .ll In the comparison of different samples the approach adopted was that of internal ratioing. This involved the comparison of the peak of interest with some other peak that arises from a chromophore assumed to have the same concentration in all the samples studied. Internal ratioing ensures that any variation in signal intensity arising from differences in the coupling of different samples to the microphone is compensated for.The method is discussed in more detail below. In PAS the light source must be modulated at some suitable frequency. Results and Discussion The spectra for dry starch and starch having two different moisture contents are shown in Fig. 1. In order to obtain useful quantitative information from these spectra some considera-tion must be given to the relationship between signal intensity and concentration in photo-acoustic spectroscopy. The intensity R of a signal at some wavelength A can be expressed as - - (1) R = A l o f ( B ) 9 * . where I is the incident light intensity andf(p) is a function relating the amount of light F 2.0 2.5 Waveiengthlpm Fig. 1. Near infrared spectra of potato starch with different degrees of moisture content 1 dried; 2 7.6 g of water and 100 g of starch; and 3 26.4 g of water and 100 g of starch.Point X is the water peak. Points C E and I; are peaks due to starch May 1983 STARCH USING NEAR IR PHOTOACOUSTIC SPECTROSCOPY 593 absorbed by the sample to the intensity of the thermal wave at the sample surface. function will be discussed in more detail below. the thermal wave intensity to the final intensity in the spectrum. having two components : This The term A is a composite term and relates I t may be considered as - * (2) A = A x A . . where A represents the amplification of the microphone and subsequent circuitry and is a constant of the instrument. The parameter A characterises the coupling of the sample to the microphone and this term will almost certainly be sample dependent.Thus A will be affected by the sample surface area and the thermal characteristics of the sample and the coupling gas. For quantitative work therefore when comparing different samples the term A must be eliminated. The simplest way of achieving this is to ratio the moisture peak height to a starch peak that is unaffected by water. Thus (3) The subscripts W and S in equation (3) refer to water and starch respectively. The intensity of light incident upon the cell is in general wavelength dependent but the ratios of these intensities at different wavelengths are a function only of the instrument and remain invariant. In practice however the actual intensities I I incident upon the sample may be affected by light scattering.8 Another problem that can arise with the ratioing method is that the intensity of the starch peak may contain some contribution from the water peak and vice versa.In order to test for sample dependent light scattering and the effects of water on the starch peaks the ratios of the starch peaks relative to each other were examined as a function of water content. The results are shown in Fig. 2. The intensities of points B C D and E (on Fig. 1) \v s 0.8 0.7 o O B 0 0 0 0 0 I I I I I -.-0 5 10 15 20 25 30 Water sorbed per 100 g of dry substrate/g Fig. 2. Variation of ratioed intensity of starch peaks as a function of water content. Points C and E ratioed to point F (Fig. 1 ) ; B ratioed to F and D ratioed to F. Solid lines represent mean values of C E and D. ratioed to point I; (within the same spectrum) are displayed as a function of water concentra-tion.The graph illustrated in Fig. 2 shows that the ratios of points C D and E are approxi-mately constant whilst point B increases dramatically with water content. This increase in the ratio of point B is to be expected in that it will clearly be affected by the proximity of the water peak (point A). It would seem therefore that problems caused by light scattering and the overlap of peaks are not significant in this sample. Point F was chosen as the reference peak (i.e. R,) in all subsequent measurements quoted 594 Analyst Vol. 108 The general form of the functionf(P) has been deduced for a homogeneous medium by Rosencwaig and Gersho quoted in reference 12 and is complicated in form.As a simplifica-tion Rosencwaig and Gersho consider six limiting instances which (apart from very black strongly absorbing solids) show f ( P ) to be of the formf(/3) = K(P) where K is a constant of the sample and P = 2.303~C (E is the molar absorptivity and C is the concentration of absorb-ing species in moles per litre). Fig. 3 shows graphs of R,/Rs veysus mass of added water for two starch samples. A similar graph for water absorbed on microcrystalline cellulose is also included. Even if one allows for the residual intensity at 1.9 pm when no water has been added to the starch then this merely shifts the intercept of the graph. Strictly Fig. 3 should be plotted with concentration rather than mass units along one axis; however the volume of the system increases by less than 10% over the whole range of added moisture and this approximation should consequently only give rise to a small deviation in linearity.It is worth noting that in some systems density effects could be important. BELTON AND TANNER DETERMINATION OF MOISTURE IN Clearly the expected linear relationship is not observed. Here we have assumed that the starch density is constant. 1.1 I = 5 10 15 20 25 0.4 Water sorbed per 100 g of substratelg D Fig. 3. Ratio of intensity of photoacoustic signal a t 1.9 pm to intensity a t point F (Fig. 1) as a function of water content. 1 and 2 two samples of starch and 3 a sample of micro-crystalline cellulose, The non-linear photoacoustic response obtained in this study can be understood on closer examination of the form off@).Poulet et al.' have developed a quite general form of the Rosencwaig - Gersho equation. The only restriction that Poulet et al. specify is that the acoustic signal should originate solely from within the sample. In practice the region from which the signal arises is determined by the thermal diffusion length p. For non-metals p is of the order 10-1 mm or less at the chopping frequency used in this study and as the sample depth is approximately 2 mm this restriction presents no problems. The relationship proposed by Poulet et al. is that In the limit of pp<1 then equation (4) reduces to It is instructive at this point to consider the magnitude of pLp for the starch - water system. If p is the density of moist starch W is the mass of water added per 100 g of starch M is the relative molecular mass of water and WT is the total mass of starch and water the May 1983 STARCH USING NEAR IR PHOTOACOUSTIC SPECTROSCOPY 595 assuming the following values for the parameters in equation (6) p = 1.53 x 103 g 1-1, W = 25 g W = 125 g M = 18 and E = 1.2 1 mol-l cm-l.For pure water p = 6.8 x 10-3 cm and consequently The value of p chosen is that of pure water and this is almost certainly not appropriate for the system described here. However it should be of the correct order of magnitude and is likely to be larger in a crystalline sample such as starch (see p. 96 of reference 12). involved in this work a useful approximation to equation (4) is to assume that (p/3)2 <(pP + 2)2. = 0.3. For the range of values of Under these conditions equation (4) becomes .. Graphs off@) veysus pJQ for equations (4) (5) and (7) are shown in Fig. 4. Equation (5) with its inherent assumptions clearly represents a serious misrepresentation of the graph to be expected from the system studied here. Equation (7) is a good approximation to the full equation and has the advantage of being more tractable mathematically. Substitution of equation (7) into equation (1) yields and thus a graph of 1/R versus l/P should give a straight line. some initial intensity (R,) at 1.9 pm before water has been added. that For the starch - water system equation (8) represents an oversimplification in that there is The intial conditions are and Rs = 2$A I S P P S (PUBS + 2) where the subscript S refers to the starch peak and I to the initial peak intensity.. . RI - I (PPSPI + 2PI) E - I (pPsP1 + 2 P s ) When water is added the total intensity R at 1.9 pm (peak A) is given by Rw = 24A.L p (Pw + PI) d P W + PI) + 2 where /3,v is the optical absorption due to added water. Thus - R\v - I1 (Pw + PI) (PUBS + 2) I s Cp(Pw + PI) + ZIPS Rs On combination of equations (9) and (10) and after some manipulation we obtain p=- Rs - _ - I s JUPs(pP1 + 2) 1 s . P s (&I + 212 I1 2(pPs + 2) + ‘?; 2Pw (pPs + 2) G-RR * ’ Thus Equation (11) is analogous to equation (8) and predicts a linear relationship for the reciprocal of photoacoustic intensity at the water peak (after correction owing to initial intensity) and I.’C’,/W,. Fig. 5 is a graph of P versm W,/W for two sets of.experimental data involving water on starch. For comparison a similar graph for microcrystalline cellulose is shown. Al 596 BELTON AND TANNER 0.3 - 0.2 s k 0.1 6 1 ’ 1 I I 5 10 15 20 25 Mass of substrate/mass of water 0 0.1 0.2 0.3 0.4 0.5 PLP Fig. 4. Plots off( /3) veysus p/3 for 1 equation (4); 2 equation (6); and 3 equation (7). 5 4 k 3 2 Fig. 5. Plots of P versus WS/WW for 1 and 2 two samples of starch and 3 a sample of microcrystalline cellulose. The lines are the least-squares fits to the data. three graphs are linear and thus there is good agreement with the theory outlined above. Slight differences in the gradients and intercepts of the two starch - water graphs can probably be accounted for by the differences in the extent to which the two sets of starch samples have been dried prior to absorbing the water.This leads to different R and 18 values and conse-quently differences in the P versus W,/W graphs. Conclusion The results presented here show that the photoacoustic response to water absorbed on starch is non-linear with the water concentration. The deviation from linearity is most severe when high concentrations of chromophore are studied. This is consistent with a previous studys where a linear response was reported but only a low range of water concentration was investi-gated. Both sets of results can be plausibly explained in terms of photoacoustic theory and a suitable linearising approximation can be made. Results for water on microcrystalline cellulose show the same non-linear behaviour and thus the non-linear response would seem to be a general phenomenon and not a peculiarity of the starch - water system.It should be noted however that the above treatment of results takes no account of the likely heterogeneous nature of the starch - water system. Different water molecules within a starch granule may have different /3 values and also the value of p may change with change in water content. If substantial amounts of water bind to the surface of the starch granule then this will almost certainly affect the coupling of the thermal wave to the microphone. Conse-quently there is a need to work on simpler and better defined systems in order to test the validity of the photoacoustic theory. This work is currently in progress. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Rockley M. J. and Devlin J. P. Appl. Spectrosc. 1980 34 407. Somoano R. B. Angew. Chem. Int. Ed. Engl. 1978 17 238. Adams M. J. Beadle B. C. King A. A. and Kirkbright G. F. Atzulyst 1976 101 553. Rosencwaig A. and Hall S. S. Anal. Chew. 1975 47 548. Adams M. J. and Kirkbright G. F. Analyst 1977 102 678. Adams M. J. Highfield J . G. and Kirkbright G. F. Anal. Chem. 1977 49 1850. Poulet P. Chanibron J. and Unterreiner I*. J . Appl. Phys. 1980 51 1738. Burgraff L. W. and Leyden D. E. Anal. Chena. 1981 53 759. Castleden S. L. Kirkbright G. F. and Menon I<. R. Amzlyst 1980 105 1076. Goddu R. F. and Delker D. A. Anal. Chem. 1960 32 140. Malkin S. and Cahen D. Anal. Chem. 1981 53 1420. Rosencwaig A. “Photoacoustics and I’hotoacoustic Spectroscopy,” John Wiley Sew I;ork 1980, Received iYovewbav Bth 1982 Accepted hrovembev 25th 1982 pp. 93-124

ISSN:0003-2654    

DOI:10.1039/AN9830800591

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst May 1983 Vol. 108 pp. 597-602 597 Substitution in Cellulose Ethers Part 11." Substituents on the Glucose Units Using High-performance Liquid Determination of the Distribution of Alkoxyl ~ ~~ Chromatography Konrad Sachse and Klaus Metzner and Thomas Welsch Kombinat V E B Chemische Werke Buna DDR-42 12 Schkopau German Democratic Karl-Marx- Universitat Leipzig Saktion Chemie Analytisches Zentrum Liebigstr German Democvatic Republic Republic 18 DDR-7010 Leipzig, A procedure is described for the quantitative characterisation of the distribu-tion of alkoxyl groups in methyl- and ethylcelluloses. The total hydrolysate is derivatised by reaction with benzoyl chloride in sodium hydroxide solution. Several favourable implications of pre-column derivatisation for the subsequent separation by high-performance liquid chromatography are discussed.Isocratic separations are carried out on LiChrosorb Si100 10 pm using chloroform with 0.1 and 0.2% V / V ethanol as the eluent. All eight individual glucose ethers present in the hydrolysates including the positional isomers of alkyl- ancl dialkylglucoses can be determined quantitatively from the chroma-tograms. Thus the procedure provides data on the distribution of substituents according to their number and position in the anhydroglucose units. Keywords A lkylcellulose ; substituent distribution ; position of substituents ; @re-column derivatisation ; high-performance liquid chromatography In Part I of this series results of the use of quantitative thin-layer chromatography (TLC) for the characterisation of the substitution in cellulose ethers have been rep0rted.l The proced-ure was shown to provide detailed information on the proportions of tri- di- mono- and unsubstituted anhydro-glucose units in the polymer.For a better understanding of the colloidal chemical properties of alkylcelluloses it is desirable to obtain additional information concerning the positions of the alkoxyl substituents in the monomeric unit. The problem of characterising quantitatively this particular distribution has not yet been solved. The time-consuming procedure of fractionation of a methylcellulose h ydrolysate by classical liquid chromatography as reported by Croon and Manley in 1963,2 does not represent a comprehensive analytical procedure. Although with the development of high-performance liquid chromatography (HPLC) in the last decade many workers have achieved fast separa-tions of mixtures of sugars and/or their deri~atives,~-ll there is no evidence in recent literature of a more efficient solution to this separation problem.The aim of this work was to determine by HPLC all of the components including the positional isomers in alkylcellulose hydrolysates thus providing an analytical procedure for the quantitative characterisation of the substituent distribution according to the positions in the glucose ring. * For details of Part I of this series see reference list p. 602 598 SACHSE et al. SUBSTITUTION Experimental Samples Analyst Vol. 108 The cellulose ethers investigated were technical-grade samples of methylcellulose and ethylcellulose.Reagents grade (VEB Laborchemie Apolda G.D.R.). chloroform was passed through a column of activated silica gel prior to use in HPLC. amount of residual ethanol in the purified solvent was determined by gas chromatography. from Koch-Light Laboratories Ltd. Colnbrook Buckinghamshire. Glucose chloroform benzoyl chloride and sodium hydroxide were all of analytical-reagent For the removal of polar components the The 3-O-Methyl-a-~-glucopyranose and 2,3-di-0-methyl-a-n-glucopyranose were purchased Hydrolysis For total hydrolysis of 100-mg amounts the treatment in 7 ml of 2% V/V sulphuric acid at 140 "C for 1 h in a sealed glass ampoule was sufficient. Hydrolytic cleavage of the samples has been described in detail previous1y.l Derivatisation For perbenzoylation 100 mg of the dry cellulose ether hydrolysate are reacted with 1 ml of benzoyl chloride in 4 ml of 15% m/m sodium hydroxide solution in a 20-ml reaction vial.While the mixture is shaken vigorously a perceptible temperature rise indicates the beginning of the reaction. Subsequently the system is cooled down in ice - water and again shaken until self-heating is detected. These two steps are repeated several times over 10 min until no more heat is liberated. To be certain that the reaction has run to completion the mixture can be stored at 0-5 "C overnight. The benzoates formed appear as a very viscous liquid or semi-solid particles. They are removed from the aqueous phase by extraction with 10ml of chloroform. After evaporating off the solvent at 60 "C a 0.5% m/V solution of the reaction products in the eluent is prepared 5-10 pl of which are injected on to the column.High-performance Liquid Chromatography Most of the separations were carried out with an apparatus consisting of a micro-dosing pump (MC 706L Mikrotechna Prague Czechoslovakia) a damping unit laboratory-built 5-and 10-p1 sample loops and a variable wavelength ultraviolet detector (VSU-2P spectrophoto-meter VEB Carl Zeiss Jena G.D.R.). Quantitative evaluation was accomplished using the commercially available chromatographic data system AS 4 (VEB Chromatron Berlin G.D.R.). For part of the work a liquid chromatograph Model 1084B (Hewlett-Packard Avondale, PA USA) was used. The columns (250 x 4 mm) were made in the laboratory from stainless-steel tubing with a low dead volume design of the inlet and outlet parts.LiChrosorb 5100 10 pm (E. Merck, Darmstadt F.R.G.) served as the stationary phase. The columns were packed using a slurry of the sorbent in carbon tetrachloride - dioxane (1 + 1 V / V ) that was pumped with heptane into the column. Packing quality was checked via the parameter A as proposed by KnoxI2 and other workers. A values of between 1 and 1.7 confirmed that good packing had been achieved by this technique. Results and Discussion Derivatisation After total hydrolysis of the cellulose ether a mixture of glucose 2- 3- and 6-O-alkylglucoses, 2,3- 2,6- and 3,6-0-dialkylglucoses and 2,3,6-0-trialkylglucose remains to be separated. Possible anomerisation during the hydrolytic process and/or subsequent reactions is not relevant in the present investigations.To solve this separation problem it appeared useful to consider pre-column ultraviolet derivatisation; there were two main reasons for this. Firstly it allows the ultraviolet detector to be employed which would be specific and more sensitive towards these substances after th May 1983 IN CELLULOSE ETHERS. PART I1 599 introduction of chromophores and generally is easier to handle in comparison with the refractive index detector normally used for underivatised carbohydrates. Secondly if all free hydroxyl groups are converted as a result of derivatisation so that the polarities of the sample molecules are decreased the alkoxyl groups would in this instance give relatively larger contributions to the over-all retention on silica gel than they would be able to provide in the non-derivatised molecules where the interaction of hydroxyl groups with the sorbent would be dominant.From this point of view the introduction of less polar groups into the glucose ether molecules can be expected to facilitate a separation of positional isomers at appropriate chromatographic conditions. Furthermore this step leads to better solubility in hydro-phobic solvents which opens up a wider range of possibilities in the choice of the mobile phase. Several known procedures for the ultraviolet derivatisation of monosaccharides have been tested.lo~l1~l3 Results have shown that perbenzoylation does fulfil all the requirements demanded for a pre-column derivatisation reaction in HPLC if the conditions are chosen according to the classical Schotten - Baumann acylation.Esterification of the glucoses is then easily carried out giving the highest yields of all the reactions tested (80% of the theoretical value) and only very small amounts of by-products are formed. Moreover the performance of the derivatisation reaction in an aqueous medium offers several advantages over benzoyla-tion in pyridine,ll as the mixture need not be heated and the final products are simply isolated by extraction. Chloroform has been used as the extracting solvent in accordance with the choice of the eluent described below. 3 Time/min Fig. 1. Chromatogram of a perbenzoylated test mixture 1 ethyl benzo-ate; 2 methyl benzoate; 3 and 4 glucose; 5 , 3-0-methylglucose ; and 6,2,3-di-O-methylglucose.Column 250 x 4 mm, LiChrosorb Si100 10 pm; eluent chloroform con-taining 0.2% V / V of ethanol; flow-rate 1.0 ml min-l; and detection a t 278 nm. The formation of a single derivatisation product of each sample component is illustrated by the chromatogram of a perbenzoylated test mixture in Fig. 1. lH nuclear magnetic resonance spectra of the benzoylation product of glucose supported this and confirmed that all five hydroxyl groups had been converted. The double peak for glucose in Fig. 1 is caused by a partial separation of Q- and P-anomers. Small amounts of ethyl and methyl benzoates can appear as by-products owing to the presence of trace amounts of ethanol in the chloroform use 600 SACHSE et al. SUBSTITUTION Analyst Vol.108 for extraction and methanol in the hydrolysate. However these substances are almost un-retained under the chromatographic conditions used here and therefore do not interfere with the determination of the components of interest. Separation of the Hydrolysates Preliminary TLC tests on silica gel suggested that chloroform should be a suitable eluent for the separation of the glucose - ether benzoates mixture. Further investigations have shown that isocratic separations on LiChrosorb Silo0 are possible if chloroform with defined small amounts of ethanol is used as the mobile phase. The selectivity of the chromatographic system depends heavily on the percentage of alcohol present. Thus 0.2% V/V of ethanol in chloroform was found to be optimum for the separation of the eight species of the methyl-cellulose hydrolysate (Fig.2) whereas for ethylcellulose 0.1 yo V/V was appropriate (Fig. 3). The alcohol’s function in the system is that of a typical moderator as described for example by Engelhardt.14 Special attention must be paid to maintain the exact eluent composition, as a 0.1% change of the ethanol concentration may cause shifts of capacity factors amounting to several units. Another necessary condition for good resolution of all eight sample components under the selected chromatographic conditions involves column performance. I t has been established that the critical pair of peaks (peaks 7 and 8 in Fig. 2) will be resolved sufficiently (&> l) if at least 3 500 theoretical plates are generated for 2,3,6-0-trimethylglucose (capacity factor 9).The complete assignment of the peaks in Fig. 2 is complicated by the fact that for most of the methyl- and dimethylglucose isomers no reference substances are available. In order to establish the number of methoxyl groups present in the various components a preparative TLC separation has been carried out (for the conditions see reference 1). The isolated fractions of glucose methyl- dimethyl- and trimethylglucoses have been perbenzoylated in the described manner and finally injected on to the column. As expected glucose (peak 1) and trimethylglucose (peak 7) showed only one peak while the methyl- and dimethylglucose fractions were both separated into three components (peaks 2 3 and 5 and 4 6 and 8 respec-ively).For the assignment of those peaks representing positional isomers we used the relative velocity constants of methoxylation of the various hydroxyl groups. Knowing that position 2 is the most reactive followed by 6 and 3,l5$l6 it should be justified to ascribe the largest of the three methylglucose peaks to 2-O-methylglucose the second largest to 6-0-methylglucose and the smallest to 3-0-methylglucose. The assignment of the latter has been confirmed by running the reference substance. The identity of the dimethylglucose peaks has been estab-lished on the same basis again supported by one reference substance (peak 4). 1 I Time/min Fig. 2. Separation of a perbenzoylated hydrolysate of inethylcellulose 1 glucose (two anomers) ; 2 3-0-methylglucose ; 3 2-0-methyl-glucose ; 4 2,3-di-O-methylglucose ; 5 6-0-methylglucose; 6 2,6-cli-O-niethylglucose ; 7, 2,3,6-tri-O-niethylglucose ; arid 8 3,6-di-0-methylglucose.Conditions as in Fig. 1 May 1983 IN CELLULOSE ETHERS. PART I1 601 The analysis of the non-derivatised samples on phase systems usually employed for carbo-hydrate separations (LiChrosorb NH, E. Merck F.R.G. ; acetonitrile - water mixtures) was attempted but no separation of the positional isomers could be achieved. Moreover it be-came obvious t hat pre-chromatographic derivatisation according to our procedure resulted in a decrease in the detection limits by 2-3 orders of magnitude. Thus for instance a few nano-grams of glucose were still detected in the ultraviolet region after perbenzoylation. I 23 0 5 10 15 20 25 Time/min Fig.3. Separation of a perbenzoylated hydrolysate of ethylcellulose 2 glucose; 1,3 and 5 ethylglucoses ; 4,6 and 8 diethylglucoses ; and 7 triethylglucose. Eluent chloroform containing 0.1 % V / I' of ethanol ; other condi-tions as in Fig. 1. At 7 min the sensitivity of the plotter was increased. Quantitative Determination For the quantitative evaluation of the chromatograms the different ultraviolet absorbances, which must be expected because of the different numbers of benzoyl groups attached to the various species had to be taken into account. The relative responses given in Table I have been determined by chromatographing derivatised test mixtures containing defined amounts of glucose 3-O-methylglucose and 2,3-0-di-methylglucose. As a result a linear dependence of the logarithm of relative responses on the number of benzoyl groups has been established.It has been assumed that the ultraviolet response is identical for positional isomers and depends ultimately on the number of benzoyl groups attached to the molecule. As for 2,3,6-0-tri-methylglucose no reference substance was available the relative response of this component has been established by extrapolation of the above mentioned dependence. In Table I1 the results of a series of determinations with a methylcellulose sample are presented. The values obtained give a detailed picture of the alkoxyl groups' distribution and TABLE I EXPERIMENTAL VALUES OF RELATIVE ULTRAVIOLET RESPONSES FOR PERBENZOYLATED COMPONENTS Component Relative response at 278 nm Glucose .. 1 Methylglucose . . 0.63 Dimethylglucose . . 0.37 Trimethylglucose . . 0.23* * Extrapolated value 602 SACHSE METZNER AND WELSCH confirm the reproducibility of our procedure as a whole as well as that of all the steps involved, i.e. hydrolysis derivatisation and HPLC. The average degree of substitution DS is calcu-lated according to the equation DS = c2 + c3 + c5 + qc + cg + c,) + 3c7 100 where c2 . . . c8 are the molar percentages of the components numbered as in Fig. 2. The calculated DS of the methylcellulose sample is in good agreement with the value of 1.86 obtained by quantitative TLC. It is again higher than the corresponding Zeisel value of 1.76. Possible reasons for this difference have been discussed e1~ewhere.l.~~ For ethylcellulose analogous information has been obtained by this procedure.TABLE I1 QUANTITATIVE CHARACTERISATION OF THE SUBSTITUENT DISTRIBUTION I N A METHYLCELLULOSE SAMPLE Component Glucose unsubstituted . . . . . . Methylglucoses-2-O-Methylglucose . . . . 3-O-Methylglucose . . . . 6-O-Methylglucose . . . . 2,3-O-Dimethylglucose . . 2,6-O-Dimethylglucose . . . . . . . . 3,6-O-Dimethylglucose . . 2,3,6-0-Trimethylglucose . . Dimethylglucoses-Trimethylglucose-Mass proportion, 10.3 % mlm 11.6 1.5 6.8 9.3 19.2 13.5 27.8 Standard deviation (n = 9) yo mlm 0.7 0.8 0.2 0.8 0.7 0.5 1.4 0.9 Degree of substitution . . 1.82 0.02 Conclusions It has been shown that HPLC offers new possibilities for a very detailed characterisation of the substitution in alkylcelluloses.The reproducibility of the whole procedure described in this paper has been confirmed. It is possible to determine quantitatively the proportions of monomeric units according to number and position of the ether substituent. The proposed procedure will serve as a valuable tool for obtaining more detailed information about the chemical structure of cellulose ethers. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. References Sachse K. Metzner K. and Welsch T. AnaZyst 1982 107 53. Croon B. I. and Manley R. S. J. in Whistler R. L. Editor “Methods in Carbohydrate Chemistry, Palmer J. K. Appl. Polym. Symp. 1975 28 237. Jones A. D. Burns I. W. Sellings S. G. and Cox J . A. J. Chromatogr. 1977 144 169. McGuinnis G. D. and Fang P. J. Chromatogr. 1978 153 107. Ugrinovits M. Chromatographia 1980 13 386. Yang M. T. Milligan L. P. and Mathison G. W. J. Chromatogr. 1981 209 316. Orth P. and Engelhardt H. Chromatographia 1982 15 91. Sinner M. J. Chromatogr. 1976 121 122. Nachtmann F. Fresenius 2. Anal. Chem. 1976 282 209. Lehrfeld J. J. Chromatogr. 1976 120 141. Knox J. H J. Chromatogr. Sci. 1977 15 352. Meyer H. “Analyse und Konstitutionsermittlung Organischer Verbindungcn,” Springer Vienna, Engelhardt H. J. Chromatogr. Sci. 1977 15 380. Croon I. Sven. Papperstidn. 1960 63 247. Rogowin S. and Deriwitzkaja W. Faserforsch. Textiltech. 1957 8 61. Sachse K. Dissertation Leipzig 1982. Volume 111 Cellulose,” Academic Press New York and London 1963 p. 281. 1938 p. 587. NOTE-Reference 1 is to Part I of this series. Received November 15th 1982 Accepted November 24th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800597

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst, May, 1983, Vol. 108, p p . 603-607 603 Development and Evaluation of a Radioimmunoassay for the Detection of Amphetamine and Related Compounds in Biological Fluids Peter Alan Mason, Tara Singh Bal, Brian Law and Anthony Christopher Moffat Central Research Establishment, Home Ofice Forensic Science Service, Aldermaston, Reading, Berkshire, RG7 4PN A radioimmunoassay has been developed for the detection of amphetamine and its analogues in blood and urine without any pre-treatment of the samples. It is based on a commercially available antiserum and a [1*5I]iodinatecl derivative of amphetamine. The assay can detect low levels of amphetamine (less than 10 ng ml-l) in very small samples (50 pl) of blood and urine. It is cheap (3 pence per test) , rapid, simple to perform and is specific for compounds closely related to amphetamine.A high, positive correlation was obtained ( r = 0.93) when results of the analyses of urine samples from volunteers who had ingested amphetamine were compared with those produced by gas chromatography - mass spectrometry. The assay has proved very useful for the detection of amphetamine and closely related compounds in biological fluids. Keywords : A nzphetainine detection ; blood ; urine ; radioinzrnunoassay Inimunoassays are now accepted as a very efficient means of screening samples of biological materials for the presence of drugs. This is particularly so in forensic science where only a small percentage of the samples tested are likely to be positive for any given drug. The reasons for the popularity of these types of assays are that they are relatively cheap, rapid and simple to perform.Immunoassays generally lack complete specificity, and although this is a draw- back in some applications, it can be a distinct advantage in screening for the presence of groups of drugs in biological samples. Amphetamine has a low relative molecular mass (135), which imposes considerable constraints on the development of immunoassays for its detection, Only two such assays are available commercially : an enzyme immunoassay (Emit Amphetamine DAU) marketed by Syva (UK) Ltd. and a radioimmunoassay (RIA) (Amphetamine Abuscreen) marketrd by Roche Diagnostics Ltd. Although both are excellent for the analysis of urine samples,l the Emit assay is not suitable for use with blood samples and the RIA is not sufficiently sensitive to detect the low levels of amphetamine likely to be present during therapy or in some instances of drug abuse.Apart from the above kits, the authors are not aware of a commercial source of anti-amphetamine antiserum. We have therefore developed and evaluated an RIA suitable for analysis of both blood and urine samples submitted for forensic analysis. Experimental Materials and Equipment Dorset. Unless otherwise stated, all chemicals were of AnalaR grade from BDH Chemicals, Poole, Radio immuiz oassa? Phosphate buffer (0.1 AT, pH 7.4) containing 0.2% m/V of bovine gamma-globulin (Cohn Fraction 11, Sigma Chemical Co., Poole, Dorset) and 0.01% m/V of sodium azide was used throughout the assay. Antiserum, obtained from an Emit Amphetamine DAU kit [Syva (UK) Ltd., Maidenhead, Berkshire] was diluted with assay buffer (1 + 99) immediately before use.Crown Copyright.604 MASON et a,?. : RADIOIMMUNOSSAY OF AMPHETAMINE Analyst, VoZ. 108 The radiolabelled amphetamine derivative N-(3-[1251]iodo-4-hydroxyphenyl)isobutyl- amphetamine (1251A), specific activity 3.32 TBq mmol-l (8.15 hlBq pg-l), was prepared by iodination of N-(4-hydroxyphenyl)isobutylamphetamine by the chloramine-T reaction2 as previously described3 and stored in ethanol at 4 "C. I t was diluted with assay buffer (typically 1 + 399) as required, to give approximately 546 Bq (0.067 ng) per 100 pl. Polypropylene microcentrifuge tubes were obtained from Hughes and Hughes Ltd., Romford, Essex . Gamma counting was carried out using an NE1600 gamma-counter (Nuclear Enter- prises Ltd., Beenham, Berkshire), which had an efficiency of 61% for [1291]iodine.The gamma- counter was interfaced to a Commodore, Model 3032, PET computer. Standard solutions of amphetamine were prepared in synthetic urine4 at concentrations of 0, 25, 50, 125, 250, 500, 1000, 5000 and 10000 ng ml-l. Polyethylene glycol (PEG, Mr 6000) was obtained from Sigma Chemical Co. and was used to prepare a 27.5% m/V aqueous solution for use as a separation reagent. Gas chromatography - mass spectrometry - mailti$Ze ion detection Gas chromatography (GC) was carried out using a Perkin-Elmer Sigma 3B gas chromato- graph. A glass column (2.0 m x 2.0 mm i.d.) was packed with 3% OV-17 on Chromosorb W HP (100-120 mesh).Helium was used as carrier gas a t a flow-rate of 25 ml min-l ; the oven temperature was 135 "C (isothermal) and the injection temperature was 260 "C. The gas chromatograph was interfaced to a VG Analytical 70/70 H mass spectrometer via a single-stage jet separator. The mass-spectrometric (MS) conditions were as follows: electron energy 70 eV with filament current 200 mA, source temperature 200 "C and interface temperature 200 "C. Multiple ion detection (MID) measurements were made by magnetic switching under electron- impact conditions to monitor the base peak and second most intense peak for the trifluoroacetyl derivatives of amphetamine (m/z 140 and 118) and methylamphetamine (m/z 154 and 118). A residence time of 200 ms was used for each ion. Methods Radioimm.unoassay Solutions of sample or standard (50 p1) were pipetted into duplicate sets of polypropylene microcentrifuges tubes.The tubes were then capped, shaken and incubated at room temperature for 30 min. Equilibrium was attained after 15 min and was stable for up to 24 h after that time. PEG solution (250 p1) was then added and the tubes were re-capped and vortexed until the contents appeared homogeneous before being centrifuged (3.5 min at 9 000 g). The supernatant was removed by aspiration and the tubes containing precipitates were counted in a gamma-counter. Urine and blood samples were assayed directly. In order to determine the background levels of cross-reactivity, 100 samples of urine and 43 samples of blood were obtained from normal subjects who were not receiving amphetamine type medication, and assayed as described above.The condition of the blood samples ranged from fresh unhaemolysed to haemolysed/putrefied. A study was also carried out to determine whether the assay was affected by the presence of preservative agents. To known blank urine samples (2.5 ml) were added phenylmercury(I1) nitrate and sodium fluoride (50 + 100mg), sodium sulphate and sodium fluoride (300 + 300 mg) and sodium azide (50 mg). Blood samples (1 ml) were dispensed into vials containing sodium fluoride and potassium oxalate (37.5 + 18.7 mg). The concentrations of preservatives in the urine samples were approximately five times greater than those recommended for forensic science use ; the concentrations in blood samples were approximately 2.5 times greater than those recommended.Cross-reactivities of a number of compounds structurally related to amphetamine (Table I) were determined by comparison of their calibration graphs with that of amphetamine (Fig. 1 ) . To these were added 1251A and antiserum (100 pl of each). GC - MS - MID Samples and standard solutions were extracted and processed according to a previously Some modifications were found to be necessary : standard solutions were described m e t h ~ d . ~May, 1983 AND RELATED COMPOUNDS IN BIOLOGICAL FLUIDS 605 prepared in synthetic urine and methylamphetamine was used as the internal standard at a concentration an order of magnitude greater than that suggested (i.e., 7 320 ng ml-l). Under the GC conditions described above, the trifluoroacetyl (TFA) derivatives of ampheta- mine and methylamphetamine had retention times of 2 min 52 s (A) and 5 min 6 s (B), respec- tively. A calibration graph was prepared by plotting the ratio of the signals of m/z 118 at retention times A and B against the concentrations of amphetamine in the standard solutions (0-5 pg ml-l).Each of the extracts was analysed in duplicate and the mean value was calcu- lated. This gave a correlation coefficient of 0.999 and intercept of 0.013 (n = 7). A similar calibration graph was prepared by plotting the ratio of the signals at m/x 140 and 154 (for amphetamine and methylamphetamine derivatives) against standard concentrations of amphetamine. The amphetamine concentrations in the samples were determined by measuring either of these ratios and comparing them with the corresponding calibration graph.This gave a correlation of 0.9993 and intercept of 0.026 (n = 7). OraL ingestion of amphetamine Four volunteers each took an oral dose of dexamphetamine sulphate (two 5-mg tablets). Venous blood was taken into Monovette syringes (lithium heparin) before and at 2 , 4 and 6 h after ingestion of the drug. Urine samples, obtained before and at 2, 4, 6, 12, 24 and 36 h after ingestion of the drug, were stored at -18 "C. The dosage level was approved by the Medical Ethics Committee, Chemical Defence Establishment, Porton Down. The blood samples were stored at 4 "C before analysis. This experiment was carried out on volunteers from laboratory staff. Results and Discussion An example of a calibration graph obtained with the amphetamine RIA is shown in Fig.1. The assay has a very wide dynamic range that extends to 10000 ng ml-l whilst still maintain- ing sensitivity to less than 25 ng ml-' of amphetamine. This is a very useful feature of an assay for amphetamine because blood and urine samples from a subject who has consumed amphetamine may show enormous differences in concentration, urine levels often being up to 1000 times higher than blood levels. Thus, the wide range of this assay allows analysis of both blood and urine samples with the same calibration graph. \ ' X X \ X 'X \ x\ x\ z 0 "0 102 103 1 04 Concentration of amphetaminehg ml-' Fig. 1. Calibration graph for amphet- amine obtained using the radioimmuno- assay. The cross-reactivities of several compounds closely related to amphetamine are listed in Table I.I t is apparent that the assay is specific for amphetamine, methylamphetamine and a few closely related compounds. In particular it is not very sensitive to /3-phenylethylamine. This is particularly important in an assay designed for forensic use, as /3-phenylethylamine is one of several amines produced during putrefaction. However, even the low level of cross- reaction seen in this assay may result in interference by this compound in severely putrefied606 MASON et al. : RADIOIMMUNOSSAY OF AMPHETAMINE Analyst, VoZ. 108 TABLE I CROSS-REACTIVITIES OF SEVERAL COMPOUNDS CLOSELY RELATED TO AMPHETAMINE Relative reactivity Relative reactivity Compound in RIA* Compound in RIA* Amphetamine . . .... Methylamphetamine . . .. Phentermine . . .. .. Benzphetamine .. .. Tranylcypromine . . . . Ephedrine . . .. .. /3-Phenylethylamine . . . . Fenfluramine . . .. .. Phenylpropanolamine . . .. Mephentermine . . .. 1 1.4 3.7 11 16 27 29 34 57 79 Chlorphentermine . . . . Phenmetrazine . . .. . . Diethylpropion . . .. . . 4-Hydroxyamphetaniine . . Methylphenidate . . . . Nialamide . . . . .. Phendimetrazine . . .. Phenelzine . . . . . . Phenylephrine . . . . . . 79 97 > 100 > 100 > 100 > 100 > 100 > 100 > 100 * Determined a t 60% depression of binding (equivalent to 660 ng ml-l of amphetamine). blood samples. It is unlikely that an immunoassay for amphetamine could be developed that did not detect P-phenylethylamine to some extent because they differ only in one methyl group.The distribution of background levels of cross-reactivity in 100 samples of urine from normal subjects was 6.4 & 6.3 (S.D.) ng ml-l. A positive/negative cut-off was set at 30 ng ml-l to ensure a 99% probability of obtaining a true positive result. Similar analysis of 43 blank blood samples showed that the background level of cross reactivity was 1.4 & 1.2 (S.D.) ng ml-l. A positive/negative cut-off was set at 6 ng ml-l to ensure a 99% probability of obtaining a true positive result. Of the preservatives tested, only the combination of phenylmercury(I1) nitrate and sodium fluoride in urine gave an elevation of background levels (equivalent to 100 ng ml-l of amphet- amine). It is therefore recommended that neither of these preservatives is used in urine samples.Background levels in the blood samples were not affected by the presence of the preservatives tested. Coefficients of variation for urine samples containing 300 and 2 500 ng ml-l of amphetamine were 12.2 and 10.Oyo intra-assay (n = 20), respectively, and 15.1 and 12.2% inter-assay (n = 12), respectively. The results of RIA analysis of blood samples from volunteers who had ingested a single dose of dexamphetamine sulphate are shown in Table 11. Amphetamine blood levels reached a plateau between 2 and 4 h after ingestion of the drug, at a mean concentration of 12.5 ng ml-l. Results of analysis of urine samples from the same experiment are shown in Table 111. In marked contrast to the blood concentrations, amphetamine concentrations in the urine samples do not follow a uniform pattern.This has been shown to be due to changes in urinary pH causing fluctuations in the rate of excretion of am~hetamine.~,' TABLE I1 BLOOD AMPHETAMINE CONCENTRATIONS, MEASURED BY RIA, OF FOUR VOLUNTEERS FOLLOWING AN ORAL DOSE OF 10 mg OF DEXAMPHETAMINE SULPHATE Time after oral dose/h Mean amphetamine blood concentration f standard deviationlng ml-l 0 2.3 f 1.6 2 10.4 f 2.1 4 12.5 f 1.0 6 12.3 f 0.6 Twelve of the urine samples were also analysed by GC - MS - MID for comparison with the results obtained by RIA. The agreement between the two methods was good (correlation coefficient = 0.93, intercept 83.4, slope 2.36, n = 12). However, from this study it is apparent that RIA over estimates the value obtained by the more specific technique by a factor of approximately 2.4.This is not surprising as it is likely that the RIA detects metabolites of amphetamine that do not interfere with GC - MS - MID analysis.May, 1983 AND RELATED COMPOUNDS IN BIOLOGICAL FLUIDS 607 TABLE I11 URINARY AMPHETAMINE CONCENTRATIONS, MEASURED BY RIA, OF FOUR VOLUNTEERS FOLLOWING AN ORAL DOSE OF 10 mg OF DEXAMPHETAMINE SULPHATE Time after oral dose/h 2 4 6 12 24 36 Amphetamine concentration/pg ml-I kubject 1 Subject 2 Subject 3 Subject 4 0.9 0.2 0.3 0.1 2.2 1.5 0.6 1.1 1.1 3.1 1.0 1.0 0.8 0.9 1.1 0.2 3.1 0.5 2.3 2.1 0.2 0.4 0.9 0.4 In conclusion, the amphetamine RIA described is suitable for screening biological fluids for It is rapid, simple to use and is In terms of reagent expense, each test costs approximately 3 pence. the presence of amphetamine and closely related compounds. very cheap. We thank Dr. R. Gleadle (Chemical Defence Establishment, Porton Down) for his help in the collection of blood samples. References 1. 2. 3. 4. 5. 6. 7. Bost, R. O., Sutheimer, C. A,, and Sunshine, I., Clin. Chem., 1976, 22, 789. Hunter, W. M., and Greenwood, F. C., Nature (London), 1962, 194, 495. Mason, P. A., Law, B., and Moffat, A. C., J . Immunoassay, accepted for publication. Rodgers, R., Crowl, C. P., Eimstad, W. M., Hu, M. W., Kam, J. K., Ronald, R. C., Rowley, G. L., and Ullman, E. F., Clin. Chem., 1978, 24, 95. Foltz, R. L., Fertimen, A. F., Jr., and Foltz, R. B., “GC/MS Assays for Abused Drugs in Body Fluids,” National Institute on Drug Abuse, Research Monograph 32, Washington DC., 1980, p . 150. Beckett, A. H., and Rowland, M., J . Pharm. Pharmacol., 1965, 17, 59. Beckett, A. H., Salmon, J . A., and Mitchard, M., J . Pharm. Pharmacol., 1969, 21, 251. Received November 17th, 1982 Accepted December 13th, 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800603

出版商:RSC    

年代:1983

数据来源: RSC