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. Dahrnen (The Netherlands)L. de Galan (The Netherlands)"G. J. Dickes (Bristol, U.K.)A. C. Docherty (Billingham, U.K.)D. Dyrssen (Sweden)*L. C. Ebdon (Plymouth)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)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.)"Members of the Board serving on the Analytical Editorial BoardEditor: P. C. WestonAssistant Editor: Ms. D. ChevinSenior Assistant Editors: Mrs. J. Brew, R. A. YoungREGIONAL ADVISORY EDITORSDr. J. Aggett, Department of Chemistry, University of Auckland, Private Bag, Auckland, NEW ZEALAND.Professor L. Gierst, Universith 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. I. Rube&ka, Geological Survey of Czechoslovakia, Malostranske 19, 118 21 Prague 1, CZECHO-Professor J. RGiiGka, 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, W1V OBN. Telephone 01 -734 9864. Telex No. 268001The Analyst (ISSN 0003-2654) is published monthly by The Royal Society of Chemistry, BurlingtonHouse, London WIV 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 f 93.50, Rest of World f 99.00. USA $201 .OO.Purchased withAnalytical Abstracts U K f 226.50, Rest of World f 238.50, USA $487.00. Purchased with AnalyticalAbstracts plus Analytical Proceedings UK f251 .OO, Rest of World f265.00, USA $539.00. Purchasedwith Analytical Proceedings UK f 11 7.50, Rest of World f 124.50, USA $253.00. Air freight and mailingin the USA by Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11 003.USA Postmaster: Send address changes to: The Analyst, Publications Expediting Inc., 200 MeachamAvenue, 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.0 The Royal Society of Chemistry, 1983. All rights reserved. No part of this publication may be reproduced,stored in a retrieval system, or transmitted in any form, or by any means, electronic, mechanical,photographic recording, or otherwise, without the prior permission of the publishers.Bruxelles, BELGIUM.bosch 7700, SOUTH AFRICA.Establishment, 21 020 lspra (Varese), ITALY.SLOVAKIA.DENMARK.Kensington, N.S.W. 2033, AUSTRALIA.U.S.A

ISSN:0003-2654    

DOI:10.1039/AN98308FX033

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

ANALAO 1 08 (1 290) 1033-1 I 62 (I 983) September 1983THE ANALYSTTHE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRYCONTENTSI033 Inductively Coupled Plasma Source Mass Spectrometry Using Continuum Flow Ion Extraction-Alan L. Gray and Alan R. Date1051 In Sifu Gaseous Pre-treatment of Liver Extracts in a Modified Carbon Rod Atomiser During theDetermination of Eadmium and Lead-John W. Steiner and Harvey L. Kramer1060 Evaluation of Selenium Determination in Biological Material by Atomic-absorption Spectroscopy-Rodney J. Mailer and James E. Pratley1067 Relative Fluorescence Quantum Yields Using a Computer-controlled Luminescence Spectro-meter-Alun T. Rhys Williams, Stephen A. Winfield and James N. Miller1072 Studies of Calcium Ion-selective Electrodes in the Presence of Anionic Surfactants-Anthony J.Frend, Gwilym J.Moody, J. D. R. Thomas and Brian J. Birch1082 Voltammetric Study of the Mercury Dissolution Reaction Mechanism at Solid Electrodes-PaulKiekens, Marc Mertens, Charles Laire and Edward Temmerman1086 Automatic Titration by Stepwise Addition of Equal Volumes of Titrant. Part VIM. Determinationof Alkalinity and Total Carbonate in Sea Water-Axel Johansson, Sten Johansson and Gunnar Gran109’1 Determination of Dimetridazole Residues in Poultry Tissues by High-performance Liquid Chromato-graphy-Anthony Hobson-Frohock and Jayne A. Reader1096 Determination of Water i n Ethanol and in Moist Air-lrshad M. Pirzada and John H. Hills1102 Reversed-phase Paper Chromatographic Studies of Some Rare Earth Elements-C.G. Yeole and1108 Solvent Extraction of Rare Earth Metals with Crown Ethers-Lin-Mei Tsay, Jeng-Shang Shih and1114 Simultaneous Determination of Choline and Betaine in Some Fish Materials-Sana E. Valdes Martinez11 20 Spectrophotometric and Analogue Derivative Spectrophotometric Determination of Micro-amounts of Iron with 2,2’-Dipyridyl-2-benzothiazolylhydrazone-Raj Bhushan Singh, TsugikatsuOdashima and Hajime lshiiSpectrophotometric Study and Analytical Applications of the Complexes of Copper(l1) and Zinc(l1)w i t h Some Sulphonated Azo Dyes-Maria Pesavento and Teresa SoldiSubstituted N-Hydroxy-NN‘-diarylbenzamidines as Selective Extractants for the Spectrophoto-metric Determination of Vanadium(V) in the Presence of Acetic Acid or an Azide-Miss AbhaRani Jha and Rajendra Kumar MishraV.M. ShindeShaw-Chii Wu11281135SHORT PAPERS1141 Spectrophotometric Determination of Phosphate in Polluted Waters by Solvent Extraction ofMolybdenum Blue-Miss Abha Chaube and V. K. Gupta1144 Spectrophotometric Determination of Some Pyrazolidinedione Derivatives in PharmaceuticalPreparations-Michael E. El-Kommos1148 Selective Complexometric Method for Palladium Determination in Alloys Using Thiocyanate asReleasing Agent-Krishna Narayan Raoot, Sarala Raoot and V. Lalita KumariSpectrophotometric Determination of Rhodium( 111) in Thermocouple Wires Using Thiocyanate andRhodamine 6G-Sambamoortny Jaya, Talasila Prasada Rao and Tiruvesaloor Venkatarama Rama-krishnaSolvent Extraction Studies of Thorium(1V) w i t h Aliquat 336 as an Extractant-Miss Mangal R.CO M M U N I CAT1 0 N SSimple Modification t o a Commercially Available Atomic Vnpour Accessory t o Reduce MemoryEffects when Determining High Levels of Arsenic in Geological Materials-Charles H. BranchSimple and Rapid Procedure for the Determination of Lead in Whole Blood by Use o f a Slotted Tubeand Discrete Nebulisation Flame Atomic-absorption Spectrometry-Andrew Taylor and Alistair A.Brown11511255Shivade and Vijay M. Shinde115811591162 BOOK REVIEWSSummaries of Papers in this Issue-Pages iv, v, vi, vii, viii, ix, x, xi, xiiPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class at New York, USA, Post Offic

ISSN:0003-2654    

DOI:10.1039/AN98308BX035

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

iv SUMMARIES OF PAPERS IN THIS ISSUE September, 1983Summaries of Papers in this IssueInductively Coupled Plasma Source Mass SpectrometryUsing Continuum Flow Ion ExtractionAn inorganic mass spectrometry system is described that uses an atmosphericpressure inductively coupled plasma (ICP) as an ion source. Solutionsamples may be introduced directly by nebulisation and the analysis time isabout 1 min. Ions are extracted from the bulk plasma without any inter-vening boundary layer so that the full advantages of the ICP as a high-temperature dissociation and ionisation medium are realised.A quadrupole mass analyser is used with a pulse counting ion detectorfollowed by a multi-channel scaler data system. General background levelsare low and a wide range of elements may be determined with detectionlimits below 1 ng ml-1.Spectra are very simple with few molecular analyte ions and only singlyand doubly charged species are found.Mass resolution is adequate to avoidpeak overlap and isotope ratio determinations may readily be made withprecision below 0.5%, with integration times of about 5 min.The operating characteristics and performance of the system are describedand illustrated and the future development potential is discussed.Keywords : Inductivelj) coupled plasma mass spectrometry ; continuum $owion extractionALAN L. GRAYDepartment of Chemistry, University of Surrey, Guildford, Surrey, GU2 5XH.and ALAN R. DATEInstitute of Geological Sciences, 64-78, Gray’s Inn Road, London, WClX 8NG.Analyst, 1983, 108, 1033-1050.In Situ Gaseous Pre-treatment of Liver Extracts in a ModifiedCarbon Rod Atomiser During the Determination of Cadmium andLeadAn in situ gaseous pre-treatment of bovine liver extracts was performed in amodified carbon rod atomiser.Xitrogen, hydrogen, oxygen and methanewere introduced into the atomiser to eliminate interference and to effectthe continuous re-coating of the atomiser with pyrolytic graphite. An opticaldevice improved the precision of the injection of the sample. The effective-ness of the in situ sample pre-treatment was demonstrated by the determi-nation of cadmium and lead in certified samples.Keywovds : Cadw&m and lead deteviizination ; lizrev extvacts ; in situ gaseouspre-treatment ; naodi$ed carboiz rod atomisev ; electvotJLevma1 atomic-absorption spectroscopyJOHN W.STEINER and HARVEY L. KRAMERDepartment of Primary Industries, Animal Research Institute, 665 Fairfield Road,Yeerongpilly, Brisbane, 4105, Australia.Analyst, 1983, 108, 1051-1059September, 1983 SUMMARIES OF PAPERS I N THIS ISSUEEvaluation of Selenium Determination in Biological Materialby Atomic-absorption SpectroscopyExisting methods of selenium determination by atomic-absorption spectro-scopy using the hydride vapour generation technique were examined.Of the factors investigated it was found that the method of digestion couldbe a source of considerable error. Stirring time in the vapour generator,acid concentration apd final solution volume were also critical factors affectingthe consistency of results.Sodium tetrahydroborate( 111) pellets were foundto have considerable advantages over a solution of the same chemical owingto their stability and the uniform rate of addition and reaction.Keywords : Selenium determination ; biological material ; atomic-absorptionspectroscopy ; hydride vapour generationRODNEY J. MAILERNSW Department of Agriculture, Agricultural Research Institute, Wagga Wagga,New South Wales 2650. Australia.and JAMES E. PRATLEYSchool of Agriculture, Riverina College of Advanced Education, Wagga Wagga,New South Wales 2650, Australia.Analyst, 1983, 108, 1060-1066.Relative Fluorescence Quantum Yields Using aComputer-controlled Luminescence SpectrometerRelative fluorescence quantum yields are determined using a computer-controlled luminescence spectrometer.The relative absorbances of thestandards and unknowns are measured using the same instrument as for thefluorescence measurements. Relative quantum yields are presented for awide range of compounds at room temperature.Keywords : Relative quantum yields determination ; computer-controlledluminescence spectrometerALUN T. RHYS WILLIAMS and STEPHEN A. WINFIELDPerkin-Elmer Ltd., Beaconsfield, Buckinghamshire, HP9 1QA.and JAMES N. MILLERDepartment of Chemistry, Loughborough University of Technology, Loughborough,Leicestershire, LEll 3TU.VAnalyst, 1983, 108, 1067-1071vi SUMMARIES O F PAPERS I N THIS ISSUEStudies of Calcium Ion-selective Electrodes in the Presence ofAnionic SurfactantsSeptember, 1983Eighteen membrane systems, based on a calcium bis(di[4-( 1,1,3,3-tetra-methylbutyl)phenyl]phosphate} ion sensor have been compared in potentio-metric studies with a membrane based on the sensor with a dioctyl phenyl-phosphonate solvent mediator with respect to interferences of calciumion-selective electrodes by anionic surfactants, especially by sodium dodecyl-sulphate (SDS) and sodium tetradecylbenzenesulphonates (ABS) .Electrodesmade from poly(viny1 chloride) matrix membranes of the sensor with trioctylphosphate solvent mediator are far superior to the other membranesystems in resisting interference by anionic surfactants and are shown toyield calcium ion levels in the presence of a wash liquor to match the expectedvalues.Some improvement over dioctyl phenylphosphonate is also offered bydecan-1-01, dodecan-1-01 and tetradecan-1-01 but the use of such solventmediators impairs calcium ion selectivity and poly (vinyl chloride) plasticisingqualities.The use of alternative polymer matrices based on poly(viny1idenechloride) and VAGH copolymer (hydrolysed vinyl chloride - vinyl acetate)offer no advantages.A nineteenth membrane, obtained from a commercial supplier, exhibitedinterference by SDS.An interesting effect is the increase (rather than the normal decrease) ine.m.f. observed for those electrodes from membranes in which the amount offree active sensor was reduced to low levels. By using optimum levels ofsensor i t was possible to fabricate electrodes exhibiting a zero e.m.f.changewhen changing the background 1 0 - 2 ~ calcium chloride to one that is also1 0 - 3 ~ in SDS. Such membranes are based on decyl phosphate grafted toVAGH copolymer.X-ray fluorescence and chromatographic studies on membranes of thesensor with dioctyl phenylphosphonate in poly (vinyl chloride) show SDS tobe a significant agent in leaching membrane components, especially ofdioctyl phenylphosphonate. Such observations are indicative that theinterference of calcium ion-selective electrodes by anionic surfactants maybe the result of the different solubilities of calcium - surfactant complexes inthe solvent mediator of the membrane.Keywords : Calcium ion-selective electrodes ; anionic surfactant interferenceANTHONY J.FREND, GWILYM J. MOODY and J. D. R. THOMASApplied Chemistry Department, UWIST, Cardiff, CF1 3XA.and BRIAN J. BIRCHUnilever Research Laboratory, Bebington, Wirral, Merseyside, L62 4XN.Analyst, 1983, 108, 1072-1081September, 1983 SUMMARIES OF PAPERS I N THIS ISSUEVoltammetric Study of the Mercury Dissolution ReactionMechanism at Solid ElectrodesviiStripping voltammetry with collection at a rotating platinum ring - glassycarbon disc electrode was used for the elucidation of the mechanism of theelectrochemical dissolution of a mercury film from a solid electrode. Incomplexing electrolytes, the dissolution of mercury gives rise to the formationof divalent mercury ions. In non-complexing electrolytes, the initial productof the electrodissolution of metallic mercury is also mercury(I1) , which reactswith as yet unoxidised mercury atoms present on the electrode surface to givemercury(1) ions.Consequently, mercury(1) ions are formed as a result of areproportionation reaction following the electrochemical oxidation of mercuryto mercury(I1).Keywords : Stripping voltammetry ; rotating ring-disc electrode ; mercurydissolution kinetics ; reproportionationPAUL KIEKENS, MARC MERTENS, CHARLES LAIRE and EDWARDTEMMERMANLaboratory for Analytical Chemistry, Ghent University, Krijgslaan 281, S-12 B-9000Ghent, Belgium.Analyst, 1983, 108, 1082-1085.Automatic Titration by Stepwise Addition of Equal Volumes ofTitrant. Part VIII. Determination of Alkalinity and TotalCarbonate in Sea WaterPrevious parts of this series have presented methods for the evaluation oftitration data.This paper shows the utility of some of these methodsand extensions of them for solving a practical problem, i.e., the determina-tion of alkalinity and total carbonate in sea water. The procedure is basedon the solution of a set of linear equations.The method presented has been tested on theoretical and experimentaltitration data and has given results that compare extremely well with thoseobtained by using non-linear curve-fitting methods.Keywords Sea water analysis ; potentionaetric titration ; linear multipleregression ; alkalinity ; total carbolzateAXEL JOHANSSON, STEN JOHANSSON and GUNNAR GRANDepartment of Analytical Chemistry, The Royal Institute of Technology, S-100 44Stockholm 70, Sweden.Analyst, 1983, 108, 1086-1090.Determination of Dimetridazole Residues in Poultry Tissues byHigh-performance Liquid ChromatographyA high-performance liquid chromatographic method has been developed forthe determination of the anti-histomoniasis drug dimetridazole.Recoveriesof the drug down to 0.01 mg kg-l “spiked” on to poultry tissues are 80% orabove and the method has been applied t o the determination of residues intissues from broiler chickens and turkeys.Keywords : Dimetridazole determination ; poultry tissue ; high-performanceliquid chromatographyANTHONY HOBSON-FROHOCK and JAYNE A. READERARC Food Research Institute, Colney Lane, Norwich, NR4 7UA.Analyst, 1983, 108, 1091-1095... Vlll SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Water in Ethanol and in Moist AirSeptember, 1983Analyses of aqueous alcohols containing as little as 0.01% of water in bothliquid and gaseous phases were performed by gas chromatography on aPorapak Q column using a thermal conductivity detector. The method issimple, quick, reproducible and only a 0.1-p1 volume of the liquid sample isneeded.It has many uses including the determination of alcohol in afermenter as it is formed, and the same column can be used to measure theabsolute humidity (0.001-0.18 or more) of atmospheric air and other gases,using a 1.0-ml sample.Selected examples of experimental results are presented and the methodis compared with other published methods.Keywords : Water determination ; ethanol ; moist air; thermal conductivitydetector ; gas chromatographyIRSHAD M.PIRZADA and JOHN H. HILLSDepartment of Chemical Engineering, University of Nottingham, University Park,Nottingham, NG7 2RD.Analyst, 1983, 108, 1096-1 101.Reversed-phase Paper Chromatographic Studies of SomeRare Earth ElementsReversed-phase paper chromatographic studies of rare earth elements suchas scandium, yttrium, lanthanum, neodymium and cerium have been carriedout on Whatman hTo. 1 filter-paper, impregnated with trioctylamine, triiso-octylamine and Aliquat 336 S as stationary phases and using organic com-plexing agents such as sodium acetate, sodium malonate and sodiumsuccinate solutions as the active mobile phase. Iiesults for the separationof ternary and quaternary mixtures are reported.Keywords : Rare earth element separation ; reversed-phase paper chromato-graphy ; acid eluentsC. G.YEOLE and V. M. SHINDEAnalytical Laboratory, Department of Chemistry, Shivaji University, Kolhapur416 004, India.Analyst, 1983, 108, 1102-1 107.Solvent Extraction of Rare Earth Metals With Crown EthersCrown ethers, such as 15-crown-5, 12-crown-4 and dibenzo- 18-crown-6, wereused to extract individual rare earth ions from aqueous solutions containingpicrate into nitrobenzene solution. The rare earth metal ion europium(I1Ijis extracted as a 2 : 1 crown - ion sandwich complex with 12-crown-4, but asa 1 : 1 complex with both 15-crown-5 and dibenzo-18-crown-6. From studiesof picric acid concentration effects on coinplesation, the extracted species ofEu(NO,), with 15-CrOwii-5 and clibetizo-l8-crowi~-G are Eu[( 15-crown-5)-(picrate),(SO,)] and Eu[(clibeiizo-18-cro~Vii-~j (picrate),(SO,j], respectively,but Eu[( 12-crown-4),(picrate),] is forniecl lvitli l%ClU\Vi1-4. Tile extractionof rare earth ions showed that TW+, I<U3+, Gc13+, Sd3+ and Yb3+ can be easilyextracted using 15-crown-6 ; however, the estraction of Ce4+, Sm3+, Dy3+ andLu3+ is more difficult.Keywords : Rave envth metals ; solvent extvnction ; c ~ o w n ethersLIN-ME1 TSAY, JENG-SHANG SHIH and SHAW-CHI1 WUDepartment of Chemistry, National Taiwan Normal University of Taipei, Taiwan 117,Republic of China.Analyst, 1983, 108, 1108-1 113

ISSN:0003-2654    

DOI:10.1039/AN98308FP089

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Se$tember, 1983 SUMMARIES OF PAPERS I N THIS ISSUESimultaneous Determination of Choline and Betaine inSome Fish MaterialsixThe assay of choline and betaine, as water-insoluble reineckate derivatives,has been adapted to the simultaneous determination of these two quaternarynitrogenous compounds. No preliminary column fractionation is required.The solution resulting from the initial extraction - hydrolysis stage is dividedinto two portions; one portion is adjusted to pH 1.0 and the other to pH 8-9before the addition of ammonium reineckate. Under acidic conditions,choline and betaine reineckate are precipitated and a t an alkaline pH onlycholine reineckate precipitates. The reineckate salts are re-dissolved inacetone to give pink solutions for spectrophotometric measurement a t 526 nm.Choline reineckate has a higher linear absorptivity than betaine reineckate andthis feature is allowed for by reference to the respective calibration graphs;the choline content of the extract is determined directly and the betainecontent by difference.Model systems containing 2-3 mg of pure cholineor betaine can be assayed but sample masses of 3-4 g of biological materialare recommended for the analysis.Keywords : Choline determination ; betaine determination ; ammoniumreineckate ; fish materialsSANA E. VALDES MARTINEZFood Science Division, Department of Bioscience and Biotechnology, University ofStrathclyde, 131 Albion Street, Glasgow, G1 1SD.Analyst, 1983, 108, 11 14-1 119.Spectrophotometric and Analogue Derivative SpectrophotometricDetermination of Micro-amounts of Iron with 2,2’ -Dipyridyl-2- benzothiazolylhydrazoneThe synthesis of 2,2’-dipyridyl-2-benzothiazolylhydrazone (DPBH) , charac-teristics, reactions with various metal ions and its application in the selectivedetermination of iron are presented. DPBH reacts with iron(I1) to form astable 1 : 3 metal to ligand complex having absorption maxima a t 427 and615nm in 0.4% Triton X-100 solution.A solution of the complex gives aconstant absorbance in the pH range 4.5-8.4 a t 427nm and 3.0-9.6 a t615 nm. The molar absorptivity, sensitivity for an absorbance of 0.001 andrelative standard deviation for 10.3 pg of iron (10 replicates) are 3.41 x lo4and 1.23 x 104 1 mol-1 cm-1, 1.64 and 4.54 ng cm-2, and 0.7 and 0.5% a t 427and 615 nm, respectively.Application of the proposed method to the deter-mination of iron in water samples and further sensitisation of the method byemploying analogue derivative spectrophotometry are also described.Keywords : Iron determination ; 2,2’-di~yridyl-2-benzothiazolylhydrazone ;analogue derivative spectrophotometry ; water analysisRAJ BHUSHAN SINGH, TSUGIKATSU ODASHIMA and HAJIME ISHIIChemical Research Institute of Non-Aqueous Solutions, Tohoku University,Katahira, Sendai, 980, Japan.Analyst, 1983, 108, 1120-1 127X SUMMARIES OF PAPERS I N THIS ISSUESpectrophotometric Study and Analytical Applications of theComplexes of Copper(I1) and Zinc(I1) with SomeSulphonated Azo DyesSeptember, 1983The complexing ability of some water-soluble azo dyes towards copper(I1)and zinc(I1) were studied in 0.1 M sodium perchlorate and perchloric acidsolutions, at 25.0 "C.Solutions with CM > CL, where CM is the concentra-tion of metal and CL the concentration of ligand, and CL> CM (CM up toM, CL up to 2 x lo-* M and pH up to the precipitation of metal hydrox-ides) were investigated spectrophotometrically .The ligands T-azo-C and 3,B-disulpho-TAN are able to complex bothcopper(I1) and zinc(II), but amaranth only copper(I1). Besides the usualcomplexes having 1 : 1 metal to ligand molar ratios, a 2: 1 complexhas been detected in the system copper(I1) - T-azo-C. The results arecompared with others obtained previously. The copper (11) complexes aremuch more stable than those of zinc(I1) and those of many other metal ions.Therefore, a method is proposed for the spectrophotometric determinationof copper(II), which is almost free from interferences.Keywords : Copper determination ; copper and zinc complex formation ;sulphonated azo dyes ; spectrophotometryMARIA PESAVENTO and TERESA SOLD1Dipartimento di Chimica Generale, Universit& di Pavia, Viale Taramelli 12, 27 100Pavia, Italy.Analyst, 1983, 108, 1128-1 134.Substituted N- Hydroxy- NW-diarylbenzamidines as SelectiveExtractants for the Spectrophotometric Determination ofVanadium(V) in the Presence of Acetic acid or an AzideThe reactions of 1 1 newly synthesised N-hydroxy-NN'-diarylbenzamidines(HOA) with vanadium(V), in the presence of acetic acid or an azide, havebeen studied spectrophotometrically to determine the effect of substituentson the complexing properties of the ligand.The investigations showed theformation of 1 : 2 : 1 [vanadium(V) : HOA : acetic acid] and 1 : 2 : 2 [vanadium-(V) : HOA : azide] complexes in chloroform. Based on the mixed complexformations, simple, rapid, sensitive and selective methods for the spectro-photometric determination of microgram amounts of vanadium(V) have beendeveloped. The methods have been applied to the determination of thevanadium content of BCS steels.Keywords : Vanadium determination ; spectvopltotometry ; N-hydroxy-NN'-diavylbenzamidines ; steel analysis ; solvent extractionMiss ABHA RANI JHA and RAJENDRA KUMAR MISHRADepartment of Chemistry, Ravishankar University, Raipur 492 010, India.Analyst, 1983, 108, 1135-1 140.Spectrophotometric Determination of Phosphate in PollutedWaters by Solvent Extraction of Molybdenum BlueShort PaperKeywords : Spectvophotometry ; phosphate ; polluted waters ; solvent extraction ;molybdenum blueMiss ABHA CHAUBE and V.K. GUPTADepartment of Chemistry, Ravishankar University, Raipur, 492 010, India.Analyst, 1983, 108, 1141-1 144September, 1983 SUMMARIES OF PAPERS I N THIS ISSUESpectrophotometric Determination of Some PyrazolidinedioneDerivatives in Pharmaceutical PreparationsShort PaperKeywords : Pyrazolidinedione derivative determination ; spectrophotometry ;p-dimethy laminocinnamaldehyde ; pharmaceuticalsMICHAEL E.EL-KOMMOSDepartment of Pharmaceutical Chemistry, Faculty of Pharmacy, University ofAssiut, Assiut, Egypt.Analyst, 1983, 108, 1144-1 147.Selective Complexometric Method for Palladium Determinationin Alloys Using Thiocyanate as Releasing AgentShort PaperKeywords ; Palladium determination ; selective complexometry ; thiocyanaterelease ; metal alloysKRISHNA NARAYAN RAOOT, SARALA RAOOT and V. LALITA KUMARIDefence Metallurgical Research Laboratory, P.O. Kanchanbagh, Hyderabad-500 258,India.Analyst, 1983, 108, 1148-1151.Spectrophotometric Determination of Rhodium( 111) inThermocouple Wires Using Thiocyanate and Rhodamine 6GShort PaperKeywords : Rhodium determination ; spectrophotometry ; thiocyanate ;Rhodamine 6G ; thermocouple wiresSAMBAMOORTHY JAYA, TALASILA PRASADA RAO and TIRUVESA-LOOR VENKATARAMA RAMAKRISHNADepartment of Chemistry, Indian Institute of Technology, Madras-600 036, India.Analyst, 1983, 108, 1151-1 155.Solvent Extraction Studies of Thorium(1V) with Aliquat 336as an ExtractantShort PaperKeywords ; Thoriunz(I V ) solvent extraction; Aliquat 336Miss MANGAL R. SHIVADE and VIJAY M.SHINDEAnalytical Laboratory, Department of Chemistry, Shivaji University, Kolhapur,416 004, India.Analyst, 1983, 108, 1155-1157.xxii SUMMARIES OF PAPERS I N THIS ISSUESimple Modification to a Commercially Available Atomic VapourAccessory to Reduce Memory Effects when Determining High Levelsof Arsenic in Geological MaterialsSeptember, 1983CommunicationKeywords : Arsenic determination ; geological materials ; hydride generationtechnique ; atomic-absorption spectrometryCHARLES H. BRANCHGeochemistry and Petrology Division, Institute of Geological Sciences, Gray’s InnRoad, London, WClX 8NG.Analyst, 1983, 108, 1158-1 159.Simple and Rapid Procedure for the Determination of Lead inWhole Blood by Use of a Slotted Tube and DiscreteNebulisation Flame Atomic- absorption SpectrometryCommunicationKeywords ; Lead determination ; slotted tube ; discrete nebulisation $ameatomic-absorption spectrometry ; whole bloodANDREW TAYLOR and ALISTAIR A. BROWNRobens Institute of Industrial and Environmental Health and Safety, University ofSurrey, Guildford, Surrey, GU2 5XH.Analyst, 1983, 108, 1159-1161

ISSN:0003-2654    

DOI:10.1039/AN98308BP095

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

September 1983 Vol. 108 No. 1290 The Analyst Inductively Coupled Spectrometry Using Plasma Source Mass Continuum Flow Ion Extraction Alan L. Gray* and Alan R. Date Department of Chemistry University of Surrey Guildford Surrey G U2 5XH Institute of Geological Sciences 64-78 Gray's Inn Road London WClX 8NG An inorganic mass spectrometry system is described that uses an atmospheric pressure inductively coupled plasma (ICP) as an ion source. Solution samples may be introduced directly by nebulisation and the analysis time is about 1 min. Ions are extracted from the bulk plasma without any inter-vening boundary layer so that the full advantages of the ICP as a high-temperature dissociation and ionisation medium are realised. A quadrupole mass analyser is used with a pulse counting ion detector followed by a multi-channel scaler data system.General background levels are low and a wide range of elements may be determined with detection limits below 1 ng ml-1. Spectra are very simple with few molecular analyte ions and only singly and doubly charged species are found. Mass resolution is adequate to avoid peak overlap and isotope ratio determinations may readily be made with precision below 0.5% with integration times of about 5 min. The operating characteristics and performance of the system are described and illustrated and the future development potential is discussed. Keywords Inductively coupled plasma mass spectrometry ; continuum pow ion extraction The use of an atmospheric pressure plasma as an ion source for atomic mass spectrometry is an attractive concept not least because of the ease and rapidity of sample introduction and exchange normally a difficult and restricting aspect of inorganic mass analysis.When an inductively coupled plasma (ICP) is used as the plasma it offers the additional possibility of high sensitivity and freedom from matrix and inter-element effects characteristics that have made the ICP such a useful source for atomic-emission spectrometry. It is not surprising, therefore that since the principle of using a plasma in this way was first demon~trated~-~ it has attracted the attention of a number of other worker^.^-^ The earlier publications in this field4,597 all referred to work with small sampling apertures, which in the flowing plasma are covered by an aerodynamic and thermal boundary layer.The resulting sample of the plasma ion population transmitted to the mass analyser was thus drawn from within this layer. The temperature in this region where the ions may reside for several microseconds was considerably lower than that in the bulk plasma. Although very low detection limits were reported the full advantages of the ICP particularly in respect of matrix tolerance and dynamic range were not achieved. One group has recently reported the results of sampling a microwave plasma6 with apertures large enough to induce continuum flow from the plasma which breaks through the inter-mediate boundary layer.8 The performance attained with this plasma however appears to be limited by the low gas temperature which is probably below that of the boundary layer formed over small apertures in an ICP.It would clearly be of great advantage if continuum flow sampling could be achieved from an ICP. Some preliminary steps towards this have also been described by Douglas et aL9 but the results so far published do not yet show all the advantages over the microwave induced plasma (MIP) that should be realised from the use of an ICP. * To whom correspondence should be addressed. 103 1034 Analyst Vol. 108 We have recently published an account of our work on the boundary layer mode of sampling from an ICPlO and this paper reports the outcome of the concurrent work on continuum flow sampling in which all the characteristics expected of the ICP as an ion source have been observed. These results have previously been reported in part on several 0ccasions.~1-1~ A standard commercial argon ICP system and torch are used and solution samples are introduced via a standard pneumatic nebuliser without desolvation.The torch and load coil mounting is modified to provide a horizontal torch position so that access to the torch mouth is simplified. Ions are extracted through an aperture typically of 0.5 mm diameter into the vacuum system and analysed in a quadrupole mass analyser. Data are recorded and displayed in a commercial multi-channel scaler (MCS) data system designed originally for X- and gamma-ray spectrometry. The ion extraction process and the ionisation characteristics of the ICP as a source are discussed and the performance obtained on solutions containing both trace and matrix elements is described.GRAY AND DATE ICP SOURCE MS USING Experimental Instrumentation The major components from which the system was assembled are listed in Table I. The complete system and its operation in the boundary layer mode have been described else-TABLE I EXPERIMENTAL SYSTEM Component Descriptions and supplier Inductively coupled plasma system . . Model ICP 2500 with APCS-1 automatic power supply and AMN-2600E automatic matching network. Output power up to 2 600 W. Frequency crystal controlled 27.12 MHz. Torch Type T1.O. Load coil 3 turns of +-in copper tubing, water cooled. Nebuliser . . . . Fixed geometry cross-flow type 09.790. Jarrel-Ash Co., Aerosol chamber local manufacture. Plasmatherm Inc. Kresson N J USA Waltham MA USA.No desolvation Vacuum system . . . . Vessel local supply. Vacuum pumps : Stage 1. Stage 2. ED 660 mechanical pump. E09 vapour pump, Stage 3. ED 100 mechanical pump. E04 vapour pump, Both vapour pumps fitted with water-cooled baffle valves. All supplied by Edwards High Vacuum Ltd. Crawley Sussex Mass analyser . . Quadupole analyser Type 12-12S mass range 0-800 8.m.u. Resolution >2.5 M. Fitted with Galileo channel electron multiplier ion detector Type 4870. VG Analytical Ltd., Altrincham Cheshire Pulse counting channel . . . . Pulse amplifier Type 9302. Ratemeter Type 449. Scaler timer Type 776. EGG Ortec Oak Ridge TN USA Data system . . . . Canberra Series 80 with TTY and X - Y plotter read-out and cassette tape storage of data.Canberra Instruments Meriden CT USA EDM 20A mechanical pump 342 1 min-'. 1200 1 s-1. 200 1 s-1. Sampling apertures . . . 0.4 and 0.5 mm diameter 0.5 mm long. Drilled direct into Skimmer . . Nickel stainless steel or copper of local manufacture. 1 .O mm Also 1.0 mm nickel skimmers copper nickel or nickel alloy cones aperture 56O external angle. by Beam Dynamics lnc. Minneapolis MN US September 1983 CONTINUUM FLOW ION EXTRACTION 1035 where.1° To sample through the boundary layer by establishing continuum flow at the gas temperatures reported for the ICP16Js requires sampling apertures of at least 0.2 mm diameter. The gas flow admitted by apertures of this diameter is just beyond the capacity of the vapour pump used in the first stage for boundary layer sampling if the pump is to operate efficiently (at a pressure below Torr).The need to increase the capacity of this stage is avoided by preceding it with a small vessel at an intermediate pressure from which the bulk of the entering gas is removed by a mechanical pump at a pressure of about 1 Torr. The operation of a molecular beam forming stage in this pressure range was first described by Camparguel' and later studies were reported by Greene et aL1* of operations between 1 and 100 Torr. A similar stage was used by Douglas and co-workers to sample from both an MIP and ICP.6pQ The pressure ratio across the sampling aperture in such a stage is still well below the critical value at which the flow becomes supersonic. A molecular beam is formed along the axis of the system the core of which is then admitted to the next stage through a skimmer aperture.No electric fields are used in this process. Once into the first vapour pumped stage the sample is focused through the differential aperture into the mass analyser just as in the boundary layer sampling system previously described. A diagram of the plasma sampling and analysis system is shown in Fig. 1 and the details of the expansion stage and sampling interface in Fig. 2. The only remaining differences from the original systemlo lie in the mounting of the plasma torch and sampling aperture on the same horizontal axis as the ion lens and mass spectrometer and the omission of the sample desolvation. 11 'P 6 i 8 Fig. 1. Arrangement of plasma and mass analyser.1 2 3 Successive vacuum stages; 4 torch housing; 5 radiofrequency matching unit; 6 7 coolant and plasma gas inlet; 8 sample (injector) gas flow; 9 ion lens; 10 quadrupole analyser; 11 ion detector; and 12 signal output. Sampling cones are machined from the solid and the aperture drilled direct in the tip. The cones have an internal angle of 120" and fit into a recess in the water cooled front plate of the expansion stage, The ICP torch is the standard type used for atomic-emission spectrometry. Its mouth projects just beyond the wall of the torch box so that the tip of the sampling cone may be positioned up to about 3 mm from the outer end of the load coil. The torch box is mounted on a conventional optical bench that provides motion for alignment in three directions.A fixed-geometry cross-flow pneumatic nebuliser is used for sample introduction. It is mounted immediately below the torch housing and sample mist is introduced direct to the torch through a short length of glass tube. Operating Conditions The plasma is normally operated at 1200 W (reflected power <5 W) with a coolant gas flow-rate of 12lmin-1 a zero plasma gas flow-rate and a sample injector flow-rate of 0.6 1 min-l. Sample uptake at this injector flow-rate is 1.8 ml min-l 1036 GRAY AND DATE ICP SOURCE MS USING m I ’ Analyst Vol. 108 Fig. 2. Ion extraction interface. 1 First (expansion) stage; 2 second vacuum stage; 3 load coil; 4 plasma torch; 6 quartz bonnet; 6 torch housing; 7 sampling cone; 8 water-cooled flange; 9 skimmer; and 10, extraction electrode.At the usual operating position of a 0.5 mm diameter aperture 10 mm from the load coil, the operating gas pressures in the system are as follows stage 1 (expansion stage) 1 Torr; stage 2 2 x lo-* Torr; and stage 3 (mass analyser) 2 x Torr. Sample Solutions For the performance studies described here sample solutions were prepared in the laboratory from analytical-reagent grade standard solutions at 1000 pg ml-l. Solutions were diluted with distilled water to the desired concentration and acidified to 1% with Aristar doubly distilled nitric acid. Operating Procedure In operating the system full advantage is taken of the facilities of the programmable multi-channel scaler. For normal spectrum display memory groups of 1023 channels are used to accumulate and store data from scans of up to 100 a.m.u.width. The mass analyser scan is triggered by the start of the MCS sweep and normally a dwell time per channel of 1 ms is used. The full scan therefore takes just over 1 s and when 60 scans are pre-set a total integrating time of approximately 1 min is obtained. The system may be programmed to make repeated integrations recording data on to tape while doing so for subsequent printout on teletype or graphics plotter for evaluation. Regions of interest may be set around any wanted peaks or background areas and total peak integrals printed out. Blank spectra may be stored before the sample run and then subtracted automatically before printing out peak integrals if desired. Operation in the scanning mode is usually most convenient for analytical work but some performance studies may be more easily carried out in the single ion mode.The mass analyser is set to the peak of the wanted analyte ion by observing the ratemeter response and the multiscaler then set to integrate the ion signal as a function of time by setting a convenient channel dwell time such as 1 s. In this way a plot of the response is obtained direct on the VDU the 1023 channels accommodating a run of about 17 min. This mode of operation is very suitable for determination of detection limits studying response time or memory effects noise levels stability and for other measurements where the response to a single ion is sufficient. Accumulated data may be printed or plotted out in the usual way. When a solution is presented to the nebuliser the response appears in a few seconds.A period of 20-30s is usually allowed for the system to equilibrate when a new solution is introduced before the start of an integration September 1983 CONTINUUM FLOW ION EXTRACTION Results and Discussion Sampling Interface Operation 1037 At the normal operating position of 10 mm from the load coil the gas temperature is reported as 7500 K15916 and this figure gives good agreement with the observed working pressures and pumping capacities. A t this temperature the aperture Knudsen number Kn, is 3 x 10-3, which is well below the value requirea to puncture the boundary layer. When the aperture is placed in the working position in the flame the boundary layer formed over the sides of the cone near the tip may be clearly seen.When water is nebulised this layer appears as a dark space but if a solution of 1000 pg ml-1 of an yttrium salt is introduced the boundary layer is made more visible by the red band emission from yttrium(I1) oxide reformed in the cooler layer. In the centre of the flame the blue emission from yttrium atoms and ions may be seen as they stream through the aperture.lg Within the expansion stage the incoming gas is formed into a molecular beam by the enhancement of the energy along the axis. A terminal Mach number of about 12 in front of the skimmer aperture is calculated from the parameters of this stage.s At a pressure of 1 Torr much higher than usual in molecular beam stages considerable scattering at the fringes of the beam is to be expected and it is not clear whether the normal theory of beam formation may be used at this pressure.However good spectra are produced at aperture -skimmer spacings greater than 5 mm with an optimum between 7 and 10 mm. At lower pressures also the gas temperature in the core of the beam should fall to about 75 K before reaching the skimmers but the skimmer tip operates well above room temperature so again this departure from theory is probably a consequence of the higher pressure. However a beam does appear to be formed as little further signal is gained as the skimmer aperture diameter is increased above 1 mm although up to that point the response is roughly pro-portional to aperture area. However in spite of the formation of a beam the efficiency of such a stage is low.By comparison with boundary layer sampling it is calculated that only 0.4% of the incoming ions pass into the next stage. This compares favourably with the figure of 0.1% reported by Campargue17 for a similar stage. Clean spectra are obtained with a simple structure similar to those obtained in boundary layer sampling and the random background count from all sources over the whole spectrum is between 10 and 20 counts s-1. An axial stop on one lens cylinder is used to obstruct direct photons from the plasma so that in spite of the in line torch and aperture these contribute a negligible amount to the background. The random count is thought to arise mainly from photon emission from decaying excited ions principally of argon within the ion lens. A similar but smaller random background has been observed in boundary layer sampling.The use of apertures large enough to induce continuum flow into a low pressure first stage, of the type used for boundary layer sampling has been found to produce an intense pinch discharge in the mouth of the aperture which contributes a very high photon count to the b a c k g r ~ u n d . ~ ~ ~ It was found in the initial exploration of continuum flow from the ICPll that this pinch discharge was more intense at low pressures behind the aperture and was completely quenched at pressures above about 4 Torr. It was also found to be more intense on the edges of the flame where the gas temperature was lower so that its intensity appears to be related to both gas density and rate of expansion through the aperture.At a working pressure in the expansion stage of about 1 Torr there is no observable pinch discharge in the aperture and no photon contribution to the background is detectable. A residual discharge may still be present as in the intense light from the plasma a weak discharge would not be visible. The temperature of the tip of the sampling cone is a function of the thermal conductivity and thickness of the cone walls. Nickel sampling cones have been dimensioned so that they operate at red heat as it was originally considered that this would minimise sample condensa-tion problems and memory effects. However in the presence of a high concentration of atomic oxygen from the nebulised water in the plasma flame nickel is found to oxidise on both internal and external surfaces where the cone is red hot.This does not interfere with the sampling process as far as can be observed but it does create a rough surface over the tip of the cone which tends to retain condensed material and cause slight memory effects. Nickel oxide is not readily removed without mechanical cleaning so some of the oxidation-resistant nickel alloys are being investigated. These unfortunately have considerably poore 1038 Analyst Vol. 108 thermal conductivity and require thicker sections. Copper cones operating below red heat have been found to work well with very low memory. Sample condensation does occur on the cooler parts of the cone at concentrations above about 10 pg rnl-l in an annulus around the aperture but does not obstruct the aperture itself.The response of the system to a 10pgml-l solution of lead is shown in Fig. 3. After a 4-min run on 208Pb+ the response takes 55 s to decay to of the peak signal after the last of the lead solution mist has reached the plasma. Metals such as lead cadmium and zinc show greater memory effects than the less volatile elements. For example the response to barium after solution ceased to reach the plasma fell to 3.2 x of the peak value in 60 s. Thus successive measure-ments with a 1-min blank run between them would show less than 0.1% of cross-contamina-tion. The lifetime of the present sampling cones is limited by sputtering in the region immediately behind the aperture. This removes metal and eventually weakens the lip of the aperture. Both copper and nickel are metals with high sputter yields so alternative materials of lower yield are being investigated.However the cones are not expensive and may be replaced in a few moments after the life of 30 h or more that is usually obtained. The principal draw-back of the sputtering is the presence of the metal peaks in the spectrum. There is also a small contribution to the spectrum from the skimmer tip although not enough to imply significant skimmer wear. If the same metal is used for both skimmer and cone no additional interference is caused. For this reason pure metals are preferred to alloys for these com-ponents. GRAY AND DATE ICP SOURCE MS USING I- 1 I I 0 1 2 3 4 5 6 Ti m e h i n 1 .o 0.8 U 0.6 d 4 0.4 2 0.2 0 i I I I I 2 4 6 8 1 0 Ion energylev Fig.3. Signal decay on 10 pg ml-l lead solution. Fig. 4. Distribution of ion energy Peak level 3.8 x 1 0 5 counts s-l on aosPb+. Arrow at entrance to quadrupole analyser at shows the 10-3 response point. 1200 w. Ion Optical Considerations The ion optical system and its operating potentials remain very similar to those used for boundary layer sampling one additional cylinder being mounted at the base of the skimmer to operate as an extraction electrode at a potential of about -200 V. In boundary layer sampling the low ion energy spread of about 2 eV appeared to be related to conditions in the boundary layer rather than in the plasma and enabled excellent resolution to be obtained from the mass analyser. Once ions had been drawn from the bulk plasma without the intervening collisional processes in the boundary layer it was expected that a much wider energy distribution would be found.However a retardation plot on cobalt ions at the entrance to the analyser produced the differential energy distribution shown in Fig. 4 which has a half width of about 7.5 eV. This is small enough to enable good mass resolution to be obtained without the use of an energy analyser. Choice of Plasma and Aperture Operating Conditions During the early stages of the development of continuum sampling based on the use of a standard ICP developed for emission analysis little variation on the normal plasma operating conditions has been explored partly because it was desired to build on atomic-emission experience and make as much use of already available data as possible.However withi September 1983 CONTINUUM FLOW ION EXTRACTION 1039 the variables available it was clearly important to determine the range of possible plasma operating powers and to establish the optimum sampling point in the flame. These studies were performed using indium as a convenient analyte. This has a low first ionisation energy (5.79 eV) and a second ionisation energy well above the first of argon so that it can be assumed to be fully ionised to singly charged ions. The l151n+ isotope abundance 95.72%, was monitored. A sampling aperture operating in the continuum flow regime disturbs the plasma upstream for a distance of 2-3 diameters. A 0.4 mm diameter aperture thus samples gas from a region of about 1 mm deep and 2.4 mm in diameter.The spatial resolution of an ion distribution plot obtained by moving the flame about over such an aperture is therefore about 2 mm. Transverse and longitudinal profiles obtained in this way using the single ion mode at a power level of 1200 W and at a solution concentration of 10 pg ml-l are shown in Figs. 5 and 6. The transverse profiles show a strong peak along the axis with a distribution that broadens with increasing aperture - load coil separation in accordance with the visible distribution of the blue emission in the flame seen when yttrium is introduced. In this particular plot the centre of the distribution may be seen to be slightly skew suggesting that the flame axis was not co-linear with the optical bench. The longitudinal profile shows a smoothly decreasing response from the torch mouth outwards.100 80 s ai 60 v) 0 a $ 40 a 20 0 5 4 3 2 1 0 1 2 3 4 5 Fig. 5. Profile of response across plasma flame a t various distances from load coil on ll5Inf. Power 1200 W. Load coil distance A 5 mm ; B 10mm; C 15mm; and D 20mm. Radial distance/mm 100 $ 60 40 al a 20 0 Axial distance from lead coil/mm Fig. 6. Variation of response along flame axis on l151n+. Power 1200 W. Clearly these profiles suggest that the greatest signals are obtained close to the torch before significant ion diffusion into the outer regions of the flame dilutes the analyte ions. This may not be so however where longer sample dwell times in the plasma are required to vaporise and dissociate samples containing a complex matrix or refractory molecules.It may be necessary to use greater aperture - load coil separations in instances where volatilisation interference appears to be present at high matrix concentrations or where incomplete dissociation of refractory oxides is observed. For normal operation a sampling position at 10 mm from the coil has been used at which matrix effects do not appear to be significant but this still requires further study for a range of matrices. Certainly the longitudinal profile shown in Fig. 7 of the metal and monoxide response from a uranium solution taken in the scanning mode suggests that there is no decrease in the degree of dissociation of this highly refractory oxide at close separations. The proportion of oxide response remains roughly constant out to 15 mm after which it increases sharply.This suggests that the majority of the oxide ions do not arise from undissociated uranium oxide molecules but are reformed in the boundary layer which i 1040 GRAY AND DATE ICP SOURCE MS USING Analyst Vol. 108 still present over the sides of the sampling cone up to the edges of the aperture from which they are then entrained into the extracted gas flow. In this instance a scan was performed at each power level over the mass range from 112 to 117 a.m.u. This enabled background channels either side of the l151n+ peak to be observed as well so that the signal to background (random) ratio could also be plotted. Plasma gas flow-rates were not changed during these runs. The l151n signal increases with power up to a plateau level at 1600 W but the signal to noise ratio peaks at 1400 W and decreases above that.As indium is expected to be fully ionised even at low power it is assumed that this increased response is a consequence of the contraction of the flame and central channel diameter which is observed close to the torch mouth as the power is increased. No problems with the plasma torch or sampling system were observed during short-term operation up to 1800 W but the increased heat transfer to the expansion stage for continuous operation at these powers would require better water cooling. There seemed to be little advantage in operating above 1400 W and a level of 1200 W was chosen for normal work. The variation of the response from indium with plasma power is shown in Fig.8. -\ 30 20,s 5 t-. 0 3 10 0 I 0 0 4 800 2 80 40___4400 0.8 1.0 1.2 1.4 1.6 1.8 Power/kW I I I I I Fig. 8. a Signal intensity I and 0 signal to background ratio I / B as. power for 1151n+ a t 10 mm from load coil. 0 5 10 15 20 Distance of aperature from load coil/mm Fig. 7. Variation of uranium metal and oxide response along flame axis at 1200 W showing uranium oxide to uranium ratio (yo). The effect of increased plasma power on the response to a uranium solution at 10 pg ml-l is shown in Fig. 9 again at 10 mm from the load coil. The scanning mode was again used to show both metal and oxide responses. The 238U+ signal increases just as l151n+ did up to 1400 W but beyond this the response falls again. Uranium has a low second ionisation energy of about 12 eV so the fall in response is probably associated with a shift in the ionisa-tion equilibria towards doubly charged ions at high plasma temperatures.The 254UO+ response however shows little change throughout which further supports an origin for the oxide ions in the boundary layer around the edge of the aperture. a I I I I I 0.8 1.0 1.2 1.4 1.6 1.8 Plasma power/kW Fig. 9. Response veYsws power for uranium and uranium oxide a t 10 mm from load coil Se$tember 1983 CONTINUUM FLOW ION EXTRACTION 1041 Background Response and Blank Spectra The experimental detection limits obtained by any analytical method depend critically on the background levels above which the response from the wanted species must be identified. The random background of the system across the whole spectrum has been found to be reasonably low between 10 and 20 counts s-l.However for samples introduced as nebulised solutions the response obtained to the blank normally 1% Aristar nitric acid in distilled water may for some parts of the spectrum be much more significant. Such a blank spectrum is shown in Fig. 10. The response is broadly similar to that obtained in boundary layer ~ a m p l i n g ~ t ~ ~ with two major groups of peaks one the oxygen and hydrogen (or water) group between 16 and 19 a.m.u. and the other at 40 and 41 a.m.u. due to 40Ar+ and 41ArH+. There are however some significant differences notably the great reduction in the third major group of peaks seen in boundary layer spectra at 30 32 and 33 a.m.u. due to the absence of the boundary layer.The important ion 30NO+ is only present as a small peak in continuum flow sampling. As this ion formed in the boundary layer has a low ionisation energy of 9.4 eV it greatly disturbs the ionisation equilibria if present at high concentration, and it can suppress the ionisation of species of higher ionisation energy in boundary layer sampling. Another ion in this group that is greatly reduced is 320,+ and 3302H+ is even smaller. However the presence in the plasma of such high concentrations of hydrogen and oxygen from the solvent does still result in substantial peaks at some inconvenient positions and while they are present some wanted ions will not be detected. Fortunately however, these are relatively few and while for some analyte ions there are small molecular background peaks resulting from ion molecule reactions between hydrogen nitrogen oxygen and argon, which increases the limits of detection for many ions only the random background is signifi-cant.Such small background peaks may be readily removed from the displayed spectrum by subtracting a blank in the data system which greatly contributes to the visibility of unexpected trace peaks. A final major difference from the boundary layer spectrum is the much smaller 80ArAr+. This residual peak is probably due to ions formed in the expansion by a polymerisation reaction but the much larger peak found in boundary layer spectra is more likely to be mostly the product of ion molecule reactions in the boundary layer. The proton attachment peak EIArArH+ found in the boundary layer spectra is absent in continuum spectra.Fig. 10. Blank spectrum on 1% nitric acid solution over the mass range 0-90 a.m.u. Integration time, 1 min; peak count rate “ArHf 9.8 x los s-l. Characteristic Spectra The spectra observed show the same characteristics of simplicity as those reported for boundary layer samplinglo However because the continuum sample is drawn from a higher temperature than exists in the boundary layer more doubly charged ions are seen and thus there are more peaks occurring at values of 9422. Apart from these doubly charged ions ion peaks occur effectively at unit mass intervals. Small oxide peaks less than 1% o 1042 GRAY AND DATE ICP SOURCE MS USING Analyst Vol. 108 the height of the main peaks are seen for elements with the most refractory oxides but apart from these simple elemental peaks are found for both metal and non-metal analytes.Typical spectra taken from the data system memory on to the graphics plotter are shown in Figs. 11-13. These were run on multi-element solutions each element being at a concentration of 10 pg ml-1 for an integration time of 1 min. Fig. 11 shows a spectrum at the bottom end of the mass range. The peaks of 1H+ and 2H+ are off-scale but a small 3H+ peak is seen and well isolated peaks of the lithium and boron isotopes are found. The boron isotopes are at about the correct isotope ratio but the apparently depleted state of the standard lithium solution with a low abundance for 6Li was a source of concern until the lithium was found to have been procured by the suppliers from a nuclear research establishment.The authors have speculated on the problems of un-suspecting users of a standard solution prepared on the basis of an incorrect relative atomic mass! Also to be seen in this spectrum are peaks of 12C+ and 13C+ probably arising from residual carbon dioxide in the argon. Even if it is present in the argon at levels sufficient to cause a significant background it may be simply removed from the response to a carbon containing analyte by blank subtraction. Direct measurements of carbon-12 to carbon-13 ratios may thus be made on solutions. This has been confirmed on 13C-enriched samples on which ratios have been measured with agreement better than 1%. Fig. 12 of about I 7Li+ I Fig.11. Spectrum of lithium and boron in solution a t 10 pg ml-l showing carbon peaks. Peak count rate 'Li+ 5.1 x lo4 s-l. shows part of a wider scan expanded in the data system to display a mass range 100-140 a.m.u. Again this was obtained on a mixed solution in this instance containing 15 elements each at 5 pg ml-1. In this part of the scan singly ionised peaks are seen for In Te I Cs and La with doubly ionised peaks of Pb and U. The resolution setting used for this run was not high enough to separate the doubly ionised Pb peaks. This plot is taken from a wide mass analyser scan from 0 to 260 a.m.u. which as an exception to normal practice was stored in 2048 memory channels in order to provide high print-out resolution for parts of the spectrum. A spectrum of a 10 pg ml-1 solution of heavy elements is shown in Fig.13 where singly ionised spectra of W Au Mg Pb Bi Th and U are seen. Only Pb Th and U doubly ionise I I Fig. 12. Expanded portion of wider scan showing Pb2+ In+, Ua+ Tef I+ Cs+ and La+ each a t a concentration of 5 pg ml-l. Peak count rate laaCs+ 6.2 x lo4 s-1 September 1983 CONTINUUM FLOW ION EXTRACTION 1043 Fig. 13. Spectrum of a mixture containing W Au Hg Pb Bi Th and U each a t a concentration of 10 pg ml-l. Peak count rate 20BBi+, 9.6 x 104 S-1. among these and Pb only slightly as its second ionisation energy is 15.03 eV. The lead response is thus almost as large in total as that for bismuth whereas the peaks for thorium and uranium which have second ionisation energies of about 12 eV are relatively small.Both Tho+ and UOf peaks are present. An expanded plot of the tungsten and mercury isotopes shown in Fig. 14 demonstrates the low background level. The clearly visible peaks of lS0W+ and lSsHg+ are each produced from a solution concentration of 14 ng ml-l. That for mercury is immediately adjacent to the lg7Au+ peak which is about lo3 times bigger but it is smaller than the tungsten peak of the same concentration as mercury is only about 50% ionised because of its higher ionisation energy. 182-4\(\1+ \ 19; Fig. 14. Expanded portion of Fig. 13 showing Isow+ and lS6Hg+ a t concentrations of 14 and 15 ng ml-l respectively. Peak count rate lsoW+ 250 s-l. Rare earth elements are usually difficult to analyse because of the complex spectra produced by multiple ionisation and by peaks from the refractory oxides.The continuum flow spectra from the ICP are remarkably simple however. All have low second ionisation energies and therefore show strong doubly charged ion peaks. First and second ionisation spectra are shown together in Fig. 15 for a mixture of 15 rare earth elements (supplied by Johnso 1044 GRAY AND DATE ICP SOURCE MS USING Analyst Vol. 108 (a) Ar.Ar 70 80 mlz 90 Tm Ho Lu HoO TmO LUO I 140 150 160 170 mlz 180 190 Fig. 15. Spectra of (a) singly and (b) doubly charged ions of a 15rare earth element mixture each at 10 pg ml-I. Upper spectrum M2+ ions mlz 69-90. Peak count rate 175L~2+ 2.4 x lo4 s-l. Lower spectrum M+ ions m/z 138-192. Peak count rate 175Lu+ 9.7 x lo4 s-l. Note *OY+ appears in (a).Matthey Chemicals) at a concentration of 10pgml-1 each. Although there are a number of coincidences between isotopes little difficulty exists identifying and quantifying each element. Although this selection of spectra include elements with ionisation energies ranging from Cs at 3.89 eV to I and Hg at over 10 eV the range of peak heights allowing for abundance, is small and illustrates well the wide element coverage. The only oxide peaks clear of the main spectrum can be seen to be very small September 1983 CONTINUUM FLOW ION EXTRACTION 1045 Quantitative Performance In contrast to other atomic ion sources the ICP source will accept fresh samples as fast as the chosen integration cycle will permit. It is there-fore possible to determine detection limits in terms of the concentration equivalent to the standard deviation of the background just as in other methods of atomic spectrometry.Detection limits determined in the single ion mode are shown in Table I1 expressed as 20 TABLE I1 COMPARISON OF DETECTION LIMITS (20 BLANK) A convenient rate is one every 2 min. Element Lithium . . Boron . . Magnesium Aluminium Titanium . . Vanadium Chromium Manganese Cobalt . . Zinc Germanium Arsenic . . Selenium . . Rubidium . . Silver . . Cadmium . . Indium . . Tellurium . . Caesium . Barium . . Lanthanum Cerium . . Tungsten . . Gold . . Mercury . . Lead Bismuth . . Thorium . . Uranium . . * .Detection limit/ng ml-l h r 3 ICP - SMS ICP - AES* FAAS? 3 1.9 2 1 3.2 1000 0.5 0.1 0.2 0.6 15 20 0.3 2.5 50 0.4 3.3 20 0.2 4.1 3 0.8 0.9 3 0.5 4 5 3 1.2 0.6 1 1 50 7 35 100 15 50 100 2 0.3 0.2 4.7 2 0.5 1.7 1 0.1 42 30 0.5 27 70 15 0.1 0.2 6.7 1600 0.2 32 0.5 20 500 0.2 11 10 0.4 17 200 0.3 28 20 0.2 23 40 0.2$ 43 0.4 170 7 000 --0.3 0.9 20 --* From reference 20. t From reference 21. 1 Doubly charged ion. values. The value of o was determined for each element by running a blank solution for ten integrations usually of 5 s each followed by an integration on a 10 pg ml-l solution of the element usually in a multi-element solution. Comparative values for ICP atomic-emission spectrometry (ICP-AES) and flame atomic-absorption spectrometry (FAAS) taken from recent literature are also shown.The ICP source mass spectrometry (ICPS-MS) values in the continuum mode can be seen to be generally higher than those previously reportedlo for the boundary layer mode but they are considerably more uniform over a wide range of elements. The isolated high values that remain are generally due to the use of an un-favourable isotope for the measurement due to a background coincidence or to the presence of a small background peak under the isotope peak used. These detection limits obtained at an early stage in the development of the continuum flow sampling technique are nevertheless very promising and compare very f avourably with other established techniques.It is known that ion collection and transmission in this system are very inefficient and offer considerable scope for improvement. It is therefore hoped that better detection limits will be achieved as the method is improved. The linear dynamic range of the plasma ion sampling interface and mass analyser is found to be greater than that of the ion detection system in its present form. The channel electro 1046 GRAY AND DATE ICP SOURCE MS USING Analyst Vol. 108 multiplier loses gain above pulse rates of 1 MHz so that ion pulses are increasingly lost below the discriminator threshold. To cover a range of more than 5 decades it is necessary to reduce the ion transmission for high sample concentrations by for example reducing the extraction electrode potential.By doing this the response has been found to be approxi-mately linear over 6 decades up to 1000 pg ml-l. The matrix effects found in boundary layer sampling were particularly severe and pre-cluded its use at total concentrations above about 10 pg ml-l. However in the continuum mode the matrix effects appear to be similar to those experienced in ICP-AES. Ion suppression by species of low ionisation energy is small as can be seen in Fig. 16 where the 10 20 50 100 200 500 1 000 Sodium concentration/pg ml-' Fig. 16. Effect of sodium addition on response to e 5sC~+; and 0 aosBi+ each at 10 pg ml-I. response of cobalt and bismuth falls only slowly with sodium concentrations up to 1000 pg ml-l. The effects of major matrix elements on the response to traces also appears to be small.Results of comparative analyses with and without a typical rock matrix show little effect on the trace responses at total solids contents over 500 pg ml-l with the matrix elements correctly recorded. However much more study is needed of this complex subject in order to confirm these initially promising results. The relatively uniform sensitivity shown by the detection limits in Table I1 over a wide range of ionisation energies suggests that the effective ionisation temperature Ti of the plasma at the sampling point is high. Measurements of the response to known concentrations of elements with a range of ionisation energies may be used to determine Ti by using the Saha equation to relate degree of ionisation to temperature and ionisation energy.The most direct method of doing this is to use the ratio of singly and doubly charged ions of the same atom but this requires an accurate knowledge of the ion transmission at values of m/z a factor of 2 apart. It was preferred therefore in this work to use two singly charged ions of similar m/z so that uniform ion transmission may be assumed. An ideal pair for this would appear to be caesium and iodine as they can be obtained in a fixed stoicheiometric ratio as caesium iodide and are close in mass at 133 and 127 a.m.u. The ionisation energy of caesium is so low (3.89eV) that it may be assumed to be fully ionised but the second ionisation energy is 25.08 eV. The 133Cs+ integral may therefore be used to represent the atomic concentration of both atoms before ionisation for comparison with the 12'1+ integral.A similar compound is rubidium bromide but this was not readily available and so a suitable solution of the two elements was used. Solutions of these two thermometric pairs were run at 10 pg ml-l at a power level of 1200 W together with blank solutions so that background could be subtracted. Two sampling positions were used and the values of Ti calculated from the ion integrals. Unfortunately values of partition coefficient for bromine were not available so a value of Zi/Z = 1 was assumed. In order to simplify the calculation for a multi-component plasma it is necessary to assume a value for the electron concentration H e . The values assumed at 1200 W were 7 x 1015 at 5 mm and 4.6 x 1015 at 10 mm.22923 In spite of the potential errors in the rubidium - bromine result the agreement shown for these two thermometric pairs in Table I11 is very reasonable although the values obtained for Ti are higher than usually reported from AES measurements.A plot of normalised response against ionisation energy for elements with a range of energies is shown in Fig. 17. For clarity only a few have been identified on the plot. The solid line shows the response calculated from the Saha equation for Ti = 9000 K Z i / Z = 1 and ne = 4.5 x 1015 cm-3. The experimental values fall close to this calculated response, which shows a high degree of ionisation at energies well above 1OeV. Even chlorine a September 1983 CONTINUUM FLOW ION EXTRACTION TABLE I11 IONISATION TEMPERATURE Ti AT 1200 W CALCULATED FROM Cs I Rb AND Br RESPONSE 1047 r Sampling position/mm Cs I Rb B; 5 9 090 8 920 10 9 060 8 930 13.01 eV is ionised to more than 5% and is readily detectable.Both halogens and non-metals appear to form positive ions normally and can be detected if the first ionisation energy is below that of argon (15.76 eV) and their response does not coincide with a major back-ground peak. As the ionisation equilibrium in the plasma is dominated by the high con-centration of argon species with ionisation energies above that of argon are not significantly ionised. However only three elements are thus excluded Ne F and He and of course argon cannot be detected if it is used as the plasma support gas. If a reasonable lower limit is set to the useful minimum degree of ionisation as lyo which is reached at 14 eV then this excludes only N (14.53 eV).The measured value of Ti of 9000 K thus appears to represent an acceptable compromise between an adequate degree of ionisation for as many elements as possible and the desire to avoid too many multiply charged ions. It excludes five elements at the penalty of double ionisation to varying degrees for 26 others. The lowest third ionisa-tion energy level is that for La at 19.17 eV so no ions with a charge state above 2 will be seen. The analyser can be set to resolve peaks only 0.5 a.m.u. apart (see Fig. 15) so that together with the low yield of molecular analyte ions the spectra remain simple and peaks are easy to identify. 1.0 $j 0.5 c x 0.2 ; 0.1 a 0.02 9) > 0) .- t;; 0.05 -0.01 ' ' ' ' ' 4 6 8 10 12 14 lo n isa t i o n en erg yleV Fig.17. Relative response zlersus ionisation energy at 10 mm and 1200 W. Graph shows response calculated from Saha equation for Ti = 9000 K and n = 4.5 x 1015 ~ m - ~ . Experimental values shown as points. Isotope Ratio Measurements The ability to resolve adjacent mass peaks sufficiently to reach background between them provides simple peak-area measurement in the data system. Random background from a blank run may be automatically subtracted and peak integral ratios printed out direct at the end of the run. The determination of isotope ratios on a routine basis on solution samples is thus straightforward. The results of such measurements with this equipment in the boundary layer mode have been reported previouslyll and recently the same technique has been described for the continuum mode.13 In both modes a precision of 0.5% or better has been demonstrated on natural galena samples and the results compared with thermal ionisatio 1048 GRAY AND DATE ICP SOURCE MS USING Analyst Vol.108 values on the same samples. The use of continuum sampling for this work offers the advantage of greater matrix tolerance but the higher random background and lower absolute count rate sensitivity were thought initially to be disadvantages. However in both modes the available precision certainly to the level of O.l% appears to be dependent only on counting statistics and continuum mode operation enables the sample concentration to be increased from the 1 pg ml-l normally used in boundary layer work to much higher levels.In practice the limit is set by counting loss in the data system. Losses of a few per cent. at high count rates caused by the finite pulse pair resolution of the system are usually ignored as insignificant but are intolerable when precise ratios between peaks of different sizes are required. Some correction is possible but loss of detector gain at high rates is not easy to determine. As a general rule therefore count rates for isotope ratio determinations are restricted to about 105 s-l and thus concentrations are limited to about 10 pg ml-l or less of the largest isotope. Even so peak integrals of well over lo6 counts may be accumulated in integration times of about 5 min provided that the scan width is limited to between 5 and 10 a.m.u.This is usually ample for the wanted isotopes. Although faster ion detectors and electronics are available they are costly and a better alternative to allow higher concentrations or better ion transmission to yield greater precision would be to use the electron multiplier in a d.c. mode and digitise the output to a lower rate that can be handled without loss. This and other d.c. techniques are being explored. However routine precisions of 0.1% are thought to be within the capability of the present system in the near future at concentrations from 1 to 10 pg ml-l. There seems to be little advantage in trying to better this unless it can be improved to the point where it approaches that achievable with thermal sources.The range l-O.l% adequately covers the needs of a wide range of analytical applications in stable tracer studies isotope dilution analysis and mineralogical prospecting. Although counting statistics may dominate precision in this range other more fundamental problems may make worthwhile further improvement much more difficult to achieve. Further Development Nebulisation of solutions is a very convenient method of sample introduction and a major step forward in mass spectrometry. It lends itself well to automatic operation by the use of a sample changer when samples blanks and calibration standards may be introduced in any desired pattern the resulting analytical data being handled by a dedicated computer. It is a very inefficient process only about 1% of the solution actually reaching the plasma, but even so in an automatic system of this type a volume of only about 3-10 ml is required for an analysis.When this is inconveniently large as for many biological samples a re-circulating nebuliser of the type described by Hulm~ton,~~ may be used with which analysis times of 30 min or more may be obtained from a 1-ml sample. This is particularly attractive for isotope ratio determinations on small samples. 0 2 4 6 8 ' Fig. 18. Single ion response to 20sPb+ for electrothermal vaporisation of a 5-111 drop of 10 pg ml-l solution (5 ng). Integral 2.4 x 106 counts. Vertical axis shows memory channel content. Ti me/s September 1983 CONTINUUM FLOW ION EXTRACTION 1049 Sample introduction to the ICP for use as an ion source remains similar to that for AES and any of the methods described for emission spectrometry may be used such as electro-thermal vaporisation laser ablation and hydride generation.The origin of most of the “permanent” background peaks in the high hydrogen and oxygen populations in the plasma, typically about 10s times the analyte concentration makes any method that avoids intro-ducing water to the plasma potentially attractive for improved detection limits. These alternative techniques are being studied. Early indications are that electrothermal vaporisa-tion of microsamples from a graphite rod offers some advantages. A system similar to that described by Gunn et aZ.25 has been used to introduce samples of 5-pl volume. The sample is desolvated in the usual way at low temperature and then flashed off at high temperature into the injector gas flow.As scan times of as little as 20 ms may be used this provides an ample number of scans over the changing signal to make isotope ratio measurements. On the other hand the scan width must be restricted to the minimum to avoid loss of signal while scanning beyond the important mass range. A typical single ion response from a 5-pl sample is shown in Fig. 18 and the spectrum obtained from 500 ng of cadmium in Fig. 19. A short pulse of ions lasting a few seconds is obtained. 114( Fig. 19. Spectrum of 500ng of cadmium by electrothermal vaporisation. Peak integral 114Cd+ 2.22 x lo5 counts. Excellent agreement has been obtained between isotope ratios determined on microsamples and nebulised solutions.Absolute detection limits ( 2 4 obtained from early runs on a few elements are shown in Table IV. The background spectrum is much simpler in the absence of water and hitherto unusable parts of the spectrum become available. The large 320,+ peak excludes the determination of sulphur by nebulisation but by thermal vaporisation the principal isotope ratios of sulphur have been determined on a single 200-ng sample to better than 1%. The promising results obtained even on very short duration samples introduced into the injector gas suggest that laser ablation from solids would be an attractive method of direct solids mass spectrometry with the very minimum of sample preparation. Direct solids introduction into a plasma ion source was always an aim of even the earliest work and it is intended to explore both laser ablation and arc aerosol generation with the ICP source.Both TABLE IV ABSOLUTE DETECTION LIMITS BY ELECTROTHERMAL VAPORISATION Measurements made on 5-pl samples. Element Detection limitlpg (20) As . . 7 Cd . . 2 Pb 1 Se . . 12 Zn . . * . 1050 GRAY AND DATE these techniques should be capable of greatly reducing the inefficient use of sample involved in preparing solutions of O.l-lyo of solid followed by the loss of 99% of the sample in the nebuliser. in sensitivity can be recovered the ultimate sensitivity of the method should greatly exceed that of any other laboratory method of comparable speed and convenience. If only part of this loss of at least This work is supported by the Institute of Geological Sciences and the R & D programme The authors thank Mr.P. J. The paper is published by on Uranium Exploration of the European Communities. Moore and Miss E. Waine for critically reading the manuscript. permission of The Director Institute of Geological Sciences (NERC). 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. References Gray A. L. Proc. SOC. Anal. Chem. 1974 11 182. Gray A. L. Analyst 1975 100 289. Gray A. L. “Dynamic Mass Spectrometry,” Volume 4 Heyden London 1976 pp. 153-162. Houk R. S. Fassel V. A. and Svec H. J. “Dynamic Mass Spectrometry,” Volume 6 Heyden, Houk R. S. Fassel V. A. Flesch G. D. Svec H. J. Gray A. L.and Taylor C. E. Anal. Chem., Douglas D. J. and French J. B. Anal. Chem. 1981 53 37. Gray A. L. and Date A. R. “Dynamic Mass Spectrometry,” Volume 6 Heyden London 1981, Steams C. A. Kohl F. J. Fryburg G. C. and Miller R. A. Nut. Bur. Stand. Spec. Publ. 1979, Douglas D. J. Quan E. S. K. and Smith R. G. Spectrochim. Acta Part B 1983 38 1/2 29. Date A. R. and Gray A. L. Analyst 1981 106 1255. Date A. R. and Gray A. L. Spectrochim. Acta Part B 1983 38 1/2 39. Gray A. L. and Date A. R. Int. J . Mass Spectrom. Ion Phys. 1983 46 7. Date A. R. and Gray A. L. Int. J . Mass Spectrom. Ion Phys. 1983 48 357. Date A. R. and Gray A. L. Analyst 1983 108 159. Fassel V. A, Pure Appl. Chem. 1977 49 1533. Human H. G. C. and Scott R. H. Spectrochim. Acta Part B 1976 31 459. Campargue R. Rev. Sci. Instrum. 1964 35 111. Greene F. T. Beachey J. E. and Milne T. A. Nut. Bur. Stand. Spec. Publ. 1979 No. 561 Volume Gray A. L. Eur. Spectrosc. News 1982 43 front cover. Boumans P. W. J. M. “Line Coincidence Tables for Inductively Coupled Plasma Atomic Emission Data Sheet 091/11/79 Instrumentation Laboratory Wilmington MA 1979. Alder J. F. Bombelke R. M. and Kirkbright G. F. Spectrochim. Ada Part B 1980 35 165. Kalnickey D. J. Fassel V. A and Knisely R. N. Appl. Spectrosc. 1977 31 137. Hulmston P. Analyst 1983 108 166. Gunn A. M. Millard D. L. and Kirkbright G. F. Analyst 1978 103 1066. London 1981 pp. 234-251. 1980 52 2283. pp. 252-266. No. 661 Volume 1 303. 1 431. Spectrometry,” Pergamon Press Oxford and New York 1980. Received Mavch 23rd. 1983 Accepted April Sth 198

ISSN:0003-2654    

DOI:10.1039/AN9830801033

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst September 1983 Vol. 108 99. 1051-1059 1051 In Situ Gaseous Pre-treatment of Liver Extracts in a Modified Carbon Rod Atomiser During the Determination of Cadmium and Lead John W. Steiner and Harvey L. Kramer Department of Primary Industries Animal Research Institute 665 Fairfield Road Yeerongpilly Brisbane, 4 105 A ustralia An in situ gaseous pre-treatment of bovine liver extracts was performed in a modified carbon rod atomiser. Nitrogen hydrogen oxygen and methane were introduced into the atomiser to eliminate interference and to effect the continuous re-coating of the atorniser with pyrolytic graphite. An optical device improved the precision of the injection of the sample. The effective-ness of the in situ sample pre-treatment was demonstrated by the determi-nation of cadmium and lead in certified samples.Keywords Cadmium and lead determination ; liver extracts ; in situ gaseous pre-treatment ; modi,fied carbon rod atorniser ; electrothermal atomic-absorption spectroscopy Despite the application of many alternative techniques to the determination of trace amounts of metals electrothermal atomic-absorption spectroscopy retains advantages. Features of this technique include high precision and sensitivity the small sample size required the limited pre-treatment of some samples the low cost of equipment and its relatively simple operat ion. During early method development the role composition and flow-rate of inert gases used for sheathing the carbon rod atomiser (CRA) were studied.lS2 The introduction of the hydrogen diffusion flame increased sensitivity and reduced interference for some Another advance was matrix modification which involved the direct addition of either chemical reagent^^.^ or chemically active gases to the at~miser.~-l~ Precision was improved through the use of mechanically guided1l?l2 and automatic sample dispensers.13 These techniques have been improved further by injecting the sample directly into the pre-heated atomiser .I4 A layer of pyrolytic graphite on the atomiser resolves the problems of memory effects and loss of analyte in the atomiser.It also increases operational life reduces interference and gives more uniform heating.15J6 In sit% pyrolysis treatment is necessary to reduce this effect. While this problem can be avoided by chelation and organic solvent extraction of the analyte,17 such techniques are tedious, time consuming and susceptible to the addition of contaminants.A new method for the in situ gaseous pre-treatment of bovine liver samples directly in the atomiser which avoids both solvent extraction and complete sample digestion is described and has not been reported previously. The precision of sample introduction is improved by the use of an optical device. The partially digested liver samples were analysed for cadmium and lead to illustrate the efficiency of the method. However this pyrolytic coat is prone to attack by highly oxidising matrices. Experimental Materials Gases Propane 60% ; propene 30% ; isobutene 10%. Gas Supply Franchises Queensland. Propane 88% ; propene 10% ; butane and butenes 2%.Handigas Commonwealth Industrial Gases Ltd. Brisbane. Methane with propane and hydrogen 8Z-8670. Natural gas Brisbane Gas Company. Methane 99.9%. Matheson 1052 STEINER AND KRAMER GASEOUS PRE-TREATMENT OF LIVER Analyst Vol. 108 Methane in nitrogen 10%. Hydrogen industrial dry. Oxygen medical BP. Nitrogen high purity. Commonwealth Industrial Gases Ltd. Commonwealth Industrial Gases Ltd. Commonwealth Industrial Gases Ltd. Commonwealth Industrial Gases Ltd. Glassware cleaning (re-distilled before use). ionic detergent. water. New borosilicate glassware was initially de-contaminated by soaking in 10% nitric acid It was subsequently maintained in a clean condition using a non-Before use the glassware was rinsed thoroughly with doubly glass-distilled Standard solutions The concentrations of the working standards were 0.002-0.006 mg 1-1 for cadmium and 0.01-0.03 mg 1-1 for lead.These were prepared daily from concentrated stock solutions (BDH Chemicals Ltd.). Standard reference material Bureau of Standards. 0.34 & 0.08 mg kg-l for lead (95% confidence limits). Freeze-dried powdered bovine liver (SRM 1577) was obtained from the US National Certified concentrations were 0.27 & 0.04 mg kg-l for cadmium and Modification of Instrument A Varian Techtron AA-175 spectrophotometer equipped with a carbon rod atomiser (CRA-go) was modified to perform in sit% gaseous pre-treatment of samples in the atomiser. Gas jets The chimney was modified as illustrated in Fig. 1. The original sheath gas entry port was removed and refitted on the opposite side of the chimney.The original orifice of this port was sealed by brazing. A 150-mm long hollow stainless-steel needle was bent at a 90" angle and inserted through the re-positioned port so as to place the end of the needle in the centre of the shims. This end was flush with the top of the shims shown in Fig. 1. The length of the needle was sufficient to extend into the polypropylene tube attached to the gas entry port. A further bend in the inserted needle allowed the needle to exit through the walls of the plastic tubing and terminate in a Luer fitting. A gas-tight seal between the needle and the plastic tube was achieved with epoxy cement. To position the end of the needle in the centre of the shim assembly the central flat shim was removed trimmed and cut in half, removing 2mm from its length.It was then re-assembled on each side of the jet. The diameter of the needle (2 mm 0.d.) ensured that the crinkle shims secured it in position. This vertical gas jet was used to introduce extra methane into the atomiser environment. - / Atorniser I ' \ Original port Luer ,/ fitting Gas-tight seal Stainless -steel tube \ Polypropylene tube \ New port Fig. 1. Vertical cross-section of chimney September 1983 EXTRACTS I N CARBON ROD ATOMISER FOR DETERMINING C D AND P B 1053 The relocation of the sheath gas entry port on the chimney was necessary to provide adequate space for the optical device. The design of the optical-device support permitted the direct attachment of two transverse gas jets (Fig.2) which were directed towards the ends of the atomiser. With this arrangement both the optical field of view and the gas transverse jets could be precisely aligned with the interior of the atomiser. The transverse gas jets were constructed from stainless-steel needles (i.d. 2 mm length 80 mm) bent at a 70" angle near the discharge jet outlet. The outlets were positioned at a distance of 5 mm from each end of the atomiser. During the atomisation process these jets were removed from the light path by raising the optical device on a vertical rail. H2 I 0 2 I Transverse gas jets Atomiser III Zone A .r t / \ Zone C Zone C t t t T \ / I I I I I I \ N 2 - CH4 - H2 I Fig. 2. Gas flow and mixing of gases in the atomiser.Gas control box The standard gas control and mixing facility for hydrogen and nitrogen in the Varian Techtron CRA-90 power control unit was replaced with a specially constructed gas control box. This control box housed rotameters fine needle valves float flow meters pressure gauges on/off valves and a gas mixing chamber. This chamber was used to mix nitrogen, hydrogen and methane before they entered the chimney via a polypropylene tube. The chamber consisted of a stainless-steel cylinder (volume 80 ml) mounted at the rear of the gas control box. Four stainless-steel 3-mm bulkhead fittings (Swagelock) were attached to this cylinder. Three of these fittings served as gas inlets and the fourth was the outlet. The flow-rates of the gases were measured with bubble-type flow meters and a stop-watch.Separate gas outlets were provided on the control box to facilitate the connection of the PTFE gas lines to the vertical and transverse jets. Optical device the optical device. the image of the internal atomiser could be seen. (GV TL-11 Olympus Optica) was focused into the atomiser via the optical device. illumination allowed a magnified image (x 3) of the interior of the atomiser to be viewed. A prism a mirror a shutter and a magnifying lens encased in a stainless-steel box comprised The top of the box was covered with transparent plastic through which A light beam from an external light source Thi 1054 Procedure Preparation of SRM 1577 and test samples The SRM 1577 and test samples of liver tissue were prepared by weighing suitable amounts (approximately 0.5 g dry approximately 3.0 g fresh) into glass-stoppered glass digestion tubes of 50ml capacity.Concentrated nitric acid (5ml) was added and the tubes were lightly stoppered before being placed in a shaking water-bath at 80 "C for 4 h. The partially digested samples were diluted to 25 ml before analysis. STEINER AND KRAMER GASEOUS PRE-TREATMENT OF LIVER Analyst VoZ. 108 In situ pyro[ysis treatment The selection of the best coating agent for the in situ pyrolysis treatment was carried out by adding small amounts of hydrocarbon gases (natural gas Handigas and Speed-e-gas the methane - nitrogen mixture or pure methane) to the sheath gas (pure nitrogen). The natural gas Handigas and Speed-e-gas were passed through an activated charcoal filter to remove impurities before mixing with the sheath gas.To determine the optimum conditions for pyrolysis treatment the hydrocarbon gases were added at different flow-rates (20-160 ml min-1) without changing either the test solution or the instrumental conditions. Under each set of conditions the sensitivity precision and repeatability of the analyses of the blank standard and sample solutions were determined. Examination of the surface coating of the atomiser by optical and electron microscopy assisted in the assessment of the optimum coating conditions. Instrwnent operation With the instrument in "total-absorption mode," the optical device and transverse gas jets were aligned with the atomiser in the atomisation compartment and then fixed in place (Fig.2). The atomiser was raised to the required temperature of 85-90 "C before injecting the sample solution using a micro-dispensing pipette (Scientific Glass Engineering Pty. Ltd., Australia) with polypropylene tip. The depth of insertion of the tip into the atomiser was monitored with the optical device. The critical balance between the rate of injection and evaporation of the sample was constantly observed. The pipette was carefully removed after injection. Ten seconds after the ashing cycle began the transverse gas jets and optical device were rapidly removed from the atomiser. The instrument was switched into "corrected-absorption mode.'' The atomisation cycle commenced on completion of the ashing cycle and the absorbance was recorded on the digital readout system.The instrument was operated in "peak mode" with continuous background correction and the instrumental conditions are listed in Table I. TABLE I INSTRUMENTAL CONDITIONS Parameter Volume of injection/pl . . Drying temperature/"C . . Ashing temperature/"C . . . . Atomising temperature/"C . . I . Drying time/s ,. ,. Ashing time/s . . Hold time/s . . Ramp rate/"C s-l . . Residence time of transverse gas jets/s . . f . PM voltage/V Cadmium 85 350 2 000 40 40 800 10 398 2.0 2.5 Lead 4.0 90 600 2 200 50 50 400 10 417 3.0 Optimam gas flowrates and lead are listed in Table 11. The optimum flow-rates of the gases used in the CRA-90 for the determination of cadmium Operating seqzcence experimental conditions and instrumental procedures were established.The effectiveness of the pre-treatment of the samples was determined after the optimum This determinatio Se@?mber 1983 EXTRACTS IN CARBON ROD ATOMISER FOR DETERMINING CD AND PB 1055 TABLE I1 OPTIMUM FLOW-RATES (MEAN & STANDARD DEVIATION) OF GASES FOR THE DETERMINATION OF CADMIUM AND LEAD Flow-rate/ml min-l 1 Position of gas introduction Gas Cadmium Lead Sheath . . . . Nitrogen 6711 f 62 6711 f 62 Hydrogen 133 f 2.2 116 f 2.4 Methane 100 f 1.3 100 f 1.3 Transverse gas jets . . . . Hydrogen 224 f 1.2 224 f 1.2 Oxygen 160 f 1.1 101 f 0.8 Vertical gas jet Methane 8.9 f 0.08 0 consisted of an exact repetition of the following operating sequence designed to monitor changes in accuracy precision and repeatability while different gases and the optical device were being used.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Install a new graphite tube and align the atomiser. Fire the atomiser once. Inject and atomise the blank solution three times. Inject and atomise each of the three standard solutions four times determine the correlation coefficient and the slope of the calibration graph. Inject SRM 1577 nine times and its duplicate three times determine the concentration. Inject a liver sample twelve times. Inject the blank solution twice. Inject each of the three standard solutions three times and determine the correlation coefficient and the slope of the calibration graph. Inject SRM 1577 nine times and determine its concentration (at this stage the tube has undergone 60 firings).Inject another liver sample three times. Inject and atomise the blank solution twice. Determination of accwacy precision and recovery After the optimum conditions for the analyses were determined the accuracy and precision of the procedure were calculated from data obtained by analysing SRM 1577 four times. The influence of the matrix on the analysis of cadmium and lead in fresh liver was examined by adding small amounts of the analytes to weighed samples before digestion and then measuring the recovery. Results and Discussion Gaseous In Situ Sample Pre-treatment in the Atomiser Pre-treatment of partially digested bovine liver samples was performed in the atomiser by adding different gases.The effect of the gases during drying ashing and atomising was demonstrated by determining cadmium and lead in SRM 1577 (Tables I11 and IV). Each reported concentration is calculated from nine absorption measurements. Ranges given are for the 95% confidence limits. Accuracy was assessed by the analysis of SRM 1577 and by comparison of the measured and certified values. The resulting difference was evaluated with the Student t-test. Poor accuracy but good precision for both determinations were observed when only nitrogen was used. As the number of firings increased the accuracy decreased further for both elements. However a significant improvement in accuracy and precision for both elements resulted from the sequential addition of different gases to the atomiser and the use of the optical device.The differences between the measured and certified values of the SRM 1577 were not significant when the pre-treatment was included in the analysis. Role of Oxidising and Reducing Gases Added to the Atomiser The addition of gases to various locations in the atomiser provided different effects. Methane performed two distinct roles as a pyrolytic coating agent; and as a reducing agent 1056 STEINER AND KRAMER GASEOUS PRE-TREATMENT OF LIVER Analyst VoZ. 108 TABLE I11 EXPERIMENTAL CONDITIONS AND RESULTS OF THE REPETITIVE DETERMINATION OF CADMIUM IN CERTIFIED STANDARD LIVER SRM 1577 The standard liver samples contained 0.27 f 0.04 mg kg-1* of cadmium. Experimental conditions Gas introduction system r A 7 Vertical Transverse Optical Sheath jet jets device - - - - - - .. . . N2 N,+ H + CH - . N + H + CH, N + H + CH, N2 + H + CH4 - . CH H + 0 + - - - N + H2 N + H + CH CH - -- CH H, CH 0 - - * - * . N + H,+ CH - * CH H,+ 0, * At the 95% confidence limit. t 19-25 firings. $52-60 firings. Determined concentration*/ mg kg-1 Period l t 7izzq-0.284 f 0.009 0.328 f 0.012 0.307 f 0.015 0.281 f 0.008 0.352 f 0.009 0.264 f 0.023 0.284 f 0.009 0.269 f 0.006 0.356 f 0.010 0.302 f 0.015 0.346 f 0.008 0.289 f 0.009 0.361 f 0.015 0.330 f 0.008 0.274 f 0.010 0.277 f 0.005 When methane approached the atomiser it was thermally decomposed into its elemental constituents. The thermal dissociation commenced at 607 O C . 1 8 Some of the resulting free carbon particles formed a thin layer of pyrolytic graphite on the surface of the incandescent atomiser.The most effective formation of the pyrolytic coat occurs at a temperature of 2000 "C.19 Some of the excess of free carbon particles arising from methane degradation flowed through the atomiser contacting metal complexes undergoing dissociation and thus effecting reduction. This was proposed by Campbell and Ottaway20 and was further supported by the experimental findings made by Ashton,21 who showed that the addition of glucose to samples resulted in a large increase in sensitivity for some metals and also increased durability of the atomiser. The complete absence of methane during the analysis of cadmium and lead resulted in rapid loss of sensitivity and precision and the development of memory effects.The proportions of carbon used for re-coating the atomiser and for chemical reduction processes have not been determined. However there is a clear indication that different amounts of free carbon are necessary for the analysis of different elements. Methane added to the sheath gas was sufficient for the analysis of lead but extra methane provided through TABLE IV EXPERIMENTAL CONDITIONS AND RESULTS OF THE REPETITIVE DETERMINATION OF LEAD IN CERTIFIED STANDARD LIVER SRM 1577 The standard liver samples contained 0.34 f 0.08 mg kg-l.* Experimental conditions Gas introduction systems r \ Transverse N2 . . . . . . . . N + H2 - . N + H + CH - N + H + CH - . 0, N + H + CH . * - * H + 0, N2 + H2 + CH4 .. * * H + 0, A r Sheath jets ---N + H + CH * * . . H, * At the 95% confidence limit. t 19-25 firings. $ 52-60 firings. Determined concentration*/mg kg-' A Optical 7 device Period l t Period 2; - 0.221 f 0.022 0.174 f 0.030 - 0.325 f 0.005 0.294 f 0.016 - 0.346 f 0.019 0.316 f 0.029 - 0.319 f 0.013 0.343 f 0.017 - 0.335 f 0.034 0.262 f 0.019 - 0.317 & 0.003 0.323 &- 0.004 + 0.328 f 0.006 0.319 f 0.00 Seibtember 1983 EXTRACTS I N CARBON ROD ATOMISER FOR DETERMINING CD AND P B 1057 the vertical jet was necessary not only for cadmium but also for molybdenum cobalt and chromium.22 The presence of methane with hydrogen and oxygen influenced the absorbance signal for each metal (cadmium and lead) differently. Signals from standard solutions in each instance were slightly en-hanced but those from SRM 1577 were severely reduced for cadmium and strongly enhanced for lead (Tables I11 and IV).The advantages of adding either oxygen or hydrogen into the atomiser for the reduction of interferences have been known for many years.'-1° These gases produce an increase in sensitivity and repeatability while lowering the ashing temperature. However the advan-tages of the simultaneous introduction of hydrogen and oxygen directly into the atomiser during injection drying and ashing of the sample have not been investigated previously. Their presence facilitates a more uniform drying and ashing of the sample solution. The results presented in Tables 111 IV and VII show that oxygen and hydrogen introduced into the atomiser for the initial 10 s of ashing reduced matrix interference and increased repeata-bility sensitivity linearity and accuracy of the determination.Hydrogen also acted as a reducing agent when added to the atomiser. The mixing ratios of these gases are highly critical. Deterioration of the Pyrolytic Acid and In Situ Pyrolysis Treatment These factors include the oxidising properties and composition of the matrix temperature of the atomiser during the ashing and atomising stage hold time ramp rate number of firings the amount of oxygen and moisture in the sheath gas and the properties of the sheath gas. However the most important factor causing deterioration of the pyrolytic coat is the highly oxidising nature of the sample m a t r i ~ .~ ~ ~ ~ To improve the resistance of the atomiser to chemical attack and oxidation continuous pyrolytic re-coating is necessary. A convenient and successful method for the continuous coating of the Perkin-Elmer heated graphite atomiser (HGA-2100) involves the use of pre-mixed gas (10% methane in nitr0gen).~5 Other hydrocarbon gases have been used successfully for in sit% coating of the Shandon Southern atomiser (A3470) .26 Our use of these methods with various hydrocarbon gases for in sit% coating of the CRA-90 was unsuccessful except with pure methane. The application of pure methane for the in sit.u pyrolysis treatment of the CRA-90 resulted in a shiny non-porous evenly distributed pyrolytic coat on the surface of the graphite tube. A combination of factors affects the lifetime of the pyrolytic coat.Gas Flows in the Modified Atomiser The traditional vertical flow of the sheath gas during injection drying and ashing of the sample in the CRA-90 was changed by the modification. The introduction and mixing of the gases used in the modified CRA-90 is illustrated in Fig. 2. Three different zones A B and C are postulated. Zone A is located within the atomiser and contains a complex mixture of oxygen hydrogen nitrogen and methane during injection drying and ashing of the test solution. This is due to the position of the transverse gas jets. The gas stream from these jets is very narrow and travels transversely with a much greater velocity than the vertical gases from the chimney so that small amounts of these gases (nitrogen hydrogen and methane) are entrained and carried into the atomiser.Once the transverse gas jets are removed from the atomiser the composition of the gas mixture in the atomiser will be identical with that surrounding the atomiser (zone B). Zone C contains a lower concentration of methane than zone B owing to the location of the vertical gas jet. However zones B and C are identical when the vertical jet is not used. Use of Optical Device During manual injection into the pre-heated atomiser the optical device was used to balance the rate of injection against the rate of evaporation so that optimum drop size was maintained. An increase in the evaporation rate caused a rapid deterioration of the drop existing between the end of the syringe tip and the wall of the graphite tube.The re-formation of this drop attached to the end of the tip and the wall often led to spitting of the re-entering solution. On the other hand if the injection rate was greater than th 1058 STEINER AND KRAMER GASEOUS PRE-TREATMENT OF LIVER Analyst VoZ. 108 evaporation rate an irregular change in the area of distribution of the particulate material within the atomiser occurred and resulted in increased error. Tables I11 and IV show only a minor improvement in the experimental results when the optical device was used for the injection of samples. This was mainly because highly experi-enced operators performed this investigation. However when inexperienced operators used the device their results were equivalent to the standard achieved by expert operators.Without the device the results of inexperienced operators were extremely variable. TABLE V ANALYSIS OF SRM 1577 LIVER Trace element contentt/mg kg-l No. of No. of analyses Element sub-samples per sub-sample* Certified Found Cadmium . . 4 4 0.27 f 0.04 0.27 f 0.002 Lead 4 4 0.34 f 0.08 0.33 f 0.009 * Performed on separate days. t At the 95% confidence limit. Evaluation of Modified Procedure The accuracy and precision obtained with the modified CRA-90 are shown in Table V while recoveries of elements added to bovine liver are given in Table VI. The determination of accuracy and precision were repeated regularly approximately every 200 analyses during the course of more than 6000 analyses. Established procedures were used each time and no variations were observed.TABLE VI RECOVERY OF TRACE ELEMENTS ADDED TO LIVER No. of Amount added/ Recovery f s.d., Element determinations PLg % Cadmium . . 4 Lead 4 0.4 94.6 f 0.9 0.5 96.8 f 1.2 Table VII compares the analytical data for cadmium and lead using the instrument “with” and “without” the in sit% gaseous pre-treatment procedure. The absolute error for the determination of cadmium was higher for samples without pre-treatment. There was no significant difference in the correlation coefficient (r) of the calibration graph. The sensitivity (slope b,) was significantly higher for the determination of cadmium when the pre-treatment was used. The absolute error for the determination of lead was significantly higher when samples were analysed without pre-treatment.There was only a slight difference in sensitivity and in the correlation coefficient in analysing the samples with or without pre-treatment. TABLE VII COMPARISON OF ANALYTICAL PARAMETERS OBTAINED WITH AND WITHOUT THE in situ GASEOUS PRE-TREATMENT Cadmium in situ Lead in situ pre-treatment pre-treatment Analytical parameter 7iiiiT- - Absolute error*/mg kg-l. . . . 0.00 +0.09 - 0.01 -0.12 Correlation coefficientlr . . . . 0.9999 0.999 3 0.999 7 0.996 7 Slope/b . . . . 32.100 25.650 5.380 5.705 value of SRM 1577. The absolute error is defined as the difference between the observed and the certifie Sefitember 1983 EXTRACTS IN CARBON ROD ATOMISER FOR DETERMINING CD AND PB Comparison of the Modified CRA-90 with the GTA-95 and HGA-500 Atomisers Even though some of the latest atomisers (Varian GTA-95 and Perkin-Elmer HGA-500) appear to have features similar to those of the modified CRA-90 there are distinct differences between them.The GTA-95 and HGA-500 atomisers carry regular fittings for adding gases into the internal and external environment of the atomiser and on/off valves are installed for alternation of these gases. However the gas supply systems in these atomisers are not as versatile as the system used in our modified CRA-90. The modifications described allow accurate control mixing and introduction of up to five gases into various locations in the atomiser and chimney. This results in uniform evaporation of the sample solution reduced matrix interference and improved control of chemical reduction processes occurring during atomisation.Commercially available atomisers permit the control of two gases so that the benefits deriving from the use of more than two gases have not been fully exploited previously. 1059 Conclusion Partially digested bovine liver samples were chemically pre-treated directly in the atomiser by using different gases. The introduction of these gases was synchronised with the over-all operation of the atomic-absorption spectroscopic instrumentation. Long-term improve-ments in sensitivity accuracy precision and repeatability of analysis were achieved. In addition the working time was reduced considerably by using the simplified sample pre-treatment. We thank Mr. B. J. Blaney and Dr. D. C. Moy for their assistance in writing this paper.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. References Alder J. F. and West T. S. Anal. Chim. Acta 1970 51 365. Brodie K. G. and MatouSek J. P. Anal. Chem. 1971 43 1557. Reeves R. D. Patel B. M. Molnar C. J. and Winefordner J. D. Anal. Chem. 1973 45 246. Araktingi Y . E. Chakrabarti C. L. and Maines I. S. Spectrosc. Lett. 1974 7 97. Garnys V. P. and MatouSek J . P. Clin. Chem. 1975 21 891. Hinderberger E. J. Kaiser M. L. and Koirtyohann S. R. A t . Spectrosc. 1981 2 1. Kundu M. K. and Prdvot A. Anal. Chem. 1974 46 1591. Delves H. T. and Woodward J. At. Spectrosc. 1981 2 65. Beaty R. D. and Cooksey M. M. A t . Absorpt. Newsl. 1978 17 53. Beaty M. Barnett W. and Grobenski Z. At. Spectrosc. 1980 1 72. Maessen F. J. M. J. Posma F. D. and Balke J. Anal. Chem. 1974 46 1445. Millar K. R. Cookson F. and Gibb F. M. Lab. Pract. 1979 28 752. Maxfield R. E. A t . Absorpt. Newsl. 1979 18 100. Manning D. C. Slavin W. and Myers S. Anal. Chem. 1979 51 2375. Sturgeon R. E. and Chakrabarti C. L. Anal. Chem. 1977 49 90. Wall C. D. A t . Absorpt. Newsl. 1978 17 61. KomArek J. and Sommer L. Talanta 1982 29 159. Walish W. and Jaenicke O. Talanta 1975 22 167. Clyburn S. A. Kantor T. and Veillon C. Anal. Chem. 1974 46 2213. Campbell W. C. and Ottaway J. M. Talanta 1974 21 837. Ashton A. Lab. Pract. 1978 27 1050. Steiner J. W. and Ryan K. M. unpublished data. Studnicki M. Anal. Chem. 1979 51 1336. Wilson D. O. Commun. Soil Sci. Plant Anal. 1979 10 1319. Manning D. C. and Ediger R. D. A t . Absorpt. Newsl. 1976 15 42. Thompson K. C. Godden R. G. and Thomerson D. R. Anal. Chirn. Acta 1975 74 289. Received 23rd March 1983 Accepted 11th April 198

ISSN:0003-2654    

DOI:10.1039/AN9830801051

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

1060 Analyst September 1983 VoE. 108 $9. 1060-1066 Evaluation of Selenium Determination in Biological Material by Atomic-a bsorption Spectroscopy Rodney J. Mailer NS W Department of Agriculture Agricultural Research Institute Wagga Wagga New South Wales 2650, Australia and James E. Pratley School of Agriculture Riverina College of Advanced Education Wagga Wagga New South Wales 2650, A ustmlia Existing methods of selenium determination by atomic-absorption spectro-scopy using the hydride vapour generation technique were examined. Of the factors investigated it was found that the method of digestion could be a source of considerable error. Stirring time in the vapour generator, acid concentration and final solution volume were also critical factors affecting the consistency of results.Sodium tetrahydroborate(II1) pellets were found to have considerable advantages over a solution of the same chemical owing to their stability and the uniform rate of addition and reaction. Keywords Selenium determination ; biological material ; atomic-absorption spectroscopy ; hydride vapour generation The determination of nanogram levels of selenium in biological material has generally been based on a fluorimetric technique with 2,3-diamin~naphthalene,~ a method that has been improved by the use of semi-automated techniques.2 Other methods available include atomic-absorption spectro~copy,~s~ radiochemical methods15 gas - liquid chromatography6 and, more recently inductively coupled plasma emission ~pectroscopy.~ Analysis by atomic-absorption spectroscopy has been shown to be a quick and simple method of routine analysis using the hydride vapour generation technique3s8p9 and is currently used on a routine basis in many laboratories.Although the literature contains several references to this method there appear to be a number of factors generally overlooked that have considerable bearing on accuracy and precision. Through the necessity to mineralise all forms of selenium and at the same time prevent loss through volatilisation at high temperatures or in reducing conditions the method of sample dissolution is a major factor in obtaining consistently accurate results. Several workers have studied the types of oxidation mixtures used in dissolution including perchloric acid - sulphuric acidlg nitric acid - urea,6 phosphoric acid - hydrochloric acidlo and most frequently nitric acid - perchloric a ~ i d .~ * ~ - ~ v ~ Conditions within the vapour-generation chamber including volume of solution acid concentration and acid composition must be consistent. The solution contains hydrochloric acid used to reduce selenium(V1) to selenium(IV) the other components will depend on the digestion method employed. Volumes of 5-50 ml have been used and hydrochloric acid concentrations within the chamber of between 5 and 50% .499910 Increasing the concentration of the acid solution or the introduction of other acids e.g. nitric acid will reduce the peak absorbance. l1 The presence of oxygen within the silica cell causes significant absorbance a t 196.0nm. A concise routine of purging and stirring is therefore necessary prior to the addition of sodium tetrahydroborate(II1) .lo Sodium tetrahydroborate(II1) has been used both in sodium hydroxide solution at con-centrations of 2.5-6yO9 and in the dry state as pellets.The solution is relatively unstable and is added with a syringe via a septum and therefore at a variable rate. Pellets have been shown to be more reproducible and more convenient owing to their stability. This paper reports on some aspects of previous methods using the atomic-absorption technique outlining sources of variability and means of minimising them MAILER AND PRATLEY Experimental Apparatus and Conditions 1061 A thermostatically controlled aluminium block with a two-step heat control and over Holes were drilled in the block to A Varian Model AA875 double-beam atomic-absorption spectrophotometer was used in temperature cut-out was used for sample digestion.accommodate 25 Pyrex digestion tubes 32 mm diameter x 215 mm tall. conjunction with a Varian Model 65 hydride generator. Atomic-absorption conditions The conditions used were as follows light source Varian hollow-cathode lamp with a deuterium background correction lamp; lamp current 10 mA ; wavelength 196.0 nm ; spectral band pass 1.0 nm; fuel - support acetylene - air; and measurement mode peak height 15 s integration time. Recorder The recorder was a Varian 9176. A chart speed of 5 cm min-l and a pen response of 10 mV full-scale deflection were used. Hydride vapour generator The following conditions were used internal dead volume 150 ml [+30 cm of 6 mm poly(viny1 chloride) tubing] ; absorption cell quartz 30-cm path length ; nitrogen supply pressure 140 kPa; and flow-rate 2 1 min-l.Reagents The digestion acid used for blood samples was a 2 + 1 V/V mixture of analytical-reagent grade nitric acid - perchloric acid. Hydrochloric acid 6 M. Distilled from borosilicate glass apparatus. Hydrochloric acid 20% V/V. Diluted from 6 M hydrochloric acid. Sodium tetrahydroborate(ll1) powder. Sodium tetrahydroborate(II1) pellets. Selenium standard solution 1000 pg ml-l. Nitric acid sfi. gr. 1.42. Perchloric acid 72% V/V. AnalaR grade from BDH Chemicals Ltd. For plant material a 5 + 2 mixture was used. Supplied by Alfa products Danvers MA USA. Supplied by Alf a products.Supplied by BDH Chemicals Ltd. AnalaR grade from BDH Chemicals Ltd. Preparation of Samples Blood samples A 1.00-ml sample of whole blood was measured into a clean test-tube using a positive displacement micropipette to ensure total dispensing of the viscous medium. Carefully, 3.0 ml of nitric acid - perchloric acid were added; the tube was covered and allowed to stand overnight. (Standing overnight was not found to be necessary with blood samples if a slow rate of heating was employed i.e. ambient to 150 "C over a period of 2 h.) The following morning the tubes were placed in a cold digestion block and heated to 150 "C. This tempera-ture was maintained until brown fumes of nitric acid had disappeared and white wispy fumes of perchloric acid had become visible.Charred samples were discarded although this was infrequent with the blood samples. The temperature of the block was raised to 200 "C and held for approximately 30 min at which time dense white fumes could be seen. This was to ensure the complete removal of nitric acid and the total conversion of any organic forms of selenium into selenous acid. The tubes were then removed and allowed to cool to ambient temperature prior to the addition of 0.2 ml of 6 M hydrochloric acid after which they were heated at 150 "C for 15 min to reduce selenium(V1) to selenium(1V). The tubes were removed and allowed to cool then 20 ml of 20% V/V hydrochloric acid were added to each. This process took about 90min. Plant samples Extra nitric acid was required to break down the fibrous material and it was necessary to add the perchloric acid separately, after the initial digest to avoid the loss of selenium due to charring.The digestion process was modified for plant material 1062 MAILER AND PRATLEY EVALUATION OF SELENIUM Analyst Vol. 108 A 1.0-g sample of dried plant material was weighed into each sample tube and 5 ml of nitric acid were added. A glass funnel was inserted into the top of the tube and was left to stand overnight. The next morning the tubes were placed in a cold aluminium digestion block and the temperature was increased slowly over a period of 30 min to 130 “C. Dense brown fumes were formed in the tubes. This temperature was maintained until the fumes became a light honey colour (approxi-mately 1 h) then 2 ml of perchloric acid were added through the funnel.After 2 min the funnels were removed and the temperature was increased to 150-170 “C and maintained until the nitric acid was removed and white wispy fumes of perchloric acid were evident. The temperature was further increased at this stage to 190-200 “C and held for a further 30 min to remove the final traces of nitric acid. The tubes were cooled to ambient temperature 0.2 ml of 6 M hydrochloric acid was added and then they were re-heated to 150 “C for 15 min. The tubes were again cooled to ambient temperature and 20 ml of 20% hydrochloric acid were added to each. Atomic-absorption Determination The method of calibrating the instrument was critical for achieving accurate and repro-ducible results. The Varian AA875 spectrophotometer has three modes of calculation the integration-repeat mode in which the average reading over a given integration time is displayed and continually updated ; the peak-height mode which indicates the highest reading over a time interval; and the peak-area mode which indicates the area under the graph during the hydride generation reaction.Although peak-area calculation has been reported to produce consistent res~lts,l~~J’ our work and that of others10 has shown peak-height calculation to be more reliable. The time required for peak area is in an excess of 25 s whereas peak height required only 10-15 s. The method of calibration was to allow the chamber to purge with nitrogen for about 2 min and then to calibrate zero in the integrate repeat mode. An integration time of 15 s was then called and the peak-height mode engaged.The instrument was zeroed using a blank solution as follows. The vapour generator lid was slightly raised above the chamber and 21 ml of blank solution (1 ml of perchloric acid 20 ml of 20y0 hydrochloric acid) were dispensed into the chamber The lid was closed quickly to avoid an excess of air from entering the chamber and the stop-watch was started. After 15 s the magnetic stirrer was turned on for 5 s before “Cal Zero” and “READ” were pressed and a sodium tetrahydro-borate(II1) pellet was immediately dispensed into the chamber. After 15 s the indicator displayed 0.000 and the instrument was zeroed. The drain plug was pressed to remove the previous sample. The lid was raised and the sample dispensed into the chamber.Immediately the lid was closed and the stop-watch started. After 15 s the stirrer was started and stirring was continued for 5 s before “READ” was pressed. Immediately a pellet was dispensed into the chamber and the absorbance value was obtained from the digital indicator 15 s later. This process was repeated for all standards and samples. A chart recording was run sequentially to ensure that a characteristic response was obtained for each sample and the peak height was within the integration time allowed. Re-calibration of zero was necessary as the recorder pen failed to return to the base line following sample reading. Rinsing the chamber between samples was found to be unnecessary. Reading of the sample solutions was essentially the same as for calibrating zero.Results and Discussion Sample Preparation Cleanliness of glassware The utmost care needs to be taken with cleanliness of glassware to avoid contamination. The glassware was new and was kept separate from all other glassware. Following each digest tubes were immediately rinsed several times with distilled water and inverted to drain dry. No detergents or acids were used for washing up. The use of selenium as a catalyst in Kjeldahl digests is a potential source of significant contamination if both analyses are undertaken in the same laboratory September 1983 DETERMINATION IN BIOLOGICAL MATERIAL BY AAS 1063 Digestion pyocess Rapid heating in the initial stages of digestion caused vigorous boiling which gave low readings from samples of known concentrations of selenium possibly due to physical loss.Charring may also result and this leads to the volatilisation of selenium. All samples showing signs of charring were discarded. Fig. 1 shows the effect of charring on duplicate samples of blood. It was found that plant samples showing slight darkening when the temperature was increased from 150 to 190 "C could be recovered by adding 0.5-1.0 ml of nitric acid and re-digesting at 150 "C. It was more reliable however to discard the digest and start anew. Complete removal of nitric acid from the digested solutions is also stressed as its presence affects the absorbance and creates erroneous results (Fig. 2). The majority of nitric acid should be removed whilst the temperature is maintained at 150 "C to avoid vigorous boiling.Final traces of nitric acid are removed at 200 "C in the final digestion step. 0.10 Q) C m e s 3 0.05 0 C Fig. 1. (A) Effects of charring during digestion on the absorbance value of a blood sample compared with (B) the same sample without charring (without background correction). (C) Blank solution. 0.10 8 $ a m e 0.05 n 0 1 2 3 Volume of nitric acid addedhl Fig. 2. Effects of nitric acid on absorbance. Sample concentration 100 ng per 20 ml. Reduction of selenium The success of the reduction of selenium(V1) to selenium(1V) is critical to the reliable determination of selenium by atomic-absorption spectroscopy. The reduction step occurs according to the reaction Se0,2- + 2HC1+ Se0,2- + C1 + H,O 6 C D E F G H Fig.3. Selenium recovery in standard solutions from 100 ng of selenium with different methods of reduction. A 0.2 ml of 6 N HC1 at 150 "C for 5 min; B 0.2 ml of 6 N HC1 at room temperature; C 20 ml of 20% HC1 at 150 "C for 15 min; D 10 ml of 20% HCl at 150 "C for 5 min; E 5 ml of 20% HC1 a t 150" C for 15 min; F 5 ml of 20% HC1 a t 150 "C for 10 min; G 5 ml of 20% HCl at 150 "C for 5 min; and H 5 ml of 20% HC1 a t room temperature 1064 MAILER AND PRATLEY EVALUATION OF SELENIUM Analyst VoZ. 108 The need to obtain a complete reduction of selenium has been indicated.14 The conditions for our determinations were 150 "C for 10-15 min as outlined elsewhere,2 although other workers have used lower temperatures99l4 and in some instances no heating at all.4 Hydrochloric acid was provided by adding 0.2 ml of 6 N hydrochloric acid.A larger volume was found to be unnecessary and required longer to reach the temperature necessary to reduce selenium(V1) to selenium(1V). Standards did not require reduction. Nine samplings containing 100 ng of selenium and reduced by different methods resulted in readings with a mean of 0.081 ng and a range of 0.080-0.084 (Fig. 3). All samples were however, subjected to the same reduction step in order to standardise conditions. Y 0 50 150 250 Seleniuming mi-' Fig. 4. Effect of background correc-tion on absorbance. A With back-ground correction; and B with no background correction. 50 2q 10 5 Volume of HCl/ml Fig. 5. Absorbance spectra from 100ng of selenium added in various volumes of 20% m/V HCI.Determination by Atomic-absorption Spectroscdpy Background correction Although background correction was not found to be essential it resulted in more repro-ducible peaks less noise and greater accuracy. Two calibrations were obtained with and without background correction (Fig. 4). The mean values of the standards were obtained (130 ng) which would be expected to give absorbance values of 0.144 and 0.223 respectively. To test the value of the two methods, 0 5 10 20 50 Acid concentration YO V/V Fig. 6. Effect of hydrochloric acid concentration on absorbance readings. Selenium concentration, 100 ng per 20 ml September 1983 DETERMINATION IN BIOLOGICAL MATERIAL BY AAS 1065 95% tolerance limits for a selenium content determination were ca1c~lated.l~ The limits obtained were (125 135) and (98 170) respectively.The recorder trace in Fig. 1 without background correction indicates the extent of background absorbance. 0.10. 0 (II e a a 0 -Volume The volume of the sample used in the vapour generator chamber is a critical factor. An increase in volume produced a wider and flatter peak (Fig. 5). This is in agreement with the findings of Brodiell for arsenic determination but conflicts with those reported by Hobbins.lo However the findings in each instance indicate the importance of consistency of volume in obtaining reproducible results. Concentration of acid Increasing acid concentration in the solution resulted in decreased absorbance (Fig. 6). In our investigations 20% hydrochloric acid gave good reproducibility and was used for all analyses.Pellets versus sodium tetrahydro borate(III) solution (2.5%) With the use of solution it was necessary to use a fine gauge needle when penetrating the septum to prevent leakage. It has also been shown that concentrations of less than 4% can cause tailing of peaksg although a 2.5% solution is commonly used.16 Pellets were more consistent despite absorbance levels being lower in solutions of equivalent selenium concentration (Fig. 7). An additional disadvantage of the solution is its instability requiring fresh preparation after 2 h. This is an important consideration when a large number of samples are to be analysed . Consequently the injection time was slow and variable. I C D I Solution Fig. 7. Comparison of the absorbance from a sample con-taining 100 ng of selenium using sodium tetrahydroborate(II1) C and D solution and A and B pellets.Purging and stirring time Purging of the sample with nitrogen to remove all traces of oxygen and stirring prior to the addition of the pellets were necessary in order to obtain reproducible results (Figs. 8 and 9). We used a purging time of 20 s the last 5 s of which the sample was stirred as recommended by Hobbins.lo Conclusion Determining deficiencies of selenium in plant material or blood often involves samples containing 20ngml-l of selenium or less. The ability to measure these values require 1066 MAILER AND PRATLEY Stirring time/s Fig. 8. Effect of stirring time on sample absorbance with no initial purging. Selenium concentration, 100 ng per 20 ml.5 20 40 80 Purging time/s Fig. 9. Effect of purging time on sample absorbance. Selenium con-centration 100 ng per 20 ml. Stirring time 10 s. extreme accuracy cleanliness and in particular consistency in the method. The importance of acid volume and concentration and the large variations in absorbance with purging and stirring times indicate the sensitivity of the method employed. The aspects of the method outlined in this paper allows precise and accurate measurements of trace amounts of selenium. The method has the advantage of speed and simplicity provided that the precautions outlined are observed. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. References Watkinson J. H. Anal. Chem. 1966 38 92. Watkinson J. H. Anal. Chim. Ada 1979 105 319. Thompson K. C. and Thomerson D. R. Analyst 1974 99 595. Clinton 0. E. Analyst 1977 102 187. Gorsuch T. T. Analyst 1959 84 135. Kurahashi K. Inoue S. Yonekura S. Shimoishi Y. and Tbei K. Analyst 1980 105 690. Pahlavanpour B. Pullen J. H. and Thompson M. Analyst 1980 105 274. Hall R. J. and Gupta P. L. Analyst 1969 94 292. Lloyd B. Holt P. and Delves H. T. Analyst 1982 107 927. Hobbins. W. B. “Varian Instruments a t Work AA-11,” Varian AA Resource Centre Park Ridge, Brodie K. G. Am. Lab. 1979 11 58. Nhve J. Hanocq M. Molle L. and Lefebvre G. Analyst 1982 107 934. Brodie K. G. “Atomic Absorption Spectroscopy Asian Seminar Series,” Varian Techtron Pty. Ltd. Springvale Australia 1980. Brodie K. G. Am. Lab. 1977 9 73. Williams E. J . “Regression Analysis,” John Wiley New York 1979. “Model 65 Vapour Generation Accessory Operation Manual,” Publication No. 85-100338-00 Varian Illinois 1981. Techtron Pty. Ltd. Springvale Australia. Received March 141h 1983 Accepted March 31st 198

ISSN:0003-2654    

DOI:10.1039/AN9830801060

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst September 1983 Vol. 108 $9. 1067-1071 1067 Relative Fluorescence Quantum Yields Using a Com puter-controlled Luminescence Spectrometer* Alun T. Rhys Williams and Stephen A. Winfield Perkin-Elmer Ltd. Beaconsfield Buckinghamshire HP9 lQA and James N. Miller Department of Chemistry Loughborough University of Technology Loughborough Leicestershire LE11 3T U Relative fluorescence quantum yields are determined using a computer-controlled luminescence spectrometer. The relative absorbances of the standards and unknowns are measured using the same instrument as for the fluorescence measurements. Relative quantum yields are presented for a wide range of compounds a t room temperature. Keywords Relative quantum yields determination ; computer-controlled luminescence spectrometer The fraction of the number of quanta absorbed by a molecule that are emitted as fluorescence is termed the fluorescence quantum yield.Its determination provides information con-cerning radiationless processes in molecules and for example the determination of the potential of fluorophores in assays. Quantum yields are measured either on a relative basis with reference to a standard or by using an absolute method. Both methods have been reviewed by Demas and Crosbyl and by Bridges2 Absolute quantum yields were measured by Weber and Teale3 by using the dipolar scattering of monochromatic light from glycogen solutions as a standard unit of quantum yield. A comparison was then made with the fluorescence from a solution with the same apparent absorbance for the excitation light.Other methods for determining absolute quantum yields include photoacoustic spectroscopy4 and calorimetric rneth~ds.~ The most widely used method of determining quantum yields is by the relative method and the quantum yield of the unknown Qx is calculated according to the following equation: EX I R & Q - Q - - _ -Ax * E R * I ' ni X - R * where QR is the quantum yield of the standard A is the absorbance of the solution E is the corrected emission intensity I is the relative intensity of the exciting light and n is the average refractive index of the solution. Subscripts R and X refer to the reference and unknown compound respectively. When a Rhodamine quantum-corrected reference system is used the relative intensity of the excitation light at different wavelengths is taken to be unity which simplifies equation (1) to The following factors can affect the measurement of relative quantum yields polarisation; refractive index changes ; re-absorption of the emission ; internal reflection ; variation of optical density with band width ; and calibration errors between ultraviolet absorption and fluorescence spectrometers.The use of an integrating sphere6 largely overcomes the errors associated with polarisation and refractive index changes. Working at very dilute absorbances or extrapolating to zero absorbance eliminates errors from re-absorption and internal reflection. The errors associ-ated with measuring the absorbance values are based upon the difference in characteristics between an ultraviolet - visible absorption spectrometer and a fluorescence spectrometer.* Presented at the Pittsburgh Conference and Exposition on Analytical Chemistry and Applied Spectro-scopy Atlantic City N J USA March 1983 1068 RHYS WILLIAMS et al. RELATIVE QUANTUM YIELDS USING Analyst Vol. 108 Unless two identical monochromators are used errors may arise from two sources. Firstly, the effect of spectral band width on the shape of the ultraviolet absorption spectrum hence absorbance value ; the shape of the fluorescence excitation spectrum hence relative intensity. For example the ultraviolet absorption spectrum of pyrene in hexane for two different slit widths is shown in Fig. 1. This illustrates the change in absorbance values and peak ratios, which occur as the spectral band width is changed.Unless the fluorescence spectrometer has an identical monochromator the excitation spectra will also be different.' 0.8 0.6 Q C 0 D i Oa4 n a 0.2 n 520 260 300 340 380 220 260 300 340 380 Wavelengthlnm Fig. 1. Ultraviolet absorption spectra of pyrene in hexane with (a) 2-nm band width and (b) 4-nm band width. Secondly wavelength accuracy is particularly important especially where sharp absorbance/excitation bands occur. For pyrene a 0.5-nm error in the excitation wavelength at 241 nm will result in a 7% intensity error. For a l-nm error the error is 15%. Several methods have been proposed to overcome these deficiencies including time-correlated single-photon counting.* Britten et aL9 presented a method where the absorbance of a solution was determined by measuring the fluorescent intensities at two points along the absorbance path.Gains and DawsonlO used absorptivity-related constants derived from the apparent approximately hyperbolic relationship between fluorescence and concentration. In this paper a method is proposed that overcomes the errors associated with measuring absorbances by using the same instrument to measure fluorescence spectra and absorbance values. A mirror placed at the sample focus is used to reflect light into the emission mono-chromator. A 10-mm path length cuvette is placed in the reflected beam and by synchron-ously scanning the excitation and emission monochromators at the same wavelengths the instrument is turned effectively into a single-beam absorption spectrometer.Quinine sulphate is still regarded as the best available although it has a number of disadvantages such as an emission spectrum that is dependent upon the excitation wavelength and quantum yields, which vary on the type and normality of acid concentration. Velapoldi and Mielenz,ll in a National Bureau of Standards publication suggest that quinine sulphate in 0.1 M perchloric acid be used in certification measurements. A quantum yield of 0.59 is suggested at an excitation of 347 nm. When measuring quantum yields in organic solvents 9,lO-diphenylanthracene in cyclohexane has been proposed.12 The choice of primary quantum standard is particularly important. This is comparable to a value of 0.546 in 1.0 M sulphuric acid. In this solvent the quantum yield is reported as being unity.Experimental All fluorescence spectra were measured on a Perkin-Elmer Model LS-5 luminescence spectrometer fitted with a red sensitive R928 photomultiplier. Data were recorded using September 1983 A COMPUTER-CONTROLLED LUMINESCENCE SPECTROMETER 1069 Perkin-Elmer Model 3600 data station with the PECLS I1 applications software. Absorption spectra were measured on a Perkin-Elmer Model Lambda 5 spectrometer and quantum yields were measured at room temperature. The Model LS-5 gives quantum-corrected excitation spectra through the use of a Rhodamine 101 quantum-corrected reference detector. Emission spectra were uncorrected and hence a correction graph must be generated. A correction graph from 250 to 630 nm is obtained in the following manner.Over the region 250-410 nm the excitation system is used as a light source of known spectral distribution. Light from the excitation monochromator is reflected into the emission monochromator by a scatterer and the excitation and emission monochromator scanned synchronously at the same wavelength. The excitation band pass is set at 5 nm with the emission band pass set at 20 nm as recommended by Melhuish.13 200 300 400 500 600 Wavelengthlnm Fig. 2. Emission correction graph for the Model LS-5 fitted with a red sensitive photomultiplier. In order to overcome errors associated with second-order radiation and the fact that the sharp emission lines of xenon at ca. 470 nm make precise correction difficult the correction graph from 410 to 630nm is generated by comparing the emission spectrum of quinine sulphate run under identical conditions with that published by Velapoldi and Mielenz .ll Both correction graphs are merged to give the emission correction graph Fig.2. The shape of this graph is determined in part by the efficiency of the emission grating and Fig. 3 shows the grating response graph with the grating blazed for maximum efficiency at approximately 450 nm. This graph was generated by dividing the correction graph generated by scanning both monochromators synchronously at the same wavelength by that obtained by placing a plane mirror in front of the emission grating and scanning the excitation monochromator. The resultant graph does not include the effect of the photomultiplier response. The latter, also shown in Fig.3 was obtained by dividing the correction graph by the grating 100 80 % 2 60 2 -$ 40 Q) [L 0 Q) -20 200 300 400 500 600 Wavelengthinm From excitation monochromator To emission monochromator 10-nm path length cuvette Plane mirror Fig. 4. Schematic diagram of the trans-mission cell holder used for the determina-tion of absorbance values. Fig. 3. (A) Grating response graph and (B) photomultiplier response graph for the Model LS-5 1070 RHYS WILLIAMS et al. RELATIVE QUANTUM YIELDS USING Analyst VoZ. 108 response graph. All of these procedures were performed on the Model 3600 data station. The Model LS-5 luminescence spectrometer was used as a single-beam ultraviolet - visible absorption spectrometer by placing a mirror at the sample focus to reflect the excitation light into the emission monochromator.The light needs to be heavily attenuated with a wire-mesh gauze to prevent overloading of the sample photomultiplier. A 10-mm path-length cuvette was placed in the reflected beam (Fig. 4) and by synchronously scanning both monochromators at the same wavelength a transmission spectrum was obtained. By dividing a sample spectrum by the solvent spectrum the transmission spectrum of the solute was obtained. Fig. 5 compares the absorbance spectrum of pyrene in hexane as measured on the Model LS-5 with the excita-tion spectrum. Table I compares the ratios of the main peaks as measured on the ultra-violet absorption spectrometer and on the Model LS-5. Excellent agreement between the ultraviolet absorption and fluorescence excitation spectra is observed.In addition the absorbance value falls midway between the values observed on the ultraviolet spectrometer The latter was converted to an absorbance spectrum. run at 2- and 4-nm band widths. O.* I 0.6 -8 0 n 5 0.4 n a 0.2 -Wavelengthlnm (a) The absorbance spectrum and (b) the excitation spectrum of pyrene in hexane as measured on the Model LS-5. Fig. 5. Results and Discussion A series of compounds in a variety of solvents were prepared and the ultraviolet absorption was measured using the method described under Experimental. The solutions were diluted to give absorbances in the range 0.02-0.35 absorbance unit i.e. within the expected linear TABLE I COMPARISON OF PEAK HEIGHTS OF PYRENE IN HEXANE USING AN ULTRAVIOLET ABSORPTION SPECTROMETER AND LS-5 LUMINESCENCE SPECTROMETER Lambda 5 ,-A- Luminescence spectrometer Band width I A \ Absorption Fluorescence Ratio of peaks 2 nm 4 nm 2.5 nm 2.5 nm 241/273nm .* 1.733 1.851 1.786 1.779 2411335nm . . 1.585 1.632 1.632 1.63 September 1983 A COMPUTER-CONTROLLED LUMINESCENCE SPECTROMETER 107 1 calibration range of fluorescence emission ‘uemws concentration. The solutions were not de-gassed before measurement. The emission spectra were measured using a 2.5-nm band width and corrected using the correction graph in Fig. 2 and the areas under the corrected spectra calculated. The quantum yields were calculated according to equation (2) and the results by this method compared with some literature values which are given in Table 11.The reproducibility of measuring and calculating the quantum efficiency of the same solution is within *3-5%. TABLE I1 ROOM TEMPERATURE FLUORESCENCE QUANTUM YIELDS The following abbreviations are used 9,10-DPA 9,lO-diphenylanthracene ; 9-MA, 9-methylanthracene; TPB tetraphenylbutadiene; QS quinine sulphate; C cyclohexane; H hexane ; MCH methylcyclohexane ; PA photoacoustic spectroscopy ; and W water. Compound Anthracene Benzene . . Coronene Fluorene . . . . . . Pyrene _ . TPB . . . . . . . . Sodium salicylate . . . . Lucifer Yellow . . . . Fluorescein . . . . . . 9,lO-DPA . . . . . . 9-MA . . . . . . . . Solvent MCH MCH MCH MCH C C C C 1 M NaOH W 0.1 M NaOH Excitation1 nm 252 254 303 265 241 262 256 346 302 430 460 Emission rangelnm 360-510 270-360 400-500 280-400 350-550 360-540 360-580 300-500 340-550 460-630 470-620 Literature values Standard I A > QR QP QP Solvent Ref.DPA (1.00) 0.31 0.33 H 14 15 DPA (1.00) 0.03 0.05 H DPA (1.00) 0.71 0.54 H 3 DPA (1.00) 0.31 0.32 C 16 QS (0.59) 1.00 1.00 C 12 DPA (1.00) 0.42 - - -DPA (1.00) 0.84 0.86 PA 4 QS (0.59) 0.25 0.25 NaOH 17 QS (0.59) 0.20 0.24 W 18 QS (0.59) 0.82 0.84 0.1 MN~OH 5 - - DPA (1.00) 0.18 -The method proposed for calculating relative fluorescence quantum yield was found to produce results that compare favourably with those previously published. The discrepancy between the observed and literature value for fluorene is probably the result of a relatively low value observed by Weber and Teale3 because Dawson and Windsor15 observed a value of 0.68.The main advantage of this technique is the ability to measure absorbance values on a fluorescence spectrometer thus eliminating errors associated with the use of an absorption spectrometer. The use of a desk-top computer greatly simplifies the experiment and routine quantum yield measurements are easily performed by automating the control of the instru-ment and data calculation. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Demas J. N. and Crosky G. A. J . Phys. Chem. 1971 75 991. Bridges J . W. in Miller J. N. Editor “Standards in Fluorescence Spectroscopy” Chapman and Weber G.and Teale F. W. J. Trans. Faraday Soc. 1957 53 646. Adams M. J. Highfield J. G. and Kirkbright G. F. Anal. Chem. 1980 52 1260. Olmsted J. J . Phys. Chem. 1979 83 2581. Ware W. R. and Rothman W. Chern. Phys. Lett. 1976 39 449. Bendig J. Kreysig D. and Schoneich R. Opt. Spectrosc. USSR 1980 49 29. Upton L. M. and Cline Love L. J. Anal. Chem. 1979 51 1941. Britten A. Archer-Hall J. and Lockwood G. Analyst 1978 103 928. Gains N. and Dawson A. P. Analyst 1979 105 481. Velapoldi R. A. and Mielenz K. D. Natl. Bur. Stand. Spec. Publ. 1980 No. 260. Heinrich G. Schoof S. and Gasten H. J . Photochem. 1974/75 3 315. Melhuish W. H. J . Res. Natl. Bur. Stand. Sect. A 1972 76 547. Guilbault G. G. “Practical Fluorescence Theory Methods and Techniques,” Marcel Dekker New Dawson W. R. and Windsor M. W. J . Phys. Chem. 1968 72 3251. Berlman I. B. “Handbook of Fluorescence Spectra of Aromatic Molecules,’’ Academic Press New Inokuchi H. Harada Y. and Kondow T. J . Opt. SOG. Am. 1964 54 842. Stewart W. W. Nature (London) 1981 242 17. Hall London 1981 p. 68. York 1973. York 1965. Received March 4th 1983 Accepted March 28th 198

ISSN:0003-2654    

DOI:10.1039/AN9830801067

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

1072 Analyst September 1983 Vol. 108 pp. 1072-1081 Studies of Calcium Ion-selective Electrodes in the Presence of Anionic Surfactants Anthony J. Frend Gwilym J. Moody and J. D. R. Thomas Afifilied Chemistry De#artment UWIST Cardifl CFl 3XA and Brian J. Birch Unilever Research Laboratory Bebington Wirral Merseyside L62 4XN Eighteen membrane systems based on a calcium bis{di[4-( 1,1,3,3-tetra-methylbutyl)phenyl]phosphate} ion sensor have been compared in potentio-metric studies with a membrane based on the sensor with a dioctyl phenyl-phosphonate solvent mediator with respect to interferences of calcium ion-selective electrodes by anionic surfactants especially by sodium dodecyl-sulphate (SDS) and sodium tetradecylbenzenesulphonates (ABS) . Electrodes made from poly(viny1 chloride) matrix membranes of the sensor with trioctyl phosphate solvent mediator are far superior to the other membrane systems in resisting interference by anionic surfactants and are shown to yield calcium ion levels in the presence of a wash liquor to match the expected values.Some improvement over dioctyl phenylphosphonate is also offered by decan-1-01 dodecan-1-01 and tetradecan-1-01 but the use of such solvent mediators impairs calcium ion selectivity and poly (vinyl chloride) plasticising qualities. The use of alternative polymer matrices based on poly(viny1idene chloride) and VAGH copolymer (hydrolysed vinyl chloride - vinyl acetate) offer no advantages. A nineteenth membrane obtained from a commercial supplier exhibited interference by SDS. An interesting effect is the increase (rather than the normal decrease) in e.m.f.observed for those electrodes from membranes in which the amount of free active sensor was reduced to low levels. By using optimum levels of sensor it was possible to fabricate electrodes exhibiting a zero e.m.f. change when changing the background 1 0 - 2 ~ calcium chloride to one that is also 1 0 - 3 ~ in SDS. Such membranes are based on decyl phosphate grafted to VAGH copolymer. X-ray fluorescence and chromatographic studies on membranes of the sensor with dioctyl phenylphosphonate in poly(viny1 chloride) show SDS to be a significant agent in leaching membrane components especially of dioctyl phenylphosphonate. Such observations are indicative that the interference of calcium ion-selective electrodes by anionic surfactants may be the result of the different solubilities of calcium - surfactant complexes in the solvent mediator of the membrane.Keywords Calcium ion-selective electrodes ; anionic surfactant interference In the absence of calcium ions the Orion 92-20 liquid membrane calcium ion-selective electrode (ISE) responds to anionic surfactants over limited concentration ranges with an anionic slope.1$2 Llenado observed that when this electrode was immersed in a solution containing calcium ions and a small aliquot of Sulframine (a linear-chain alkylbenzene sulphonate mixture with Cll 42.5%; C,, 34.6%; C,, 13.8%; and <C9 and >CIS 9.1% in the alkyl chain) was added there was a rapid negative shift in e.m.f. This same e.m.f. shift has also been observed for poly(viny1 chloride) matrix membrane calcium ISES.~ This work is concerned with further studies of this effect and of possible means of reducing the interference.Experimental Electrodes Poly(viny1 chloride) and other polymer matrix membrane electrodes with inner solution FREND MOODY THOMAS AND BIRCH 1073 of 10-1 M calcium chloride were assembled by the general procedure described previou~ly4~~ and the membrane compositions and electrode characteristics are summarised in Table I. Membrane No. I I1 I11 IV 17 VI VI I VIII IX X XI XI1 XI11 XIV xv XVI XVII$$ XVIII$$ XIX TABLE I COMPOSITION OF MASTER MEMBRANES AND ELECTRODE CHARACTERISTICS Composition* Slope at Limit of 25 "C/mV detectiont/ decade-1 M x 0.36 g DOPP$ + 0.17 g PVCY .. . . 0.36 g decan-1-01 + 0.17 g PVC . . . . 0.36 g dodecan-1-01 + 0.17 g PVC. . . . 0.36 g tetradecan-1-01 + 0.17 g PVC . . . . 0.36 g hexadecan-1-01 + 0.17 g PVC . . 0.36 g trioctyl phosphate + 0.17 g PVC . 0.36 g tripentyl phosphate + 0.17 g PVC . . 0.36 g dioctylphosphite + 0.17 g PVC . . 0.18 g decan-1-01 + 0.18 g DOPP + 0.17 g PVC . . 0.18 g dodecan-1-01 + 0.18 g DOPP + 0.17 g PVC . . 0.18 g tetradecan-1-01 + 0.18 g DOPP + 0.17 g PVC . . 0.18 g hexadecan-1-01 + 0.18 g DOPP + 0.17 g PVC . . 0.36 g DOPP + 0.085 g poly(viny1idene chloride) + 0.085g PVC . . . . . . (vinylidene chloride) + 0.085 g PVC . . 0.085gPVC phosphoric acid (VAGHPl) + 0.17 g PVC . . 0.36 g DOPP + 0.17 g PVC .. Philips IS 561/SP Ca2+ membrane. . . . . . 0.36 g DOPP 4- 0.085 g VAGH + 0.085 g PVC . . . . 0.18g DOPP + 0.18g decan-1-01 + 0.085g poly-0.18g DOPP + 0.18g decan-1-01 + 0.85g VAGH + 0.36 g DOPP + 0.01 g VAGH grafted with monodecyl-30.9 27.5 26.5 28.1 29.7 30.1 31.0tt 33.4 22.0 31.8 28.8 27.3 35.6 32.1 30.2 26.2 30.4 (A) 27.8 -1.8 3.0 5.0 5.2 9.5 7.5 6.0t t 8.5 8.0 4.0 4.5 7.2 3.1 4.2 5.1 3.2 6.0 (A) 2.0 -kE:Ns: 2.411 0.25 0.61 0.53 0.55 0.043** 0.021tt $1 1.4 0.46 0.42 0.43 0.21 0.31 0.48 0.52 1.2 (A) 0.31 -* Including 0.04 g of calcium bis(di [4-( 1,1,3,3-tetramethylbutyl)phenyl]phosphate}. t Determined using calcium ion buffers based on STP (see Table 11).$ "a+] = 5 x 10-3M. DOPP = dioctyl phenylphosphonate. 7 PVC = poly(viny1 chloride). 1) Data from reference 6. ** "a+] = 0 . 0 5 ~ . t t Data from reference 7. $ $ too low for measurement. $5 Details of VAGHPl in reference 8. The 6 membranes (A-F) prepared for each of the compositions XVII and XVIII each contained 0 (A) 0.001 (B) 0.005 (C) 0.01 (D) 0.015 (E) and 0.02 (F) g of calcium bis {di [4-( 1,1,3,3-tetrarnethylbutyl)phenyl]phosphate} respectively. The master membranes were cast in the normal way4s5 except that those of the highest alcohols (membranes III-V) were redissolved in tetrahydrofuran and recast in order to obtain membranes of homogeneous composition. Poly(viny1idene chloride) or VAGH co-polymer alone that is without the additional support of poly(viny1 chloride) as in membranes XIII-XVI gave membranes of insufficient mechanical strength when used with normal proportions of dioctyl phenylphosphonate solvent mediator while smaller amounts of the phosphonate gave membranes of gel-like consistency Such matrix membranes of poly-(vinylidene chloride) or VAGH copolymer were therefore not evaluated in this study.Electrodes were conditioned overnight in 10-1 M calcium chloride solution and calibrated with serially diluted calcium chloride solutions in the 10-1-10-5 M range and at below M (0.1 M ionic strength) with sodium tripolyphosphate (STP) buffers (Table 11) set at pH 9 1074 FREND et al. STUDIES OF CALCIUM ION-SELECTIVE Analyst Vol. 108 TABLE I1 COMPOSITION OF STP-BASED (1.472 g 1-1) CALCIUM ION BUFFER SOLUTIONS OF 0.1 M IONIC STRENGTH The pH was adjusted to 9.0 with ammonia solution.[NaCl]/g 1-1 Total [Ca2+]/~ Free [Ca2+]/~ Ca2+ activity/M 3.57 5 x 10-4 1.02 x 10-6 4.08 x 10-7 3.62 1.0 x 10-3 2.41 x 9.64 x 3.68 1.5 x 10-3 4.37 x 10-6 1.70 x 3.74 2.0 x 10-3 7.31 x 2.92 x 3.85 3.0 x 10-3 2.09 x 8.36 x 3.80 2.5 x 10-3 1.21 x 10-5 4.84 x 10-6 Reagents and Materials All materials except the following were of the best analytical grade available. Calcium bis{ di [4- (1,1,3,3-tetramethylb~tyl)phenyl]phosphate) ,9 which has the formula Ca{ [CH3.C( CH,) ,.C,H,.O] 2.P( 0) 0) , and dioctyl phenylphosphonate,1° which has the formula C,H,.P(O) [O(CH,),CH,], were prepared as described previously. Dioctyl phenylphosphonate is also listed by Lancaster Synthesis and in the Alfred Bader Rare Chemicals List of the Aldrich Chemical Company.VAGH copolymer (a partially hydrolysed vinyl chloride - vinyl acetate copolymer of relative molecular mass ca. 23000) from Union Carbide (UK) Ltd. and poly(viny1idene chloride) homopolymer from ICI Ltd. were purified by first dissolving in tetrahydrof uran. The polymer was then precipitated by the dropwise addition of the solution into methanol. The separated precipitate was homogenised in a household homogeniser filled with de-ionised water. The powdered polymer was then filtered and dried at 60 “C under vacuum. Specially pure sodium dodecylsulphate (SDS) was obtained from BDH Chemicals Ltd., while sodium tetradecylbenzenesulphonate (ABS) was obtained from Unilever Research Laboratory.A “model soap powder’’ with the following component composition (yo mass in parentheses) was prepared (using water to bring the total of components to 100%) ABS (7.0) ; tallow soap (2.0) ; sodium silicate (10.0) ; sodium sulphate (10.0) ; STP (35.0) ; and sodium perborate (25.0). Procedures Efect of anionic swfactants on calcium ISEs The various electrodes (Table I) were tested for interference to SDS and ABS as appro-priate in conjunction with an Orion Model 90-02-00 double-junction reference electrode. The e.m.f. measurements were made with a Corning EEL Model 112 digital millivoltmeter -pH meter used in conjunction with a Servoscribe Model RE 4541 potentiometric chart recorder. The various test solutions were maintained at 25 & 0.1 “C. Normally the effect of added surfactant was examined in four background solutions, namely lo- and lo- M calcium chloride solutions 10-1 M sodium chloride solution and a calcium ion buffer solution formulated to give a calcium(I1) concentration of 1.70 x lo- M (pCa = 5.77) (Table 11).M) were added to the particular background solution (25 cm3) in which the calcium reference electrode pair had been equilibrated to a steady response. E.m.f. readings were noted for each aliquot added and further aliquots added until the background solution had reached a 10-3~ con-centration in surfactant. Each full run was normally performed six times by using a fresh membrane from the parent master membrane for each run. Except for electrodes from membranes 111 IV VI and XI calibrations for calcium ions following exposure to calcium ions were not possible; electrodes from membrane VI were the most robust and re-calibration with calcium gave responses near to those existing prior to contact with the surfactant.Thus aliquots (0.05 cm3) of surfactant solution (5 x Efect of anionic surfactants on re ference-type electrodes In checks on the effects of anionic surfactants on reference-type electrodes the Orion, Model 94-07 fluoride ISE was used as reference in conjunction with the electrode under study namely the EIL Model 1070030 pH electrode Orion Model 90-02-00 double September 1983 ELECTRODES IN THE PRESENCE OF ANIONIC SURFACTANTS 1075 junction reference electrode and a silver - silver chloride reference electrode from an EDT Research ISE body. An electrode made from membrane I (Table I) was used as the control.The e.m.f.s were monitored with an Orion Model 901 millivoltmeter fitted with two high impedance inputs in conjunction with a potentiometric chart recorder. The background solutions to which aliquots (0.05 cm3) of anionic surfactant solutions (5 x 1 0 - 2 ~ ) were added were pH 7 buffered (B) (solutions prepared in Radiometer Type S1326 pH 7.00 buffer solution) and unbuffered (A) potassium fluoride solutions M) that were also 10-5 M in potassium chloride. Analysis of calcium ions in detergent solutions The previously calibrated calcium ISE/Orion 90-02-00 double junction electrode pair were placed in a solution (25 cm3) of model soap powder of selected concentration. The solution was spiked with aliquots (0.025 cm3) of 1 M calcium chloride until the concentration of calcium chloride added was 5 x 1 0 - 3 ~ .Equilibrated e.m.f.s were noted for the various spikes. The calcium ion calibrations were compared with the calculated values by allowing for the effect of the model soap powder components. Mechanistic Experiments X-ray fluorescence studies Master membranes of membrane I composition (Table I) were cut into quadrants and each quadrant recast to full size. The resulting thin membranes were suspended for 48 h in appropriate stirred liquids (25 cm3) namely doubly de-ionised water and SDS solutions (5 x After rinsing thoroughly with de-ionised water and air-drying the surfaces of untreated and treated membranes were examined with a Model TN 2000 Tracer Europa energy-dispersive X-ray fluorescence spectrophotometer (Unilever Research).Additional master membranes were first suspended in 10-1 M SDS for 48 h and similarly examined. The membrane compositions were as follows (a) calcium bis(di[4-(1,1,3,3-tetramethylbutyl)phenyl]phosphate} (0.04 g) 2-nitrophenyl phenyl ether (Alfred Bader Library of Rare Chemicals of Aldrich Chemical Company) (0.36 g) and poly(viny1 chloride) (0.17 g); and (b) dioctyl phenylphosphonate (0.36 g) and poly(viny1 chloride) (0.17 g). Confirmatory thin-layer chromatography (TLC) experiments for leached components were carried out on the above suspension solutions following evaporation to dryness and dissolution of the residue in tetrahydrofuran (1 cm3). Chromatograms of the tetrahydrofuran extracts were run on silica gel with benzene - acetic acid (9 + 1) and detected with iodine vapour.and 10-1 M). Gas chromatogyaphy A conditioned electrode of membrane I (Table I) with a double-junction reference electrode were equilibrated to constant e.m.f. in 10 cm3 calcium chloride solution M). The solution was spiked with 0.02 cm3 SDS (5 x M) and when the e.m.f. had again stabilised (after ca. 5 min) the electrode pair was removed. The aqueous solution was extracted with dichloromethane (10 cm3) and the dichloromethane extract reduced in volume (to 2 cm2) by vacuum evaporation. This was chromatographed in a Perkin-Elmer Model F11 gas chromatograph fitted with a Carbowax 20M column using a flow-rate of 20 cm3 min-l at a temperature of 130 "C and the instrument injection temperature setting at No.8. The output was recorded on a Servoscribe Model 4541 potentiometric recorder with a sensitivity of 100 mV f.s.d. and a chart speed of 300 mm h-l. Standards of dioctyl phenylphosphonate and SDS were prepared in dichloromethane and similarly chromatographed. Results and Discussion The responses of calcium ISEs based on a calcium bis [di(4-octylphenyl)phosphate] sensor with a dioctyl phenylphosphonate mediator in poly(viny1 chloride) to added surfactants have already been d~cumented.~ The general effect of added anionic surfactant even at about 2 x M is to lower the e.m.f. response of the calcium ISEs when in contact with the various test solutions regardless of whether or not calcium is present in the cell test solution.s Similar behaviour has been observed in the present studies for calcium ISEs with a sensor o 1076 FREND et al.STUDIES OF CALCIUM ION-SELECTIVE Analyst Vol. 108 an isomer of calcium bis [di(4-octylphenyl)phosphate] namely calcium bis(di [4-( 1,1,3,3-tetramethylbutyl)phenyl]phosphate) (membrane I in Table I). Calcium ISEs based on either of these sensors behave ~imilarly,~Jl and these anionic surfactant interferences are undesirable. In these studies aimed at designing an electrode to diminish the undesirable effects, particular attention has been given to SDS but with some reference also to ABS. It is convenient to express the results of the investigation and their discussion under three main headings namely the effect of anionic surfactants on calcium ISEs with membrane compo-sitions as in Table I the effects of anionic surfactants on reference type electrodes and mechanistic studies based on X-ray fluorescence and chromatographic experiments.Effect of Anionic Surfactants on Calcium ISEs The membrane compositions summarised in Table I have been designed to demonstrate the extent of anionic surfactant interference on calcium ISEs based on various solvent mediators and polymer matrix compositions. Experiments on membranes XVII and XVIII were designed to determine any role by the calcium ion sensor itself. In all instances it is convenient to express the interferences in terms of the changes in e.m.f. on bringing the various background solutions up to M in anionic surfactants (Tables I11 and IV) a level approximating to that normally used in practice.Calcium ISEs with diferent solvent mediators Table I11 summarises the changes in e.m.f. for the respective electrodes according to membrane type and background solutions for added SDS and ABS. In general these show the interferences for the M calcium chloride solutions to be similar for each type of membrane for added SDS and is greatest for an electrode made from a commercial calcium ISE membrane (membrane XIX). E.m.f. changes for 10-1 M sodium chloride solutions are generally less pronounced and slightly less so again in the STP buffer system. The striking feature for added ABS is the considerably more pronounced e.m.f. changes for the 1 0 - 4 ~ and TABLE I11 ADDED SURFACTANT INTERFERENCES FOR CELLS WITH CALCIUM ELECTRODES IN VARIOUS SOLUTIONS AE (s.d.for n = 6) caused by 10-3 M surfactant in solutions/mV Membrane /- A . No. lo- M CaC1, SDS surfactant data-I -70 (3.4) I1 -24 (2.1) I11 -0.4 (0.1) IV -1 (0.5) V -45 (5.0) VI -3 (0.7) VI I -61 (3.6) VIII -75 (2.9) I X -30 (1.4) X -24 (3.6) XI -3 (0.3) XI1 -58 (3.8) XI11 -76 (4.0) XIV -81 (3.1) xv -86 (6.1) XVI -34 (3.1) XIX -113 (5.2) ABS surfactant data-I -65 (3.0) I11 -1 (1.1) VI -2 (1.0) XI -3 (0.1) X -21 (6.1) XI1 -60 (2.1) M CaC1, - 68 (4.1) -25 (3.0) -3 (1.7) -50 (3.7) -2 (0.4) -65 (2.9) -65 (6.1) -40 (2.1) -27 (2.9) -52 (2.9) -79 (3.6) -82 (5.2) -44 (2.7) -70 (3.6) -2 (1.0) -2 (1.2) -95 (3.9) -91 (4.0) -72 (2.1) -61 (1.7) -124 (5.1) -99 (6.1) -94 (4.7) 10-1 M NaCl -40 (2.5) -10 (1.3) -3 (2.5) -32 (2.1) -34 (1.7) -36 (3.1) -12 (1.7) -8 (1.3) -2 (0.4) -59 (1.5) -50 (2.9) -60 (3.8) -38 (3.1) -2 (1.1) -3 (0.9) -35 (3.7) --43 (2.1) -10 (1.4) -3 (0.5) -18 (3.1) -15 (2.1) -42 (3.1) STP buffer system' -32 (6.9) -2 (1.5) -2 (0.6) -20 (3.0) -28 (4.1) -44 (5.1) -15 (2.1) -8 (2.0) -29 (3.1) - 11 (2.0) -3 (1.0) -1 (1.0) ------23 (2.1) -11 (2.3) -2 (0.3) -13 (2.9) -7 (1.5) -20 (2.9 September 1983 ELECTRODES IN THE PRESENCE OF ANIONIC SURFACTANTS 1077 calcium chloride solutions compared with those for the This is a reflection of the considerably increased interference occurring at ca.3.5 x 1 0 - 4 ~ ABS level and observed for all membranes studied in a Such increased interference led to S-shaped plots for e.m.f.ve~sus [ABS]. The replacement of dioctyl phenylphosphonate solvent mediator with C1,,-Cl4 alkan-1-01s (membranes 11-IV) significantly reduces SDS interferences with improvement as the alkyl chain is lengthened (Table 111). Also there are significant transient increases in e.m.f. for the c1&6 alkan-1-01s (membranes 11-V) with associated lengthening of response times for reaching the equilibrium e.m.f. The reduced interferences for the alkan-1-01 mediator electrodes and previously noted for decan-l-ol,S suggest that the interference effect may be due to the different solubilities of calcium - surfactant complexes in the solvent mediator of the membrane. It is also possible that the interference is related to the leaching of the solvent mediator from the membrane surface by the surfactant in solution.The supporting evidence for this view is the stiffening of membrane I on soaking in SDS solution thus indicating the leaching of plasticising dioctyl phenylphosphonate from the membrane and is discussed here with the mechanistic data. Of considerable significance is the relatively small interferences shown by membrane VI for the circumstances of Table 111 that is except for M calcium chloride solution with added ABS. This is a considerable gain over the previously reported3 reduction of inter-ference by the partial replacement of dioctyl phenylphosphonate with decan-1-01 especially as the replacement of dioctyl phenylphosphonate by decan-1-01 is at the expense of lost calcium ion selectivity.6 Erosion of selectivity is less likely for the single solvent mediator of membrane VI and previous studies’ have shown the trioctyl phosphate solvent mediator to be compatible with good calcium over inorganic ion selectivity.The relatively small surfactant interference of membrane XI does not have the assured supportive quality of normal calcium ion selectivity nor of a single solvent mediator. M solutions. M calcium chloride background. Such “transients” are not observed at 35 “C. Calcium ISEs with diferent polymer matrices The data for membranes I and XIII-XVI (Table 111) indicate that the partial replacement of the poly(viny1 chloride) polymer matrix with either VAGH or poly(viny1idene chloride) does not offer any improvement in electrode behaviour.Calcium ISEs with diferent levels of sensor This series of experiments was based on membranes XVII and XVIII for SDS added to The notable effect here is that the level of calcium ion sensor has an effect on the level of interference by SDS to the extent that for low levels of free sensor-grafted sensor ratios (membranes XVII) and of free sensor to dioctyl phenyl-phosphonate ratios (membranes XVIII) the interference is manifested by increases in e.m.f. on adding M calcium chloride (Table IV). M SDS to the background M calcium chloride solution. TABLE IV ADDED SDS ( M) INTERFERENCES (e.m.f CHANGES) FOR CELLS WITH CALCIUM ELECTRODES I N M CALCIUM CHLORIDE Membrane type (see footnote $5 in Table I) A B C D E F A r \ E.m.f. changes for membranes XVII/mV .. . . 45 14 -16 -37 -51 -63 E.m.f. changes for membranes XVIII/mV . . . . 14 10 -46 -64 -68 -68 Based on the data for membranes XVII it was possible to design an electrode with the optimum free sensor to grafted sensor ratio. For this the experimental change in e.m.f. on making the background 1 0 - 2 ~ calcium chloride solution equivalent to 1 0 - 3 ~ in SDS was +1 mV. The null point for the system of membranes XVIII occurs at a free sensor to dioctyl phenylphosphonate ratio of 6 x but such a membrane is more difficult to reproduce and the electrodes from the membranes so fabricated gave experimental change 1078 FREND et d. STUDIES OF CALCIUM ION-SELECTIVE AndJJSt VOZ. 108 in e.m.f. of +3 to -4mV on making the background 1 0 - 2 ~ calcium chloride up to 1 0 - 3 ~ in SDS.Membrane XVII(A) is based on a grafted decyl phosphate sensor and matches the calcium ISEs based on free calcium bis(didecy1phosphate) with dioctyl phenylphosphonate in poly-(vinyl chloride)8J2 in general ISE performance. Because the grafted and ungrafted sensors give electrodes of similar lifetimess9l2 there is no advantage obtained by the long synthetic procedure in producing the grafted sensor membrane. Unfortunately calcium (mono [4-(1,1,3,3-tetramethylbutyl)phenyl]phosphate ] grafted to VAGH copolymer gives electrodes of poor calcium over sodium selectivity.8 TABLE V ADDED SURFACTANT RESPONSES FOR CELLS WITH REFERENCE TYPE ELECTRODES veYsm AN ORION 94-07 FLUORIDE ELECTRODE E.m.f. changes caused by M surfactant in solutions A and B/mV Orion 90-02-00 r A 7 Orion pH double-junction Silver - silver electrode electrode chloride electrode 7-7 * * Surfactant added A* Bt A B A B Sodium butylsulphate (SBS) .. . . 0 0 1 -0.7 9 5 Sodium dodecylsulphonate (SDS) . . 3 2 12 1.0 14 -49 Sodium tetradecylbenzene sulphonate (ABS) . . 2 4 35 0 25 -21 Sodium tetradecylsulphate (STS) . . . 20 -0.5 0 0 6.0 0.5 * A 10-3 M potassium fluoride + 10-5 M potassium chloride (unbuffered). t B as for A but buffered to pH 7. Effects of Anionic Surfactants on Reference Type Electrodes It is not easy to draw conclusions from the data on e.m.f. changes of various reference type electrodes used in conjunction with the Orion Model 94-07 fluoride ISE brought about by 10-3 M surfactant in solutions A and B (Table V).However the Orion Model 90-02-00, double-junction electrode used as reference in this work is insensitive to added surfactant in the pH buffered solutions B that is assuming that the fluoride electrode is also insensitive and does not suffer from a compensating interference. The responses in the unbuffered solutions are not as reassuring. Analysis of Calcium Ions in Detergent Solutions Calcium ion determination was carried out by spiking calcium chloride into solutions of model soap powder at various levels of added soap powder. The calcium ion levels deter-mined using calcium ISEs with selected membranes (I VI VIII IX X and XI) for 1 2 3, 4 and 5 mM of added calcium chloride are compared with calculated values13 in Table VI. TABLE VI FREE Ca2+ ACTIVITIES IN SIMULATED WASH LIQUORS FOUND BY CALCIUM ELECTRODES OF VARIOUS MEMBRANES Experimental aca8+ (s.d.for n = 6) determined with electrodes of various membranes M I* VI+ VIII* IX* X* XI* M aCa'+ (calculated)/ [Caz+] added] r- -3 g 1-1 model soap powder added-1 1.6 (1.5) x 10-77 1.2 (0.6) x lo-' 1.9 (1.3) x lO-?t 2.3 (0.4) X 3.0 (0.6) X lo-?? 4.1 (1.1) x lo-?? 1.1 x lo-" 2 4.5 (2.1) x lo-?? 3.6 (1.3) x lo-' 5.9 (2.0) x 10-lt 6.9 (0.9) x lo-'? 2.0 (0.9) x 2.5 (1.5) x 3.6 x 3 6.0 (1.0) x l O - 7 t 3.6 (1.3) x 8.3 (2.7) x lO-?t 4.1 (0.7) x 9.0 (1.5) X 1.1 (0.3) x 4.3 x 4 1.2 (0.6) x 10-6 1.0 (0.4) x 2.1 (0.6) x 10-6 1.0 (0.6) x lo-'. 2.3 (0.8) x 3.1 (1.0) x 1.6 x lo-' 5 1.7 (0.8) x 2.1 (0.6) x lo-' 3.1 (0.9) x lo-" 1.9 (0.8) x lo-* 4.6 (1.6) x 5.9 (1.1) x 1.9 x 5 g 1-1 model soap powder added-1 1.0 (0.4) x 10-7t 2.7 (1.0) x 10-0 1.1 (0.6) x 10-7t 4.9 (1.9) x 10-7t 1.8 (0.5) x 10-7t 2.6 (0.8) x 10-7t 8.9 x 10-7 2 1.5 (0.7) x 10-7t 3.5 10.51 x 10-8 2.1 10.71 x 10-7t 7.2 12.51 x 10-7't 4.4 (0.9) x 10-7t 4.9 (1.01 x 10-7t 2.5 x 10-0 3 6.9 (i.6j x 10-7+ 4.9 (0.4j x 10-6 6.5 (i.7j x lo-'+ 1.0 (0.2j x 10-6' 1.0 (0.3j x 10-6' 1.6 (0.5j x 10-6' 5.4 x 10-6 4 8.7 (0.4) x lo-'? 8.7 (2.4) x 10-6 9.0 (2.7) x lo-?? 1.9 (0.1) x lo-' 2.5 (0.9) X 3.4 (0.9) x 1.4 x lo-& 5 1.1 (0.6) x 10-o 2.0 (0.8) x 1.5 (0.6) x lo-" 5.3 (0.9) x 6.1 (0.2) x 10-o 7.5 (2.0) X lo-" 7.7 x Figures in parentheses are s.d.(n = 6). t Determined by extrapolation of calibration line September 1983 ELECTRODES IN THE PRESENCE OF ANIONIC SURFACTANTS 1079 Electrodes based on membranes I and VIII perform poorly in the simulated wash liquors.Improved performance is observed when the membranes contain mixtures of dioctyl phenyl-phosphonate with decan-1-01 (IX) dodecan-1-01 (X) and tetradecan-1-01 (XI). However the outstanding membrane is the one based on the trioctyl phosphate solvent mediator thus confirming the superiority observed in the solvent mediator parameter studies as reported under Results and Discussion. The electrode gave measured calcium ion levels close to the calculated values especially at the 3 g 1-1 level of powder. Mechanistic Studies The various experiments concerning interferences of calcium ISEs by anionic surfactants indicate apart from any other factors that the solvent mediator is significantly involved.Further Oesch and Simon1* have indicated that the loss of solvent mediator and/or electro-active species from the polymeric membrane phase into solution determines the lifetimes of electrodes. In view of the important role of dioctyl phenylphosphonate solvent mediator in imparting calcium ion selectivity when used with organophosphate calcium ion sensors it is pertinent to determine whether the anionic surfactant interference may be aggravated by the loss of these membrane components. This has been attempted here by X-ray fluorescence and chromatographic studies. (A) Expansion I h (B) Expansion (C) Expansion I I P CI Ca Energy+ Fig. 1. XRF spectra of recast Membrane I quadrants leached by A de-ionised water; B, 5 x 10-3 M SDS; and C 10-1 M SDS 1080 FREND et at.STUDIES OF CALCIUM ION-SELECTIVE Analyst VOl. 108 X-ray fluorescence (XRF) The XRF spectra of recast membrane I quandrants subjected to leaching by de-ionised water 5 x Fig. 2 shows the XRF spectra of specially fabricated membranes of (a) calcium sensor + 2-nitro-phenyl phenyl ether + poly(viny1 chloride) and (b) dioctyl phenylphosphonate + poly(viny1 chloride) after exposure to de-ionised water and to 10-1 M SDS. Fig. 1 shows a progressive decrease in the phosphorus content of the membrane surface in passing from (A) to (C). It is only in (C) that the calcium peak is decreased significantly. Fig. 2 confirms these observations. Considering the spectra together in relation to the normal levels of surfactant in solution (10-3~) in the main experiments above it can be deduced that the leaching of the solvent mediator is likely to be significant.However it has to be noted that the treatments in the XRF studies have been severe and prolonged. Neverthe-less a change in the dioctyl phenylphosphonate concentration at the membrane surface can be significant in altering the boundary potential. The thin-layer chromatography experiments on leached residues from membrane I confirm that the surfactant promotes the leaching of dioctyl phenylphosphonate (RF = 0.83) and sensor (R = 0.21) from the membrane at 10-1 M SDS. Under the milder conditions pro-vided by 5 x M SDS and 10-1 M SDS are shown in Fig. 1 (A Band C respectively). M SDS only dioctyl phenylphosphate can be detected in the leachate. P CI Ca P CI Fig.2. XRF spectra of specially fabricated membranes of (a) calcium bis{di [-4- ( 1 1,3,3-tetramethylbutyl) phenyllphosphate} and 2-nitrophenyl phenyl ether in poly(viny1 chloride) and (b) dioctyl phenylphosphonate in poly(viny1 chloride) after exposure to (1) de-ionised water and (2) 10-1 M SDS. Energy ___+ Gas chromatography studies These were directed to determining leaching of dioctyl phenylphosphonate from membrane I under normal interference conditions rather than the prolonged exposures of the XRF and TLC studies. Standard dioctyl phenylphosphonate yielded two gas - liquid chromatographic peaks with retention times of 5.0 and 12.0 min respectively. Peaks of the same retention times were obtained for the calcium - SDS extract as described under Experimental but no such peaks were observed when SDS was absent.Such peaks confirm the synergistic effect of SDS in leaching dioctyl phenylphosphonate. Conclusion This work shows the composition of the poly(viny1 chloride) matrix membrane system organophosphate-based calcium ISEs to be critical when anionic surfactants are present. The studies directed to the normal calcium ISEs based on membrane I that is the calcium bis{ di [a-( 1,1,3,3-tetramethylbutyljphenyl]phosphate) sensor with dioctyl phenylphosphonate in poly(viny1 chloride) indicates that the presence of anionic surfactant at the M leve September 1983 ELECTRODES IN THE PRESENCE OF ANIONIC SURFACTANTS 1081 has a detrimental effect on membrane components not otherwise observed in the absence of surf act ant.An important outcome of the potentiometric studies is the considerably reduced anionic surfactant interferences by calcium ISEs based on calcium bis{ di[4-(l11,3,3-tetramethyl-butyl)phenyl]phosphate) with trioctyl phosphate as solvent mediator (membrane VI). This can mark a significant step foward in the determination of free calcium ions in the presence of anionic surfactants. The authors thank the Science and Engineering Research Council for a studentship (to A. J.F.) under the scheme for Co-operative Awards in Science and Engineering in conjunction with Unilever Research Laboratory Port Sunlight; Mr. R. Lee of Unilever Research Labora-tories for obtaining the XRF spectra; and Union Carbide (UK) and ICI Ltd. are thanked for their gifts of VAGH copolymer and poly(viny1idene chloride) homopolymer respectively. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References Birch B. J. and Clarke D. E. Anal. Chim. Acta 1973 67 387. Llenado R. A. Anal. Chem. 1975 47 2243. Craggs A. Moody G. J. Thomas J. D. R. and Birch B. J. Analyst 1980 105 426. Moody G. J. Oke R. B. and Thomas J . D. R. Analyst 1970 95 910. Craggs A. Moody G. J. and Thomas J . D. R. J . Chem. Educ. 1974 51 541. 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. R. Analyst 1978 103 68. Hobby P. C. Moody G. J. and Thomas J . D. R. Analyst 1983 108 581. Craggs A. Delduca P. G. Keil L. Key B. J. Moody G. J. and Thomas J. D. R. J . Inorg. Nucl. Craggs A. Delduca P. G. Keil L. Moody G. J. and Thomas J. D. R. J . Inorg. Nucl. Chem., Moody G. J. and Thomas J. D. R. Ion-Sel. Electrode Rev. 1979 1 3. Keil L. Moody G. J. and Thomas J. D. R. AnaZyst 1977 102 274. Birch B. J . personal communication. Oesch U. and Simon W. Anal. Chem. 1980 52 692. Chem. 1978 40 1483. 1978 40 1943. Received January 26th 1983 Accepted April 15th 198

ISSN:0003-2654    

DOI:10.1039/AN9830801072

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

1082 Analyst September 1983 Vol. 108 PP. 1082-1085 Voltammetric Study of the Mercury Dissolution Reaction Mechanism at Solid Electrodes Paul Kiekens," Marc Mertens Charles Laire and Edward Temmerman Labovatory for Analytical Chemistry Ghent University Krijgslaan 28 1 S-12 B-9000 Ghent Belgium Stripping voltammetry with collection a t a rotating platinum ring - glassy carbon disc electrode was used for the elucidation of the mechanism of the electrochemical dissolution of a mercury film from a solid electrode. In complexing electrolytes the dissolution of mercury gives rise to the formation of divalent mercury ions. In non-complexing electrolytes the initial product of the electrodissolution of metallic mercury is also mercury(I1) which reacts with as yet unoxidised mercury atoms present on the electrode surface to give mercury(1) ions.Consequently mercury(1) ions are formed as a result of a reproportionation reaction following the electrochemical oxidation of mercury to mercury(I1). Keywords Stripping voltammetry ; rotating ring-disc electrode ; mercury dissolution kinetics ; reproportionation Mercury has widespread use in electroanalytical chemistry. In stripping voltammetry the mercury film on solid electrodes offers great advantages and has been used in many studies. However the electrodissolution kinetics of mercury itself from a solid electrode have been much less commented on in the literature. In particular no unambiguous answer can be found concerning the problem of the valence of the mercury ions formed during mercury electroionisation.The oxidation may produce either mercury(I) or mercury(I1) or both ions. This study was undertaken to try to elucidate this problem and to obtain an insight into the kinetics of mercury electrodissolution. We used mainly the rotating ring-disc technique.l Mercury(I1) ions are reduced and deposited on the disc of the electrode followed by stripping of the deposit using a linearly varying potential. A fraction (No) of the material stripped at the disc is transported to the surrounding ring electrode where a selective detection is possible at a fixed potential. As the potential of the ring is constant almost no capacitive current is flowing and the base line which is not influenced by the scan rate at the disc is essentially flat. The collection efficiency No,2 represents the fraction of the amount of electroactive species produced at the disc that theoretically reacts at the ring.Part of the reaction product at the disc escapes into the bulk of the solution during the crossing of the gap between ring and disc. No can also be calculated from the dimensions of the ring-disc electrode. Experimental Reagents Experiments were performed in solutions of hydrochloric perchloric sulphuric and nitric acids in solutions of sodium perchlorate sodium thiocyanate sodium sulphate and potassium nitrate at various concentrations and pH values and in 0.1 M acetate buffer solution (pH 4.66). Solutions were prepared from analytical-reagent grade reagents (Merck) and water freshly generated by a Milli-Q system (Millipore Inc.).A standard 0.1 M mercury(I1) solution was prepared from mercury(I1) nitrate and stabilised by adding a suitable amount of concentrated nitric acid (Merck Suprapur) to obtain a final acid concentration of 0.1 M. Cell solutions were carefully de-aerated with nitrogen containing less than 1 p.p.m. of oxygen. Instrumentation A home-made ring-disc electrode with a glassy carbon (Tokai Electrode Mfg. Co. Tokyo) disc and a platinum ring was used. The disc radius and inner and outer ring radii were 0.225, 0.246 and 0.268 cm respectively. This gives a theoretical collection efficiency No of 0.21. * Research Assistant of the NFSR (Belgium) KIEKENS MERTENS LAIRE AND TEMMERMAN 1083 The collection efficiency determined with the copper(I1) - copper(1) system in 0.5 M potassium chloride solution was 0.22.3 A saturated calomel electrode (S.C.E.) served as the reference electrode.A water-bath allowed the temperature to be held constant at 25 & 0.1 "C. The cell assembly apparatus and other equipment have been described el~ewhere.~ Electrode Pre-treatment The glassy carbon ring-disc electrode system was polished in the usual way the final polish-ing being effected with 0.05-pm alumina on Buehler microcloth to a mirror-like finish. Ultra-sonic vibration served to remove any alumina or other impurities (e.g. platinum from the ring electrode) that might adhere to the disc electrode surface. The electrochemical pre-treatment of the disc consisted of a continuous cycling of the poten-tial between fixed values in the working-electrolyte at a scan rate of 0.1 V s-l until a repro-ducible and very low background was obtained.The ring was pre-treated in the same way. During this pre-treatment nitrogen was passed through the cell solution. After the de-aeration and electrode activation period a nitrogen atmosphere was maintained above the cell solution. Electrolysis Procedure From an initial disc potential that was positive enough to avoid any mercury(I1) reduction, pre-electrolysis was started by switching the potential of the disc to the desired deposition potential Edep. Before the expiration of the electrolysis a potential that was positive enough to collect (oxidise) mercury(1) ions that might leave the disc electrode during mercury electro-dissolution was applied to the ring electrode. After deposition the disc potential was scanned linearly in the anodic direction and the mercury dissolution curve (disc) and collection curve (ring) were recorded.Results and Discussion A survey of literature data5-ls about mercury electrodissolution indicates that mercury(I1) ions are mainly formed in complexing electrolytes while mercury(1) ions are found in non-complexing electrolytes. Some data are summarised in Table I. One exception can be noticed according to Combet and Dozol electrodissolution of mercury in perchloric and nitric acid at a glassy carbon electrode produces only divalent mercury ions.12 According to Brainina and Neiman,14 a linear dependence is expected between the anodic peak potential E, of the metal stripping curve and the logarithm of the potential scanning rate v.Such a relationship is found for mercury ionisation in various electrolytes. Fig. 1 illustrates the linear behaviour between E and logv for 0.1 M sodium thiocyanate + 0.01 M perchloric acid and for 0.1 M perchloric acid. From the slopes14 of these plots the values of /3n (/3 = anodic transfer coefficient; n = number of electrons involved) were found to be 1.44 TABLE I ELECTRODISSOLUTION PRODUCTS OF MERCURY AT VARIOUS ELECTRODES IN DIVERSE ELECTROLYTES5-13 Electrode Platinum Glassy carbon Platinum Gold . . Platinum Mercury . . Graphite Graphite Glassy carbon Mercury . . ,. * . Electrolyte KNO, HClO, HClO, HNO, H2S04 KSCN KSCN KSCN HClO, HNO, NaCN H2S04 HClO, Reference 5 6 7 8 7 9 10 11 12 13 *Ed <0.6V.1 Ed >0.85 v 1084 Analyst Vol. 108 and 1.74 respectively. Other values obtained are 1.31 in 1 M sulphuric acid and 1.69 in 0.1 M acetate buffer solution (pH = 4.66). These results suggest that the valency of the mercury ions formed during electro-oxidation must be two independent of the electrolyte used to obtain a reasonable value for /3.15 However ring-disc experiments in perchloric and sulphuric acid solutions demonstrated the existence of monovalent mercury ions during electrodissolution. This finding is also in accordance with the experiments by Allen and .Johnson.6 The mercury(1) ions produced a t the disc may react at a sufficiently positive potential at the ring to give mercury(II) or they may also react at a negative ring potential to give metallic mercury.This is shown in Fig 2. The two ring collection peaks at 1.4 V (oxidation) and -0.2 V (reduction) are almost equal, indicating that the deposited mercury is mainly stripped as mercury(1). Of course this is in contradiction with the results obtained from simple semi-logarithmic analysis of the mercury ionisation peak at the disc as mentioned before. KIEKENS et al. VOLTAMMETRY OF THE MERCURY 0.47 4 0.46 c! tn 6 0.45 z 0.44 0.43 $ -2.0 0.04 0.03 4 tn c! 0.02 0.01 $ 0.00 Fig. 1. Dependence of the anodic stripping peak potential ED of mercury dissolution on the logarithm of the electrode potential scanning rate v in (A) 0.1 M perchloric acid and (B) 0.1 M sodium thiocyanate + 0.01 M per-chloric acid.Concentration of mercury(I1) is 2 ELM. 0.6 0.5 0.4 0.3 EJJ VS. S. C. E. Fig. 2. Current - potential graphs for electrodissolution of mercury at a glassy-carbon disc and simultaneous collection of stripped mercury ions at the platinum ring of a rotating ring-disc electrode in 0.1 M perchloric acid. Concentration of mercury-(11) = 10 ELM; deposition time = 1 min; rotation speed of the electrode = 188 rad s-l; E d e p = -1.0 V against the S.C.E.; v = 0.1 V s-1; Er = ring potential; and Ed = disc potential. These conflicting experimental results can be explained by accepting the occurrence of a special type of chemical reaction following mercury electrodissolution Le. reproportionation. The first formed mercury(I1) ions are supposed to react with as yet unoxidised mercury atoms still present on the electrode surface leading to the formation of mercury(1) ions.This explanation is acceptable as the equilibrium constant K for the reaction Hg(0) + 2Hg(I) or KgZ2+ indicates that monovalent mercury ions are stabilised (K = 166).16 These mercury(1) ions (possibly in a dimeric state) reach the surrounding ring electrode where they can be oxidised or reduced. This proposed reaction mechanism for mercury dissolution also explains the fact that the stripping peaks in thiocyanate solutions or generally in complexing electrolytes are abou September 1983 DISSOLUTION REACTION MECHANISM AT SOLID ELECTRODES 1085 twice the size of the dissolution peaks obtained in perchloric acid or generally in non-complex-ing electrolytes.In non-complexing electrolytes half of the mercury deposit may be stripped from the electrode surface in a non-electrochemical way because of the reproportionation reaction. In a complexing electrolyte our experiments indicate that the newly formed mercury(I1) ions at the electrode surface are involved in a subsequent chemical reaction, stabilising the mercury(I1) ions as complex ions. Investigations in an electrolyte solution of constant ionic strength containing x M sodium thiocyanate + (2-x) M sodium perchlorate + 0.01 M perchloric acid suggest that Hg(SCN) is the primary product of mercury electro-dissolution (0.1 < x < 2). This co-ordination number of 3 was derived from the dependence of the anodic peak potential on the thiocyanate concentration.ll The co-ordination number in the bulk of the solution is reported to be 4.lopl7 The oxidising properties of complexed mercury(I1) species towards mercury atoms can be calculated to be considerably lowered in comparison with the action of no or weakly complexed mercury(I1) ions.In complexing electrolytes the absence of monovalent mercury ions is also confirmed by ring-disc experiments. At the ring only a reduction signal of stripped divalent mercury ions can be obtained at sufficiently negative potentials. Finally we can conclude that the complexing properties of the electrolyte towards mercury ions are of considerable importance during mercury electrodissolution. This is also true for the concentration of the complexing agent. This was observed in hydrochloric acid solutions at various concentrations.At a low hydrochloric acid concentration e.g. 0.01 M both an oxidation and a reduction peak of mercury(1) ions produced at the disc can be obtained a t the ring. With increasing chloride ion concentration both the ionisation peak at the disc and the reduction peak at the ring grow while the oxidation peak at the ring diminishes. This means that more and more mercury(I1) ions are produced instead of mercury(1) ions and that reproportionation becomes less important because of the increased formation of mercury( 11) chloride complexes. At a hydrochloric acid concentration exceeding 2 M the mercury dis-solution peak reaches its maximum value and an oxidation peak at the ring can no longer be obtained.Hitherto only electrodissolution experiments a t a glassy carbon electrode have been de-scribed. However electro-oxidation measurements at a rotating ring-disc electrode with a disc of gold or iridium1* confirm the foregoing results obtained at a glassy carbon electrode. In non-complexing electrolytes the mercury ionisation again seems to be influenced by a repro-portionation reaction. This indicates the exclusive formation of divalent mercury ions at the disc. M.M. and Ch.L. thank the IWONL and P.K. the NFSR for financial support. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Albery W. J. and Hitchman M. L. “Ring-disc Electrodes,” Clarendon Press Oxford 1971. Albery W. J. and Bruckenstein S. Trans.Faraday SOC. 1966 62 1920. Napp D. T. Johnson D. C. and Bruckenstein S. Anal. Chem. 1967 39 481. Kiekens P. Verbeeck R. M. H. Donche H. and ‘I’enimerman E. J . Electroanal. Cltem. 1983 147, Hartley A. M. Hiebert A. G. and Cox A. J. J . Electroanal. Chem. 1968 17 81. Allen R. E. and Johnson D. C. Talanta 1973 20 799. Hassan M. Z. Untereker D. F. and Bruckenstein S. J . Eleclroanal. Chem. 1973 42 161. Lindstrom T. R. and Johnson D. C. Anal. Chem. 1981 53 1855. Eluard A. and TrBmillon B. J . Electroanal. Chem. 1967 13 208. Kartushinskaya A. I. Stromberg A. G. and Kolpakova N. A. Elektrokhimiya 1971 7 1243. Brainina Kh. Z. and Neiman E. Ya. Zh. Anal. Khim. 1971 26 875. Combet S. and Dozol M. Electrochim. A d a 1979 24 1283. Kirowa-Eisner E. Talmor D. and Osteryoung J. Anal. Chem. 1981 53 581. Brainina Kh. Z. “Stripping Voltammetry in Chemical Analysis,” John Wiley New York 1974. Verplaetse H. Kiekens P. Temmerman E. and Verbeek F. J . Electvoanal. Chem. 1980 115 235. Cotton F. A. and Wilkinson G. “Advanced Inorganic Chemistry,” Third Edition Interscience, Murray R. W. and Gross D. I. Anal. Chem. 1966 38 392. Laire Ch. unpublished results. 235. New York 1972 p. 509. Received February 25th 1983 Accepted March 18th 198

ISSN:0003-2654    

DOI:10.1039/AN9830801082

出版商:RSC    

年代:1983

数据来源: RSC