摘要:

Modern Spectroscopyby J. MICHAEL HOLLAS, University of Reading, UKIntroduces a wide range of spectroscopies, including the backgroundtheory and the applications to structure determination and chemicalanalysis.CONTENTS INCLUDE: Some Important Results in QuantumMechanics; Electromagnetic Radiation and its Interaction with Atomsand Molecules; Molecular Symmetry; Rotational Spectroscopy;Vibrational Spectroscopy; Electronic Spectroscopy; Photoelectron andRelated Spectroscopies; Lasers and Laser Spectroscopy.0471911216 408 pages December 1986 (cl) f9.95/$16.950471911313 408 pages December 1986 (pr) f9.95/$16.95Quantitative Analysis usingChromatographic Techniquesby ELENA KATZ, Applied Research Department, Perkin Elmer Corp, Norwalk,Connecticut, USACovers the quantitative aspects of all the chromatographic methodsand provides a unique single source of up-to-date information.Individual chapters are written by experts in the field, and specialfeatures include detailed treatment of the principles of quantitativedetection and sources of error in data processing procedures.Series: Separation Science.CONTENTS INCLUDE: Detection in Quantitative Liquid Chroma-tography; Quantitative Analysis by Liquid Chromatography; Detectionin Quantitative Gas Chromatography; Quantitative Analysis by GasChromatography; Quantitative Thin-Layer Chromatography; Chro-matography as a Quantitative Tool in Pharmaceutical Analysis; IsAutomation the Future of Quantitative Chromatography?; Physico-Chemical Information from Peak Shape and Width in LiquidChromatography.0471914061 446 pages January 1987 f37.50/$63.95Principles of Electrochemistryby J.KORYTA, I. Heyrovski Institute of Physical Chemistry & Electrochemistry,Prague, Czechoslovakia andJ. DVORAK, Faculty of Science, Charles University, Prague, CzechoslovakiaProvides the basic theory and applications of electrochemistry, which isbecoming increasingly important in a number of fields. Electrontransfer theory and double layer theory are included in understandableterms and mathematics is kept to a basic minimum.CONTENTS INCLUDE: Equilibrium Properties of Electrolytes;Transport Processes in Electrolyte Systems; The Electrical Double Layer;Process in Heterogeneous Electrochemical Systems; MembraneElectrochemistry and Bioelectrochemistry.0471912115 460 pages March 1987 f49.50/$84.40Telephone your credit card order: (0243) 829121, Customer Service Dept.Freefone - Dial 100 and ask for Freefone 3477 (UK only)We will refund your payment without question if you return any unwantedtitle to us in re-saleable condition within 30 days.All books also available from your booksellerJohn Wiley & Sons MdBaffins Lane, Chichester, Sussex PO19 lUD, EnglandANALYTICAL CHEMISTRY BYOPEN LEARNINGGas Chromatographyby JOHN E.WILLElT, Wolverhampton Polytechnic, UKCONTENTS INCLUDE: The Working Gas Chromatograph: The CarrierGas, The Oven, The Column, The Injection System, Detectors;Columns; Materials, Liquid Stationary Phases, Solid Stationary Phases,The Support; Choosing the Other Parameters: Length of the Column,Temperature Programming, Sample Size, Attenuation, Flow Rate,Injection Heater Temperature; The Gas Chromatography of LessVolatile Samples; Qualitative Analysis by Gas Chromatography;Quantitative Analysis by Gas Chromatography.0471913316 272 pages March 1987 (cl) f28.00/$47.750471913324 272 pages March 1987 (pr) f9.95/$16.95Samples and Standardsby BRIAN W.WOODGET, Hatfield Polytechnic andDEREK COOPER, North Staffordshire Polytechnic, UKCONTENTS INCLUDE: The Analyst’s Approach; Introduction toSampling; Design of a Sampling Procedure; Methods of TakingSamples; Standardization and Calibration; Analytical Standards andCalibration Curves; Standard-addition Methods; Internal-standardMethods; Calibration by Computational Means; Comparison ofCalibration Procedures; Monitoring the Performance of AnalyticalProcedures.0471912891 318 pages January 1987 (cl) €32.00/$54.600471912905 318 pages January 1987 (pr) f11.50/$19.60Principles of Electroanalytical Methodsby TOM RILEY and COLIN TOMLINSON, Brighton Polytechnic, UKCONTENTS INCLUDE: Basic Principles of Solution Chemistry andElectrochemistry: Ions in Solution, lon-ion Interactions, IonicMigration in Electrolyte Solution, Electrodes and the ElectrochemicalCell; Galvanic Cells: Electrode Types, The Nernst Equation, The LiquidJunction, EMF Measurements, PH and its Measurement, Poten-tiometry; Electrolysis; Review of Methods of ElectroanalyticalChemistry: Classification and Relationships, Principles of the MoreImportant Methods, Present Usage.0471913294 272 pages February 1987 (cl) f28.00/$47.750471913308 272 pages February 1987 (pr) f9.951816.95Radiochemical Methodsby WILLIAM J.GEARY, Department of Chemistry, Sheffield City Polytechnic,CONTENTS INCLUDE: Introduction: Nuclear Properties and Radio-active Decay; Radio-Isotopes and Labelled Compounds: Preparationand Availability; Practical Aspects: Detection and Counting by GasIonization, Scintillation and Semi-conductor Methods; RadioAnalytical Methods: Direct Determination, Tracer Investigations,Activation Analysis; New Applications: Literature Examples, PotentialApplications of Radiochemical Techniques.0471911178 250 pages October 1986 (cl) f28.00/$47.750471911186 250 pages October 1986 (pr) f9.95/$16.95Circle 002 for further informatioSPECTROCHIMICAACTAMOCECULAR sremuoscory SPECTROCHIMICAACTAPart B: Atomic SpectroscopyEditor-in-Chief: P W J M BOUMANS,Ph ilips Research Laboratories,Department of Spectrochemistry, PO Box80.000,5600 JA Eindhoven, TheNetherlandsEd i to rs : W S LAW N , Perkin- ElmerCorporation, USA and H OECHSNER,Univerity of Kaiserslautern, FederalRepublic of GermanyAssistant Editor: J A C BROEKAERT,lnstitut fur Spektrochemie undA ng e wan dte Sp e ktrosko pie (ISA S),Federal Republic of GermanySpectrochimica Acta, Part 8, covers topicsfrom rapidly expanding areas in atomicspectroscopy, mass spectroscopy forinorganic analysis, and surface, interface,thin film and micro analysis. The articlesdeal with: theory and fundamentals,methodology development,instrum en ta ti0 n, an d applications.Recent years have seen the publication ofan ever increasing number and variety ofarticles contributed by the leading authorsin spectroscopy who have understood theimpact of Spectrochimica Acta, Part B, onthe development of atomic spectroscopyand related fields. Therefore the journal isan indispensable source of information forall analytical spectroscopists.Authors canfeel assured that the submittance of theirmanuscripts is followed by fast, thoroughand efficient refereeing, substantialeditorial advice and rapid publication of therevised text in the journal with the longesttradition in spectroscopy and the unrivalledstandard.Patents Section -The journal containsabstracts and illustrations of recentlyissued United States Patents and publishedpatent applications filed from over 30countries under the Patent Co-operationTreaty.Subscription InformationPublished monthly (Volume 42)Annual subscription (1987) DM1245.00Two-year rate (1 987/88) DM2365.50A selection of papersAn automated direct sample insertionsystem for the inductively coupled plasma,W E PETTIT & G HORLICK.A steady-state approach to evaluation ofproposed excitation mechanisms in theanalytical ICP, G D RAYSON &G M HIEFTJE.The determination of trace elements ingeochemical exploration samples by ICP-MS, A R DATE & D HUTCHISON.Interference minimization using secondsurface atomizer for furnace atomicabsorption, T M RETTBERG &J A HOLCOMBE.Spatial resolution enhancement for linearp h otod i ode a rra y atomic spectrometry,S W McGEORGE & E D SALIN.Experimental control of the solvent load oflCPs and effects of chloroform plasma loadon their analytical performance,F J M J MAESSEN etal.Influence of the generator frequency andthe plasma gas inlet area on torch design inJ M MERMET.An improved interface for ICP-MS,D J DOUGLAS & J B FRENCH.Scanning electron microscopy studies onsurfaces from electrothermal AAS, B WELZet al.Qualitative and semi-quantitative ICAP-AESanalysis using a computer-controlledmonochromator, P D P TAYLOR &J DE DONDER.ICP-AES, E MICHAUD-POUSSEL &Pergamon Press FREE SAMPLE COPIES AVAILABLE ON REQUESTAdvertising rate card available on request. Back issues andcurrent subscriptions are also available in microform.The Deutsch Mark prices shown include postage and insur-ance.For subscription rates in Africa, Asia, Australasia, UKand Eire and the Americas amlv to vour nearest PerqamonHeadington Hill Hall, Oxford OX3 OBW, UKFairview Park, Elmsford, New York 10523, USAoffice. Prices are subject to 'change 'without notice. - 5&17'10/8THE ANALYST READER ENQUIRY SERVICE M AY'87For further information about any of the products featured in the advertisements in this issue, please writethe appropriate number in one of the boxes below.Postage paid if posted in the British Isles but overseas readers must affix a stamp.rPLEASE USE BLOCK CAPITALS LEAVING A SPACE BETWEEN WORDS Valid 12 months1 NAME2 COMPANY ~ I l l l l I i l l l l i l i l I l i I I l ~ i I I i i I I ~ I il I l l l 1 l l ; I l l l l l l i l l -11 I 1 I I I I 1 1 m PLEASE GIVE YOUR BUSINESS ADDRESS IF POSSIBLE.IF NOT, PLEASE TICK HERE3 STREET4 TOWN I l l l l l l l l l l l l l l l l l i I l l l l l l l l l I l lH i i i I l i 1 I I I I l l I 1 H i I I I I I I I I I I I I i 5 COUNTYPOST CODE6 COUNTRY I7 DEPARTMENTIDIVISION i l l l l l l l l l i i i l l l l l i l l I l l l l l l ! i l l lPOSITION8 YOUR JOB TITLEI I Postagewill bepaid byLicenseeDo not affix Postage Stamps if posted in Gt. Britain,Channel Islands, N. Ireland or the Isle of ManBUSINESS REPLY SERVICELicence No. WD 106Reader Enquiry ServiceThe AnalystThe Royal Society of ChemistryBurlington House, PiccadillyLONDONWIE 6WFEngland1I IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII1IIIIIIII ;IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII!IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII!IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

ISSN:0003-2654    

DOI:10.1039/AN98712BP013

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

The AnalystThe Analytical Journal of The Royal Society of ChemistryAdvisory Board*Chairman: J. D. R. Thomas (Cardiff, UK)J. F. Alder (Manchester, UK)D. Betteridge (Sunbury-on-ThanE. Bishop (Exerer, UK)*C. Burgess (Ware, UK)D. T. Burns (Belfast, UK)G. D. Christian (USA)*M. S. Cresser (Aberdeen, UK)L. de Galan (The Netherlands)*A. G. Fogg (Loughborough, UK)*C. W. Fuller (Nottingham, UMV. D. Goldberg (London, UK)T. P. Hadjiioannou (Greece)W. R. Heineman (USA)A. Hulanicki (Poland)*C. J. Jackson (London, UK)*P. M. Maitlis (Sheffield, UK)E. J. Newman (Poole, UK)T. B. Pierce (Hanuell, UK)E. Pungor (Hungary)J. RSiiEka (Denmark)R. M. Smith (Loughborough, UK)W. I. Stephen (Birmingham, UK)M. Stoeppler (Federal Republic of Germany)K. C.Thompson (Sheffield, UK)*A. M. Ure (Aberdeen, UK)A. Walsh, K.B. (Australia)G. Werner (German Democratic Republic)T. S. West (Aberdeen, UK)*P. C. Weston (London, UK)J. D. Winefordner (USA)Yu. A. Zolotov (USSR)P. Zuman (USA)7es, UK)"Members of the Board serving on the Analytical Editorial BoardRegional Advisory EditorsFor advice and help to authors outside the UKDr. J. Aggett, Department of Chemistry, University of Auckland, Private Bag, Auckland, NEWDoz. Dr. sc. K. Dittrich, Analytisches Zentrum, Sektion Chemie, Karl-Marx-Universitat, Talstr.Professor L. Gierst, Universite Libre de Bruxelles, Faculte des Sciences, Avenue F.-D.Professor H. M. N. H. Irving, Department of Analytical Science, University of Cape Town,Dr. 0. Osibanjo, Department of Chemistry, University of Ibadan, Ibadan, NIGERIA.Dr.G. Rossi, Chemistry Division, Spectroscopy Sector, CEC Joint Research Centre,Dr. 1. RubeSka, Geological Survey of Czechoslovakia, Malostranske 19, 118 21 Prague 1,Professor K. Saito, Coordination Chemistry Laboratories, Institute for Molecular Science,Professor M. Thompson, Department of Chemistry, University of Toronto, 80 St. GeorgeProfessor P. C. Uden, Department of Chemistry, University of Massachusetts, Amherst,Professor Dr. M. Valcarcel, Departamento de Quimica Analitica, Facultad de Ciencias,Professor Yu Ru-Qin, Department of Chemistry and Chemical Engineering, Hunan University,ZEALAND.35, DDR-7010 Leipzig, GERMAN DEMOCRATIC REPUBLIC.Roosevelt 50, Bruxelles, BELGIUM.Rondebosch 7700, SOUTH AFRICA.EURATOM, lspra Establishment, 21020 lspra (Varese), ITALY.CZECHOSLOVAKIA.Myodaiji, Okazaki 444, JAPAN.Street, Toronto, Ontario M5S I A I , CANADA.MA 01003, USA.Universidad de Cordoba, 14005 Cordoba, SPAIN.Changsha, PEOPLES REPUBLIC OF CHINA.Editor, The Analyst:Philip C.WestonSenior Assistant Editors:Judith Egan, Roger A. YoungAssistant Editors:Anne Horscroft, Harpal MinhasEditorial Office: 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, BurlingtonHouse, Piccadilly, London, WIV OBN. Telephone 01-437 8656. Telex No. 268001The Analyst (ISSN 0003-2654) is published monthly by The Royal Society of Chemistry,Burlington House, London WIV OBN, England.All orders accompanied with payment shouldbe sent directly to The Royal Society of Chemistry, The Distribution Centre, Blackhorse Road,Letchworth, Herts. SG6 IHN, England. 1987 Annual subscription rate UK f160.00, Rest ofWorld f179.00, USA $315.00. Purchased with Analytical Abstracts UK f364.00, Rest of Worldf403.00, USA $709.00. Purchased with Analytical Abstracts plus Analytical Proceedings UKf411 .OO, Rest of World f455.00, USA $801 .OO. Purchased with Analytical Proceedings UKf200.00, Rest of World f224.00, USA $394.00. Air freight and mailing in the USA b)Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11003.USA Postmaster: Send address changes to: The Analyst, Publications Expediting Inc., 20CMeacham Avenue, Elmont, NY 11003.Second class postage paid at Jamaica, NY 11431. Alother despatches outside the UK by Bulk Airmail within Europe, Accelerated Surface Posioutside Europe. PRINTED IN THE UK.Information for AuthorsFull details of how to submit material forpublication in The Analyst are given in theInstructions to Authors in the January issue.Separate copies are available on request.The Analyst publishes papers on all aspects ofthe theory and practice of analytical chemistry,fundamental and applied, inorganic andorganic, including chemical, physical, biochem-ical, clinical, pharmaceutical, biological, auto-matic and computer-based methods. Papers onnew approaches to existing methods, newtechniques and instrumentation, detectors andsensors, and new areas of application with dueattention to overcoming limitations and to un-derlying principles are all equally welcome.There is no page charge.The following types of papers will be con-sidered:Full papers, describing original work.Short papers: the criteria regarding origin-ality are the same as for full papers, but shortpapers generally report less extensive investi-gations or are of limited breadth of subjectmatterCommunications, which must be on anurgent matter and be of obvious scientificimportance.Rapidity of publication is enhancedif diagrams are omitted, but tables and formulaecan be included. Communications receive pri-ority and are usually published within 5-8weeks of receipt.They are intended for briefdescriptions of work that has progressed to astage at which it is likely to be valuable toworkers faced with similar problems. A fullerpaper may be offered subsequently, if justifiedby later work.Reviews, which must be a critical evaluationof the existing state of knowledge on a par-ticular facet of analytical chemistry.Every paper (except Communications) will besubmitted to at least two referees, by whoseadvice the Editorial Board of The Analystwill beguided as to its acceptance or rejection. Papersthat are accepted must not be published else-where except by permission. Submission of amanuscript will be regarded as an undertakingthat the same material is not being consideredfor publication by another journal.Regional Advisory Editors.For the benefit ofpotential contributors outside the United King-dom, a Panel of Regional Advisory Editorsexists. Requests for help or advice on anymatter related to the preparation of papers andtheir submission for publication in The Analystcan be sent to the nearest member of the Panel.Currently serving Regional Advisory Editors arelisted in each issue of The AnalystManuscripts (three copies typed in double spac-ing) should be addressed to:The Editor, The Analyst,Royal Society of Chemistry,Burlington House,Piccadilly,LONDON WIV OBN, UKParticular attention should be paid to the use ofstandard methodsof literaturecitation,includingthe journal abbreviations defined in ChemicalAbstracts Service Source Index. Wherever pos-sible, the nomenclature employed should fol-low IUPAC recommendations, and units andsymbols should be those associated with SI.All queries relating to the presentation andsubmission of papers, and any correspondenceregarding accepted papers and proofs, shouldbe directed to the Editor, The Analyst (addressas above). Members of the Analytical EditorialBoard (who may be contacted directly or via theEditorial Office) would welcome comments,suggestions and advice on general policy mat-ters concerning The Analyst.Fifty reprints of each published contribution aresupplied free of charge, and further copies canbe purchased.0 The Royal Society of Chemistry, 1987. Allrights reserved. No part of this publication maybe reproduced, stored in a retrieval system, ortransmitted in any form, or by any means,electronic, mechanical, photographic, record-ing, or otherwise, without the prior permissionof the publishers

ISSN:0003-2654    

DOI:10.1039/AN98712FX017

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALAO 1 12(5) 557-71 6 (1 987)The AnalystMay 1987The Analytical Journal of The Royal Society of Chemistry55757358158759560160961561 9623627631637645649653657661665673679687689693CONTENTSSystematic and Random Errors in Known Addition Potentiometry. A Review-Derek MidgleyTemperature Compensation in Potentiometry: lsopotentials of pH Glass Electrodes and Reference Electrodes. Part 1.Temperature Compensation in Potentiometry : lsopotentials of pH Glass Electrodes and Reference Electrodes. Part II.Use of Lipophilic Additives for the Improvement of the Characteristics of PVC Membrane Lithium-selective ElectrodesFlow Injection Determination of Inorganic Bromide in Soils with a Coated Tubular Solid-state Bromide-selectiveVoltammetric Study of Coppertll) Dialkyldithiophosphates Formed by the Interaction of Dialkyldithiophosphates withVoltammetry of Copper(1) 0,O'-Di(1-methylethy1)phosphorodithioate-Miles J.Hutchings, G. J. Moody, J. D. R.On-line Determination of Ethanol During Fermentation Processes Using a Fuel Cell Sensor-W. James Criddle, Keith W.Application of a Photodiode Array Detector t o Multi-component Determination by Flow Injection Analysis-Marcel0A New Way of Organising Spectral Line Intensity Ratio Fluctuations of Different Elements-50 ThelinDetermination of Tungsten in Ores and Concentrates by Atomic Absorption Spectrometry: Suppression ofAtomisation Interferences from Calcium-Chow Chong, Nik MeriamDetermination of Low Concentrations of Tungsten and Molybdenum in Geological Materials Using InductivelyCoupled Plasma Atomic Emission Spectrometry with Pre-concentration on Activated Charcoal-Gwendy E.M.Hall, Jean-Claude Pelchat, K. Nimalasiri de SilvaProof of a Pine Wood Origin for Pitch from Tudor (Mary Rose) and Etruscan Shipwrecks: Application of AnalyticalOrganic Chemistry in Archaeology-Neil Robinson, Richard P. Evershed, W. James Higgs, Katherine Jerman,Geoffrey EglintonSpectrofluorimetric Determination of Beryllium in Rocks, Alloys and Steels with Nuclear Fast Red-Francisco Salinas,Arsenio Muhoz de la Peiia, Francisco Muiioz de la PehaFluorimetric Determination of Aluminium and Beryllium in Mixtures by Synchronous Derivative Spectrometry-Francisco Garcia Sanchez, Jose C. Marquez Gomez, Miguel Hernandez LopezSpectrophotometric Determination of Pyridoxine Hydrochloride-V. Nirmalchandar, Rajganesh Viswanathan, N.Balasu bramanianGas Chromatographic Method for the Determination of Isomeric Bromoisobutyric Acids and Their Chlorides-AnnaMayer, lstvan Simonyi, Jozsef ReiterHigh-performance Liquid Chromatographic Determination of Artemisinine (Qinghaosu) in Human Plasma andSaliva-S h is h a n 2 h aoDetermination of Inorganic Anions by Non-suppressed Ion Chromatography with Indirect Ultraviolet AbsorptionDetection-Frank G. P.MullinsInter-laboratory Calibration for the Quality Control of Pesticide Residue Analyses (198&85)-Pieter R. de Beer, Louis P.van Dyk, Susan M. Prinsloo, Awie J. Viljoen, Laurraine H. LotterTheory-Derek M idgleyPerformance of Commercial Electrodes-Derek MidgleyBased on Non-cyclic Neutral Carriers-Tatsuhiro Okada, Kazuhisa Hiratani, Hideki SugiharaElectrodeJacobus F. van StadenCopper Salts-Miles J.Hutchings, G. J. Moody, J. D. R. ThomasThomasParry, Thomas P. JonesBlanco, Jordi Gene, Hortensia Iturriaga, Santiago MaspochREPORT OF THE ANALYTICAL METHODS COMMITTEERecommendations for the Conduct and Interpretation of Co-operative TrialsSHORT PAPERSRapid Enzyme-linked lmmunosorbent Assay (ELSA) with a Visual End-point for Detecting Opiate Narcotics inSelective and Sensitive Extraction Spectrophotometric Method for the Determination of Palladium in Titanium BaseSpectrophotometric Determination of Platinum(1V) with Potassium Butyl XanthateNepal Singh, Arvind K.GargUrin-David Laurie, Andrew J. Manson, Andrew Mounsey, Frederick J. Rowell, John SeviourAlloys-Yerramilli Anjaneyulu, Chandra S. Kavipurapu, Manda Raviprakasa Reddy, B. V. Raocontinued inside back coverTypeset and printed by Black Bear Press Limited, Cambridge, Englan697 Determination of Nitrazepam and Flunitrazepam by Flow Injection Analysis Using a Voltammetric Detector-ElisaRuiz, Manuel Hernandez Blanco, Encarna Lorenzo Abad, Lucas Hernandez701 Determination of Sodium N-Methyldithiocarbamate (Metham Sodium) and Methyl lsothiocyanate in AqueousSamples by High-performance Liquid Chromatography Using a Micellar Mobile PhaseFrank G. P. Mullins, (thelate) Gordon F. Kirkbright705 Rapid Titrimetric Method for the Determination of Captan and Folpet in Fungicide Formulations-Balbir C.Verma,Narendra K. Sharma, Miss Anila Sud, Hari K. Thakur, Davender K. Sharma709 Determination of Mercury in Lake Sediments Using a Gold Film Mercury Analyser-Alena Mudroch, Ellie Kokotich711 BOOK REVIEWSERRATUM715 Differential-Pulse Cathodic Stripping Voltammetric Investigation of Cr04’, Moo4*, W04‘ and V03--M. Rasul Jan,W. Franklin SmythNEWCertified Reference MaterialsPublicationsBureau of Analysed SamplesCatalogue No. 550Overseas Reference Materials List No. 555for copies of these publications pleasewrite, telex or telephone to:BAS Ltd., Newham Hall, Newby,Middlesbrough, Cleveland, TS8 9EATelex: 587765 BASRIDTelephone: (0642) 317216Reprint of a review published in ChemicalSociety ReviewsJOHN JEYES LECTUREThe EnvironmentalChemistry of RadioactiveWaste Disposalby John R. Duffield and David R.WilliamsDept. of Applied Chemistry, UWISTOver the past forty years there has been a revolution in the wayin which man fulfils his energy requirements. In this period wehave moved from a predominantly fossil-fuel based powereconomy to one in which nuclear fission plays an increasinglysignificant role. This transition has placed new and potentiallyvery serious stresses on the environment and associatedecosystems. This review considers the environmental chemistryproblems that the disposal of radioactive waste has generatedand how they might be tackled.Brief Contents:Introduction; The Threat to Man;Contamination Pathways; TheChemistry of Waste Containment;Groundwater; Aqueous Speciationof Radionuclides; Sorption; RiskAssessment; Models and SimulationTechniques; Calculation Procedures;Databases; Verification andValidation; Conclusions and aStrategy for the Future17pp €2.00 ($4.00)PAYMENT MUST ACCOMPANY ORDER (Cheques made pay-able to “The Royal Society of Chemistry”)Orders should be sent to:K. J. Wilkinson, Books Department, The Royal Society of Chemistry,Burlington House, Piccadilly, London W1V OBN, UK.Circle 001 for further informatio

ISSN:0003-2654    

DOI:10.1039/AN98712BX019

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 557 Systematic and Random Errors in Known Addition Potentiometry A Review Derek Midgley CEGB, Central Electricity Research Laboratories, Kelvin Avenue, Leatherhead, Surrey KT22 7SE, UK Summary of Contents I n t roduction Theory Basic equations Influence of the calibration slope on accuracy Ionic strength effects Effect of complexing Liquid junction potential Measurements in the non-Nernstian response region Random errors Use of Nernstian calibrations Use of known dilution calibration Random errors Activity coefficients Effect of complexing Non-Nernstian calibrations Interferences and reagent blanks Discussion Conclusions Ref ere n ces Keywords: Known addition potentiometry; systematic and random errors; review Introduction Known addition potentiometry is a method that aims to eliminate the effect of the sample matrix and to remove the need for the calibration curves required in determinations by direct potentiometry.These aims are obviously desirable and the method has therefore been described in many texts on potentiometric analysis. 1-6 These descriptions list the ideal conditions necessary for the application of the method, but the analytical errors caused by failure to conform to these conditions are never quantified. Even the review by Mascini7 does not elaborate these points. The ideal conditions for the use of the method are as follows: (i) Activity coefficients should not vary. (ii) The fraction of free, uncomplexed, determinand should not vary. (iii) The slope factor of the electrode should not vary. Ths implies that (a) the temperature is constant and (b) the electrode calibration is linear throughout the range in question.(iv) Liquid junction potentials are constant. Conditions (i), (ii) and (iv) are difficult to achieve in principle, but in practice conditions can often be arranged so that the variations are negligible. By use of the known dilution method of obtaining the calibration slope, it is often possible to obtain accurate results even when the above conditions are not met. This review aims to quantify the errors that arise from non-fulfillment of the above conditions and to show how far they can be avoided either by the use of the known dilution method of finding the slope factor or by adjusting the conditions of measurement. The multiple known addition method is not considered.Theory Basic Equations The technique depends on measuring the change in potential of an A-selective electrode immersed in a known volume, Vo, of solution containing an unknown concentration of the determinand, A, when a known volume, V , of a standard solution of A (concentration C,) is added. The e.m.f. is related to the activity, {A}, of the determi- nand by equation (1) E=Eo+Slog{A} +Ej-E,,f . . . . (1) where Eo is the standard potential of the electrode, Ej is the liquid junction potential between the solution and the reference half-cell, Eref is the reference half-cell potential, and S is the calibration slope of the electrode, nominally equal to the Nernst slope factor Rnn( 10)lzF. The activity of A is related to its total concentration, CA, by {A}=fp,aAc~ .. . . - * (2) where f~ is the activity coefficient for A and = [A]/CA is the ratio of free to total metal ion concentration. Hence, for the initial e.m.f. we obtain, El = E0 + SlOgfA + SlogaA + slog CA + Ej - Eref - . (3) After the addition, the e.m.f. is E2 = Eo + S'lOgfA + s'log ai + s'log ci + Ej) - Eref (4) where the primed symbols represent the new values of S , fA, The successful application of the known addition method &A, CA and Ej. depends on the validity of the following approximations: S = S ' . . . . . . * * ( 5 ) Ej"Ej' . . . . . . * * (6) fA"fA . . . . . (7) aA"aL . . . . . . . * (8) Approximations ( 5 ) and (6) are usually very good, but approximations (7) and (8) may fail, as will be discussed558 ANALYST, MAY 1987, VOL. 112 below.Even so, it is usually possible either to treat the sample solution so as to make approximations (7) and (8) valid or to compensate for the inequalities by adjusting the slope factor. Assuming that all the approximations are valid, therefore, we obtain from equations (3) and (4) Now Therefore, which can be concentration . * . . (11) CA(v0 + v) rearranged to give an expression for the total of determinand CA may be obtained from equation (12) not only by arithmetical calculation but by the use of nom0grams.8~9 The calculation is often made simpler by the use of tabulated values of the term in braces in equation (12) at fixed values of VlVo and S, as in Orion electrode manuals. If the analytical procedure requires the addition of reagents to adjust the pH of the sample, to maintain a constant ionic strength or to fix the degree of complexation of the determi- nand, let VR be the total volume of all reagent solutions added prior to the measurement of e.m.f., El.Then, progressing as from equation (10) to equation (12), we obtain CA = V G . .(12a) If V << Vo, equation (12) may be simplified to give the approximation (13) CA 2: . . . . (13) If V << Vo and VR << Vo, equation (12a) also reduces to equation (13). Tabulated values of the term in square brackets have been presented5 for various values of E2 - El and S. Volume error The errors incurred by the use of the simplified equation (13) instead of equation (12) are shown in Table 1 for a range of volume additions. The concentration is always overestimated if the dilution of the sample by the standard addition is ignored and the errors increase with the volume of solution added.The errors, ACA, also depend on the ratio, @, of moles added to moles originally present and they can be approximately represented by ACA V l + @ --- - CA -vOi @ ) where @ = C,V/CAVo. The relative errors are independent of concentration. If equation (13) is to be used, it is desirable to keep VlVo < 1%, but with the availability of programmable calculators and microcomputers equation (13) no longer gives a significant saving of time compared with the accurate equation (12): its use should, therefore, be discouraged. Influence of the Calibration Slope on Accuracy The calibration slope, S, is ideally equal to the Nernst slope factor RZln(lO)lzF, where R is the gas constant, T the absolute temperature, Fthe Faraday constant and z the charge Table 1.Relative errors (YO) caused by neglecting the volume correction VlV,, % 4) 0.1 1 2 5 0.5 0.30 3.0 6.2 16.7 1 0.20 2.0 4.1 10.5 2 0.15 1.5 3.0 7.7 on the ion. In practice, S is determined empirically. The method used to find S must be related to the nature of the sample solution if the errors are to be minimised and the two procedures described below will be considered later in regard to the ionic strength of the sample solution and the presence of complexing agents. Calibration by known dilution After the addition of the standard solution and the measure- ment of E2 [equation (4)], a volume, VD, of diluent is added and the new equilibrium e.m.f., E3, is measured.E3 = EO + S" logf;( + s" log dA+ S' log cl( + E;' - Eref (14) Assuming a set of conditions equivalent to equations (5)-(8), i.e., S" = S' = S, E;'= Ej',fd: =fk, al( = ax, we obtain Now E ~ - E ~ = S ~ O ~ C L - S ~ O ~ C ~ . . . . (15) Therefore, E2 - E3 = Slog ( vo ioy2) . . . . (17) from which S can be calculated. If reagents are added prior to measurement, we have, corresponding to equation (17) as equation (12a) corresponds to equation (12), If S is to be determined with reasonable precision, the e.m.f. change, E2 - E3, should not be too small: convention- ally the solution is diluted to twice its original volume, so that E2 - E3 is about 20 mV for a univalent determinand and 10 mV for a divalent determinand.With such a large dilution it is important that the diluent is at the same temperature (f 1 "C) as the sample solution, otherwise it cannot be assumed that S" = S'; Table 4 shows the errors arising from this source alone, but in the dilution procedure there are additional errors arising from changes in EO and Eref with temperature. Moreover, where aA # 1, some change of aA with tempera- ture should also be expected. The quantification of the errors associated with changes in Eo and Eref is possible through use of the isopotential concept familiar in pH measurements.10 Isopotential data are almost completely lacking for ion-selective electrodes, although there is little experimental difficulty in determining them. Over a limited range of temperature (k 10 "C) it is valid to express the e.m.f.by E EI + S(l0g CA - log CI) where EI contains all the temperature-invariant terms and log CI the temperature-dependent terms. Log CI is the isopoten- tial point.ANALYST, MAY 1987, VOL. 112 559 Table 2. Effect of temperature variations during the known dilution procedure for an electrode with an isopotential point at log C = -4 Log Ck Temperature change on dilutionPC Parameter 3 4 5 6 2 $IS Relative error in CA, YO Relative error in CA, YO Relative error in C,, YO Relative error in CA, YO 1 SIS -1 SIS -2 s/s 0.98 -2.7 0.99 -1.5 1.01 1.6 1.02 3.3 1.0 1.02 1.04 0 2.7 4.4 1 .o 1.01 1.02 0 1.6 2.8 1 .o 0.99 0.98 1 .o 0.98 0.95 0 -1.5 -3.2 0 -3.3 -6.7 Table 3. Dilution factors required to restore solution to approximately its original concentration after known addition 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 0.2 0.3 0.4 0.5 0.6 0.8 0.9 1 .o 1.2 1.4 1.6 1-.8 2.1 2.3 2.6 2.9 3.2 3.5 3.9 0.4 0.6 0.9 1.2 1.6 2.1 2.6 3.2 3.9 t t t t t t t t t t * If reagents are added the ratio is VDl(Vo + VR). t Use of such large e.m.f. differences is not recommended and the large dilutions are usually impracticable. After the known dilution, therefore, we would have instead of equations (4) and (14) E2 = EI + S' (log CA - log CI) E3 = EI + S" (log C i - log CI) . . . . (18) . . (19) . . If no account is taken of the temperature variation, we obtain from equation (15) the apparent slope factor S. s = (E2 - E#og (CLlCA) From the equations with isopotential corrections, (18) and (19) 9 s S' S' = [E2 - E3 + (S'"") log C1]/(log C'- - log Cl;) Hence S' AS AS logC' S S' log (CJCl;) where AS = S' - S"and A E = E2 - E3.In evaluating equation (20), if the temperature changes by 1°C on dilution, AS = 0.2/2 mV decade-1 and if the dilution, as usual, halves the concentration, then AE == S'log(CA/Ci) = l8/2 mV. Sub- stituting in equation (20) and taking, as an example, log CI = 4, a 1 "C change in temperature would give the results shown in Table 2. At C i = CI, no error is introduced, but the chances of this condition prevailing are very small. If log (Cl;/CI) = f 1, the errors in concentration are about three times larger than the errors introduced by using a calibration slope obtained independently at temperatures 1 or 2 "C different from that at which the known addition was performed.If log (C;/C1) = f 2 , the errors are six times larger. It follows that temperature control may be very important if the known dilution procedure is used, especially as log( CYC,) increases. The value of log CI chosen for the example was arbitrary and could vary by several units, depending on the nature of the electrode membrane's internal contact (electrolytic or solid state) and the choice of reference electrode. In some instances (e.g., to compensate for ionic strength effects, see below), it is advantageous to dilute the sample so that E3 = El. The required degree of dilution can be obtained by taking a conveniently rounded figure close to antilog [(Ez - El)/(S/z)], S/z being a reasonable guess at the calibration slope. Dilution factors corresponding to values of IE2 - Ell are given in Table 3, calculated with S = 58 mV decade-'. If the sample solution contains a background electrolyte or a strong complexing agent , the diluent should be matched to the sample, as described in the sections on ionic strength and complexing effects .Calibration with standard solutions As in direct potentiometry, S may be found as the slope of a plot of e.m.f. against the logarithm of the concentration of a series of standard solutions. It is generally desirable to keep the conditions of the calibration as identical as possible to those obtaining in the sample solution so that systematic errors may be minimised; in particular, reagents added to the sample should also be added to the standard solutions and the standard solutions should span the expected concentration range of the samples.The temperature at which the calibration is carried out should be within 1°C of the sample temperature but in practice, and especially in field work, this may be impractic- able. In this event, the calibration slope at the sample temperature T, "C may be approximated by S = S, (273 + T,)/(273 + T,) . . . . (21) where S, is the slope determined at temperature T, "C. Before such corrections are used generally, it is desirable to test the accuracy of equation (21) at a minimum of two temperatures different from the calibration temperature and preferably including the extremes of the sample temperature range. The errors caused by calibrating at 25 "C and analysing at another temperature, T, without correcting by means of equation (21) are shown in Table 4.The errors caused by analysing at 25 "C after calibrating at temperature Tare equal (to two significant figures) but opposite in sign. It is important to note that the calibration plot is defined in terms of concentration and not thermodynamic activity. This means that in certain circumstances S will differ significantly from the thermodynamic slope factor, but as far as the known addition method is concerned this will compensate, to some extent, for deviations from the desired conditions represented by equations (6)-(8). This approximation will only be valid over a limited range of concentrations, particularly if CA > 10-3 mol 1-1 and there is no background of indifferent electrolyte.560 ANALYST, MAY 1987, VOL.112 Ionic Strength Effects In general, the known addition method involves a change in ionic strength and equation (7) is, therefore, only an approxi- mation. The inaccuracy introduced by this inequality depends on the concentration of the determinand, the charge on the determinand, the presence of background electrolyte and the value adopted for the calibration slope. A consideration of these factors will show how errors may be reduced or even avoided. The systematic errors caused by neglecting variations in activity coefficients were estimated by using equations (3), (4) and (14) to generate theoretical e.m.f.s for the three stages of the known addition procedure (with activity coefficients being calculated as below) and then using these e.m.f.s in equation (12), which assumes that the activity coefficients are constant, to give the apparent determinand concentration. Equation (12) was evaluated first with S equal to the Nernstian slope factor then with S derived from equation (17).These calculations were carried out for many combinations of the parameters Vo, CA, V, C, and VD for determinands with :barge z = 1 and z = 2 and with background media of various ionic strengths p in the sample solution, p’ in the standard solution and p” in the diluent solution. The total ionic strength, Z = 0.5Ez;Ci was calculated at each stage from the above parameters. Thus if the determinand and its counterion Table 4. Errors caused by calibrating at 25 “C and analysing at T “C Error in CA, TPC Yo 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 4.7 4.2 3.8 3.3 2.8 2.3 1.8 1.4 0.9 0.5 0.0 -0.5 -0.9 -1.4 -1.8 -2.3 have the same charge (in both the sample and standard solutions) the ionic strengths corresponding to equations (3), (4) and (14) are given by equations (22)-(24), respectively.. . . (22) Activity coefficients were calculated11 from equation (25). - logf= Az2 ( sfi- 0.3Z) . . . . (25) where A here is the Debye - Hiickel coefficient. The e.m.f.s were calculated on the assumption that the electrodes had ideal Nernstian responses at 25 “C, i.e., S = 5 9 . W ~ mV decade-1. Correspondingly, in equation (25) A = 0.511 at 25 “C. In all instances V, = Vo + V, CA varied from 10-5 to 10-1 mol 1-1, VdV from 1000 to 20, VC,lVoCA from 0.5 to 2.0 and p varied from 0 to 10-1 moll-1, whereas p’ = 0 or p and p” = 0 or p.Table 5 summarises the main conclusions of this, and Fig. 1 shows the error as a function of determinand concentration and background ionic strength for singly charged ions. Table 5 shows that the results obtained with the calibration slope calculated by known dilution are almost free from errors provided that the diluent has the same background ionic strength as the sample; this is so even if the ionic strength is not the same at all stages of the procedure, because the calibration slope so obtained compensates for the change in activity coefficients. Fig. 2 shows how the calibration slope varies with CA, p and p”. Good compensation is obtained by using the known dilution procedure in these examples, because Ci = OSCL = CA, but in practice it is not possible for the addition always to double the initial concentration and compensation then becomes less efficient for a fixed dilution.Taking the worst instance (p = 0), if Ck, = 1.5CA, i.e., VC, = 0.5 VoCA, the errors are three to ten times larger, but still ~ 1 % . If CL = 3cA, i.e., VC, = 2vOcA, the errors are twice as large, but in the opposite direction. It is therefore better to arrange the addition so that VC, c VoCA as the initial concentration will then lie within the Table 5. Concentration errors caused by variations in activity coefficients. Errors are overestimates unless marked by an asterisk Conditionst Slope (S) from calibration graph Slope (S) from known dilution$ Singly charged ions: No background electrolyte Dilute background electrolyte (p = p’ = p” = 0) p = c ., (~10-4) 1.~ = CA 010-4) -1% at CA = 10-4 -3% at CA = 10-3 e l % 2-5 Yo <1% if prf = 0; <*0.05% if prf = p 2-5% if p” = 0; C*0.05% if p” = p Concentrated background electrolyte 1OocA > 3 1OcA <l% if VdV2 100 <0.05% if 20 S VdV < 100 p 1OoCA <*0.1% if VdV2 100 2-10% if prr = 0; <*0.01% if pf‘= p <*0.7% if 20 3 VdV C 100 2-9% if p’’ = 0; <*0.01% if pr’ = p Doubly charged ions: No background electrolyte Dilute background electrolyte p = p’ = pff = 0) CA (~10-4) p = CA 010-4) 7 yo 2040% <1% if prr = 0; <*0.05% if prr = p >8% if prr = 0; <*0.05% if p’’ = p *8-30% if pfr = 0; <*0.2% if pfr = p *&30% if p” = 0; C*o.3% if p” = p t Variations of pr between 0 and p caused ~ 0 .1 % error in any instance.Concentrations in moll-1. $ V , = V + V,. Errors increase as VdV decreases, but the difference is <0.1% in the range 20 6 VdV d 1000.ANALYST, MAY 1987, VOL. 112 561 12 0.0 10 8 6 4 2 8 $ 0 t l .- P O g -2 4- - -4 -6 -8 -10 I 1 1 I I ( b ) - - - 0.0 0.1 0 Fig. 1. Errors caused by ionic strength variations in determination of univalent species: (a) Nernstian calibration; (b) known dilution calibration with VD = Vo + V and p” = 0. Background ionic strength p = 0 (0), 10-4 (+),.lO-3 (A), 10-2 (X) and 10-1 moll-1 (0 . Lines are drawn through points of equal CM/p and labelled accordngly. V = O,OlV,; vc, = VOC, 59 57 r I al D V : 55 > E s $ 59 . c r 0) P v) - 57 55 100 97.5 C 95 .; f z 92.5 8 I 100 g! 97.5 ; a Q) r a 95 92.5 Fig.2. Effect of ionic strength on slope factor by known dilution (univalent determinand) V, = Vo + V, V = O.OIVo, VC, = VoCA. (a) Same background in sample and diluent, p” = p. (b) No background in diluent, p” = 0. p = 0.0 (A), 10-4 (B), 10-3 (C), 10-2 (D) and 10-1 moll-’ (E). Solid lines join points with errors <l%, broken lines join points with errors >1% range over which the calibration slope is determined. Alterna- tively, the correct dilution to ensure that Ci = CA may be obtained from the e.m.f. change on addition (Table 3) and good compensation for ionic strength effects should ensue. If a background electrolyte of sufficient concentration (p 3 1oocA) is present in the sample (either naturally or by the addition of an ionic strength adjustment buffer), the use of the calibration slope obtained from standard solutions will give good results. The slope obtained by known dilution should not be used unless p“ = p.Provided that the total ionic strength is low (p + z~CA < 5 X 10-4) either type of calibration slope may be used with less than 1% error for univalent determinands, but for accurate work with divalent determinands at any ionic strength only the slope obtained by known dilution is recommended. A background electrolyte is unnecessary in the standard solution added to the sample. Ideally, v’ + Z’C,“ p + z~CA . . . . . . (26) as the activity coefficients would then be virtually constant. If a background electrolyte is added so that p >> CA, it is relatively simple to satisfy equation (26), but the improvement in accuracy is <0.1%.If the samples’are known to have a moderate, but constant, background ionic strength, it may be more convenient to satisfy equation (26) than to add reagent to each sample solution to give a high ionic strength. If equation (26) is satisfied, the calibration slope obtained with separate standard solutions will give accurate results. Effect of Complexing So far it has been assumed in calculating CA from equation (12) that aA = @A. When this assumption breaks down, however, gross errors may arise, depending on the concentra- tion of determinand, CA, the concentration of ligand, CL, and the stability constant, p, of the complex formed between them. In this context, ligand is defined as any substance that forms a complex with the determinand.With anion-selective electrodes, therefore, the ligand is usually a metal ion. The presence of complexing agents calls for great care in the use of the addition - dilution method, because even if a A = aA, d i may be considerably different. Let us consider a simple system containing a single type of ligand that does not participate in side reactions, e . g , protonation, hydrolysis or complexing with other species in the sample solution. Assuming that the activity coefficients are constant, the stability constant is defined by equation (27) P=[AL]/[A][L] . . . . . . (27) The total determinand and ligand concentrations are given by equations (28) and (29) CA = [A] + [AL] = [A] (1 + p[L]) . . . . (28) CL = [L] + [AL] = [L](l+ p[A]) . . . . (29) Now, a,=-=- [A1 1 .. . . . . (30) cA l+p[L1 From equation (30) it can be seen that any process changing the free ligand concentration, [L], will change aA, e.g., dilution or the addition of more determinand. Hence, when a sample solution contains a complexing agent, aA and dA will never be identical with aA, but in practice there are wide ranges of conditions in which the differences are negligible. Calculation with the Nernst slope factor Fig. 3 shows as functions of p, CA and CL the theoretically calculated errors arising from the use of equation (12) with S equal to the electrode’s Nernst slope factor; the conditions are VC, = V&A and VlV, = 0.01. The errors are small (<2%) if PCA d 10-2, but increase with pCA, becoming unacceptably large unless CL/CA 3 102.If (xA and aA in equations (3) and (4) cannot be equated, we obtain instdad of equation (12)562 ANALYST, MAY 1987, VOL. 112 Strongly complexing conditions, p[L] >> 1 The condition p[L] = p[L]‘ >> 1 corresponds to the presence of either a strong ligand (p > 104) at moderate to high concentrations or a weak ligand at high concentrations. To obtain [L] = [L]’, the addition step must involve no significant reduction in the free ligand concentration; keeping this reduction to acceptable levels depends on the ratio of ligand to determinand and on the degree of dilution. (i) Ligand - determinand ratio. Assuming that dilution is negligible, [L] = [L]’ only if the amount of complex formed is negligible compared with the free ligand concentration, i. e., [AL] << [L] = CL and [AL]’ << [L]’ = CL.In these strongly complexing solutions, [AL] = CA and [AL]’ = CL, hence we require the conditions CA << C L and CL << Ci. If the analyst has no expected values for CA and CL, a test based on the initial e.m.f., E l , can be applied, provided that a calibration graph has been prepared with solutions containing no complexing agent. The graph used to find the calibration slope should be adequate for this purpose, even if it has to be extrapolated. From equation (27), [AL] = P[A] [L], but [AL] << [L] and, therefore, p[A] << 1. The free determinand concentration, [A], is estimated from the calibration graph. If P[A] d 10-3, the ligand - determinand ratio is high enough, provided that the addition step is of normal proportions, i.e., VC, = VOCA. A more conservative procedure is to make the addition and test @[A]’ d-10-3, where [A]’ is estimated from the e.m.f., E2.The test @CA d 10-3 for weakly complexing systems is not appropriate here. (ii) Dilution. In strongly complexing conditions with a large ratio of ligand to determinand, [L] = C L and [L]’ = Ci. From equation (30), if p[L] >> 1 and p[L]’ >> 1, \ Fig. 3. Errors caused by complexin (Nernstian calibration). V = O.OlV0, VC, = VOC,, CL/CA = 0.1 (b), 1 (A), 10 (D), 100 (E) and (F) If eA denotes the apparent concentration calculated from equation (12), we have - CAx CAEA/E~ . . . . . . (32) Whereas the approximation (32) breaks down if E~/(YX deviates greatly from unity, especially if antilog (E2 - E1)/S is less than about 1.5, it is useful as a basis for showing when the use of equation (12) cannot be justified.This is important because it may be inconvenient to do the extra calculation involved in equation (31) or the calculation may be impractic- able unless a separate determination is made of the ligand concentration. In many instances the errors may be reduced by a simple modification to the analytical procedure (see below). From equations (30) and (32) we have Now [L]’ < [L] because of dilutign and complexing with the added determinand. Therefore, CA < CA, but equation (33) shows that there a_re two ranges of extreme conditions where EA/EL = 1 and CA is a good approximation to CA. These favourable conditions occur when either p[L] << 1 and p[L]’ << 1 or when p[L] = P[L]’ >> 1, i.e., when complexing is either very weak or very strong.Weakly complexing conditions, p[L] << 1 With weak ligands (p d 103) in moderate to low concentra- tions, the degree of complexation is low and [L] = CL. Equation (12) will give a <0.1% error if PCL < 10-2, and if pCL S 10-1 the error should not excged 1%. If CL is unknown, the answer given by equation (12), CA, may be tested directly. It follows from equation (29) that if [L] =r CL, P[A] <,< 1. AKsuming that with weakly complexing ligands, [A] CA, if pCA S 10-2 the errors will be small. For conditions in which VC, = VoC, and V/Vo 6 0.01,a maximum error of 2% may be expected (Fig. 3); even if pCA = 0.1 the error should be no more than 5%.The condition p[L] << 1 may also be satisfied at very low concentrations of strongly complexing ligands, such that CA > 1OOCL. With these ligands, if CA > CL, [AL] =r CL and the error in CA is -100(1 - CL/CA)%. Because CL = [AL] >> [L], the test pCL d 10-2 used for weakly complexing lbands is very conservative, although still valid. The pCA test, however, is not valid with strong ligands. In equation (12), therefore, dilution will cause eA to be in error by - lOOV/( Vo + V)% and it is desirable for V/Vo not to exceed 0.01. Complexing conditions such that aA/aL f 1 When 10-2 < p[L] < 102, equality of aA and (YL cannot be assumed and the approximate equation (32) becomes unreli- able. Fig. 3 shows the large errors that occur as pCA increases when 10-1 d CL/CA S 10.Errors may be reduced by three procedures, depending on conditions. (a) Diluting the sample so that pCA d 10-2 reduces the errors to less than 2% if V/Vo = 0.01. This procedure is more likely to be useful with weak ligands than strong ligands because in the latter instance the ratio CL/CA also determines the size of the errors but is unaffected by dilution. If CA is small, care should be taken that dilution does not reduce the determinand concentration to such an extent that the elec- trode is operating in its non-Nernstian response region. (b) By adding more ligand, we can attain the conditions CL >> CA and p[L] >> 1 required for the successful use of equation (12) in strongly complexing conditions. A 100-fold excess of ligand can be recommended (see Fig.3). Instead of increasing the concentration of the ligand actually present in the sample solution, it is often permissible to add an excess of a second strongly complexing ligand, e.g., EDTA for transition metal determinands. If the analyst has to deal with a wide variety of samples it may be convenient always to add the same ligand, hence avoiding having to prepare a range of different solutions. Mixtures of ligands are discussed below. (c) As an alternative to dilution, the degree of complexing of the determinand may be reduced by adding a reagent that reacts preferentially with the ligand. This amounts to intro- ducing a conditional stability constant, PC < p into the previous equations, so that pcCA d 10-2. This effect will often be brought about by a change of pH.This subject is discussed below.ANALYST, MAY 1987, VOL. 112 563 Compared with dilution, this procedure has the effect of increasing, rather than decreasing, the free determinand concentration, which may be advantageous near the elec- trode's limit of detection; on the other hand, interferences are more likely to be introduced. Calibration by known dilution No ligand in diluent. Fig. 4 shows as functions of p, CA and CL the theoretically calculated errors arising from the use of equation (12) with S determined from equation (17); the conditions are VCslVoCA = 1, V/Vo = 0.01, V, = Vo + Vand no complexing agent in either the additive or diluent solutions. Unless conditions are such that complexing is scarcely significant, the errors are unacceptable for most purposes.If CL C cA, the maximum error possible is 1oo(cA - CL)/CA, but if CL 2 CA, errors of almost - 100% can occur. The reason for the errors when CL 2 CA is that the calibration slope obtained from equation (17) is inaccurate, because ak f d.; the deviation of the slope from the true Nernstian value is indicative of the size of the analytical error (see Fig. 5), except where CL = CA (line A). In this instance the slope factor goes 0 - 20 8 2 ti -40 9) .- c. 3 -60 a - 80 -100 -5 -4 -3 -2 -1 0 1 2 3 Log (PCA) through a minimum as increases and may approach Nernstian values at large values of @ even though the error is large. The errors shown in Fig. 4 are reduced if the ratio VCJVOCA is decreased or if V/Vo is increased, but such reductions are too small to be of practical use.Fig. 5 shows that if the slope factor is less than 95% of its expected value (a deficiency of 312 mV decade-' for a determinand of charge z ) the results should be regarded with caution: this rule of thumb will even catch most instances where CL = CA (line A in Fig. 5). Ligand in diluent. If the diluent contains the same concen- tration of ligand as the original solution, the errors in the use of known dilution to calculate S can be largely compensated. Fig. 6 shows that for VD = Vo + V and VCs/V&A = 1 (the optimum conditions), the error is at most 1% when V/Vo = 0.01 and CL 3 1OcA. For CL = 0,lCA the error is very small. The errors are approximately proportional to V/Vo. Fig. 7 shows how the slope factor appears to be greater than Nernstian, the discrepancy being greatest at CLICA = 1.The apparent deviation from Nernstian response increases as V/Vo increases. 0 8 2 L' & -0.5 9) .- 4- - 9) a -1 Fig. 6. Error caused by com lexing (known dilution calibration with li and in diluent: V, = V + 10; VoCA = VC,). CL/CA = 0.1 (A), 1.0 (i), 10 (C), 100 (D) and lo00 (E) Fig. 4. Errors caused by complexing in addition - dilution method with no li and in diluent. VD = Vo + V, V = O.OIVo, VC, = V0CA, C,/CA = %.l (G), 1 (A), 10 (D), 100 (E) and lo00 (F) I B 0 - 20 8 t -40 9) .- + -60 a - 80 -100 80 60 40 20 Slope factor, O h of Nernstian Fig. 5. Effect of corn lexing. Relationship between relative error and slope factor in aBdition - dilution method with no ligand in (B), 3 (C), 10 (D) and 2100 (E) diluent.VD = Vo + V, V = O.OlV0, VC, = VOCA, C,/CA = 1 (A), 2 I C C i r t I I I I 115 .i 4- E Z w- 111 8 L c. 0 m 9) * n 105 100 -4 -2 0 2 4 6 8 Log (BCA) Fig. 7. Effect of complexing on slope by known dilution with li and in diluent (V, = V + Vo; VoCA = VC,). CJCA = 0.1 (A), 1.0 (BT, 10 (C), 100 (D) and lo00 (E)564 ANALYST, MAY 1987, VOL. 112 The method can sometimes even cope with the condition CL = CA. For PCA < 103, the error is <1%. For PCA > 1, the slope factor will be a multiple of the Nernstian value because the dilution step involves going from a state in which the determinand is in excess to one where it is almost completely complexed. If the slope factor is greater than 70/)z1 mV decade-1, the result should be treated with suspicion, as errors may be very large for PCA 3 103.If CL = CA and pCA 2 103, it is desirable either to increase the ligand concentration so that CL b 1OcA or to dilute the sample so that p c ~ < lo3. The results are more diverse in less ideal conditions, where the dilution step does not restore the solution almost to its original state, i.e., if, for V, = Vo + V , VC, # VOCA. When VC, > V0CA the apparent slope factor exceeds the Nernstian value by even more than that shown in Fig. 7, and conversely for VC, < VoCA. The effect on the analytical result can be seen in Table 6. A high value of CL/CA is needed to ensure that errors are no more than about 2%. If CL = CA and VC, > VoCA the errors may approach -100% if pCA > 1; this is, however, indicated by a slope factor that may be double the Nernstian value.If CL = CA and VC, < VOC,, the slope factor will again be greater than Nernstian if PCA > 1, with errors varying from about -5% at pCA = 10 to +15% at pCA > 104. Using Table 3 to calculate the dilution necessary to restore the original concen- tration of the determinand would keep these errors to a minimum. In view of the above results, whenever the slope factor is greater than 70/1z1 mV decade-1 the conditions should be adjusted so that either there is a large excess of strong complexing agent at all times or the value of pCA is reduced by dilution. Multi-ligand systems Real sample solutions, as opposed to the simple models discussed so far, are likely to contain more than one ligand.If the ith ligand, Li, forms a complex with a stability constant pi we would have instead of equations (28) and (30) CA = [A] + [ALi] + [AIq] + . . . = [A](l+ 2:(3i[Li]) . . (34) 1 The arguments developed from equations (27)-(33) for systems containing only one ligand are transferable to the multi-ligand example. Instead of equation (33), therefore, we have equation (36) and equation (12) will give good results if either (a) Zpi[Li] << 1, ZPi[Li]’ << 1 or (b) Cpi[Li] = Epi[Li]’ >> 1. With weakly complexing ligands (p < lO3), condition (a) is met if EpicL, d 10-3 moll-1 and for many purposes XPiCLi < 10-2 would give acceptable results (<1% low using the Nernst slopefactor). If the total ligand concentjations, CLi, are unknown, CA may be tested directly.If Pmax.CA G 10-2, where Pmax. is the largest of the pi, errors will be <2%. With strongly complexing ligands (P > lO4), the simplest approach is to treat the system as if it contained only one ligand, that giving the largest value of the product piCLi. This is a good approximation if (PiCLJmax. 2 1Ois . P i c ~ i . . . . (37) max The treatment for systems containing one ligand then gives adequate guidance to the accuracy. In strongly complexing solutions the ligand for which picLi is maximal must be in a large excess over the determinand. If (cLi)mm, < 1oocA errors may be large even though other ligands are in large excess over the determinand. If aA # aA, the simplest remedy is to add a ligand in excess so as to satisfy condition (37).The ligand added need not be present naturally in the sample. Table 6. Limiting error (YO) under different experimental conditions for V,, = V, + V and VlV, = 0.01 VCJV,C* 10 100 1000 0.5 -4.1 -2.3 -2.0 1.0 -1.1 -1.1 -2.0 2.0 6.4 0.2 -0.2 If the calibration slope is determined by known dilution, matching the composition of the diluent to the sample may be impracticable unless condition (37) is satisfied either naturally or by the deliberate addition of an excess of one ligand to the sample. Before this procedure is used with real samples thorough testing of model systems is recommended. Effect of side reactions Equations (28)-(37) were developed on the assumption that the ligands reacted only with the determinand. In practice, the ligands are likely to react with other species in solution and particularly with hydrogen ions. When such side reactions occur, the equations containing stability constants, p, can be re-written with conditional stability constants,lz pc = P/aL, where (xL includes terms for all the other interactions of the ligand L.For example, if L forms a complex with ion X having stability constant px, and participates in protonation equilib- ria having constants K1, K2, etc., @L = 1 + K1[H] + K2[H]2 + . . . +px[X] As [HI and [XI will change on the addition of the standard and diluent solutions, aL and, hence, pc will also change. Values of aL for the more common metal ions and ligands are available in tabular12 and graphical13 form. The change in [HI depends not only on dilution but also on the buffer capacity of the system, making a general treatment of errors arising from this source too complicated to be of much use.If the ligand participates in protonation equilibria, the pH should be kept constant. Changes in pH are undesirable in any instance, because they are usually the biggest single influence on the liquid junction potential, which should always be kept as constant as possible. Calculations with Nernstian slope. Errors arising from variations in [XI depend on the experimental conditions. If the determinand is weakly complexed (p[L] << l), variations in [XI are of no significance; in so far as the formation of the XL complex reduces [L] , errors will be smaller than in the absence of x. If the determinand is strongly complexed, the accuracy of the known addition method depends on maintaining a high ratio of CL to CA.The formation of the XL complex is equivalent to reducing this ratio and a more conservative procedure is to keep a high ratio of CL - Cx to CA, where Cx is the total concentration of X. The test P[A] S 10-3 should be replaced by Pc[A] G 10-3. Calculations with slope from known dilution. If the determi- nand is weakly complexed, the presence of a competing substance reduces the errors compared with those discussed above. If the determinand is strongly complexed, the presence of a competing substance is equivalent to reducing the ratio CL/CA and it is desirable to have C,CX S 1OcA if the errors are to be within the limits described above. Liquid Junction Potential The liquid junction potential depends on the concentration and mobility of all ionic species in both the sample and reference junction solutions, and, therefore, even the qualita- tive effect of inconsistency of the liquid junction potential during known addition is very variable.In practice the factorsANALYST, MAY 1987, VOL. 112 influencing the activity coefficients and complexing would simultaneously affect the liquid junction potential; however, the precautions necessary for controlling errors from these sources would also tend to keep the junction potential constant, i.e., the maintenance of a high, constant, ionic strength, a constant pH and an effectively constant ratio of ligand to determinand. As an example of the errors that could be caused by liquid junction potentials, calculations by means of the Henderson equation have been made for the most commonly used reference electrolyte (saturated potassium chloride) forming a junction with a sample solution containing only the salt of the determinand ion.The effect of variations in activity coeffi- cients has been neglected. The size of the volume increment affected the error by ~ 0 . 1 % for V/Vo d 0.05 and the results were valid throughout the range of concentration increments tested (0.5 d VC,lVoCA d 2 ) . Calibration by known dilution [equation (17)] compensated for changes in the liquid junction potential so well that errors exceeded 0.1% only for high concentrations of determinand (210-1 mol 1-1). The calibration slopes were sub-Nernstian (57-58 mV decade-') for cation determination and super- Nernstian (-60 to -61 mV decade-1) for anion determina- tion.The deviation from Nernstian response increased with increasing concentration when the anion in the sample was more mobile than the cation, but decreased when the cation was more mobile. These trends in the results for calibration slope are valid only for reference junctions in which the anion is more mobile than the cation, e.g., potassium chloride. If the cation were the more mobile ion in the reference junction (e.g., potassium nitrate) the trends would be reversed. The use of the slope from a separate calibration graph should give accurate results if the conditions of the calibration match those of the analysis, otherwise errors of several per cent. may occur. Using the ideal Nernstian response with the data calculated above for a saturated potassium chloride junction produced errors of 2-5% (overestimates for cations, underestimates for anions).If the cation in the reference junction were more mobile than the anion, the procedure would underestimate cationic concentrations and over- estimate anionic concentrations. Measurements in the Non-Nernstian Response Region At low concentrations, the responses of ion-selective elec- trodes cease to be Nernstian.14 Two types of deviation will be considered here, that caused by the solubility of the elec- trode's material and that caused by the blank associated with the electrolyte added to the sample to maintain the ionic strength. Deviations caused by solubility This type of behaviour is usually associated with electrodes incorporating sparingly soluble isovalent salts, e.g., AgCl, AgBr and AgI, and this will be the only instance considered.The non-Nernstian responses of many solid-state ion-selective electrodes are not governed by the solubility product mechan- ism alone and the following treatment would not apply, e.g., Cu2+, Cd2+, Pb2+, F- and S2- electrodes. Consider an electrode, consisting of a sparingly soluble isovalent salt, immersed in a medium of constant ionic strength. Let the solubility product of the salt be Ks. The total concentration, Ct, of determinand at the surface of the electrode depends on the concentration in the sample solu- tion, CA, and the concentration that has dissolved from the electrode. It has been shownl4J5 that this concentration is given by equation (38) .. . . (38) 565 The systematic errors caused by solubility were estimated by using equations (3), (4) and (14) to generate theoretical e.m.f.s for the three stages of the known addition procedures, but CA, C i and cli in these equations were replaced by Ct, C{ and C":, calculated from equation (38). The e.m.f.s so calculated were used in equation (12) to give the apparent determinand concentration and in equation (17) to give the calibration slope. Ej, aA and fA were assumed to remain constant throughout. Errors calculated using the calibration slope obtained by known dilution are shown in Fig. 8 as a function of the ratio CA/K,I for the experimental conditions VdV = 100, VD = V + Vo and VCs/V&A = 0.5 (line X), 1.0 (line Y) and 2.0 (line Z).Errors calculated using the Nernst slope factor are shown as line Y', corresponding to the conditions for line Y: the errors in conditions corresponding to lines X and Z are not shown but are slightly larger and slightly smaller, respectively, than those in line Y'. Using the Nernst slope factor, errors of less than about 2% can be expected only if CA > 1OK,1. Using the calibration slope obtained by the known dilution method, errors of less than 1% are achieved if the dilution step restores the determinand concentration to within 1% of its original value in the sample. This would rarely be achievable in practice and errors of less than 1% are obtained only if CA 2 10Ka when the final determinand concentration is 75 or 150% of the initial concentration.In general, therefore, results indicating CA < 10K$ should be treated with caution. As a rule of thumb, the error, A, in the determination of CA < 2K$ is given by equation (39) A X i(E3 - El) ~K&CA% . . . . (39) where i = k1 and is positive when the e.m.f. becomes more positive with increasing concentration. For CA > 2K$ the error is smaller than that estimated from equation (39). Equation (39) is valid for ratios of VlVo up to at least 5% : if VC, d VOCA, the error is minimised by keeping V/Vo small (-0.1%); but for VC, > V0CA values of V/Vo = 5% would be preferred. The effect of the volume ratio V/V, is small (1-2%) compared with that of the ratio VC,/VoCA (see Fig. 8 ) . Only when VC, = VoCA does V/Vo became significant in determin- ing the total error; in such instances V/Vo = 1% is a good compromise value.In conclusion, at concentrations where the solubility of the salt in the electrode is significant, the addition - dilution method gives smaller errors than the addition method using -20 ' I I I 0 5 10 15 CAf Ks+ Fig. 8. Error caused by solubility product. Known dilution calibra- tion (VD = Vo + V) with VCsIVoCA = 0.5 (X), 1.0 (Y) and 2.0 (Z). Nernstian calibration (Y') with VCs/VoCA = 1 .O566 ANALYST, MAY 1987, VOL. 112 the Nernst slope factor and also indicates that results may be biased because the calibration slope so obtained deviates greatly from the Nernstian value (Fig. 9), although similar deviations are caused by reagent blanks (cf. Fig. 11). Deviations caused by reagent blanks Reagents added to the sample, e.g., for maintaining the ionic strength, may contain material that would cause the calibra- tion of the e.m.f. against the logarithm of the nominal concentration of determinand to deviate from linearity at low concentrations.This material may include interfering sub- stances or determinand carried in the reagents as an impurity. Such interferences can often be the limiting factor in the response of glass and liquid ion-exchange electrodes , whereas determinand introduced as an impurity will affect all kinds of electrodes. Note that the effects described here are different from those caused by interferences present in the sample itself. When an electrode is influenced by reagent blanks the e.m.f. may be expressed by equation (40) E = + SlogaA + slog fA + Slog(cA + B) + Ej - &ef.* (40) where B represents the reagent blank and may be resolved into components as follows (bjf;. &i )zAtzi UAfA B = bo+Z Kj . . . . (41) where bo is the concentration of determinand introduced as an impurity and, for the ith interferent, zi is the ionic charge, bi is the concentration,f;. the activity coefficient, aj the ratio of free to total concentration and Ki is the selectivity coefficient of the electrode for that interferent. The systematic errors caused by reagent blanks were estimated by using equations (3), (4) and (14), with CA, C i and Ck replaced by CA + B, Cb, + B' and C'A + B", to generate theoretical e.m.f.s for the three stages of the known addition procedure. Ej, &A and fA were assumed to remain constant throughout.The e.m.f.s so calculated were used in equations (17) and (12) to give the calibration slope and the apparent determinand concentration. This was done for the instances where (a) only the sample, (b) sample and diluent and (c) sample, solution and diluent contained the reagents that gave rise to the blank. Calibration slope obtained by known dilution. Errors calculated using the calibration slope obtained from equation (17) when the reagents are absent from the standard solution but present in the diluent are shown in Fig. 10 as a function of the ratio CAB for the experimental conditions VdV = 100, V , = V + Vo and VCs/VoCA = 0.5 (line X), 1.0 (line Y) and 2.0 (line Z). When the reagents are present in the standard solution also, the errors are much the same, being slightly smaller when VCs/VoCA > 1 , slightly larger when VC,/VOCA < 1 and equal but opposite in sign when VCs/VoCA = 1.If the reagents are included in neither the standard solution nor the diluent, equation (17) gives a calibration slope equal to the Nernst slope factor for the electrode and equation (12) gives an apparent determinand concentration equal to CA + B , i. e., the blank appears as a positive bias. For the addition - dilution method to work successfully, therefore, the reagents must be added to the diluent, but there is no point in adding them to the standard solution. This conclusion broadly matches those reached for ionic strength effects, the control of which is one of the main reasons for adding reagents, but is much more important.An indication of the importance of the composi- tion of the diluent is the large deviation of the calibration slope from the Nernst slope factor (Fig. 11). If VC, = VOC,, the error is roughly proportional to the volume addition ratio, V/Vo, but deviation of VCJVoC, from unity (which itself has a much greater effect on the errors than variations in V/Vo) obscures any simple relationship. V/Vo = 0.01 gave the best results over a range of VCJVOCA. 60 7 50 U 0 % s o > E 2 \ 30 - Q 20 rn % ru - - 10 0 5 10 15 CAI Ks' Fig. 9. Effect of solubilit roduct on slope by known dilution (VD = Vo + V). VCJVOCA = O.JrX), 1.0 (Y) and 2.0 (Z) 100 75 8 50 L' UI 2 25 0 -10 I 1 1 I 0 5 10 15 CAI8 Fig. 10. Error caused by reagent blank with Nernstian (N) and known dilution (X, Y, 2) calibration.VCs/VoCA = 0.5 (X), 1.0 (Y) and 2.0 (Z); VD = Vo + V Fig. 11. vo + v) c C Q .- c. E - 75 z" + 8 -50 0 cn C - - 25 '@ rr - 9 d ~ _ _ 0 5 10 CAI8 Effects of rea ent blank on slope b known VCJVOCA = %.5 (X), 1.0 (Y) and $0 (Z) dilutionANALYST, MAY 1987, VOL. 112 567 Because the calibration is not Nernstian, Table 3 cannot be used to predict the desired degree of dilution in the known dilution step. Within the range 2 2 VC,/VoCA 3 0.5, if the calibration slope is greater than 85% of the Nernstian value, errors will not exceed 5% for a fixed 2-fold dilution. At lower values of the calibration slope, large (>lo%) errors should be expected unless lEg - Ell < 1/1z1 mV. Calibration with the Nernst slope factor.The use of the Nernst slope factor in equation (12) gives an apparent determinand concentration which is equal to CA + B when the reagents are not present in the standard solution (line N, Fig. 10) and only slightly smaller when the reagents are present in the standard solution. The Nernst slope factor should not be used unless an error of 100 B/CA'XO is acceptable. The use of the Nernst slope factor also has the disadvantage that it does not indicate that bias is possible, whereas the calibration slope from equation (17) deviates considerably from the theoretical value when B is of the same order as CA (or when there is a significant solubility effect). If a calibration graph is prepared properly, i.e., using the same procedure as applied to the samples, the range over which the Nernstian response is valid becomes self-evident.Source of the blank. Calculation of the apparent determi- nand concentration from equation (12), as above, corresponds to the reagents being present in the sample solution, e.g., when a solid sample is dissolved or a gaseous one absorbed. When reagents are added to the sample solution, equation (12a) would be appropriate, but all the above arguments would still be valid. Figs. 10 and 11 would then represent the errors and calibration slopes obtained in conditions exactly corresponding to those above, i.e., (Vo + VR)/V = 100, and V , = Vo + V + VR, provided that B is regarded as the notional blank in the original volume of sample solution Vo. This is readily seen if the mixture of sample and reagent solutions is thought of as aAvolume, v0 = Vo + VR, of a new sample of concentr+tion CA = CAV,-Jvo with a blank 6 = BV,-Jvo, because CA/B = CA/B.The accuracy with which the above treatment can predict the error caused by reagent blanks depends on the nature of the blank. If B = bo, i.e., no interferents are present, the predictions of the theory should be accurate. When interfer- ents contribute to the blank, however, the outcome is less predictable because selectivity coefficients often vary with concentration. Bias caused by interferences in the sample Interfering substances present in the sample influence the e.m.f. of the electrode according to equation (42). E=EO+logaA+logfA+Slog(CA+eI) . . (42) , and Ki is the selectivity where eI = X coefficient of the electrode for the ith interferent of concentra- tion Ci and charge zi with activity coefficient h, (Y; being the ratio of free to total interferent.Calibration with the Nernst slope factor. The determinand concentration calculated from equation (12) will be over- estimated, so that Ki( c$iCYi)zAlzi a A f A - CA - CA = eI . . . . . . (43) Equation (43) is accurate for ZA = zi, provided that Ki, f i and (xi are constant at all stages of the procedure. In regard to f i and ai this requirement imposes no additional condition, as fA and aA must also be constant. Ki may change with concentra- tion, but this is unlikely to be significant over the approxi- mately 2-fold concentration range covered in the known addition procedure. Even if ZA # z i , equation (43) gives a good (k YO) estimate of the error in concentration provided CA 2 81.As the interference term increases beyoEd CA, equation (43) becomes less accurate. If zA < zi, CA is less than expected from equation (43) and conversely for zA > zi. Calibration by known dilution. If zA = zi, equation (43) is valid, with the same provisos as before. If zi = 22A, use of the known-dilution calibration reduces the interference effect, so that CA - CA = 0.6 01. The slope factor becomes increasingly sub-Nernstian as eI increases relative to CA. If zA = 2zi, the interference effect is increased by the use of the known dilution calibration, which becomes increasingly super-Nernstian as €II increases, so that CA - CA 1.7 01 If known interfering substances are systematically present in the sample solution at known concentrations, the error may be reduced in the known addition - dilution procedure by using a diluent that matches the sample solution in the concentrations of interfering substances present. The error will then be the same as that obtained when the interferents are introduced with the reagents.Random Errors The random errors in the calculation of the determinand concentration from equation (12) arise from uncertainties in measuring the e.m.f. of the electrode, in determining the calibration slope, in delivering the volumes Vo and V and in the concentration, C,, of the standard solutions. With ordinary laboratory techniques, the errors in VO, V and Cs should be small, both in absolute terms (0.1-0.2%) and relative to the errors associated with the electrode.Let equation (12) be rewritten as follows C A = p / s = p / ( q r - 1) . . . . . . (44) (E2 - El) wherep = C,V/Vo, q = 1 + V/Vo and r = antilog . Let the standard deviation of a single determination of a quantity, x , be denoted by a(x) and the corresponding relative variance be denoted by R(x) = u2(x)/x2. Then from equation (44) and the rules for combination of errors, we have R(cA) = R(p) + R(s) 42'2 = R(p) + [R(q) + R(r)] . . . . (45) Now R(p) = R(Cs) + R(V) + R(Vo) . . . . (46) (47) ( = R(r) = (ln10)2u2 Progression from equations (45)-(48) requires some assumptions to be made about the source of the variability of (E2 - El). In all instances the error in (E2 - E l ) will be treated as being independent of the errors in V , Vo and C,, because these quantities are invariably known much more precisely than the e.m.f.s or slope factor and make a negligible contribution to R(E2 - El).The first treatment assumes that the variability of the e.m.f. is the dominant source of error. It would be applicable if the uncertainty in the e.m.f.s arose from the difficulty of actual measurement, e.g., because of electrical noise caused by a bad liquid junction, a damaged sensing electrode or inadequate screening, or simply because a meter of inadequate sensitivity was used. If the e.m.f. can be measured with good precision but established with imperfect reproducibility, the first treatment becomes dubious, because the absolute value of El is irrelevant to known addition potentiometry; this example may be typically envisaged as one where the liquid junction cannot be re-formed exactly as before or the temperature has changed.A different treatment is presented for this instance and then the two are compared with experimental data.568 ANALYST, MAY 1987, VOL. 112 Error mainly caused by the impreekion of e.m.f. It has been assumed7J6 that u2(E2 - El) = u2(E1) + u2(E2) and in most instances it is reasonable to assume that a(E1) = a(E2) = UE if the potentials are not too different, e.g., the electrodes cited by Midgley.17 Thus R(E2 - El) = 202/(E2 - E1)2, giving Combining equations (45)-(47) and (49), we obtain In conditions typical of the known addition procedure, i. e., CAvoandr=l+- cs v V << Vo and CA << C,, qr/s = 1 + - cs v CA VO Substituting in equation (50), 2 [(l + S) hl1ol2 5 + [k + F) log (1 +%) ln10l2 R(S) (51) With normal laboratory techniques it should be possible to obtain values of R(C,) = R(V) = R(Vo) = 10-6, i.e., with a precision of -0.1 % , which is much better than can usually be obtained for potentiometric measurements.Neglecting terms in R(C,), R(V) and R(Vo), therefore, we obtain 2 [ (1 + z) log (1 + s) In lo] R(S) (52) This equation may be compared with those derived by Ratzlaff16 and Mascini,’ who have overestimated the depen- dence of R(CA) on u 2 by a factor of 2 and with that of Horvai and Pungor,lg who assumed that R(S) = 0. Equation (52) shows that the analytical error in CA increases with CAVdC,V, i.e., the precision is worse for a small increase in concentration leading to a small change in e.m.f. Increasing the amount of determinand added will improve the precision, but very large increases in concentra- tion are more likely to cause bias by changing activity coefficimts, liquid junction potentials, etc., as discussed above. The accepted best compromise is a doubling of concentration in the addition step, and for this instance equation (52) can be more closely evaluated, i.e., for C,VdC,V = 1 and S = 58/zA where zA is the charge on the determinand, R(C,) = 0.0126~A2~2 + 5.7 x ~O-‘ZA~U~(S) .. (53) R(CA) appears to depend more on oE than on a(S), which itself has a dependence on uE expressed by where a, is the standard deviation of the log (concentration) values of n(>2) standard solutions used to calibrate the electrode (n may include replicates).If n is large and the solutions have a wide concentration range, a(S) << UE, leading to a minimum possible value R( CA) = 0.0126 z A ~ . . . . . . (55) If only the simplest, two-point, calibration is used, equation (54) is invalid. Instead, we have uZ(S) = 2u2/(logV)2 . . . . . . (56) where V is the ratio of the two concentrations. By experience and consensus, there should not be less than a 2-fold difference between the solutions and a practical maximum for u2(S) can be established. In this instance, therefore, we have &(S) = 2u2/(log 2)2 = 22a,2 . . . . (57) R(CA) = 0.0126~A2 + (22 X 5.7 X 1 0 - 4 ) ~ ~ 2 = O.025~A2 (58) The practical maximum errors predicted by equation (58) are twice the minimum possible errors predicted by equation (55).Errors calculated from equation (57) are given in Table 7 and show the need for precise readings of e.m.f. Meters reading to no better than +1 mV are scarcely suitable. Errors caused mainly by the imprecision of the slope factor If it is assumed that u(E2 - El) is directly dependent not on uE, but on u(S), we obtain from E2 - El = S log(ci/cA) ci CA u ( E ~ - El) = log - u(S) . . . . (59) which means that R(Ez - El) = R(S) Substituting in equation (45) we obtain Neglecting R(C,), R(V) and R(Vo), and substituting for q, r and s as before [for equation (52)], we obtain For a typical example with CAV0 = C,V and S = 58/zA mV decade-’, R(CA)=0.0011zA2U2(S) . . . (62) a(S) may be expressed in terms of oE as in equations (55)-(57). For the least precise (two-point) practical calibra- tion, we obtain from equation (57) R(CA) = 0.024 Z A ’ C F ~ .. . . (63) Columns 3 and 4 of Table 8 show the errors predicted by equation (62) at various values of u(S). As a(S) decreases below 0.2 mV decade-l, the errors in Cs, Vo and V are no longer negligible. The final column shows the effect of Table 7. Maximum relative random errors in known addition potentiometry with CAVo = C,V, assuming uE is dominant source of error. Two-point calibration with solution concentrations differing by a factor of two. Minimum possible errors are half those in the table Relative error for charges, % oE/mV lZAl = lZAl = 0.1 1.6 3 0.2 3 6 0.5 8 16 1 16 32 2 32 63ANALYST, MAY 1987, VOL. 112 569 including these terms in equation (60).For comparison, the value of aE that would produce each value of a(S) when calibrating with two solutions with a 2-fold difference in concentration is shown in column 2; in that instance the two treatments are equivalent [cf., equations (58) and (63)]. In general, however, these calculations predict that the random errors can depend less on uE than the previous treatment implied. Equation (61) may be compared with Ratzlaff'sl6 equation (16). Apart from changes in notation, Ratzlaff's estimate of R(CA) is only half that given by equation (61) because he assumed that a(E2 - E l ) = 0. E2 - El, however, depends on S and a(E2 - E l ) is non-zero, as shown by equation (59). Ratzlaff's equation (17) which should differ from his equation (16) only in notation, contains either a typographical or algebraic error.Errors caused by imprecision of both (E2 - El) and the slope factor Rice19 assumed that the error in E2 - El could be independent of the error in the slope factor. The analytical error can be expressed by equation (64), assuming also that R(V), R(Cs) and R(Vo) are negligible In calculating R(E2 - El), a(E2 - El) is measured directly and not assumed to be equal to d2aE [in which instance equation (64) would be identical with equation (52)]. Table 8. Random errors in known addition potentiometry calculated assuming u(S) is dominant source of error Relative error for CAVo = C,V, YO ~~ ~ Equation Equation (62) ( W t lZAl = 1 lZAl = 2 IZAl = 1 UE*/ 4 S Y mV decade-' mV 0.05 0.01 0.16 0.33 0.23 0.1 0.02 0.32 0.66 0.36 0.2 0.04 0.66 1.3 0.68 0.5 0.11 1.6 3.3 1.6 1 .o 0.21 3.3 6.6 3.3 * Value of uE giving stated a(S) from calibration with two standard t Evaluated with R(C,) = R(V) = R(Vo) = solutions differing in concentration by a factor of two.Rice showed that at a constant value of a(E2 - E l ) and a(S) = 0.1-0.5 mV decade-1, the relative error in CA had a minimum at values of C,V/CAVO around 1-2. Comparison of observed and calculated errors Table 9 shows published data for the precision of analysis by known addition potentiometry together with values of UE, a(E2 - E l ) and a(S). The values of a(S) were obtained over at least one decade of concentration and usually from several repeat runs. Most types of ion-selective electrode are rep- resented: solid-state (chloride, bromide , fluoride), liquid ion-exchange (nitrate, caffeine and phenothiazines) , gas-sens- ing (ammonia) and redox - solid-state (chlorine).The errors predicted from aE by equations (58) and (53) are always too large, whereas those predicted from a(S) by equation (62) show good agreement with the observed values in a majority of instances. a(E2 - El) is not only less than v2aE but less than UE itself, showing that the first of the above treatments is not appropriate in any of the instances for which results have been reported. If a(E2 - E l ) is less than g2aE, so should be a(& - E3) in the known dilution procedure for finding the slope factor. The values of a(S) for this procedure would then be much less than those predicted by equations (56) and (57) and in consequence equations (58) and (63) would overestimate the random error in the determinand concentration. This could not be con- firmed as no data on known dilution were found in the literature.In two of the three instances for which known addition data were available (chloride and chlorine), equation (64) predic- ted the error very well. As the same data were used to calculate a(E2 - El) and CA, however, the agreement may be better than would be expected for the prediction of future results. In the third instance (fluoride), equation (62) gave a better approximation. It is likely that with a good electrode, constant conditions and a stable solution the assumption that the variance in E2 - El depends directly on the variance in S is a valid one. Equation (62) would give a good prediction in such instances, e.g., Midgley's fluoride data.20 With the chlorine data,27 however, the solution was far less stable, because the determinand was both reactive and volatile.Additional sources of variance in E2 - El were possible, therefore, and equation (64) was the more appropriate. Even with one type of electrode, the relevance of equations (62) and (64) may depend on circumstances, e.g., sample matrices can vary over many decades of concentration of concomitant substances and require various chemical treat- ments before measurement and even an abnormally wide range of ambient temperature during the measurements may invalidate equation (62). Table 9 shows, however, that Table 9. Observed and calculated precision in known addition potentiometry UE/ u(E,-El)/ a(S)/ Determinand Sample mV mV mVdecade- Fluoride20 .. . . Fluoride19 . . . . Chloride21 . . . . Chloride22 . . . . Nitrate23 . . . . Nitrate24 . . . . Ammonia25 . , . . Ammonia26 . . . . Residual chl0rine2~ . . . . Bromide28 . . . . Chlorpromazine29 Promethazine29 . . Perphenazine29 . . Caffeine30 . . . . Boric acid solution Sodium hydroxide Paper machine water Water Drinking water Pickling baths Water Kjeldahl digests solution Sea water Peach extract Solutions and tablets Solutions and tablets Solutions and tablets Analgesic tablets 0.44 - 0.39 0.39 0.5 - - 0.53 0.75 - - - - - 0.24 0.08 0.24 0.49 0.26 0.20 - - 0.17 1 0.32 0.43 0.65 0.7 Relative standard deviation in concentration, YO Observed Eqn. (62) Eqn. (53) Eqn. (58) Eqn. (64) 0.7 0.8 5.0 6.8 1.5 - - - 0.3-0.8 0.3 2.1 0.8 4.4 6.1 2.5 2.3 1.8 4.5 6.2 - - - 7.9 - 2.3 0.6 0.9 1 .o 0.7 1.1 - - - - - - - 8.4 - - 5.3 1.2 8.4 12 5.3 3 3.3 1 .&2.6 1.1 1.1-1.9 1.4 2.1 2.1 1.2-2.6 2.5 - - - - - - - - - - - - - -570 ANALYST, MAY 1987, VOL.112 equation (62) is a generally good predictive tool and its relative simplicity makes it attractive for this purpose. Discussion A number of general rules for the successful application of the known addition method may be extracted from the preceding detail. In discussing electrode performance, sub- and super-Nern- stian refer to empirical slope factors that are , respectively, below and above the theoretical value, although the linear relationship between the e.m.f. and the logarithm of the concentration or activity may still be valid.A non-Nernstian response is one where this relationship has been established as having broken down. Use of Nernstian Calibrations Applying a fixed value of the slope factor is liable to cause errors of several per cent. except in fairly restricted circum- stances. The temperature of analysis should be within 1 "C of the calibration temperature. The sample should contain, or have added to it, a background electrolyte about 100 times the concentration of the determinand if activity coefficient effects are to be avoided; this same step will also reduce the effects of liquid junction potentials. If the determinand is strongly complexed there should be a 100-fold molar excess of ligand over determinand. The method will not in any way avoid the errors caused by the presence of interferences in the sample or by the introduction of interferences or additional determinand in reagents added to the sample.Use of the Nernst slope factor at concentrations where the electrode response deviates from ideality will cause errors: the determinand concentration should be ten times the solubility of the electroactive component in the membrane. Use of Known Dilution Calibration If the slope factor is obtained by this means, many of the errors associated with potentiometric analysis can be minimised. This can only be achieved if the diluent is appropriately matched to the sample (including any reagents added). The temperature should not differ by more than 1 "C. The diluent should contain the same background ionic strength as the sample and the same concentration of ligand. The errors caused by reagent blanks can be much reduced, but the effects of interfering substances present in the original sample are not.Even if the electrode is operating in its non-Nernstian response region, errors are much smaller than those obtained with a fixed value for the calibration slope: note, however, that direct potentiometry would generally be preferred in such circumstances . A feature of the known dilution calibration is that it alerts the analyst to the reliability of the result. (a) If the slope factor agrees with the electrode's Nernst slope factor, systematic errors in concentration should be small, unless there are interferents in the sample. Even if the slope factor differs from the Nernstian value, the concentra- tions should be accurate provided that the e.m.f.after dilution (E3) is approximately the same (+1 mV) as in the original sample ( E l ) and the diluent has been matched to the sample as above. Interferents present in the original sample, however, will always bias the result. (b) If E3 # El k 1 mV, and the slope factor is non-Nernstian, the result should be regarded with suspicion. Except near the electrode's limit of detection, the degree of dilution needed to make E3 = El can be obtained from Table 3 (provided that the diluent and sample are matched with respect to background ionic strength, ligand concentration and temperature). If ionic strength effects are significant, the known dilution calibration is sub-Nernstian even if the ionic strength in the diluent is at the correct level for effective compensation.Having less ligand in the diluent than in the sample causes the slope factor to be low when complexing is a significant factor. Reagent blanks and solubility product effects always produce sub-Nernstian responses. The known dilution calibration is super-Nernstian if the diluent contains too much ligand, or even the correct (non-zero) concentration, or if the ionic strength in the diluent is too high. The ligand concentration is the more significant factor in producing super-Nernstian responses. Random Errors The analytical precision is more dependent on the precision of the slope factor than on the precision of the standard potential. Equations (62) and (64) for the prediction of errors are more appropriate than equation (58), unless the measure- ment of e.m.f.is imprecise because of some deficiency in the equipment. Equation (58) and its related formulae are applicable to analysis by separate measurements of e.m.f. in a sample and a spiked sample, in which the mathematics of calculating the concentration are the same as for known addition. Known addition should be the more precise tech- nique because of its lower dependence on a ~ , the standard deviation of the e.m.f. Activity Coefficients If the Nernst slope factor is used, errors caused by variations in the activity coefficients are small (< 1 %) if the sample contains a background electrolyte that is about 100 times the determi- nand concentration [Fig. l(a) and Table 51.If the background is not present naturally, it can be added, as is usual in direct potentiometry . If the known dilution calibration is used, the errors are small (<0.5%) if the concentration of the background electrolyte is the same in the sample and the diluent, but in other instances it may be large, especially for divalent ions (Table 5). Effect of Complexing The presence of complexing agents with stability constants, p, such that pCA > 102 will cause errors, which may be very large, except in certain circumstances (Figs. 3 and 4). If a strong ligand (p > 104) is in large molar excess (- 100-fold) the errors are less than 5%0 with a Nernstian slope factor (Fig. 3) and less than 2% with a slope factor obtained by known dilution (Fig. 6), but the diluent must contain the same concentration of ligand as the sample.If the ligand is weak (p < 103) and its concentration, CL, is not too high S lO-I), the error obtained with the Nernst slope factor is 4 % (Fig. 3). The slope obtained by known dilution gives errors of <1% if the diluent contains the same concentration of ligand (Fig. 6), but if the diluent contains no ligand the errors are <5% only if pCL < 10-1.5 (Fig. 4). The prediction of errors for intermediate cases is difficult and it may be best to modify the sample so that it conforms to one of the two instances above. Solutions containing fairly high concentrations of weak ligands can be diluted as long as the determinand concentration is not taken below the limit of Nernstian response of the electrode.More of the strong ligand can be added to bring its concentration up to the required excess. If the sample solution is inadequately characterised, the results can be tested for the likeliness of errors,For calculation with the Nernstian slope, pCL < 10-2 or pCL 9 10-2 with weak ligands and P[A] s l O - 3 with strong ligands indicate that errors should be small. With the known dilution calibration a slope factor exceeding 70/1z1 mV decade-1 is a sign that the sample requires some treatment. Allowance can be made for systems containing more than one ligand and for ligands that complex species other than the determinand.Table 10. Sources of systematic error for ion-selective electrodes Type of Ionic strength electrode Species sensed effect Solidstate .. . . . Glass and liquid ion exchange . Gassensing . . . F- c1- Br- I- CN- SCN- S2- Residual chlorine cu2+ Pb2+ , Cd2+ Ag+ Na+ Li + K+ NH4+ CaZ+ Ba2+ NO3- CIOe-, BF4- NH3 COZ NO, SO2 Small Small Small Small Small Small Significant * Significant Significant Small Small Small Small Significant Significant Significant 1 Small Small Nil Significant * Significant * Significant * I Principal complexing species H+,* AP+,* Fe3+,* H3B03 Hg2+, Cd2+ , TP+ Hg2+, TP+, Cd2+ Hgz+ C d 2 + , Pb2+ Transition metals Fe3+, Hg2+ H+ * C1-, Br- Many organic and inorganic compounds Carboxylic acids EDTA H+ ,* transition metals H+* H+* H+,'HCHO Principal interferences OH- S032-, CN-, Br-,t I-,t S2-t S032-, CN-, I-,t S2-t CN- , S2- t I-, S2-t Br-, I-,t S2-t -SH groups Oxidants Hg2+,t Ag+t Hg2+,t Ag+,t Cu2+t Hg2+ t H+, Li+, Ag+ H+, Na+ , Ag+ K+, Na+ Transition metals Large anions H+, NH4+, Ag+ Causes of non- Nernstian response Kinetic Solubility Oxidation of S2- Trace organics Oxidation of membrane Adsorption? Mainly reagent blanks but solubility effects are possible Alkylamines Kinetic? or blank so2 Blank c0z7 Atmospheric COz blank Blank (SO,) 4(ix), >6 (glass) 4(ix), 5 (glass) 5-6 s5 4 >4 4 <6 4-5 >5 5 4-5(ix) * In these instances it is very unusual to omit sample treatment that nullifies the effect.t These have catastrophic interferences and must be absent. t These are limits in optimum conditions at about 25 "C.572 ANALYST, MAY 1987, VOL. 112 Non-Nernstian Calibrations Factors causing the electrode to have a non-Nernstian response will produce large errors if the Nernst slope factor is used, but will be smaller if the known dilution calibration is used and this will at least indicate that the interpretation of the results may be problematic (Figs.8 and 9). If the e.m.f. after dilution is returned to the original value in the sample, the errors may be kept within reasonable limits (5%). The factors responsible are usually the solubility of the membrane material or the presence of a reagent blank. The known dilution calibration may also give a non- Nernstian value because it is compensating for variations in complexing , activity coefficients or liquid junction potentials. In this region of the electrode’s response, direct poten- tiometry with a carefully prepared calibration graph will usually give the best results.Interferences and Reagent Blanks Interfering substances present in the original sample will interfere in known addition potentiometry as in direct potentiometry. A good example is given by Hassan et aZ.31 for measurements on calcium - magnesium mixtures with a variety of liquid ion-exchange electrodes. One possible exception occurs if the concentration of the interfering substance is known: it can then be added to the diluent for the known dilution calibration and treated as if it were a reagent blank. Errors caused by a reagent blank (interferents or additional determinand in reagents added to the sample and diluent) can be minimised by the correct use of the known dilution calibration, but not when the Nernst slope factor is used (Fig.There may also be an option of removing the interference by processes such as precipitation, oxidation, complexation and volatilisation, as appropriate to the interfering substance and the electrode. 10). Conclusions The systematic errors ensuing from the use of an ideal (Nernstian) slope factor in non-ideal conditions have been demonstrated for the following instances: (a) variations in activity coefficients; (b) the effect of complexing; (c) the solubility of the electrode membrane; (d) the presence of a reagent blank; and (e) the presence of interferences. Table 10 gives a qualitative classification of the sources of error affecting the known range of commercially available electrodes. The errors would arise in both direct poten- tiometry and known addition potentiometry, although not necessarily to the same extent. In many instances the errors can be eliminated by suitable treatment of the sample solution. The lists of complexing and interfering substances are indicative rather than exhaustive, as the effects depend on the concentration and many other substances could be significant in appropriate circumstances. The known dilution method of obtaining the calibration slope has been shown to be capable of largely compensating for errors otherwise expected in the examples (a)-(d) above. The conditions necessary to achieve compensation have been determined. Errors arising from procedural variations have been investi- gated for temperature and volume addition. The precision of the slope factor is of paramount impor- tance in the over-all precision if analysis is carried out by this method. The known addition method is not, as is sometimes implied, a cure-all for difficult potentiometric analyses, but its intel- ligent application can often yield results inaccessible to direct potentiometry . Published by the permission of the Central Electricity Generating Board. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. References Moody, G. J., and Thomas, J. D. R., “Selective Ion Sensitive Electrodes,” Merrow, Watford, 1971. Camman, K., “Das Arbeiten mit ionenselektiven Elektroden,” Springer, Berlin, 1973. Koryta, J., “Ion-selective Electrodes,” Cambridge University Press, Cambridge, 1973. Bailey, P. L., “Analysis with Ion-Selective Electrodes,” Heyden, London, 1976. Midgley, D., and Torrance, K., “Potentiometric Water Analy- sis,” Wiley, Chichester, 1978. Vesely, J., Weiss, D., and Stulik, K., “Analysis with Ion- Selective Electrodes,” Ellis Horwood,Chichester, 1978. Mascini, M., Ion-Sel. Electrode Rev., 1980, 2, 17. Karlberg, B., Anal. Chem., 1971,43, 1911. Ip, S. Y., and Pilkington, N. H., J . Water Pollut. Contr. Fed., 1978,50,2778. Mattock, G., “pH Measurement and Titration,” Heywood, London, 1961, p. 190. Davies, C. W., “Ion Association,” Butterworths, London, 1962. Ringbom, A., “Complexation in Analytical Chemistry,” Inter- science, New York, 1963, p. 35. Kragten, J., “Atlas of Metal - Ligand Equilibria in Aqueous Solution,” Ellis Horwood, Chichester, 1978. Midgley, D., Ion-Sel. Electrode Rev., 1981,3, 43. Bardin, V. V., Zavod. Lab., 1962,28,910. Ratzlaff, K. L., Anal. Chem., 1979, 51, 232. Midgley, D., Analyst, 1979, 104, 248. Horvai, G., and Pungor, E., Anal. Chim. Acta, 1980,113,287. Rice, T. D., Anal. Chim. Acta, 1983, 151, 383. Midgley, D., Analyst, 1985, 110, 841. Korhonen, J., and Lumme, P.O., Pap. Puu, 1977, 59,558. Neupert, L., Acta Hydrochim. Hydrobiol., 1982, 10, 557. Hulanicki, A., Lewandowski, R., and Maj, M., Anal. Chim. Acta, 1974, 69,409. Burman, J. O., and Johansson, G., Anal. Chim. Acta, 1975,80, 215. Vandevenne, L., and Oudewater, J., Trib. CEBEDEAU, 1973,26, 127. NubC, M., Van den Aarsen, C. P. M., Giliams, J. P., and Hekkens, W. T. J. M., Clin. Chim. Acta, 1980, 100, 239. Dimmock, N. A., and Midgley, D., Talanta, 1982,29, 557. Austin, R. K., and Phillips, D. J., J . Agric. Food Chem., 1985, 33, 1165. Cosofret, V. V., and Buck, R. P., Analyst, 1984, 109, 1321. Hassan, S. S. M., Ahmed, M. A., and Saoudi, M. M., Anal. Chem., 1985, 57, 1126. Hassan, S. K. A. G., Moody, G. J., and Thomas, J. D. R., Analyst, 1980, 105, 147. Paper A61451 Received November 26th, 1986 Accepted December lst, 1986

ISSN:0003-2654    

DOI:10.1039/AN9871200557

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 573 Temperature Compensation in Potentiometry: lsopotentials of pH Glass Electrodes and Reference Electrodes Part 1. Theory Derek Midgley CEGB, Central Electricity Research Laboratories, Kelvin Avenue, Leatherhead, Surrey KT22 7SE, UK The reasons for the non-linearity of the temperature response of glass pH electrodes are explored theoretically and the relationship between pHisor the isopotential pH (at which the e.m.f. is invariant with temperature), and the different components of the potentiometric cell is demonstrated. In particular, the temperature coefficient of pH of the glass electrode's filling solution is shown to be paramount in designing electrodes with characteristics convenient for instrumental temperature compensation. The pH of most filling solutions varies parabolically with temperature and the linear correction applied by the temperature compensation circuits of pH meters will have a limited useful range.Filling solutions with linear temperature characteristics are proposed, but the resultant electrodes would be incompatible with the many pH meters having a fixed isopotential setting of 7. Liquid junction potentials are shown to make a small and essentially linear contribution to the temperature dependence of the cell. Keywords: pH; glass electrode; reference electrode; temperature compensation; isopotential Although the most accurate measurements of pH require sample and standard solutions to be at the same temperature, some applications have to cope with varying temperatures. Temperature compensation is, therefore, a normal feature of pH meters and is also found on some process analysers for use with ion-selective electrodes.Modern instruments may incor- porate digital thermometers reading to 0.1 "C and those incorporating microprocessors may have settings for the slope factor and isopotential pH (enabling corrections to be made for the temperature variation of the standard potential) as precise as 0.01 mV pH-1 and 0.001 pH, respectively. Instrumentally, therefore, it should be possible to compensate for temperature more accurately than before, provided that the thermal characteristics of sensing and reference electrodes can be defined with adequate precision and the theory implicit in the design of the temperature compensation circuitry is correct. In this paper, the temperature dependence of pH cells is examined theoretically and in Part I1 the performance of commercially available glass pH and reference electrodes will be considered.All pH meters incorporate some form of temperature compensation, although its exact nature is not always made explicit by the controls on the meter or the operating manual. In the equation relating e.m.f. to pH, both "constants" (the standard potential, Ecell, and the slope factor, k ) vary with temperature. The temperature coefficient of the slope factor should be constant and predictable for all electrodes and compensation for this is a feature of almost all pH meters, under either manual or automatic control. The temperature coefficient of the standard potential is the source of various approximations, errors and misapprehen- sions.In principle, it can be calculated for known reference elements in known solutions inside the glass and reference electrodes; in practice, the analyst does not know the composition of the solution inside the glass electrode. An indication of this temperature coefficient is sometimes sup- plied by the manufacturer in the form of the isopotential pH, but unfortunately this is often quoted with insufficient rigour (not being defined with respect to a named reference electrode) and may even be confused with the zero-point pH (at which the e.m.f. = 0 mV). For temperature compensation with most pH meters it is desirable that these points should coincide in practice, but they are theoretically distinct and may differ considerably. E=ECeI1-kpH .. . . * * (1) Even with electrodes whose temperature characteristics are well defined, errors may arise from the limitations of pH meters. Many pH meters lack an isopotential pH adjustment and work on the basis (often unstated in the instruction manual) that the isopotential pH is 7; errors are unavoidable if the electrodes have different characteristics. Commercially available pH meters invariably apply a correction that assumes that the standard potential varies linearly with temperature, but the non-linearity of the temperature variation of the standard potential has long been known.1 Isopotential pH values are usually determined experimen- tally and applied as if they were linear temperature coeffi- cients. Theoretical treatments have expressed isopotentials only in empirical coefficients.1-3 This paper relates the isopotential pH to the internal reference electrode and filling solution of the glass pH electrode and to the external reference electrode and its liquid junction with the test solution.The sources of non-linearity in temperature compen- sation are considered. Factors analogous to isopotential pH can also be derived for ion-selective electrodes394 and in these instances better linearity may be obtained. Theory The conventional potentiometric cell with a glass electrode for measuring pH may be represented as below. 1 MX I solution containing I 1 test solution I 1 X-ions,ax I I pH M'X I M' I glass I innersolution I /membrane1 dx,pH' 1 1 In practice, X- is the chloride ion and M and M' are either silver or mercury or, more rarely, thallium.If the membrane responds to hydrogen ions with perfect selectivity, the e.m.f. of the above cell may be represented as in equation (2). E = (E" - kloga'x + kpH')glass - kpH + E, -(E" - klog QX)ref . . (2) where E"' and E" are the standard potentials of the two reference elements M'j M'X and MIMX, Ej is the liquid junction potential and k is the Nernst slope factor, equal to RTlnlOIFwhere R is the gas constant, T K is the temperature and F is Faraday's constant. The terms in brackets are constant at a given temperature and Ej is assumed to be so for574 ANALYST, MAY 1987, VOL. 112 the sake of calibrating the pH meter (the reference electrode is chosen so that this is a reasonable assumption).Equation (2) can be re-expressed as equation (1), where EOcell contains all the constant terms. pH meter temperature compensation circuits2,s assume that EOcell and k vary linearly with tempera- ture. For k it may be seen that this is correct because Elk RlnlO aT- F ---- -constant . . . . . In the case of ITcell, however, this assumption is an approxi- mation. Subtracting equation (1) from equation (2) and differentiat- ing, we obtain ak + kapH' -1oga'x - - (aZZ' kaloga'x aT aT aT log ax - Rearranging equation (4) gives Examination of the terms in equation (5) shows the conditions necessary for obtaining an EOcel, that varies linearly with temperature. (i) The terms grouped as (-log + pH' + log ax) ak/aT will vary linearly because of equation (3).(ii) The standard potential terms dEOldT and dEO'ldT are not necessarily linear, but the curvature is negligible over a range of +15 "C for calomel and silver - silver chloride electrodes. With matching reference elements in the glass and reference electrodes, however, these terms cancel provided that the cell is isothermal. (iii) On a molal scale, the terms kaloga'x/aT and kalogaxl aT are approximately zero, provided that the solutions are not saturated. With saturated solutions, the introduction of a temperature-dependent solubility makes the terms non-zero and, generally, non-linear, e.g., with saturated potassium chloride solutions aaCl/a T is approximately linear, making (iv) The liquid junction term aEj/a T should be small for any junction likely to be used in a pH cell and it will be shown below that any variation should be almost linear over a range of 30 "C for concentrated, but not saturated, potassium chloride bridge solutions.If the junction contains a saturated reference solution, additional non-linear terms may arise. (v) The term kapH'/aT depends on the nature of the solution inside the glass electrode. With a solution of strong acid (pH <4), apH'/aT = 0 and this approximation is still fairly good for many weak acid buffers over a moderate range of temperature, e.g., acetate buffers. However, because the circuitry of some pH meters restricts the choice of pHiso, glass electrodes generally contain buffers closer to neutrality and for such solutions apH'/aT is both non-zero and non-linear. This will be discussed after the consideration of isopotentials.It can be seen that careful consideration needs to be given to points (ii)-(v) if a linear variation of EOcell with temperature is to be achieved. Points (ii)-(iv) come readily within the control of the analyst, but point (v) does not. Isopotential Correction The temperature coefficient of the standard potential is usually expressed in pH methodology as the isopotential pH, i.e., the pH at which the e.m.f. of the electrode pair is invariant with temperature. aiogacl/a T curved. From equation (l), the e.m.f. at the isopotential point, Eiso, is given by By the definition of pHiso, aEi,,4aT = 0. Hence, Eiso = EOcell - @Hiso . . . . . . (6) aEOcell/aT=pHisOak/aT . . . . (7) Substituting from equation (3), 3EOCell/aT = pHi,J?lnlO/F .. . . (8) Equation (8) and the definition of PHis0 imply that EOcell varies linearly with temperature, hence EOc-11 may be expressed as where E, is the value of EOcell at reference temperature, T,. Hence Comparison of equations (6) and (9) shows that RTsln10 F Eiso = E, - - PHiso Substituting in equation (1) gives an expression that relates the experimental variables (E, pH) at (ideally) all temperatures by means of two temperature-independent constants (Eis0, pHis,) and the theoretically predictable slope factor (k) with its simple linear dependence on absolute temperature. E = Eiso - k(pH - pHis,) . . . . (10) Equation (10) forms the basis of temperature compensation in pH meters. Once Eiso, k and pHiso have been set at one temperature, through the "buffer," "slope factor" and "iso- potential'' controls, respectively, the meter thereafter requires adjustment only of the temperature setting for it to be used to measure pH at other temperatures.The analysis of the temperature-dependent components of the cell potential [equation ( 5 ) ] indicated that, in general, aE"cell/aT is not constant, and it follows that, in terms of the idealised pHiso defined above, equation (8) can only be an approximation. pHiso may be obtained graphically by plotting e.m.f. against pH for a number of solutions at a series of temperatures. If aE"cell/aT were a constant, all the isotherms would intersect at pHiso. In practice, non-linearity of the temperature dependence means that the intersecting iso- therms form a zone covering a range of pHiso.The minimum data for calculating pHiso are e.m.f.s in two standard buffer solutions at two temperatures, enabling intersecting isotherms to be drawn or solution by simultaneous equations of the form of equation (10). Hence, E - E' + kpH - k'pH' k - k' . . . . PHiso = where k and k' are obtained from AEIApH at temperatures T and T', pH and pH' are known and E and E' are measured. The precision of pHiso calculated in this way can be estimated from the standard deviation of the e.m.f. by the usual rules for combination of random errors. Neglecting random errors in the pH values of standard buffer solutions we obtain from equation (11) + . . . . . . " (k - k')2 where a(x) is the standard deviation in quantity x . Evaluating equation (12) for an ideal case with PHis0 = 7.0,ANALYST, MAY 1987, VOL.112 575 determined from measurements with potassium hydrogen phthalate buffer (pH 4) at 15 and 25 "C and with the standard deviation of all e.m.f. measurements assumed to be 0.1 mV, we obtain a (E - E') = 0.14 mV, a(k) = 0.05 mV decade-1 (for determination over a 3 pH span) and E - E' = -5.4 mV. Substituting in equation (12), -- a2(pHiso) - 5.2 x 10-4 + 2.47 x 10-3 49 Hence a(pHiso) = 0.38 and is dominated by the error ink - k' [the second term on the right-hand side of equation (12)]. Alternatively, equation (10) can be transformed into E + kpH = Eiso + kpHiso . . . . (13) If e.m.f.s are recorded in a buffer solution over a range of temperatures, the left-hand side of equation (13) can be plotted against k.Ideally the graph should be linear, with a slope equal to pHiSO, but it will generally be curved in practice. This method has the advantage that the curvature gives a qualitative indication of the range of usefulness of PHiso. pHi, as an Electrode Characteristic pHiso is a characteristic of the electrode pair in the poten- tiometric cell, but if the body of the reference electrode is separated from the junction by a long tube and held at a constant temperature, the cell pHiso can be equated with the glass electrode's pHiso. Such an assumption neglects the contribution to pHiso from the temperature coefficient of the liquid junction potential (discussed below). Provided that a remote junction reference electrode of the above sort is used, the same pHiso should be found for the glass electrode, regardless of the nature of the reference electrode or its temperature (provided that it is constant).With a non- isothermal cell of this kind, therefore, pHiso can be regarded as a characteristic of the glass electrode. If the glass electrode in the non-isothermal cell is replaced by a second reference electrode and the temperature of the test solution is varied, the change in e.m.f. can be attributed to the second reference electrode alone and its temperature coefficient calculated. This temperature coefficient can be expressed as an isopotential factor characteristic of the reference electrode; in a pH cell this factor would be formally equivalent to a pHiso value, although it is unrelated to any real solution pH. This terminology is adopted below: a graph of E against k gives pHiso for the reference electrode as the slope of the line represented by equation (14).E = Ei, + kpHiso - - . . . . (14) pHiso for an isothermal cell consisting of glass and reference electrodes may now be calculated as the difference of individual pHiso values obtained as above. pHiso(cell) = pHiso(glass) - pHiso(reference) Relationship between the Isopotential and Zero Points Glass electrodes should6 be marked with their zero point, denoted by E X , with respect to a stated reference electrode at 25 "C, i.e., Xis the pH at which the glass - reference electrode pair gives 0 mV. Almost all commercial electrodes are nominally E7. Manufacturers rarely quote pHis0 and it is often assumed that pHiso = 7 also. However, these quantities are not identical.Hence pHiso for a non-isothermal cell with a remote junction exceeds that for the corresponding isothermal cell by pH 2.2 or 0.4 for reference electrodes with calomel or Ag - AgCl elements in 3 moll-1 potassium chloride solution, respectively. With the remote junction electrode at 25 "C, the zero-point pH is the same for both configurations of cell. From equation (2), at E = 0 mV the zero-point pH is given by equation (15), where AEO = E"' - E". AE'+Ei a'x k ax -log-++H'. . . . (15) pH" = From equations (4) and (7), aAE" + aEj ~ k apH' k alog ax k alog a'x a T +-- a T a T aT a T PHiso = akia T . . . (16) - a x ~ The relationship between pH" and pHiso will be considered for a range of electrode types. Isothermal cells with non-saturated inner and reference solu- tions In this instance, where fx is the activity coefficient.Concentrations defined as molality are independent of temperature and activity coeffi- cients' are observed to vary by about 0.01% K-1, which is equivalent to 0.0025 mV K-1 and less than can be detected with a pH meter reading to 0.1 mV. Similarly, aloga'x/aT = 0. Further, any small non-zero components of the two terms tend to cancel. On a molar scale alogax/aT # 0, and this would involve additional terms in the density of the solutions and different aE"IaT terms from those which follow. The tem- perature coefficient of the e.m.f. is itself independent of the concentration scale and use of the molal scale gives simpler expressions. The NBS standard buffers are defined on a molal scale.Combining equations (15) and (16) and eliminating alogax/ a T terms, aAE' + aE, + kapH' If the two reference elements are of the same type, AE' = 0 and aAE"/aT = 0. Hence In practical pH cells, Ej is arranged to be small and its variation with temperature will make only a small contribution (see below). As an approximation, therefore, apH' a T pHiso pH" + T- . , . . (19a) or 1 apH' pHiso == pH" + - - . . (19b) lnlO alogT ' . Equation (19a) shows that even with identical inner and external reference electrodes and favourable assumptions about liquid junction potentials, pHiso will not coincide with pH" unless apH'/aT = 0. The latter condition could only be expected from a solution of a strong mineral acid. From equation (15), with a 3 moll-1 KCl reference electrolyte, AE = 0 and Ej = 0, it follows that for pH"(= pHis,) = 7.0 as desired, the internal filling solution of the glass electrode would have to be 4 x 10-4 mol 1-1 hydrochloric acid (or a solution having the same activity of hydrochloric acid).576 ANALYST, MAY 1987, VOL. 112 With non-identical reference elements, equation (18) shows that pHiso = pH" only if (again setting Ej = 0 and aEj/aT = 0) aAE" kapH' +- AE" a T i3T - k ak/a T I .. . Equation (19b) shows that pHis, is a constant only if pH' changes linearly with the logarithm of the absolute tempera- ture. However, this is not a property of real solutions (see below), except in the trivial case apH'/alogT = 0. The present state of pH meter technology treats pHiso as a constant and hence meters have a limited range in which temperature compensation can be applied with a given accuracy.The next stage of sophistication (which would require a microproces- sor-based pH meter only slightly more complicated than present models) would be to apply a linear correction to pHis, by equation (19a): a prerequisite for this would be a glass electrode whose internal pH changed linearly with tempera- ture. There are probably no such electrodes available at present, although suitable solutions could be devised. With present pH meters, requiring pH" = pHiso = 7, ideal conditions inside the glass electrode can be calculated from equations (15) and (20). This has been done in columns (3) and (4) of Table 1 for calibration at 25 "C versus reference electrodes with 3 moll-1 KC1 reference electrolyte (neglecting liquid junction potentials). Microprocessor-based meters should be able to cope with a much wider range of electrodes, because the equalities pH" = pHis, = 7 are unnecessary, although some microprocessor meters imitate the limitations of older meters.Table 1 gives mathematical solutions, but it does not follow that real chemical compounds exist to fulfil them, particularly when apH'/aT is required to be constant over the desired range of temperature. Even when the reference elements are identical and apH'/aT = 0, the concentration of hydrochloric acid required may be considered too low for chemical stability inside the electrode. Isothermal cells with at least one saturated solution Because the solubility of potassium chloride varies by about 0.04 mol kg-1 K-1, we have aloga&T = 0.005 (equivalent to an increase of 0.14 in pHis,).Reference solution saturated with potassium chloride Derivations can be carried out exactly as before, except that the term apH'/aT in equations (17) and (18) is replaced by (apH'/a T + aloga& 7). The conditions required for PHis0 = pH" = 7.0 with a saturated potassium chloride reference solution can be obtained from Table 1 by (i) subtracting 0.17 from the pH' - loga'a column and (ii) subtracting 5 X 10-3 from the isothermal apH'/aT column. Inrier solution saturated with potassium chloride The derivation proceeds as before except that in equations (16) and (17) apH'/aT is replaced by (apH'/aT - aloga'cl/ 32'). The conditions required for pHiso = pH" = 7.0 are obtained from Table 1 by (i) adding 0.40 to the pH' - loga'cl column to obtain pH' at 25 "C and (ii) adding 5 X 10-3 to the isothermal apH'/a T column.Both inner and reference solutions saturated with potassium chloride In this instance alogacl/a T = aloga'cl/i3 T and equations (16) and (17) are still valid. Conditions for pHiso = pH" = 7.0 are obtained from Table 1 by (i) adding 0.23 to the pH' - loga'cl column to give pH' at 25 "C and (ii) retaining the isothermal apH'/a T values. Non-isothermal cells The reference electrode is kept at a constant temperature as that of the rest of the cell is varied. The temperature coefficient of the e.m.f. is obtained by differentiating equation (2), omitting the terms in the ( )ref parentheses. ak kapH' ak l o g a ' x z + + pH'-) a T dass Non-saturated inner and reference solutions As before, aloga'x/aT = 0.At pH = pHis, we have, by definition, aE/aT = 0 at constant pH, hence aE. ak apH' ak a T a T a T aT a T aE"' + - - (loga'x - pH') + k - - pHis, - = 0 Therefore, (23) ak/a T pHis, = pH' - logalx + At E = 0, equation (16) is still valid for the non-isothermal cell, hence substituting in equation (23) gives Evaluating equations (16) and (24) for electrodes calibrated at 25 "C and with neglect of liquid junction potentials gives the results in columns 3 and 5 of Table 1. Inner solution saturated with potassium chloride For aloga'&T = 0.005 and wherever apH'aT appears in equations (22)-(24) it should be replaced by (apH'/aT - aloga'&T). The conditions for pHiso = pH" = 7.0 are obtained from Table 1 by (i) adding 0.40 to the pH' - loga'cl column to obtain pH' at 25 "C and (ii) adding 5 x 10-3 to the non-isothermal apH'/a T column. Table 1.Conditions required inside glass electrodes to give pHiso = pH" = 7.0. (Calibrated at 25 "C versus reference electrodes with 3 moll-' KC1 reference solutions) Reference element apH'Ja T apH'JaT required for required for Glass Reference isothermal non-isothermal electrode half-cell pH' - loga'cl cell cell AgCl . . . . . . AgCl 6.77 0 -2.7 x 10-3 AgCl . , . . . . HgzC12 7.54 3.5 x 10-3 -5.3 x 10-3 HgzC12 . . . . . . Hg2C12 HgZC12 . . . . . . AgCl 5.99 -3.5 x 10-3 -6.2 x 10-3 6.77 0 -8.8 x 10-3ANALYST, MAY 1987, VOL. 112 577 Reference solution saturated with potassium chloride As the temperature of the reference electrode is constant, pHiso is unaffected, but 0.17 is subtracted from column 3 to give pH' - logalcl (non-saturated inner solution) or 0.23 added to give pH' at 25 "C (saturated inner solution).Influence of the Liquid Junction Potential on the Isopotential Point The contribution of the liquid junction potential, evident in equations (18) and (24), has so far been neglected, mainly because it was expected to be small for practical liquid junctions, but also because the liquid junction potential of an unknown sample is not directly calculable. However, evalua- tion of the liquid junction contribution in some instances is instructive. Equations (18) and (24) for expressing the isopotential pH each contain the terms -Ej/k and (aEj/aT) (aT/ak).Calcula- tions by Picknett8 show that for a reference electrode with a concentrated potassium chloride electrolyte, E, is likely to vary within the range -3 to -7 mV for a wide range of sample electrolytes in the concentration range 10-2-10-6 mol 1-1. (Although the basis of these calculations is not strictly rigorous,9 experimental e.m.f.s in dilute solutions were predicted to within k0.5 mV.) The approximate maximum effect on pHiso, therefore, is 7/k = 0.12, but as only the variation of Ej with respect to the calibrating solution can be discerned, the effect is unlikely to exceed 0.06 units of pHiso. If PHis0 is determined from measurements in the same buffers as are used for calibration, the contribution to pHiso from this term should be even smaller.However, at the worst the Ej/k term is a fraction of the experimental variation in determining An estimate of aEj/aT may be obtained by differentiating the Henderson equation for the liquid junction potential, which is conveniently simple but involves assumptions that do not always apply to real junctions in pH measurements.10 PHiso. where hi, Ci and zi are the molar conductivity, concentration and charge (with sign) of the ith ion, respectively, and the superscripts R and L denote the right- and left-hand sides of the liquid junction. Letting P, Q, X and Y equal the sums in the numerators and denominators of the pre-logarithmic and logarithmic terms, we obtain Hence P a Q a T + -. - (25) aQ - P- I A aT Q2 - --- *-log-- k log-. a T Q Y \ Y P Y Q X Y2 + --.-ln10 - The only temperature-dependent variables in P, Q, X and Y are the molar conductivities.Hence aEj ak Ej ( ncp; PY lnlO J \--/ Y2 Equation (26) can be evaluated by replacing ahi/a T by ALi/A T from tabulations by Rpbinson and Stokes7 which show that conductivities vary approximately linearly with temperature over a range of about 30 "C. With concentrated (but not saturated) potassium chloride electrolyte in the reference electrode, both Ei/k and the variation in log X / Y are negligible and with dilute test solutions Cf. >> CiR, so that the terms P and Q are influenced only by CKL and C&. Hence With a 3 mol 1-1 potassium chloride reference solution at 25 "C, a Ej -(mVK-1)=0.0410gXhiCiR- 0.10 aT Table 2 shows the terms in equation (25) for two different test solutions.The dominance of the middle term is obvious. From equations (18) or (24) the variation of junction potential with temperature contributes about -0.5 unit to pHiso for 10-4 mol 1-1 hydrochloric acid solutions or -0.2 for 0.1 mol 1-1 acetic acid - acetate buffer. In the empirical determination of PHis0 these contributions would only be apparent if pHiso were determined in solutions different from the calibration solu- tions. The results in Table 2 show that pHiso determined in a buffer solution is likely to be about 0.3 unit too high for application to dilute sample solutions. From equation (26) and the known conductivities we may also infer that aEj/aT makes an almost linear contribution to the over-all temperature coefficient provided that the tem- perature range does not exceed about 30 "C.The first term on the right-hand side of equation (26) is truly linear, but is dominated by the others in ahi/aT. The calculations show that liquid junction potentials make a small contribution to the temperature coefficient of the e.m.f. of a cell. However, empirically determined values of isopoten- tial pH should largely compensate for this effect and the likely error in pH is small compared with that from other sources. Temperature Variation of pH of Electrode Filling Solutions The appearance of the term 3pH'IaT in equations (18)-(24) shows that the internal filling solution has a significant effect on the thermal characteristics of an electrode. Restricting consideration to the conditions required to produce Table 1, it can be seen from column 3 of that table that pH' = 7 + logdcl k 1.The practical limits for a'cl are about lo-3-100 moll-1, hence the limits of pH' are 3-8. Inspection of the properties of standard buffer solutions11 shows that most carboxylate and phosphate buffers go through a minimum in pH somewhere in the range 0-50 "C, i.e., neither pH' nor 3pH'IaT is constant. Over a wide range of temperature (040 "C), a 0.1 moll-1 acetic acid + 0.1 moll-' sodium acetate buffer comes nearest to constancy, with aANALYST, MAY 1987, VOL. 112 578 Table 2. Components of temperature coefficient of liquid junction potential and contribution to isopotential pH x a P ? aT k Y a T Q Q aT Y Test solution mV K-1 mVK-1 mVK-1 mV K-1 to pHiso -k- p a -logXt/ -1 8Ej ak. “I -/clog- - (-) I aT Contribution 10-“moll-1 hydrochloric acid .. . . -0.015 -0.086 -0.002 -0.10 -0.5 0.1 moll-1 sodium acetate + 0.1 moll-’ acetic acid . . . . . . -0.006 -0.037 +0.001 -0.043 -0.2 * Reference solution 3 moll-1 potassium chloride. t Symbols refer to equation (25). maximum spread of 0.05 pH, but in the ambient and sub-ambient range (0-30 “C) 0.05 moll-1 potassium hydrogen phthalate has a spread of only 0.012 pH. Buffers in which apH‘IaT is more nearly constant are those with amino groups of suitable strength, e.g., N-tris(hydroxy- methyl)methyl-2-aminoethanesulphonic acid (TES) and N-2- hydroxyethylpiperazine-N’-2-ethanesulphonic acid (HEPES) buffers12 (pH = 7.5). 3-(N-Morpholino)propanesulphonic acid (MOPS) would seem to be a good candidate for a pH 7.1 buffer.13 However, these buffers generally would give too negative a value of apH’/aT in Table 1 for PHis0 = pH” = 7 to be achieved.Other buffers of this type are known” but they have not been studied at different temperatures. The more acidic buffers might be especially worthy of study in relation to glass electrode filling solutions. Amine buffers in aqueous glycerol have also been suggested for this purpose.14 Calculation of apHtaT The only instance considered is that of a solution of a single monobasic acid, HA, partially neutralised by sodium hydrox- ide. The system is defined by the mass and charge balance equations, the protonation constant, K = [HA]/[A][H], and the autoprotolysis constant, K,. For convenience, charges have been omitted and activity coefficients neglected.Total acid = TA = [HA] + [A] = [A](l + K[H]) Total base = TB = [Na] Charge balance: [HI + [Na] = [OH] + [A] i.e., Rearrangement followed by differentiation with respect to temperature gives { 3K[H]2 + 2[H](l+ KTB) + TB - TA - KKw] = aT aK aKW aT aT --([HI’ + [HI’ TB - [HI K,) + - (1 + K[H]) Expressing the temperature differentials in logarithmic form gives 3[H]2 + 2[H] (K-1+ TB) + (TB - TA) K-1- K , (27) Equation (27) shows that the variation of pH with tempera- ture is fairly complicated, as alogK,/aT is almost linear whereas 3logKla T may be parabolic. In certain circumstances, the above general equation can be greatly simplified. For moderately concentrated solutions of moderately weak acids (such that K2KW < 1 and KTA >> 1) [HI = K-1 at half-neutralisation and then apWa T = alogK/a T.Discussion The relationship between the zero point and the isopotential point, and the conditions necessary to make them coincide, have been demonstrated. The non-linearity of an electrode’s temperature coefficient is shown to depend mainly on the temperature coefficient of the pH of the solution inside the glass electrode, but the use of saturated solutions in the glass or reference electrodes is another cause. The contribution of the liquid junction potential (and its temperature coefficient) to the isopotential correction has been shown to be small for the usual concentrated potassium chloride bridging solutions. In an alternative approach (the Ross electrode), the non- linearity of the pH buffer’s temperature response is compen- sated by an oppositely-responding reference electrode based on a redox couple.15 With the factors listed above, the approximate nature of the linear temperature corrections applied by pH meters becomes apparent.For linear characteristics to be achieved, new filling solutions need to be devised for glass electrodes, but such solutions are unlikely to give pHiso = 7 and a meter with an adjustable isopotential control would be necessary. Over a limited temperature range, certain weak acid filling solutions have attractive properties for electrodes intended for use in the environmental range (0-30 “C) and could give pHiso = 7. Non-ideal Electrodes Midgley and Torrance16 observed with symmetrical pH cells, in which the glass electrode was filled with a more concen- trated potassium chloride solution than usual, that the temperature-dependent terms did not cancel as expected unless the test solution also contained concentrated potassium chloride.As the anomaly was observed only with very dissimilar solutions on each side of the membrane, it is inferred that differences in hydration, ion exchange or adsorption between the inner and outer surfaces of the membrane could be induced by the nature of the two bathing solutions and give rise to an additional source of temperature dependence which has not been considered in this paper or elsewhere. This work was carried out at the Central Electricity Research Laboratories and is published by permission of the Central Electricity Generating Board. 1. 2. 3. 4. 5. 6. References Jackson, J., Chem. Znd. (London), 1948, 7. Mattock, G., “pH Measurement and Titration,” Heywood, London, 1961. Covington, A. K., CRC Crit. Rev. Anal. Chem., 1974,3, 355. Negus, L. E., and Light, T. S., Znstrum. Technol., 1972, 19 (Dec.), 23. Westcott, C. C., “pH Measurements,” Academic Press, New York, 1978. British Standard BS 2586, “Glass and Reference Electrodes for the Measurement of pH,” British Standards Institution, London, 1979.ANALYST, MAY 1987, VOL. 112 579 7. Robinson, R. A., and Stokes, R. J., “Electrolyte Solutions,” Second Edition (Revised), Buttenvorths, London, 1965. 8. Picknett, R. G., Trans. Faraday SOC., 1968, 64, 1059. 9. Covington, A. K., “Specialist Periodical Report, Electro- chemistry,” Volume 1, Chemical Society, London, 1970, p. 72. 10. MacInnes, D. A., “The Principles of Electrochemistry,” Reinhold, New York, 1939, p. 231. 11. Perrin, D. D., and Dempsey, B., “Buffers for pH and Metal Ion Control,” Chapman and Hall, London, 1974. 12. Bates, R. G., Vega, C. A., and White, D. R., Anal. Chem., 1978,50, 1922. 13. Sankar, M., and Bates, R. G., Anal. Chem., 1978, 50, 1922. 14. Simon, W., and Wegmann, D., US Pat., 3445363, 1969. 15. Ross, J. W., US Pat. 4495050, 1985. 16. Midgley, D., and Torrance, K., Analyst, 1982, 108, 1297. Paper A61281 Received August 14th, 1986 Accepted December 1 Oth, I986

ISSN:0003-2654    

DOI:10.1039/AN9871200573

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 581 Temperature Compensation in Potentiometry: lsopotentials of pH Glass Electrodes and Reference Electrodes Part II.* Performance of Commercial Electrodes Derek Midgley CEGB, Central Electricity Research Laboratories, Kelvin Avenue, Leatherhead, Surrey KT22 7SE, UK Commercial glass electrodes were shown to have non-linear temperature characteristics with isopotential pH values up to 2 pH below the expected PHis0 = 7.0. The use of calomel rather than silver - silver chloride reference electrodes causes a reduction of about 2 pH in the PHis0 of an isothermal cell. Each unit by which PHis0 is lower than the value assumed by the meter causes an error of -0.0035 pH K-1. The design of some reference electrodes makes it almost impossible for them to achieve rapid thermal equilibrium with the test solution and, hence, for proper temperature compensation to be applied.Cells containing such electrodes are not isothermal and may approach the character of cells with remote junction reference electrodes. Such cells may suffer from drifting e.m.f.s and give different temperature errors in solutions above and below the standardisation temperature. The liquid junction potential is shown to make a small contribution to the cell’s temperature coefficient, so that the value of PHis0 obtained for electrodes calibrated in standard buffers is 0.3 pH too high for use with dilute acid solutions. Keywords : pH; glass electrode; reference electrode; temperature compensation; isopotential In Part I1 the components of pH cells incorporating glass electrodes were examined theoretically with respect to tem- perature dependence and compensation.In this paper the performance of some commercial glass and reference elec- trodes is assessed in regard to their suitability for pH measurements in samples of various temperatures. Theory The theory has been given in Part 11 and will be repeated only in summary. pHis, (notionally the pH at which the cell e.m.f. is invariant with temperature) is used as a linear compensation factor for the variation of the standard potential with temperature. Hence [equation (10) of Part I] E=Eiso-k(pH-pHiso) . . . . (1) where k is the Nernst slope factor. With Eiso and pHis, known, the temperature dependence of the E - pH relationship can be calculated from the regularly varying coefficient of k [equation (3) of Part I].ak/aT = RlnlO/F = constant . . . . (2) pHiso can be calculated from measurements in two solutions at two temperatures, T and T’, by equation (3). pHiso = ( E - E’ + kpH - k’pH’)/(k - k’). . (3) The precision of pHiso is then obtained from equation (4). a2(pHiso) (PHiso)2 4 ( E - E’) + (pH)202(k) + (pH‘)2+(k’) + ~- - ( E - E’ + kpH - k’pH)2 (4) a(k) + a2(k’) (k - k’)2 where a(x) is the standard deviation in quantity x . Alternatively, pHiso can be obtained by plotting the left-hand side of equation ( 5 ) against k over a range of temperature. E + kpH = Eiso + kpHiso . . . . ( 5 ) Similarly, the reference electrode’s contribution to the over- all pHiso is obtained by plotting the e.m.f. of a non-isothermal * For Part I of this series, see page 573.cell containing two reference electrodes against k [equation (611. . . . . E = Ei,, + @Hiso * * (6) In this instance pHiso is a formal coefficient unrelated to any real solution pH. Equations (3)-(6) correspond to equations (1 1)-( 14) of Part 1. It is necessary to distinguish between pHiso and the zero-point pH (at which E = 0 mV). Experimental Apparatus Electrodes were purchased from leading manufacturers. The glass electrodes (Table 1) were all nominally E”7. The reference electrodes were of the usual calomel and silver - silver chloride type (Table 2). E.m.f.s were measured with a Table 1. Glass electrodes under test pHatE=OmV Internal Type* element Quoted Found? XIVB . . . . . . Ag-AgC1 7 5 0 . 5 6.58 XVC . . . .. . Ag-AgC1 7.0 5.75 XVIC . . . . . . Calomel 7.0 6.46 XVIID . . . . . . Ag-AgC1 - 6.02 * Letters indicate different manufacturers. t vs. 3 mol 1-1 KC1 calomel electrode at 25 “C. Table 2. Reference electrodes under test Internal Type* element VIA . . . . . . . . Calomel VIIB . . . . . . Calomel VIIIE . . . . . . Calomel XVIIIE . . . . . . Calomel XIXE . . . . . . Calomel IA . . . . . . . . Ag-AgCI Filling solution concentration/ moll-’ KC1 Junction 3 Frit 3 Frit 4 Frit 3 Remote frit 3.8 Glass sleeve 3.8 Glass sleeve * Letters indicate different manufacturers.582 ANALYST, MAY 1987, VOL. 112 Corning 110 digital pH meter reading to 0.1 mV. Techne C-100 thermocirculators were used to control the temperature of the water-jacketed glass test cell. Reagents NBS standard phosphate and phthalate buffer solutions (pH 6.865 and 4.008, respectively, at 25°C) were used as test solutions. With the reference electrodes, 3 moll-1 potassium chloride solutions were also used.Procedure Measurements were first made with the electrodes in solution at 25°C. The temperature was then changed in 5°C steps within the range 5-50 "C and the e.m.f. was recorded when it was steady. Results Glass Electrodes A graph corresponding to equation (5) is shown in Fig. 1 for data obtained with electrode XVIID between 5 and 50 "C. It is evident that the graph is non-linear, and this was so for all the Tem peratu rePC 10 25 40 390 r I 1 I 1 380 - > 370 - E 9 360 - + LU 350 - 340 - 55 60 65 Slope factor/mV (decade)-' Fig. 1. Graph for determining Hiso of glass electrode XVIID in potassium hydrogen phthalate buPfer (pH 4.008 at 25 "C) glass electrodes tested, although the curvature was less obvious in the other instances.The data were fitted empiric- ally by least-squares regression to first-, second- and third- order graphs. For all electrodes the best fit was obtained with the second-order graph. Isopotentials were calculated at a series of temperatures by differentiating the empirical second- order equations of the graphs with the results shown in Table 3. The slope of the linear least-squares fit is shown for comparison. The spread of the pHiso values indicates the degree of curvature of the graph and it is evident that the electrodes differ considerably in this respect. Even at the standard temperature of 25"C, some of the pHiso values differ considerably from the expected value of pH 7.These values were obtained with respect to a remote reference electrode and are larger than would be expected in isothermal cells with commonly used reference electrodes. These were new electrodes and the results were checked (at only two temperatures) after an interval of 4 months to see if any significant change had occurred. Whereas the results for electrodes XIVB and XVC increased by about 0.3 pH, those for the others had changed by less than 0.05 pH. A total of 10 determinations of pHiso for electrode XIVB over a period of 2 months by three different operators gave a mean of 7.8 with a standard deviation for a single result of 0.35, which is very close to the value of 0.38 calculated1 from equation (4) with a(E) = 0.1 mV.In each instance pHi, was calculated from equation (3) using two temperatures. It is obvious that pHiso is not a very precise characteristic, even of an individual electrode. Variations in manufacture must add to the uncertainty of pHiso for a particular type of electrode. Reference Electrodes Temperature coefficients expressed as pHk, The data were plotted in accordance with equation (6). The graphs (Fig. 2) were non-linear and the data were fitted empirically by least-squares regression to first- and second- order curves. Isopotential pH values were obtained at round temperatures by differentiating the empirical second-order equations of the graphs, with the results shown in Table 4. The value of pHiso obtained from the linear fit is shown for comparison.Measurements were made in the range 5-50 "C but e.m.f. values of the calomel electrodes at the extremes of the range were suspect and the 5 and 50 "C readings were disregarded. The reasons for this probably lay in the tempera- ture hysteresis of the calomel electrode, which caused errors Table 3. Isopotential pH values for glass electrodes with respect to a remote junction reference electrode. pHiso values from second-order fit refer to the temperatures indicated Temperature of second-order fit of pHi,JC pHiso from Electrode linear graph 5 10 15 20 25 30 35 40 45 50 XIVB . . . . . . 7.31 6.90 6.97 7.05 7.14 7.22 7.37 7.44 7.54 7.62 7.76 xvc . . . . . . 4.98 4.09 4.31 4.49 4.65 4.84 5.11 5.28 5.49 5.67 5.85 XVIC . . . .. . 6.17 5.38 5.55 5.71 5.87 6.04 6.34 6.50 6.64 6.74 6.99 XVIID . . . . . . 5.55 3.67 4.18 4.48 4.85 5.27 6.04 6.35 6.64 6.96 7.40 Table 4. Formal isopotential pH values for reference electrodes (with respect to a remote junction reference electrode) pHiso from Temperature of second-order fit of pHisoPC Electrode Solution* graph 5 10 15 20 25 30 35 40 45 50 VIA calomel (3 moll-' KC1) . . A 2.23 - 1.83 1.98 2.12 2.27 2.41 2.58 2.70 2.85 - B 2.16 - - - 2.20 2.19 2.17 2.16 2.15 2.14 - VIIB calomel (4 moll-' KC1) . . A 2.04 - 1.76 1.85 1.94 2.04 2.13 2.22 2.31 2.41 - B 2.02 - - - 1.58 1.72 1.87 2.02 2.16 2.31 - IA Ag - AgCl(3 moll-' KCl) . . A 0.42 0.49 0.47 0.46 0.45 0.43 0.42 0.41 0.40 0.38 0.37 B 0.44 - - - 0.57 0.53 0.48 0.44 0.39 0.34 0.30 * A, 3 moll-' potassium chloride solution; and B, in phosphate buffer solution (pH 6.685 at 25 "C).ANALYST, MAY 1987, VOL.112 583 that would be most evident at the extremes. No such problem was observed with the silver - silver chloride reference elec- trode. Two series of tests were carried out. The first, over the full temperature range, used 3 moll-1 potassium chloride solution but a second series was carried out over a limited range of temperature (20-50 "C) in phosphate buffer solution to check that the solution and, therefore, the liquid junction were not significant factors in determining pHiso. In both series of tests the electrodes were more fully immersed and given longer to equilibrate than might be usual for pH measurements. Internal temperatures of reference electrodes In the preceding sections, care was taken to see that the reference electrodes took up a definite temperature.Elec- trodes were either fully immersed, so that the element was at the same temperature as the test solution, or the elements were kept at a constant temperature and joined to the test solution by a long "remote junction" tube. However, in 5 > E . ui I ui I + E ui I 1 -5 TemperaturePC 5 25 45 55 60 65 Slope factorlmv (pH)-' Fig. 2. Effect of temperature on calomel (A) and silver-silver chloride (B) reference electrodes in 3 mol 1-1 KC1 solution versus a remote 3 moll-' KCl calomel electrode at 25 "C common practice the reference electrode temperature will often be indeterminate if the solution differs markedly from air temperature because (a) manufacturers give no instruc- tions as to the immersion depth, unless it is to recommend that there is a sufficient head of reference electrolyte solution, and (b) the design of the electrode may allow more or less convection to occur within the body and involve considerable variation in the position of the reference element. The temperature inside a variety of reference electrodes subjected to different temperature gradients was measured with a 0.6 mm diameter thermistor which could be man- oevered into various positions within the body.Initially, two types of electrode were studied, representing extremes of design for "standard" commercially available laboratory electrodes, i.e., having bodies about 12 mm in diameter and 10 cm long and excluding remote junction electrodes and electrodes designed to withstand high temperatures or pres- sures.Electrode IA (similar in construction to calomel electrode VIA) had the reference element set low in the body with a baffle immediately above it to restrict convection. A 4-cm immersion covered the element completely. Electrode XVIIIE (now discontinued) had no constraints on convection, with the element placed mid-way down the body with wide clearances all round. The results in Table 5 show that electrode IA's baffle, by limiting convection, shortened the response time in the immersed part of the electrode and enabled this part of the electrode, including the element, to approach closely the temperature of the solution outside. Provided that the electrode was immersed up to the level of the baffle, the element inside attained the temperature of the solution outside.The temperature above the baffle changed only In electrode XVIIIE convection caused the temperature inside the electrode to become almost uniform in a warm solution, but in a cold solution the temperature gradient was almost the same as in electrode IA. When the temperature was uniform, it was lower than that of the warm test solution (and pH electrode). A reference element in such a body could not be guaranteed to form part of an isothermal cell, but neither would it resemble a remote junction electrode in its thermal characteristics. Thermal equilibrium in such elec- trodes could also contribute to long response times or drifting e.m.f.s. Other examples of reference electrodes between the above extremes were tested, but only in warm solutions, as it was slowly.Table 5. Temperature variations inside reference electrodes Electrode External solution temperature/ "C IA, baffle 4 cm from end (air temperature 24 "C) . . . . 34.3 34.3 34.3 16.5 16.5 16.5 XVIIIE (air temperature 24 "C) . . . , . . . . . . 34.3 34.3 16.5 16.5 XIXE, spacer 6 cm from end (air temperature 22 "C) 35.1 35.1 VIIB (air temperature 22 "C) . . . . . . . . . . 35.1 35.1 IVC, spacer 3.5 cm from tip (air temperature 22 "C) 35.1 35.1 * Height above the tip of reference electrode. t At these heights the thermistor was alongside the reference element. . . . . Electrode immersion depth/ cm 1 1 2 2 3.5 3.5 3 3 3 3 1.5 5 c o . 5 4.5 3.5 3.5 Thermometer height*/ cm 1 8 3t 3t.1 8 2t 8 1.5 8 3 . 3 3 . 3 3t 3t 3t 4.5t Internal temperature/ "C 34.3 27.7 32.7 20.2 16.8 24.1 33.3 33.1 16.8 23.7 29.3 34.5 25.9 34.9 35.1 32.4 Time to equilibrium/ min 3 25 15 15 4 20 6 18 8 25 10 2 25 3 5 10584 ANALYST, MAY 1987, VOL. 112 evident from the earlier tests that in a cold solution the temperature gradients were much the same in all electrodes. A current model (electrode XIXE) from the same manufacturer was tested; this had a perforated spacer 6 cm from the tip and above the reference element. Table 5 shows that this electrode was intermediate in performance between electrodes IA and XVIIIE. It appears that the spacer was not as effective a barrier as the baffle in electrode IA, as the element remained below the temperature of the exterior solution, even when fully immersed.Electrode VIIB had no baffle, but the element was set in a glass tube that narrowly cleared the walls of the electrode body. This seemed to restrict convection sufficiently for the element to reach the external solution temperature in a fairly short time, provided that the electrode was immersed to a depth of 4.5 cm so as to cover the element. When only the glass tip carrying the ceramic frit junction was immersed, the temperature inside the electrode rose only slowly and its performance then approached that of a remote junction electrode. Electrode IVC had a perforated spacer mid-way up the reference element. Provided that the electrode was immersed up to the level of the spacer (3.5 cm), the solution around the lower part of the element was at the same temperature as the external solution but that around the upper part was 2-3 "C lower.As the spacer was half-way up the element, it was likely to worsen the temperature gradients in the element, which was so long that it could not be adequately immersed in the thermostated vessel. Apparent pHho values in non-ideal conditions The above results show that thermal equilibrium between a reference electrode and a sample solution cannot be readily assumed without more care being taken than is implied in standard analytical procedures. Most reference electrodes are of a design likely to produce the indeterminate type of thermal characteristics found above and in such circumstances the PHis0 values obtained in Table 4 may not be reproduced. Electrode VIA was tested at various depths of immersion.This calomel electrode has a baffle above the element so that isothermal behaviour can be expected if the electrode is immersed sufficiently (cf. electrode IA in Table 5). The results in Table 6 show that partial immersion results in values of PHis0 considerably below those obtained with a fully immersed electrode. Different values were also obtained for increases and decreases in temperature, which may be attributed to hysteresis of the calomel electrode. Although all the e.m.f.s were apparently steady values, even in the most favourable circumstances (full immersion, temperature increase) pHiso differed from that obtained after careful equilibration (Table 4). In the worst instance, pHiso differs by 2 pH from the equilibrium value, which is equivalent to a reading error of 0.07 pH over a 10 "C range.The discrepancies in pHiso are much worse for the calomel electrode than for the silver - silver chloride electrode, which shows a difference of only 0.4 in pHiso between fully immersed and completely remote positions and is far less prone to hysteresis. This further strengthens the preference for silver - silver chloride reference electrodes when measurements are made at different temperatures. Temperature coefficient of the slope factor Equation (2) expresses the theoretical variation of the slope factor with temperature. However, the slope factor itself often deviates from the ideal value and the possibility that its temperature coefficient might differ from that given by equation (2) was tested.Four glass electrodes were tested over a period of 6 months with a variety of reference electrodes (remote junction and immersion types), all of which had 3 mol 1-1 potassium chloride reference solutions. The slope Table 6 . Effect of immersion depth on isopotential pH of calomel reference electrode VIA Isopotential pH Temperature Immersed up to Immersed half-way to baffle changePC baffle 25-35 1.6 35-25 0.8 25-15 0.4 0.6 0.5 0.15 Table 7. Temperature coefficient of the slope factor Standard (akla n k J s deviation Electrode Rln 10/F (single result) No. of results XIVB . . . . 1.001 0.007 11 xvc . . . . 1.OOo 0.002 6 XVIC . . . . 0.998 0.004 8 XVIID . . , . 1.001 0.001 4 factors were determined from measurements in pH 4 and pH 7 buffer solutions at temperatures between 5 and 35°C.The slope factors showed small (d%) negative deviations from the ideal values but the temperature coefficients of the slope factors agreed with equation (2) within 0.2%, as shown in Table 7. The deviations from ideality are small compared with the uncertainty in the isopotential pH, or even with the precision of setting the temperature on many pH meters. No problems are expected with the temperature coefficient of the slope factor, at least for electrodes with acceptable slope factors (398% of Nernstian). No predictions can be made about seriously non-Nernstian electrodes. Effect of the liquid junction potential pHiso was determined from measurements at 25 and 35 "C for electrode XIVB vs. a remote calomel reference electrode. Potassium hydrogen phthalate buffer solutions (pH 4.01,0.05 mol kg-1) gave pHiso = 7.3 whereas 5 x 10-5 mol kg-1 sulphuric acid solution (pH 4.00) gave pHiso = 7.0.This difference is similar to that predicted in Table 2 of Part I for dilute acid and buffer solutions, and would cause a reading error of 0.01 pH over a 10 "C range. However, the difference is within one standard deviation for the empirical determination of pHiso and so the contribution of the liquid junction potential to the over-all pHiso may reasonably be neglected. Discussion Glass Electrodes The results for glass electrodes confirm the theoretical prediction1 that the temperature dependence of their e .m.f .s is non-linear, i.e., the isopotential pH is not a true constant. The significance of errors caused by applying a linear correction to a curved function can only be judged in relation to a specific requirement for accuracy over a given tempera- ture range.The error in pH measured at T "C with an electrode calibrated at Ts "C with an incorrect isopotential setting is given by equation (7). r-r (7) For the worst electrode in Table 3, the error caused by non-linear temperature characteristics is about 0.03 pH over a 10 "C range, but for most electrodes the error is only ca. 0.01 A further source of error is the deviation from the expected value of pHiso = 7.0. Some pH meters can correct for such deviations, but in general electrodes are supplied without acknowledgement that their pHiso may be other than 7.0. PH.ANALYST, MAY 1987, VOL. 112 585 Treating the electrodes in Table 3 as having pHiso = 7 would cause errors of up to 0.07 pH over a 10°C range in temperature in the worst case.The reference electrode can also contribute to this error, e.g., changing from a remote to an isothermal calomel reference electrode without adjusting the meter would cause an error of 0.07 pH over a 10 "C range. Reference Electrodes Both calomel and silver - silver chloride reference electrodes behaved regularly, provided that the inner reference element could readily reach thermal equilibrium with the test solution. Three factors impede the attainment of this condition. (a) The element is of such a size or so positioned that the electrode cannot easily be immersed to a depth sufficient for good thermal contact. Most of the electrodes tested required more than 4 cm immersion, which is attainable in about 75 ml of solution in a 100-ml beaker or 150 ml in a 250-ml beaker.A small reference element positioned near the tip of the electrode is an advantage. (b) Convection within the body of the electrode prevents equilibrium being attained. This is best controlled by means of a baffle placed above the electroactive part of the reference element. Few proprietary electrodes are well designed in this respect. (c) The calomel element is subject to temperature hyster- esis. If temperature changes are large or rapid, silver - silver chloride reference electrodes are preferable. A way of avoiding these three problems is to remove the reference element completely from the influence of the test solution by using a remote junction arrangement.Calomel and silver - silver chloride elements are both suitable. This type of electrode is at least as convenient. as other types for laboratory work, but may be less so for field measurements. Many proprietary reference electrodes have characteristics intermediate between those of isothermal and remote elec- trodes, indicating that they are poorly designed with respect to factors (a) and (b). For silver - silver chloride electrodes the difference in pHiso between the two extremes is fairly small, which is a further reason for preferring them to calomel electrodes for use at various temperatures. In most laboratory measurements the solutions and electrodes are close to air temperature and errors from the above sources are probably negligible.The liquid junction potential makes a small contribution to pHiso and can be the cause of small errors in temperature compensation if the ionic strength of the test solution differs substantially from that of the buffers used for calibration. Many liquid junctions, especially if they do not permit a free outflow of electrolyte solution, behave anomalously even at constant temperaturez-6 and limited studies of the thermal properties of such junctions2fj indicate that they could affect the temperature compensation more than has been observed in this work. Avoiding Errors in Temperature Compensation The best practice is to eliminate temperature compensation by bringing calibrating and test solutions to the same tempera- ture. In the laboratory this is easily done in a water-bath, with little effect on the over-all time of analysis.For field measurements it is usually more convenient to immerse the buffer solutions in the lake or stream for 10-15 min, preferably in portions of no more than about 100 ml. However , in some circumstances, temperature compensa- tion is unavoidable. A remote junction reference electrode has many advantages, but, if convenience of handling demands an immersion type of electrode, a silver - silver chloride element would be preferred because it does not suffer from hysteresis and its pHiso varies in only a small range, even if ideal conditions of thermal equilibrium are unattainable because of the electrode's design. The labelling on glass electrodes is sometimes misleading and pHis0 should be determined empirically (for a glass - reference electrode pair).Many pH meters can only accommodate a PHis0 of 7 (k0.5) and so restrict the choice of electrodes. The most stringent requirement for temperature compensa- tion in on-line pH equipment in power stations is for an error not exceeding 0.05 pH over a +30 "C range from 20 "C. This implies, by equation (7), a pHiso set to within 0.5 pH of its true value. The results show that many commercial glass and reference electrodes would not enable the specification to be met if the meter had a fixed isopotential setting of 7.0. Considering the fairly low precision with which pHiso could be determined, the requirement may be hard to fulfil. However, the more general requirement is only half as stringent (with a k15 "C range), and with improved sample conditioning in modern power stations the actual range of temperature should be even narrower. The temperature of environmental samples is, in field work, usually outside the analyst's control, yet the author is unaware of any agency's requirements for the accuracy of temperature compensation in this application. I thank Mr. C. Gatford for performing most of the practical work. This work was carried out at the Central Electricity Research Laboratories and is published by permission of the Central Electricity Generating Board. References 1. Midgley, D., Analyst, 1987, 112, 573. 2. Midgley, D., and Torrance, K., Analyst, 1976, 101, 833. 3 . Midgley, D., Amos. Environ., 1987, 21, 173. 4. Davison, W., and Woof, C., Anal. Chem., 1985,57, 2567. 5. Brezinski, D. P., Analyst, 1983, 108,425. 6. Brezinski, D. P., Anal. Chim. Acta, 1982, 134, 247. Non-Reference 1 is to Part I of this series. Paper A61282 Received August 14th) 1986 Accepted December loth, 1986

ISSN:0003-2654    

DOI:10.1039/AN9871200581

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 587 Use of Lipophilic Additives for the Improvement of the Characteristics of PVC Membrane Lithium-selective Electrodes Based on Non-cyclic Neutral Carriers Tatsuhiro Okada, Kazuhisa Hiratani and Hideki Sugihara Industrial Products Research Institute, Ministry of International Trade and Industry, Yatabe, Tsukuba, lbaraki 305, Japan Favourable effects were found for poly(viny1 chloride) (PVC) membrane lithium-selective electrodes based on synthesised non-cyclic neutral carriers to which a lipophilic salt in combination with organophosphorus compounds had been added. Tetrabutylammonium tetraphenylborate as a lipophilic salt appreciably increased the selectivity for lithium over alkali metal ions when used in combination with trioctylphosphine oxide in the ion-sensitive membrane.The selectivity coefficient -log increased from 1.5 to 2.2 for Li+ over Na+ and from 2.0 to 3.0for Li+ over K+ by the addition of these compounds. The results obtained here are of practical use in developing lithium-selective electrodes, as the proposed additives have been shown to exert positive effects on several kinds of non-cyclic lithium-selective carriers in PVC matrix ion-selective electrodes. Keywords: Lithium determination; ion-selective electrode; poly(viny1 chloride) membrane; lipophilic additives; tetra bu t ylammonium tetrap hen ylbo ra te Considerable attention has recently been focused on the development of ion-selective electrodes for the determination of lithium ion activities. The monitoring of Li+ in biological systems is very important as lithium is used in the treatment of maniacal psychosis,l-3 and a sensitive Li+-selective electrode is required for this purpose.The electrode should exhibit at least a 103-fold selectivity for Li+ over Na+ and a detection limit for Li+ of 10-4 mol dm-3.4 Unfortunately, it is very difficult to obtain Li+-selective electrodes that satisfy these requirements commercially. A great deal of research has been directed towards polymer matrix ion-selective electrodes that use electrically neutral carriers as sensors.5 There are very few naturally occurring antibiotics that have a marked selectivity towards small ions such as Li+, and therefore efforts have been devoted to the synthesis of cyclic6-9 and non-cyclic4JoJl neutral carriers with sufficient ability to complex Li+.Recently, a novel type of Li+-selective carrier was synthesised and was found to exhibit excellent properties in ion-selective transport12 and in lithium- selective electrodes.13 In developing polymer matrix ion-selective electrodes, efforts have been focused so far on the design of carriers that have satisfactory complexing ability.5J4 It is evident that the change in conformation that occurs during the complexation of carriers with ions, and hence the change in the solvent atmosphere, could greatly affect the thermodynamic and kinetic properties of ion-extraction and complex-formation processes.15 Hence it can be assumed16--18 that the choice of solvent atmosphere could stabilise the primary complex formation or destabilise the complexation of interfering ions, and that it should greatly affect the ion selectivity of liquid membrane electrodes.Such effects could be enhanced in non-cyclic camers as these have more freedom in their molecular structure than cyclic carriers. In the work reported here, attempts were made to examine the solvent effects of poly(viny1 chloride) matrix lithium- selective electrodes based on synthesised non-cyclic neutral carriers by the addition of solvent mediators (plasticisers) and several lipophilic additives to the ion-sensitive membranes. The selectivity for lithium and other electrode characteristics were improved when lipophilic salts were used in the membrane in combination with organophosphorus additives. Experimental Materials 3,3-Bis(8-quinolyloxymethyl)oxcetane (I) was utilised as a netural carrier for Li+-selective PVC membrane electrodes.It was synthesised as reported recentlyl2.13 and used after purification by column chromatography (alumina - chloro- form) followed by recrystallisation from cyclohexane. 0 0 I The plasticisers used were o-nitrophenyl octyl ether (NPOE), dioctyl phenylphosphonate (DOPP), dioctyl phtha- late (DOP), bis(2-ethylhexyl) sebacate (DOS), tris(2-ethyl- hexyl) phosphate (TOP), trioctyl phosphite (TOPi), 2,3- dimethylnitrobenzene, dibutyl sebacate (DBS), decan-1-01, 2-ethylnitrobenzene7 acetophenone, nitrobenzene (NB) and 2-nitro-p-cymene. These were available commercially and were used without further purification. Poly(viny1 chloride) (PVC, average degree of poly- merisation 1100), which was also commercially available from Wako Pure Chemical Industries, was purified by precipitation from tetrahydrofuran (THF) in methanol.The lipophilic additives used were potassium tetrakisb- chlorophenyl) borate (KTCPB) (anionic sites) and trioctyl- phosphine oxide (TOPO), TOP, DOPP and TOPi as organo- phosphorus additives, which were used as received. Series of tetraalkylammonium tetraphenylborate compounds were also used as lipophilic salts. Table 1 shows these compounds, together with their abbreviations. Tetraalkylammonium chloride or bromide was mixed with an equimolar amount of sodium tetraphenylborate in dichloromethane at room tem-588 ANALYST, MAY 1987, VOL. 112 Table 1. Lipophilic additives used in the experiments Lipophilic additive Abbreviation TOP0 TOP DOPP TOPi perature and the reaction was completed overnight.The product was filtered and the solvent removed from the filtrate by evaporation. The residue was recrystallised from ethyl acetate. For the tetramethyl- and tetraethylammonium chloride, which were not soluble in dichloromethane, the solid reaction product obtained after filtration was washed with pure water and then dichloromethane and the residual solid was dried and used. Membrane Preparation The carrier (5 mg), plasticiser (250 mg), PVC (100 mg), KTCPB (3 mg), organophosphorus compounds and/or lipo- philic salts (3-10 mg) were dissolved in 4 ml of THF. The mixture was then poured into a petri dish (42 mm diameter), which was kept horizontally on a mercury pool, and the THF was allowed to evaporate slowly at room temperature. A polymeric membrane of ca.0.2 mm thickness was obtained. A disc (5 mm diameter) was cut from the PVC membrane and was incorporated into a Teflon electrode body (13 mm in diameter, 120 mm in length) for e.m.f. measurements. E.m.f. Measurements The e.m.f. measurements were carried out using the following electrochemical cell: Ag - AgClI 10-2 mol dm-3 LiCllImem- branellsample solution1 10-1 mol dm-3 NH4N031saturated KC1 - AgCl - Ag. The sample solutions were prepared from analytical-reagent grade chlorides of alkali and alkaline earth metals and de-ionised water (specific conductivity <5 X 10-7 Q-1 cm-1). A computer-aided automatic testing apparatus for the measurement of the e.m.f. response of the electrode and for data processing based on separate solutions methods19 was prepared and utilised in this experiment.Fig. 1 depicts the apparatus. The test-tube (120 ml volume) containing sample solutions with different concentrations of metal ions was placed in position by a fraction collector. A unit composed of the ion-selective electrode, the reference electrode and a stirrer was introduced into or pulled out of the sample solution by means of a pulse motor. The e.m.f. of the solution was measured with a Keithley Model 614 electrometer (input impedance >5 X 1013) and was input into an NEC PC8801 microcomputer through an analogue - digital converter. A flow chart showing the sequence of measurements made is given in Fig. 2. After the initial set-up period, the cationic species Z was settled and the test-tubes were programmed to be placed so that the concentration, J, of the relevant cationic I 9 Electrometer A/D Fraction collector Pulse motor I PC 8801 Driver - Control - Printer I/O I Fig.1. Illustration of the automated testing apparatus for ion- selective electrodes ( Menusetting ) - * J = 0 to NS(I) Stirrer off En 1 S 0 .- +-I .- fn 3 .- s d m t l Fig. 2. Flow chart of the measuring and data processing program for testing ion-selective electrodes species fell in a 10-fold ascending order. The test-tube containing the sample solution to be measured was put into place by the “NEXT SAMPLE” routine, the electrode was placed in the solution, the stirrer was turned on and an e.m.f. value was input into the computer every 10 s until a steady-state value was attained.The stirrer was then turned off and the e.m.f. was measured. This procedure was repeated for each concentration of solution. After a series of measure-ANALYST, MAY 1987, VOL. 112 589 ments for a cationic species the electrode was rinsed with pure water and the process was repeated for all the cationic species. All experiments were performed at 25 "C, and all the measurements were controlled by the microcomputer, the data were saved on a floppy disk and the results were printed out. The potentiometric selectivity coefficients were calcu- lated from the calibration graphs obtained. Results and Discussion The solvent medium of the PVC membrane exerted an appreciable effect on the characteristics of the electrode. In Fig.3, the logarithm of the selectivity coefficient of Li+ relative to the ion M (log kf'&) for membranes consisting of various plasticisers is plotted against the dielectric constants of the plasticisers. The electrode responded fairly rapidly to changes in Li+ activity; the response time ranged from 30 to 50 s, depending on the plasticiser used. Very poor Nernstian responses were obtained for the membrane with DBS, decan-1-01, o-nitroethylbenzene, acetophenone and NB plas- ticisers. According to the electrostatic mode15914 in which the ion - ligand complex is treated as a sphere, the smallest differences in ion selectivity among monovalent cations are expected to be caused by the change in polarity of the membrane. The observed change in Li+ - alkali metal ion selectivity in Fig.3 appears to be due to the slight change in the size of the coordination sphere of the non-cyclic carrier caused by a change in the membrane solvent. For divalent cations, no specific change in the selectivity was observed. Among the plasticisers tested, NPOE gave the best membrane charac- teristics and was utilised in the following experiments. The e.m.f. response of the membrane containing the carrier and the lipophilic salt 4C4NTPB with the plasticiser NPOE are shown in Fig. 4(a)-(c). Fig. 4(a) shows the calibration graph for the PVC membrane containing 1.4% mlm carrier, 70% mlm NPOE and 0.8% mlm KTCPB, for which a good Nernstian response of 59 mV decade-1 and a detection limit of 5 X mol dm-3 for Li+ were observed. The membrane with the addition of 1.1% mlm 4C4NTPB and no KTCPB had a low e.m.f.and a poor electrode behaviour, as seen in Fig. 4(6). However, if KTCPB and the lipophilic salt 4C4NTPB were incorporated in the membrane, the electrode characteristics were significantly improved and the interference from divalent cations was suppressed. The selectivity coefficients of the membrane containing various amounts of additives are shown in Fig. 5 . KTCPB improved the Nernstian response of the electrode, but an excess amount of it increased the K+ interference. 4C4NTPB together with 0.8% mlm KTCPB improved the Li+ selectivity, especially against divalent cations. In this instance an opti- mum concentration was shown to exist for the appearance of 0 0 A B C D E F G H * 01-A-A-A-M ' I A- A- m -3 0 5 10 15 20 25 & Fig.3. Selectivit coefficients of PVC membranes plotted against dielectric constants of plasticisers composing the membranes. Membrane composition: 1 . 4 d mlrn carrier (I), 70% mlm plasticiser, 0.8% mlrn KTCPB. Plasticisers: (A) DOS; (B) DOP; (C) TOPi; (D) DOPP; (E) decan-1-01; (F) TOP; (G) 2-nitro-p-cymene; and (H) NPOE. 0, H+; A , Li+; V, Na+; 0, K+; A , Mg2+; and V, Ca2+ > E Gi 250 200 150 100 50 -5 -4 -3 -2 -1 250 200 150 100 50 300 250 200 150 100 -5 -4 -3 -2 -1 -5 -4 -3 -2 -1 Log aM Fig. 4. Calibration graphs for electrodes with PVC membranes containing carrier (I) and lipophilic additives. Membrane composition: (a) 1.4% mlm carrier, 70% mlrn NPOE, 0.8% mlrn KTCPB; (b) 1.4% mlm carrier, 70% mlm NPOE, 1.1% mlm 4C4NTPB; and (c) 1.4% mlm carrier, 69% mlm NPOE, 1.1% mlm 4C,NTPB, 0.8% mlm KTCPB.0, H+; A , Li+; 'J, Na+; 0, K+; A , Mg2+; and V, Ca*+590 ANALYST, MAY 1987, VOL. 112 3 2 1 a f s o nA * Dl -1 -2 -3 I I I I I I 0.8 1.7 0.8 0.8 0.8 KTCPB, O/om/m ~ ~~~ ~ 0.6 1.1 2.7 4C4NTPB, %m/m Fig. 5. Selectivity coefficients of PVC membranes containing various amounts of li ophilic additives. Membrane composition: 1.4% m/m carrier (11, &-7oyo m/m NPOE 3 2 1 - 2 0 .-. n2 * m 0 J -1 -2 -3 good electrode characteristics, In Fig. 6, the Li+ selectivity is shown for various kinds of lipophilic additives. The presence of the butyl group as the alkyl moiety of tetraalkylammonium tetraphenylborate proved to be the best for improving the Li+ selectivity of the electrode. Ethylene glycol, which is a polar, non-ionic additive was tested in comparison with lipophilic salts and did not have any effect on the Li+ selectivity.The e.m.f. response of the membrane incorporating organophosphorus additives is shown in Fig. 7(a)-(c). TOPO has been reported to enhance the Li+ selectivity of neutral carrier based ion-selective electrodes.16 Our results also indicated that TOPO improved the selectivity for Li+ over other alkali metal ions, especially over Na+, which is essential for the use of the electrode in biological applications [Fig. 7(a)]. However, too much TOPO caused a detrimental interference from divalent cations [Fig. 7(b)]. TOPi proved to be better than TOPO, as the interference from divalent cations was much less for the TOPi [Fig 7(c)]. The selectivities obtained with various kinds of organophosphorus additives are shown in Fig.8. Some organophosphorus compounds were shown to exert positive effects on the Li+ selectivity of the electrode, especially TOPi. However, all the compounds had optimum concentrations in the membrane, above which the electrode characteristics deteriorated. I t- No R4 = 4C1 4C4 C123C1 2C142C1 2C182C1 Ethylene RaNTPB glycol Fig. 6. Selectivity coefficients of PVC membranes containin various kinds of tetraalkylammonium tetraphenylborate as lipophilic salts. Membrane composition: 1.4% m/m camer (I), 6%70% m/m NbOE, 0.8% mlrn KTCPB, 1.1% mlm &NTPB > E Gi 250 200 150 100 50 0 200 150 100 50 0 -5 - 4 -3 - 2 -1 -5 - 4 -3 -2 - 1 -5 - 4 -3 - 2 -1 Fig. 7. Calibration graphs for electrodes with PVC membranes containing carrier (I) and organophosphorus additives.Membrane composition: (a) 1.4% mlm carrier, 69% mlrn NPOE, 0.8% mlrn TOPO, 0.8% mlrn KTCPB; ( b ) 1.4% mlrn carrier, 69% mlrn NPOE, 1.6% mlrn TOPO, 0.8% mlm KTCPB; and (c) 1.4% mlrn camer, 69% m/m NPOE, 1.6% mlm TOPi, 0.8% mlrn KTCPB. 17, H+; A, Li+; V, Na+; 0, K+; A, Mg*+ and V, Ca2+ANALYST, MAY 1987, VOL. 112 591 3 2 1 s 5 c n o aJ -1 -1 - 2 -3 Fig. 8. Selectivity coefficients of PVC membranes containing various kinds of organophosphorus compounds as lipophilic additives. Membrane com osition: 1.4% mlm carrier (I), 68-70% mlm NPOE, 0.8% rnlm KTCgB 250 100 50 200 150 1 00 50 0 -5 -4 -3 -2 -1 -5 -4 -3 - 2 -1 Log aM Fig. 9. Calibrations gra hs for electrodes with PVC membranes containing camer (I) ancflipophilic additives.Membrane composi- tion: (a 1.4% mlm carrier, 68% mlmNPOE, 0.8% mlmTOPO,l.l% rnlm 4d4NTPB, 0.8% mlm KTCPB; and ( b ) 1.4% mlm camer, 67% NPOE, 1.6% rnlm TOPO, 2.2% mlrn 4C4NTPB, 0.8% mlrn KTCPB. 0, H+; A, Li+; V, Na+; 0, K+; A , Mg2+; and V, Ca2+ Table 2. Effect of various additives on the selectivity of the electrode for lithium Additive, YO mlm Log kE,L Carrier TOPO 4C4NTPB KTCPB H+ Na+ K+ Mg*+ Ca2+ I1 0.8 -1.8 -1.6 -4.0 -3.8 0.8 0.8 -2.6 -1.9 -1.2 -3.6 -3.4 I11 0.8 1.8 -1.9 -1.6 -2.7 -0.80 0.8 0.8 0.38 -2.3 -2.4 -3.5 -1.7 0.8 0.8 0.8 1.2 -2.0 -2.3 -2.3 -1.3 The e.m.f. response of the membrane when both TOPO and 4C4NTPB were present as additives is shown in Fig. 9(a) and (b). In this instance favourable effects were observed, and the Li+ selectivity was enhanced whereas the interference from divalent cations remained suppressed.Fig. 10 shows the selectivity of the membrane when various kinds of organo- phosphorus compounds are incorporated together with 4C4NTPB. The coexistence of TOPO and the lipophilic salt appreciably increased the Li+ selectivity of the electrode and optimum ratios of the additives were shown to exist. Combina- tions of TOPO with tetraalkylammonium tetraphenylborates of various aliphatic chain lengths were tested (Fig. 11) and the best electrode characteristics were obtained with the butyl group as the alkyl moiety. In order to investigate the effects of lipophilic additives, the change in ionic selectivity of the camer-free membrane containing various kinds of lipophilic additives was examined.The results are shown in Fig. 12. The plasticiser NPOE is known to show K+ selectivity in itself, as seen in the figure. At higher concentrations of TOPO the electrode responded to Li+ and to divalent cations to a greater extent than either Na+ or K+. This trend corresponds to the change in ion selectivity of the carrier-containing membranes. Other organophosphorus compounds incorporated in the carrier-free membranes gave similar trends in ion selectivities to those seen in carrier-containing membranes, although the effects were less distinct than those obtained with TOPO. These results could be explained if organophosphorus com- pounds could be shown to stabilise the lithium - carrier complex, perhaps owing to their interaction with the com- plexed cation as a monodentate ligand, or by acting as Li+ carriers in the membrane.The lipophilic salt 4C4NTPB, incorporated in a carrier-free membrane, brought about little change in ion selectivity, except for changes in the H+ response. 4C4NTPB appears to act differently to the organophosphorus compounds, at least no ligating interaction with the primary cation. 4CfiTPB was applied as the additive to the membrane based on the reported cyclic (I99 and non-cyclic (III)ll Li+-selective neutral carriers (membrane composition: 1.4% mlm carrier, 6849% mlm NPOE, 0.8-2.4% rnlm additives). The effectiveness of 4C4NTPB was again observed in the non-cyclic carrier III, U II Y Ill where Li+ selectivity over other alkali or alkaline earth metal ions was improved significantly with the addition of as little as 0.8% mlm 4C4NTPB.These results are shown in Table 2. One possible mechanism for its interaction might be the reduction of the electrical resistance of the membrane owing to the presence of lipophilic salts,ls although it might also be considered that 4C4NTPB as a lipophilic salt in the organic phase could prevent the uptake of cations into the membrane and hence the complexation of interfering ions that were not well suited to coordination by the carriers. Non-cyclic carriers have in general much more freedom in the conformation of their structures than cyclic camers which have coordination sites in a fixed cavity size. The steric interaction between the cation and the coordinants could vary with these non-cyclic carriers, depending on the atmosphere of the solvent, and thus favourable effects caused by the lipophilic additives were observed in this experiment.Further studies are required for clarification of the reaction mechan- isms.592 ANALYST, MAY 1987, VOL. 112 3 2 1 i; m 0 -I -1 -2 -3 0.8 0.8 1.6 2.7 0.8 1.6 0.8 1.6 0.8 1.6 1.6 TOPO TOP DOPP TOPi 1.1 2.2 2.2 2.1 1.1 2.2 1.1 0.8 0.8 0.8 2.2 4C4NTPB %m/m Fig. 10. Selectivity coefficient of PVC membranes containing various kinds of organophosphorus compounds in combination with 4C4NTPB. Membrane composition: 1.4% mlm carrier, 6 M 8 % mlm NPOE, 0.8% mlm KTCPB 3 2 1 iif * o r u m -1 -1 - 2 -3 .... . . . . . . H' -. ..... -.. ... .- ... .- -.-.. ....- - .. ..... ...... ..-...... - ......-..... -... H+ .- Li+ Li + - ......-......-......-..... .- ......-...... - ...... - ...... - ...... - ...... - .....-...... - . K+ ...- ...... - .......-.. . ...-... ..-... . .. I I I I I I I I I K+ K+ K+ ...-... K+ -...... -... K+ Ca*+ I I Fig. 11. Selectivity coefficients of PVC membranes containing various kinds of tetraalkylammonium tetraphenylborate in combination with TOPO. Membrane composition: 1.4% mlm carrier, 69% mlm NPOE, 0.8% mlm TOPO, 1.1% mlm hNTPB, 0.8% mlm KTCPB Table 3. Comparison of selectivity coefficients for the PVC matrix Li+-selective electrodes based on neutral carriers Membrane components Log kL% Carrier Plasticiser H+ Na+ K+ ETH149 . . . . . . . . . . TEHP -0.1 -1.3 -2.1 ETH1644 . . . . . . . . NPOE 0.7 -2.1 -2.2 Crownether . . . . . . . . NPOE -3.5 -1.0 -0.84 Crownether .. . . . . . . TEHP 0.67 -1.0 -1.7 Crownether . . . . . . . . NPOE -3.4 -2.2 -2.0 Thiswork . . . . . . . . NPOE 2.5 -1.5 -2.0 WithTOPOand4C4NTPB . . NPOE 2.2 -2.2 -3.0 ETH1810 . . . . . . . . NPOE 0.95 -2.45 -2.6 Requiredvalue . . . . . . 2.1 -4.3 -2.8 NH4+ -1.3 -2.1 -2.0 -0.47 -3.0 -2.5 - - - Mg2+ Ca2+ -3.7 -3.3 -3.2 -3.0 -3.7 -4.0 -2.8 -2.9 -4.6 -4.3 -4.0 -2.7 -1.9 -2.0 -1.5 -1.1 -3.4 -3.6 Reference 10 11 7 8 9 20 13 4 - (Year) (1975) (1981) (1982) (1982) (1984) (1984) (1986)ANALYST, MAY 1987, VOL. 112 3 t .* K+ . . . . .. ._---Ca2+ Mg2+ ’.---- I 0.8 1.7 1.7 1.7 0.8 TOPO TOP TOPi 4C4NTPB o/o mlm Fig. 12. Selectivity coefficients of carrier-free membranes contain- ing various li ophilic additives. Membrane composition: 70-71 % mlm NPOE, 0.82 mlm KTCPB Conclusion The use of tetrabutylammonium tetraphenylborate as a lipophilic additive gave a remarkable increase in the Li+ selectivity of PVC membrane electrodes based on non-cyclic carriers, especially when it was incorporated in the membrane in combination with organophosphorus additives such as TOPO.The optimum composition of the membrane is 1.4% mlm carrier, 67% mlm NPOE, 1.6% mlm TOPO, 2.2% mlm 4C4NTPB and 0.8% mlm KTCPB. In Table 3 comparisons are made between selectivity coefficients for the reported PVC matrix Li+-selective elec- trodes based on neutral carriers. In this work, Li+ selectivity over Na+ and K+ was enhanced 5- and 10-fold, respectively, by the use of lipophilic additives in the membrane. This could compare with the synthesis of a novel Li+ carrier.The results obtained are an indication that the characteris- tics of ion-selective electrodes based on non-cyclic carriers would be considerably improved by using a solvent atmo- sphere in which the carriers are incorporated. Although there are still difficulties in the application of sensor systems to real samples to be overcome,21 the effect of the solvent atmo- sphere should be of practical aid in developing lithium- selective electrodes. 593 The authors are indebted to Professor Toshiyuki Shono and Dr. Keiichi Kimura of Osaka University for supplying compound I1 and for helpful discussions. This work was supported in part by grants from the Science and Technology Agency. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.References Amidsen, A., Dan. Med. Bull., 1975,22,277. Canessa, M., Adragna, N., Salomon, H. S., Connolly, T. M., and Tosteson, D. C., N. Engl. J. Med., 1980,302,772. Tosteson, D. C., Sci. Am., 1981, 244, 164. Metzger, E., Ammann, D., Asper, R., and Simon, W., Anal. Chem., 1986,58, 132. Freiser, H., Editor, “Ion-Selective Electrodes in Analytical Chemistry,” Volume 1, Plenum Press, New York, 1978, Chapters 3 and 4. Schindler, J. G., Stork, G., Striih, H.-J., and Schal, W., Fresenius Z. Anal. Chem., 1978,290,45. Olsher, U., J. Am. Chem. SOC., 1982, 104,4006. Aalmo, K. M., and Krane, J., Acta Chem. Scand., Ser. A , 1982,36,227. Kitazawa, S . , Kimura, K., Yano, H., and Shono, T., J. Am. Chem. SOC., 1984, 106,6978. Guggi, M., Fiedler, U., Pretsch, E., and Simon, W., Anal. Lett., 1975,8, 857. Zhukov, A. F., Erne, D., Ammann, D., Guggi, M., Pretsch, E., and Simon, W., Anal. Chim. Acta, 1981, 131, 117. Hiratani, K., Taguchi, K., Sugihara, H., and Okada, T., Chem. Lett., 1986, 197. Hiratani, K., Okada, T., and Sugihara, H., Anal. Chem., in the press. Ammann, D., Morf, W. E., Anker, P., Meier, P. C., Pretsch, E., and Simon, W., Ion-Sel. Electrode Rev., 1983, 5 , 3. Dunitz, J. D., Editor, “Structure and Bonding,” Volume 16, Springer-Verlag , Berlin, 1973. Imato, T., Katahira, M., and Ishibashi, N., Anal. Chirn. Acta, 1984, 165,285. Hara, H., Okazaki, S., and Fujinaga, T., Bull. Chem. SOC. Jpn., 1980, 53,3610. Pretsch, E., Wegmann, D., Ammann, D., Bezegh, A., Dinten, O., Laubli, M. W., Morf, W. E., Oesch, U., Sugahara, K., Weiss, H., and Simon, W., in Kessler, M., Harrison, D. K., and Hoper, J., Editors, “Ion Measurements in Physiology and Medicine,’’ Springer-Verlag, Berlin, 1985, p. 11. Srinivasan, K., and Rechnitz, G. A., Anal. Chem., 1969, 41, 1203. Metzger, E., Ammann, D., Schefer, U., Pretsch, E., and Simon, W., Chimia, 1984,38,440. Gadzekpo, V. P. Y., Moody, G. J., and Thomas, J. D. R., Analyst, 1986, 111, 567. Paper A6119 Received January 20th, I986 Accepted October 21st, 1986

ISSN:0003-2654    

DOI:10.1039/AN9871200587

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 595 Flow Injection Determination of Inorganic Bromide in Soils with a Coated Tubular Solid-state Bromide-selective Electrode Jacobus F. van Staden Department of Chemistry, University of Pretoria, Pretoria 0002, South Africa Inorganic bromide can be determined in soil extracts by flow injection potentiometry at a rate of 80 samples per hour with a standard deviation of 1.6%. Soil-extracted samples (30 pl) are injected into a 1 mol dm-3 potassium nitrate carrier solution containing 100 mg dm-3 chloride as an ionic strength adjustment buffer. The sample - buffer zone formed is transported through a laboratory-made coated tubular solid-state bromide-selective electrode on to the reference electrode. The method is suitable for the determination of bromide in the range 1-5000 mg dm-3.Keywords: Soil; bromide determination; flow injection analysis; ion-selective electrodes; flow-through electrodes The use of brominated pesticides for the fumigation of soils has greatly increased in the past few years. Organically bound bromine is normally not taken up by plants from fumigated soil. However, these compounds are readily degraded to liberate inorganic bromide to the soil. The latter is easily taken up by plants. From both toxicological and control viewpoints, a knowledge of inorganic bromide concentrations in soil is therefore required. There are several methods available for the determination of inorganic bromide in soils, plants and foodstuffs following fumigation with brominated pesticides. 1-3 However, these methods are all tedious and require extensive analytical facilities.The development and use of halide ion-selective electrodes is more attractive than other detection techniques and offers the possibility of simpler and faster methods for the determination of halide ion concentration.4-6 Bromide ion-selective electrodes have been used for the determination of brominated vegetable oil concentrations in soft drinks,7 of bromide ion in wine8 and of bromide in soils.9 However, these are all manual methods and the batchwise mode involved is time consuming for laboratories where bromide levels are to be monitored continuously. Since its introduction in 1974-75, flow injection analysis (FIA) has established itself as an analytical technique that is suitable for increasing sample output in most ar,alytical laboratories.1 0 ~ 1 There are several considerations in the design of flow-through electrochemical detectors that can be interfaced with FIA. Potentiometric and amperometric detec- tors have included wall-coated (or open) tubular, packed-bed (or porous) tubular, wire, cascade, wall-jet and thin-layer designs. Slanina et a1.12 used the cascade configuration for the flow injection determination of bromide. However, some novel flow-through tubular arrangements for ion-selective electrodes have also been used as sensors13-17 in flow injection systems. In this geometric mode the sample solution is channelled through the tubular configuration across the sensing membrane in a kind of open path. The incorporation of a tubular ion-selective electrode into the conduits of a flow injection system seems an ideal design as the hydrodynamic flow conditions can be kept constant throughout the flow system.This approach opens new dimensions and was recently extended to the concept of coated tubular solid-state ion- selective electrodes18-21 incorporated into the conduits of flow injection systems. In the work described here, the combina- tion of FIA with a coated tubular solid-state bromide-selective electrode forms the basis of a study of the determination of inorganic bromide in soils. Experimental Reagents and Solutions All reagents were prepared from analytical-reagent grade chemicals unless specified otherwise. Doubly distilled, de- ionised water was used throughout. The water was tested beforehand for traces of chloride and bromide.All solutions were de-gassed before measurements by use of a water vacuum pump. The main solutions were prepared as follows. Ionic-strength adjustment reagent (ISA) Dissolve 202.22 g of potassium nitrate and 0.4206 g of potassium chloride in 1500 cm3 of distilled water in a 2-dm3 calibrated flask. Dilute this solution to 2 dm3 with distilled water. This gives a 1 mol dm-3 potassium nitrate solution containing 100 mg dm-3 chloride. (For a 0.1 mol dm-3 potassium nitrate solution, 20.222 g of potassium nitrate is used.) Standard bromide solutions Dissolve 29.7860 g of dried potassium bromide in 2 dm3 of distilled water to give a stock solution with a bromide concentration of 10 000 mg dm-3. Standard working solutions are prepared by dilution of appropriate aliquots of the stock solution with the ISA solution (containing 0.1 mol dm-3 potassium nitrate and 100 mg dm-3 chloride) to cover the range 5-5000 mg dm-3. Apparatus Coated tubular flow-through electrode construction The basic design of the coated tubular flow-through solid-state bromide-selective membrane electrode was the same as that used for the construction of a chloride-selective electrode previously described.18 The unit consisted of 0.025 mm thick silver metal foil wound around two pieces of Tygon tubing at both ends.An inner wire of a shielded cable was wound around the outside body of the silver metal cylinder to ensure electrical contact between the electrode and the Ionalyzer instrument. The whole unit was isolated with Araldite epoxy resin.The silver - silver bromide electrode was activated by anodic deposition of silver bromide as a fine membrane on the inner wall of the tubular silver cylinder. The coating was carried out at a current density of about 20 mA cm-2 and 0.1 mol dm-3 potassium bromide solution was circulated at a rate of 1.6 cm3 min-1 through the tubular cylinder.596 ANALYST, MAY 1987, VOL. 112 Fto w injection system A schematic diagram of the flow injection system used is illustrated in Fig. 1. A Carle microvolume two-position sampling valve (Carle No. 2014) containing two identical sample loops was used. Each loop has a volume of 30 pl. A Cenco sampler unit was used to supply a series of samples to the sampling valve system. The timing of the sampler unit was 45 s for sampling with zero wash time and valve actuation at 43 s.A Cenco peristaltic pump operating at 10 rev. min-1 supplied the carrier and reagent streams to the manifold system; the sampling valve system was synchronised with a Cenco sampler unit. Tygon tubing (0.51 mm i.d.) was used to construct the manifold; coils were wound round suitable lengths of glass tubing (15 mm 0.d.). The tubular flow-through bromide-selective electrode was incorporated into the conduits of the flow injection system as shown in Fig. 1. The potentials were measured at room temperature with an Orion Research (Model 901) micro- processor Ionalyzer. The detector output was recorded with a two-channel Cenco recorder (Model 34195-041). The con- structed flow-through tubular indicator electrodes were used in conjunction with an Orion 90-02 double-junction reference electrode with 10% mlV potassium nitrate as the outer chamber filling solution.Procedure The carrier stream (1 mol dm-3 potassium nitrate) is pumped at a constant flow-rate of 3.90 cm3 min-1 (Fig. 1). A pulse suppressor coil (200 cm x 0.51 mm i.d.) is incorporated between the peristaltic pump and the sampling valve. Samples taken from the turntable of an automatic sampler are injected automatically from a 30-pl sampling loop into the carrier stream by means of a two-position valve. Whereas one loop serves the carrier stream, the other draws the sample through at a constant flow-rate of 2.00 cm3 min-1. Injected samples are mixed with the carrier stream in a 105-cm mixing coil.Potassium nitrate (1 mol dm-3) is added at a flow-rate of 1.40 cm3 min-1 further downstream for improvement of hydrody- namic flow and mixed in a second mixing coil (160 cm) before the potential is measured in the coated tubular indicator electrode. To eliminate chloride interference in the determi- nation of inorganic bromide in soils, 100 mg dm-3 chloride is included in the 1 mol dm-3 potassium nitrate solutions. A 45-s Fig. 1. Manifold and sampling rate, 80 h-1 Sample cycle sampling time is used, giving a capacity of 80 samples per hour. The valve system is actuated on a time basis that is correlated with the sampler unit; the sampling valve is actuated every 43 s. Preparation of Soil Samples Extraction efficiency studies were conducted as follows. Sandy soil (92% sand, 5.5% silt, 2.5% clay, carbon content 0.19%), loamy sand soil (84% sand, 10% silt, 6% clay, carbon content 0.31%), sandy loam soil (66% sand, 27% silt, 7% clay, carbon content 0.39%) and loam soil samples (53% sand, 35% silt, 12% clay, carbon content 0.42%) were pre-treated and tested for traces of bromide to ensure that the amount of bromide initially present was negligible further on.Bromide was added as potassium bromide solutions to 25-g portions of the above-mentioned soil samples to give a bromide concen- tration range of 3, 5, 10, 20, 60, 100, 250 and 1000 pg of bromide per gram of oven-dried soil. Portions (25-g) of the soil samples with no bromide added were used as blank samples. The samples were dried overnight at 105 "C and ground to pass through a 2-mm sieve.The samples were transferred into screw-capped glass jars. ISA solution (25 cm3), containing 0.1 mol dm-3 potassium nitrate and 100 mg dm-3 chloride, was added to each sample. The jars were shaken vigorously on a reciprocating shaker for 30 min. The suspensions were centrifuged, the supernatant liquid filtered (Whatman No. 41 filter-paper) and bromide was determined in the filtrate. Results and Discussion The preparation procedure for laboratory-made tubular electrodes is simple and easy to conduct. However, experimental results revealed that greater care had to be taken in the preparation of suitable homogeneous coated tubular bromide-selective electrodes from AgBr alone compared with the preparation of a suitable chloride-selective electrode previously described.18 Maximum contact area is obtained by using a tubular electrode that is well coated.The maximum sensitivity was obtained when the electrode was coated, tested, left in ca. 500 mg dm-3 bromide solution, re-coated, etc., until maximum response was obtained, and then condi- tioned overnight in 20 mg dm-3 bromide solution. It was also necessary to carry out an actual test run of about 10 min every microprocessor lonalyzer flow Coated l i II tubular ff I1 Reference flow-through 11 11 electrode electrode ,! /La( b cm3 min-1 I I 1 I I I I L _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ J diagram of the FIA system. Valve loop size, 30 yl; sampling time, 45 s; wash time, 0 s; valve actuation at 43 s;ANALYST, MAY 1987, VOL. 112 597 morning under the same conditions as would be used for FIA, before actual work could be started for the rest of the day.A series of preliminary tests were conducted to optimise the flow injection system parameters and the performance of the electrode itself. These experiments revealed the same results as obtained for the chloride-selective electrode18 concerning the design, contact area and volume of the tubular assembly and the FIA parameters. The only difference was an increase to 1.40 cm3 min-1 during the addition of the reagent solution further downstream in order to maintain a reasonable sample throughput. An electrode (volume 15.7 pl) with an inside diameter of ca. 2 mm and a length of 5 mm gave the best results for the type of flow system shown in Fig. 1. Typical calibration graphs, illustrating the linear response of the electrode, are presented in Fig.2. The linear response range of the electrode was measured by using the following flow system parameters. Aspiration of standard working bromide solutions in 1.0 mol dm-3 KN03 via a 3.90 cm3 min-1 carrier stream into a single-line manifold system to the detector was carried out until a steady state was just obtained. The electrode showed a linear response between about 5 and 5000 mg dm-3. The calculated Nernstian response of the tested electrode is 57 k 1 mV decade-1 [Fig. 2(A)] with a correlation coefficient of 0.9996. The linear range and slope also depend on the pre-treatment of the electrode and care should be taken during the preparation of each new coated tubular electrode.The dynamic linear response range was less [Fig. 2(B)] than that shown in Fig. 2(A) when a sample volume of 30 p1 was injected into a single-line flow injection system with a carrier stream of 3.90 cm3 min-1. However, the deviation of the graph at low concentrations is the opposite of the deviation obtained for chloride-selective18 and iodide- selective20 electrodes. Real sample throughput of bromide samples and carryover and reproducibility of results are dependent on the FIA system (Fig. l ) , the main contribution arising from the practical response time of the coated tubular bromide-selec- tive electrode. The results indicated that the practical response time of the bromide-selective electrode is slower than that of the chloride-selective electrode18 with KN03 as -100 r 0 - 1 - 2 -3 -Log ([Br-l/mg d ~ n - ~ ) Fig.2. Experimental calibration graphs giving the electrode response range. (A) Aspiration of standard working bromide solu- tions in 1.0 mol dm-3 KN03 into a single-line manifold system via a 3.9 cm3 min-1 carrier stream. Aspiration was continued until a steady-state reading was just obtained on the detector. (B) Single-line manifold FIA system with 3.9 cm3 min-1 carrier stream of 1.0 mol dm-3 KN03 and a sample volume of 30 pl the ionic strength adjustment buffer. However, the practical response time is fast enough to give the flow system used (Fig. 1) .a sample throughput of about 80 samples per hour. The pH of soil samples can vary considerably and various contaminants that may interfere are usually found in soil.The coated tubular bromide-selective electrode was also subjected to these interferences. Interferences were evaluated by conducting a series of experiments using flow systems. However, it was clear from the experimental results that the FIA system was much less affected than the continuous sample-flow method. High levels of ammonia (>lo00 mg dm-3) in soil samples destroyed the electrode response rapidly, owing to the formation of soluble diammine silver complexes. It is therefore not advisable to use the electrode for the determination of bromide in freshly fertilised soil. However, the electrode performance was not influenced when the pH in the samples was changed from 2 to 12. Iodide, cyanide, sulphide and arsenate, which form silver salts more insoluble than silver bromide, interfere.However, the con- centration of these interferents in soil is negligible and there was no noticeable effect on the electrode system. Chloride levels in soils varied considerably. The interference of chloride with the tubular bromide-selective electrode increased linearly with increasing chloride concentrations, in solutions containing no bromide, which confirmed the results obtained by Onken et aZ.9 The results obtained with mixed standard solutions containing bromide and chloride are summarised in Table 1. It is clear that interference from chloride became noticeable in solutions containing less than 100 mg dm-3 bromide, which again confirmed the results of Onken et aZ.9 However, applying the t-test to the two sets of data (Table 1) obtained from mixtures of 20-250 mg dm-3 bromide with 0 and 1000 mg dm-3 chloride showed no statistical difference between them (95% probability).This implies that only bromide concentrations lower than 20 mg dm-3 are affected by chloride present at up to 1000 mg dm-3. A series of experiments were therefore performed to try to eliminate chloride interference. The results revealed that the incorpora- tion of 100 mg dm-3 chloride in an ionic strength adjustment Table 1. Interference studies of standard solutions containing combinations of bromide and chloride. Studies conducted with the FIA system in Fig. 1 with 1.0 mol dm-3 potassium nitrate as the ionic strength adjustment solution. Values given are peak heights measured in mm Chloride concentration in Bromide mixture/mg dm-3 concentration/ mg dm-3 0 lo00 500 300 150 80 250 110 110 110 110 110 110 100 95 95 94 95 94 95 60 86 88 88 88 87 87 40 76 80 80 79 80 79 20 62 68 67 68 67 66 10 45 59 55 54 54 54 Table 2. Performance and electrode response stability over a period of time.All results are the mean of triplicate determinations Period Dayl* Day2 Day7 Day14 Concentration of injected bromide solutions/mgdm-3 . . . . . . 10 9 8 8 40 39 38 37 100 98 96 93 250 248 245 241 500 498 495 490 1000 995 986 979 5000 4991 4979 4968 * Calibrations standards started. Days 2, 7 and 14 refer to the response obtained on day 1 as calibration.598 ANALYST, MAY 1987, VOL. 112 Table 3. Recovery of inorganic bromide in soil samples fortified with potassium bromide Bromide recovered Sandy soil Loamy sand soil Bromide added Mass*/ Mass*, Mass*/ Mass*, CLg g-l pg g-1 Yo pg g-1 Yo - <1 0 <1 - 3 3.1 k0.05 103 3.1 f0.05 103 5 5.2k0.08 104 5.1f0.08 102 10 9.6k0.13 96 9.5f0.14 95 20 19f0.25 95 19f0.27 95 60 59k0.70 98 58f0.74 97 100 9 6 f 1.2 96 9 7 f 1.1 97 250 246k2.1 98 247f2.0 99 lo00 988k4.6 99 989f4.4 99 * Average of five soil replicate samples f standard error.Sandy loam soil Loam soil Mass*/ Mass*, <1 - 2.9f0.05 97 4.9k0.07 98 9.5f0.14 95 19k0.26 95 57k0.72 95 9 4 f 1.2 94 246f2.0 98 979f4.6 98 pg g-1 Yo Mass*/ <l 2.8 f 0.04 4.8 f 0.09 9.4 f 0.15 18 k 0.26 56 f 0.74 9 2 f 1.4 244 f 2.0 974 k 4.5 Pg g-l Mass*, YO 93 96 94 90 93 92 98 97 - A , mples Fig. 3. Ty ical strip-chart recording for the determination of bromide wit[ the FIA system of Fig.1 and 1.0 mol dm-3 KN03 as ionic strength adjustment solution. From left to right: 5000-5 mg dm-3 standard bromide solutions followed by samples. Standard solutions were injected four times each; samples in duplicate. Recorder paper speed = 2 mm min-1. Recorder range = 20 mV. Bromide concentrations: (A) 5000; (B) 4000; (C) 3000; (D) 2000; (E) 1OOO; (F) 500; (G) 250; (H) 100; (I) 80; (J) 60; (K) 40; (L) 20; (M) 10; and (N) 5 mg dm-3 (p.p.m.) buffer containing 1.0 mol dm-3 potassium nitrate gave the best results for the proposed FIA system (Fig. 1). Thus interference of chloride in bromide solutions is reduced to a level of less than 20 mg dm-3 bromide, where the interference of 1000 mg dm-3 chloride became noticeable. Inclusion of 100 mg dm-3 chloride in the ionic strength adjustment buffer also gave a faster response time for solutions containing 1000-5000 mg dm-3 of bromide.A typical representative recorder output for the determina- tion of bromide with a FIA system (Fig. 1) at a sampling rate of 80 determinations per hour is illustrated in Fig. 3. Determinations performed in a random order shows that carryover from one sample to another is negligible with no base-line drift experienced. It can be seen from the typical recorder output (Fig. 3) that the coated tubular solid-state bromide-selective electrode gave a very stable base line. The performance and stability of electrode response were evaluated over a period of time (Table 2). It is clear that the Table 4. Performance and reproducibility of the proposed flow injection method (FIA) for the determination of bromide in soil. Comparison of inorganic bromide concentrations (pg g-1) recovered from soils using the proposed FIA method against a standard iodimetric titration method Soil sample 1 2 3 4 5 Loamy sand soil: 1 2 3 4 5 Sandy loam soil: 1 2 3 4 5 1 2 3 4 5 Sandy soil: Loam soil: Titration 3.0 9.5 58 247 987 5.2 19 96 248 990 2.9 19 56 95 247 4.7 9.5 57 245 977 FIA 3.1 9.6 59 246 988 5.1 19 97 247 989 2.9 19 57 94 246 4.8 9.4 56 244 974 Bromide concentratiordpg g-1 Coefficient of variation,* Yo 1.59 1.48 1.27 0.82 0.43 1.56 1.40 1.28 0.81 0.44 1.58 1.41 1.28 1.09 0.81 1.57 1.48 1.28 0.81 0.44 * Mean result of 15 tests in each instance with relative standard deviation for the flow injection method.response was very stable.Due to oscillations of about 1 mV per day in electrode potentials, calibration on a daily basis was necessary. An electrode stored for 10 months maintained a slope of 57-55 mV, which indicated a relatively long lifetime. The flow injection system (Fig. 1) with an incorporated coated tubular bromide-selective electrode was also applied t o the determination of bromide in soils. KN03 (1.0 mol dm-3) containing 100 mg dm-3 chloride was used as the ionic strength adjustment buffer. Soil samples were fortified with potassium bromide solutions. The extracted bromide was then determined using the flow injection system. The extraction and measurement procedure gave satisfactory results for inorganic bromide determination in soil. The flow injection tubular electrode system measured over 90% of the bromide added (Table 3).The lower limit of detection was 1 mg dm-3. The performance and reproducibility of the proposed flow injection potentiometric method are shown in Table 4. In addition to a high sample throughput (80 h-1) over a wideANALYST, MAY 1987, VOL. 112 599 concentration range, the procedure is characterised by good reproducibility ( 4 . 6 % ) . Direct flow injection potentiometric measurement of bromide ion in soil gave results fairly similar to those obtained by a standard iodimetric titration method (Table 4) with thiosulphate.22723 Conclusion A coated tubular bromide-selective flow injection electrode system is suitable for the determination of extracted inorganic bromide in soils at a rate of about 80 samples per hour with a coefficient of variation of better than 1.6%.The system offers certain advantages over conventional sensor endcaps and manual methods with the elimination of chloride interference. The author thanks the Council for Scientific and Industrial Research, Pretoria, and the University of Pretoria for financial support. He also thanks Mr. C. C. P. Wagener and Miss M. L. Aveling for assistance in performing some of the experiments. 1. 2. 3. 4. 5. 6. References Turner, A., J. Sci. Food. Agric., 1964, 15, 265. Beckman, H., Allen, P. T., Crosby, D. G., Gauer, W. O., and Mourer, C., J. Food. Sci., 1967,32, 595. Getzendaner, M. E., Doty, A. E., McLaughlin, E. L., and Lindgren, D. L., J. Agric. Food. Chem., 1968, 16,265. LaCroix, R. L., Keeney, D. R., and Walsh, L. M., Soil Sci. Plant Anal., 1970, 1, 1. Hipp, B. W., and Langdale, G. W., Soil Sci. Plant Anal., 1971, 2 , 237. Smart, R. St. C., Thomas, A. D., and Drover, D. P., Soil Sci. Plant Anal., 1974, 5 , 1. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Turner, D. L., J. Food Sci., 1972,37,791. Graf, J. E., Vaughn, T. E., and Kipp, W. H., J. Assoc. Off. Anal. Chem., 1976,59,53. Onken, A. B., Hargrove, R. S . , Wendt, C. W., and Wilke, 0. C., Soil Sci. SOC. Am. Proc., 1975,39, 1223. RifiEka, J., and Hansen, E. H., “Flow Injection Analysis,” Wiley, Chichester, 1981. “Flow Injection Analysis Bibliography,” Tecator, Hoganas, Sweden, 1985. Slanina, J., Lingerak, W. A., and Bakker, F., Anal. Chim. Acta, 1980, 117, 91. van der Linden, W. E., and Oostervink, R., Anal. Chim. Acta, 1978, 101,419. Meyerhoff, M. E., and Kovach, P. M., J. Chem. Educ., 1983, 60,766. Frend, A. J., Moody, G. J., Thomas, J. D. R., and Birch, B. J., Analyst, 1983, 108, 1357. Mascini, M., and Palleschi, G., Anal. Chim. Acta, 1978, 100, 215. Alegret, S., Alonso, J., Bartroli, J., Paulis, J. M., Lima, J. L. F. C., and Machado, A. A. S . C., Anal. Chim. Acta, 1984, 164, 147. van Staden, J. F., Anal. Chim. Acta, 1986, 179, 407. van Staden, J. F., Anal. Lett., 1986, 19, 1407. van Staden, J. F., Fresenius 2. Anal. Chem., 1986,325,247. van Staden, J. F., Analyst, 1986, 111, 1231. Kolthoff, I. M., and Belcher, R., “Volumetric Analysis, Volume 3, Titration Methods,” Interscience, New York, 1957, pp. 256. Kempton, R. J., and Maw, G. A., Ann. Appl. Biol., 1972,72, 71. Paper A61356 Received September 22nd, 1986 Accepted November 24th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9871200595

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 601 Voltammetric Study of Copper(l1) Dialkyldithiophosphates Formed by the Interaction of Dialkyldithiophosphates with Copper Salts Miles J. Hutchings, G. J. Moody and J. D. R. Thomas* Department of Applied Chemistry, Redwood Building, UWIST, P.O. Box 13, Cardiff CFl 3XF, UK The in situ formation in organic media of copper(l1) 0,O'-dialkylphosphorodithioates [dialkyldithiophosphates (DDPs), where alkyl = 2-methylethyl (isobutyl) or I-methylethyl (isopropyl)] can be detected voltammetrically at the mercury electrode. The results show that the electrode reactions involve irreversible, diffusion-controlled processes with adsorption phenomena featuring significantly in studies involving DDP (alkyl = isopropyl) species. From a quantitative analytical perspective, the measurement of the Cull reduction peak height may prove useful for the determination of CuDDP complexes.Voltammetric wave assignments and a scheme for the processes occurring are presented. Finally, the log(stabi1ity constant) of copper(l1) - DDP (alkyl = isobutyl) is calculated to be 17.3. Keywords : Dialkyldith iop h osp h a tes; 0,O ' -dia Ik ylp h osp h o rodith ioa tes; volta m m e try; copper(l1) dia Ikyl- dithiophosphates Metal dialkyldithiophosphate (0, 0'-dialkylphosphorodithio- ates)(MDDPs) are added to lubricating oils to serve as antioxidants. In addition, owing to the formation of protective films on metal surfaces, they prevent the corrosive attack of oxidation products and also exhibit very useful load-carrying properties.1 Zinc dialkyldithiophosphates (ZDDPs) have been used as lubricating oil antioxidants for many years and their inhibition of hydrocarbon autoxidation was first investi- gated by Kennerly and Patterson2 in 1956.A wealth of information has since been accumulated on the mechanism of action2-6 and on the analysis of ZDDPs.7-21 Other metal DDPs in addition to zinc have received attention although their performance as antioxidants is not so well documented. Copper dialkyldithiophosphates (CuDDPs) have been reported to exhibit chain-breaking, antioxidant activity in the radical initiated oxidation of a hydrocarbon such as cumene,3 and it is proposed that this may arise owing to the participation of a copper(1) species in the inhibitory action. Recently, findings involving the antioxidant behaviour22 and electron spin resonance spectra23 of CuDDPs have been published, but they appear to present conflicting information with the oxidation state of the copper species in solution being a contentious issue.A voltammetric, investigation of the speciation of CuDDP systems was therefore undertaken. CuDDP formation was achieved by the following inter- actions: (i) Copper(I1) 9-octadecanoate (copper oleate) and zinc di( 1,l-dimethylethy1)dithiophosphate (ZDDP, alkyl = isobutyl). (ii) Copper(I1) perchlorate and ZDDP (alkyl = isobutyl). (iii) Copper(I1) oleate and zinc di( 1-methylethy1)dithio- phosphate (ZDDP, alkyl = isopropyl). (iv) Copper(I1) perchlorate and ZDDP (alkyl = iso- (v) Copper perchlorate and NH4DDP (alkyl = isopropyl). ProPY9 - Experimental Reagents and Solutions Either copper(I1) perchlorate hexahydrate (BDH Chemicals, laboratory-reagent grade) or copper(I1) oleate (a gift from Esso Chemicals, Abingdon) was used as a source of copper for the in situ formation of CuDDP in solution. Solvent media used were heptane (BDH Chemicals, laboratory-reagent grade), absolute ethanol (James Burrough, analytical-reagent * To whom correspondence should be addressed.grade) or an American Society for Testing and Materials (ASTM) solvent. The ASTM solvent consisted of propan-2- 01, toluene and water (50 + 49.5 + 0.5 V/V). Solutions of ZDDP (alkyl = isobutyl), ZDDP (alkyl = isopropyl) and NH4DDP (alkyl = isopropyl) (gifts from Esso Chemicals) were prepared in absolute ethanol. Supporting electrolyte solutions were 0.1 M sodium perchlorate (BDH Chemicals, analytical-reagent grade) in absolute ethanol.Instrumentation All voltammetric experiments were carried out with a PAR 174A polarographic analyser coupled to a PAR 303A static mercury dropping electrode (SMDE) assembly. Voltammo- grams were recorded using either a Hewlett-Packard 7040A or a Bausch and Lomb Omnigraphic 2000 x - y recorder. Procedure Voltammetric determinations were conducted under quies- cent conditions at ambient temperature. Supporting elec- trolyte solutions (0.1 M; 8 cm3) contained within the polaro- graphic cell were purged for 8 min with oxygen-free nitrogen. Small volumes (5-50 mm3) of concentrated analyte were then spiked into the cell using a Finnpipette and the measurements made. For differential-pulse and d.c.polarographic experiments, a drop-time of 0.5 s and modulation amplitude of 50 mV were used throughout. Linear sweep and cyclic voltammetric experiments were carried out at a hanging mercury drop electrode (HMDE) with an area of 0.020 cm2. Scan rates and scan ranges are indicated on the voltammograms presented. All potentials were recorded versus a silver - silver chloride reference system with an internal filling solution of 1 M lithium chloride dissolved in ethanol - water (1 + 1 V/V). Results and Discussion Copper(II) Oleate - ZDDP (Alkyl = Isobutyl) This interaction was studied by differential-pulse polaro- graphy, direct-current polarography, linear sweep voltam- metry and cyclic voltammetry in various solvents. Differential-pulse polarography Heptane solvent.Copper(I1) oleate in heptane in the presence of ZDDP (alkyl = isobutyl) results in the formation of a CuDDP complex. This is apparent on examination of theANALYST, MAY 1987, VOL. 112 602 ~ Table 1. Polarographic waves arising from the interaction of copper(I1) oleate and ZDDP (alkyl = isobutyl) Wave I Wave I1 Wave111 Wave IV -Ed Id V CLA -Ed Id V CLA 0.41 0.16 +O. 180 0.10 0.125 0.37 0.435 0.48 0.915 0.22 -Ed Id V CLA Copper(I1) ZDDP/ Ed Id oleate*/mM mM V CrA 0.41 0.32 +O. 165 0.34 0.145 1.01 0.465 0.72 0.925 0.52 * Copper(I1) oleate (0.41 m ~ ) has a peak potential (E,) of +0.255 V and a peak current (Ip) of 0.29 W. 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1.0 PotentialN Fig. 1. Differential- ulse polarogram at the SMDE of 0.41 mM cop r(I1) oleate (in feptane) (A) and of 0.41 m~ cop er(I1) oleate $n eptane) with added 0.32 mM zinc diisobutyldithiopiosphate (B).can rate: 10 mV s-1 0.3 -0.1 -0.5 -0.9 -1.3 PotentialN Fi . 2. (A) Differential-pulse and (B) sampled d.c. polarograms of O.!l m~ zinc diisobutyl&thiophosphate (in ethanol) at the SMDE. Scan rate: 10 mV s-1 polarographic profile of copper(I1) oleate in both the absence and presence of ZDDP (alkyl = isobutyl) (Fig. 1). The broad peak in Fig. 1 arises from the reduction of copper(I1) in copper(I1) oleate to copper metal at the SMDE. On adding ZDDP (alkyl = isobutyl), a CuDDP reduction wave appears (Fig. 1, B, wave 111) together with three other waves (Fig. 1, B, waves I, I1 and IV). The peak potentials (Ep) and peak currents (Ip) of these waves are given in Table 1.Waves I and I1 may be attributed to the formation of mercury(I1) complexes from interaction of the SMDE with the DDP anion. Bond et a1.24 have characterised these waves for L I I 1 I I I I 0.3 0.1 -0.1 -0.3 -0.5 -0.7 -0.9 -1.1 -1.3 PotentialN Fig. 3. Sam led d.c. polarogram of 0.79 m~ cop r(I1) oleate (in heptane) in tEe resence of 0.2 m~ zinc diisobutyEthiophosphate. Scan rate: 10 m5s-1 studies involving N&DDP in acetone and proposed the following electrode processes: 2HgO + 6DDP- 2HgIIDDP3- + 4e- - (1) 2HgIIDDP3- + HgO 3HgIIDDP2 + 2e- . . (2) Indeed, the sampled d.c. polarogram of ZDDP (alkyl = isobutyl) (2.1 X 10-4 M in ethanol) shows three distinct processes (Fig. 2). Waves I and I1 may be assigned to the oxidation processes (2) and (l), respectively.The third process (wave IV) corresponds to the reduction of Zn2+ to ZnO at the mercury electrode and the peak potential of - 1.05 V versus silver - silver chloride for ZDDP (alkyl = isobutyl) observed here in an ethanolic medium agrees with the observations of Shafiqul Alam et aZ.21 who noted a peak potential of -0,900 V versus silver - silver chloride for the differential-pulse polarographic determination of ZDDP (alkyl = isopropyl) in a medium containing 0.1 M tetraethyl- ammonium perchlorate in dimethylformamide. On adding copper(I1) oleate to a solution of ZDDP (alkyl = isobutyl), there is interaction and wave I11 appears (see Fig. 1, B). After interaction, the sample d.c. polarographic trace of wave I1 appears as a cathodic process (Fig.3); in the absence of copper(I1) oleate it is an anodic electrode reaction. As a reversal in the electrode reaction can be explained by considering reaction (l), if DDP- is removed by complexation with copper(II), then a cathodic process is required to redress the equilibrium. Likewise, a cathodic process would also be expected for process (2) (wave I) but this is not observed as wave I still appears anodic (Fig. 3). Bond et aZ.24 observed a reversal in both electrode processes when HgIIDDP2 was added to a solution of DDP- at the DME, and this is consistent with processes (1) and (2). The half-wave potentials of waves I and I1 cited24 are +0.14 and -0.121 V versus silver - silver chloride for 5 X 10-4 M NHJIDP (alkyl = ethyl). Peak potentials of waves I and I1 for ZDDP (alkyl = isobutyl) (2 x 10-4 M) in this study occurred at +0.16 and -0.14 V, respectively, versus silver - silver chloride.TheseANALYST, MAY 1987, VOL. 112 0.50 so-@ I 0.46 603 - (a) - V - 0 V 0 0 V 0 0 0 V 8 0 0 0 0 0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8 -0.9 -1.0 -1.1 PotentialN Fig. 4. Differential-pulse polarograms of co per(I1) oleate (in heptane) (A = 0.14, B = 0.28 and C = 0.75 m.5 in the presence of 0.79 mM zinc diisobutyldithiophosphate at the SMDE. Scan rate: 10 mV s-1 Table 2. Parameters for waves I1 and I11 for different concentrations of copper(I1) oleate in the presence of 0.79 mM ZDDP (alkyl = isobutyl Wave I1 Wave111 Copper( 11) oleate/mM 0.14 0.28 0.41 0.52 0.64 0.75 0.86 0.95 0.14 0.28 0.41 0.52 0.64 0.14 0.28 0.41 0.52 0.64 - - -Ed Id V C L A 0.090 2.20 0.102 2.05 0.112 1.93 0.119 1.82 0.125 1.63 0.130 1.59 0.133 1.55 0.138 1.52 0.092 2.85 0.105 2.75 0.120 2.62 0.128 2.35 0.135 2.31 0.142 2.20 0.132 2.33 0.140 2.30 0.145 2.15 0.148 1.97 0.151 1.82 0.155 1.62 WJ -Ed Id mV v PA 172 0.445 0.175 176 0.555 0.425 179 0.465 0.675 182 0.470 1.OOO 180 0.480 1.200 182 0.485 1.350 176 0.490 1.600 176 0.500 1.800 160 0.460 0.200 156 0.465 0.425 165 0.475 0.675 160 0.495 1.050 158 0.500 1.400 136 0.465 0.200 152 0.475 0.425 140 0.485 0.775 144 0.490 1.030 160 0.499 1.250 - 156 - - 128 - WJ mV 72 72 70 64 62 64 60 58 76 72 69 60 59 72 70 66 64 60 - - observations are in good agreement considering the different supporting media employed, namely acetone and ethanol.It is wave.111 that is ascribed to the formation of the CuDDP complex and is of importance in this work.Fig. 4 shows the differential-pulse polarographic profiles corresponding to increasing concentrations of copper(I1) oleate (dissolved in heptane) interacting with ZDDP (alkyl = isobutyl) (0.79 mM) in an ethanolic medium. Table 2 lists the values of peak potentials (Ep), peak currents (Ip) and peak widths at half-peak height (W,) obtained from the mean of three experiments involving copper(I1) oleate (dissolved in hep- tane) and ZDDP (alkyl = isobutyl) (0.79 mM). The variations of Ep and I with the concentration of copper(I1) oleate are representecfgraphically in Fig. 5(a) and (b), respectively. It is obvious from these graphs that the peak height of wave I11 is more reproducible than the peak potential for the copper(I1) oleate - ZDDP (alkyl = isobutyl) system.It follows, there- fore, that whereas the measurement of peak heights may have analytical implications, peak potentials cannot be used for quantification. 0.44 2.0 0 1 I I I 0 0 0.2 0.4 0.6 0.8 1.0 [Copper oleatellmhn Fig. 5. Graph of (a) eak potential and (b) peak current versus concentration of copperbI oleate (in heptane) at constant concentra- tion of zinc diisobutyldithiophosphate (0.79 mM) for wave I11 in differential-pulse polarography. Scan rate: 10 mV s-1. 0, 0 and V represent separate expenments 0 -0.1 -0.2 -0.3 -0.4 -0.5 Potentia I/V Fig. 6. Time-dependent differential-pulse polarograms of 0.69 mM copper(I1) oleate (in ASTM solvent) in the presence of 0.14 mM Zinc diisobutyldithiophosphate.A, B and C were run at 7-min intervals. Scan rate: 10 mV s-1 ASTM solvent. Experiments were also conducted with copper(I1) oleate dissolved in ASTM solvent. The results differ in one major respect from those involving heptane in so far as a time dependence of waves I1 and I11 is encountered (Fig. 6). With successive scanning, the peak height of wave I1 decreases whereas that of wave 111 shows a definite increase. The visible absorption band, attributed to CuIIDDP,25 cen- tred at 418 nm for this interaction also shows a time dependence and increases with time suggesting that wave 111 is due to the reduction of a copper(I1) species and the concentration of this species increases with time. It is not until after ca. 150 min that equilibrium is reached, giving a constant absorbance (A = 418 nm) of CuIIDDP (Fig.7). Direct-current polarography The d.c. polarographic behaviour of 0.69 mM copper(I1) oleate (dissolved in heptane) in the presence of ZDDP (alkyl = isobutyl) (0.14 mM) is illustrated in Fig. 8 together with the differential-pulse polarogram for comparison. From an analy- sis of the d.c. polarograms, it is possible to ascertain values of604 ANALYST, MAY 1987, VOL. 112 A ‘I1 I I I P-J-J 0 50 1 00 150 18x19 2124 2544 Time after mixing/min Time/h Fig. 7. Variation of spectral absorbance (Amax. at 418 nm) with time for copper(I1) oleate (in ASTM solvent) in the presence of zinc diisobutyldithiophosphate -0.1 -0.3 -0.5 -0.7 PotentialN Fig. 8. Differential-pulse and d.c. polarograms of 0.69 mM copper(I1) oleate (in heptane) in the presence of 0.14 mM zinc diisobutyldithiophosphate.Scan rate: 10 mV s-1 Table 3. D.c. polarographic data for copper(I1) oleate (0.69 mM) in the presence of ZDDP (alkyl = isobutyl) (0.14 mM) Half-wave Limiting potential/ current/ Solvent V (= Et) PA( = I,) an,* an,? Heptane . . -0.464 0.88 0.86 0.88 ASTM . . -0.477 0.93 0.94 0.98 * an, determined from graphs of log (Id - I)/I. t an, determined from IEf - Etl = 51.7/ana mV. the half-wave potential (EJ and the product (an,) of the transfer coefficient (a) and the rate-determining number of electrons (n,). The results for copper(I1) oleate (0.69 mM) dissolved in heptane and ASTM solvent in the presence of ZDDP (alkyl = isobutyl) (0.14 mM) are shown in Table 3. Heptane solvent studies Linear sweep voltammetry.Linear sweep voltammograms of 0.69 mM copper(I1) oleate (dissolved in heptane) in the presence of ZDDP (alkyl = isobutyl) (0.14 mM) are illustrated in Fig. 9 at three different scan rates. Table 4 lists the peak potentials (EJ and peak current (Ip) values for these experiments together with values of an,. ZdV exhibits a certain degree of constancy, whereas ZdV values vary considerably over the scan ranges used (Table 4). It is therefore possible to infer that the reduction process giving rise to wave I11 is diffusion controlled and a negative shift of Ep with scan rate is diagnostic of an irreversible process.26 Consequently, the electrode process can be explained by the equation for an irreversible electrode reaction at a HMDE.27 -0.3 -0.4 -0.5 -0.6 PotentialN Fig.9. Linear sweep voltammo ram of 0.69 mM copper(I1) oleate (in heptane) in the presence of 8.14 mM zinc diisobutyldithiophos- phate at the HMDE (area 0.020 cm2) at scan rates of 10, 20 and 50 mV s-l for A, B and C, respectively Table 4. Linear sweep voltammetric data for copper(I1) oleate (0.69 mM) in the presence of ZDDP (alkyl = isobutyl) (0.14 mM) Scan I p t ! IJV! rate/ IEp - Ep/& PA ClA mVs-1 -Efl I&A mV an,* V-+S* V-ls 10 0.504 0.55 37 1.29 5.5 55.0 20 0.513 0.91 53 0.90 6.4 45.5 50 0.518 1.54 39 1.22 6.4 30.8 100 0.525 1.65 40 1.19 5.2 16.5 * an, determined from IE, - Ep,21 = 47.7/ana mV. 1- Mean IdVa = 6.0; s = 0.78; CV = 13.1%. Cyclic voltammetry. The cyclic voltammetric profile of 0.69 mM copper(I1) oleate (dissolved in heptane) in the presence of ZDDP (alkyl = isobutyl) (0.14 mM) is shown in Fig.10. The forward scan involves the electrochemical reduction of the CuDDP complex and the reverse scan is the oxidation profile where the oxidation of CuO to Cu2+ occurs. The difference between the anodic and cathodic peak potentials (Epa - Epc) is 53 mV and the ratio of the cathodic and anodic peak currents (&/Ipa) is 1.4. These values confirm the linear sweep observations that the electrode process is irreversible. ASTM solvent studies The stationary electrode behaviour of copper(I1) oleate dissolved in ASTM solvent differs from that for heptane as solvent in so far as an additional reduction wave is observed at potentials more negative than the main CuDDP reduction profile [Fig.11(A-C)]. From Fig. 11 (A-C), it is evident that the postwave IIIa becomes more pronounced with increasing scan rate. Also, if the linear sweep voltammetric experiment is delayed, wave IIIa intensifies and becomes larger than the CuDDP reduction wave I11 [Fig. 11 (a-c) J. These observations are consistent with a process involving adsorption of an electroactive species, as at larger scan rates adsorption waves develop more than the associated diffusion controlled waves.28 Increasing the delay time (td) before carrying out the voltammetric scan has the effect of enabling more of the electroactive substance to become adsorbed at the mercury surface and therefore produce a large peak current.29 The cyclic voltammogram of the above system (Fig. 12) possesses a similar oxidation profile to that already discussed for studies involving copper(I1) oleate dissolved in heptane.Also, as encountered with the linear sweep experiments, the presence of an adsorption peak is obvious (Fig. 12, wave IIIa).ANALYST, MAY 1987, VOL. 112 605 I Reduction 31 t/ Oxidation 1 I I I PotentialN -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 Fig. 10. Cyclic voltammogram of 0.69 mM copper(I1) oleate (in heptane) in the presence of 0.14 mM zinc diisobutyldithiophosphate at the HMDE (area 0.020 cm2). Scan rate: 10 mV s-l -0.4 -0.5 -0.6 PotentialN Fig. 11. Linear sweep voltammograms of 0.69 mM copper(I1) oleate (in ASTM solvent) in the presence of 0.14 mM zinc diisobutyldithio- phosphate. Scan rates of 20, 50 and 100 mV s-1 for A, B and C, respectively; scan rate of 100 mV s-1 for delay times of 5,20 and 35 s for a, b and c, respectively Copper(I1) Perchlorate - ZDDP (Alkyl = Isobutyl) The polarographic and stationary electrode behaviour pattern of copper(I1) perchlorate complexing with ZDDP (alkyl = isobutyl) is, as expected, very similar to the interaction discussed above for copper(I1) oleate.The characteristics of the CuDDP reduction wave for both studies are in good agreement and comparisons of peak potential (Ep), peak current (Ip) and peak width at half-peak height (WJ for (i) 0.14 mM copper(I1) perchlorate - 0.78 mM ZDDP (alkyl = isobutyl) and (ii) 0.14 mM copper(I1) oleate (heptane) - 0.79 mM ZDDP (alkyl = isobutyl) are as follows: (i) copper(I1) perchlorate: Ep = -0.452 V; Ip = 0.15 PA; W, = 62 mV; (ii) copper(I1) oleate: Ep = -0.460 V; Ip = 0.20 PA; W, = 72 mV.Copper(I1) Oleate - ZDDP (Alkyl = Isopropyl) Pola rograp hy The major difference between the voltammetric response of interaction between copper(I1) oleate and ZDDP (alkyl = -0.2 -0.4 -0.6 -0.8 PotentialN Fig. 12. Cyclic voltammogram of 0.69 mM copper(I1) oleate (in ASTM solvent) in the presence of 0.14 mM diisobutyldithiophosphate at the HMDE (area 0.020 cm*). Scan rate: 50 mV s-1 0.1 -0.1 -0.3 -0.5 -0.7 -0.9 -1.1 PotentialN Fig. 13. Differential-pulse polarograms of 0.50 mM zinc diisopro yl- dithiophosphate (A) and 0.42 mM copper(I1) oleate (in heptaney in the presence of 0.33 mM zinc diisopropyldithiophosphate (B) at the SMDE. Scan rate: 10 mV s-1 isopropyl) as opposed to interaction with ZDDP (alkyl = isobutyl) is the presence of adsorbed species.Both ZDDP (alkyl = isopropyl) and the CuDDP complex formed from the interaction possess adsorption post-peaks (Fig. 13) and these have been designated waves IIa and IIIa. Bond et al.24 also observed such an adsorption - desorption process for NH4DDP (alkyl = ethyl) at the DME and attributed the adsorption wave to a HgIIDDP species. Linear sweep voltammetry Linear sweep profiles of the copper(I1) oleate - ZDDP (alkyl = isopropyl) interaction clearly show the adsorption phe- nomena mentioned above. With reference to Fig. 14, for 0.14 mM copper(I1) oleate, two principal reduction waves are observed (waves IIa and IIIa). These can be ascribed to the adsorption of HgIIDDP (IIa) and CuDDP (IIIa) complexes. The fact that the peaks are symmetrical about the peak potential (Ep) suggests that they are adsorption waves30 and a time dependence is also characteristic of adsorption.29 At this low concentration of copper(I1) oleate, an increase in the delay time results in an increase in the adsorption wave IIa and a decrease in the adsorption profile of wave IIIa.These observations can be explained in terms of the adsorbed HgIIDDP species displacing the adsorbed CuDDP complex from the surface of the mercury electrode.606 ANALYST, MAY 1987, VOL. 112 I Ila I Ila / / I I 1 -0.3 -0.4 -0.5 -0.6 Potentia IN heptane) in the presence of zinc dhsopropyldithiophosphate at t P e Fig. 14. Linear sweep voltammo~rams of copper(I1) oleate HMDE (area 0.020 cm2) at different delay times. Upper traces: [Cu" oleate] = 0.14 a, [ZDDP] = 0.44 m, delay times of 14 and 16 s,, res ectively, for A and B.Lower traces: [Cu" oleate] = 0.42 mM, Ii&DP] = 0.33 mM, delay times of 10, 31, 60 and 92 s, respectively, or C, D, E and F. Scan rate: 100 mV s-1 An increase in copper(I1) oleate concentration to 0.42 mM yields the unadsorbed CuDDP reduction wave I11 (Fig. 14). Both adsorption waves IIa (Ep==-0.35 V) and IIIa (Ep==-0.53 V) occur at more negative potentials than the waves attributed to the reduction of their bulk species, I1 (EpO.25 V) and I11 (Ep*0.48 V). This is indicative of the reactant, and not the product, being involved in the adsorp- tion electrode reaction in both instances.% The peak potential separation between the adsorbed wave IIIa and the main unadsorbed profile (wave 111) for the CuDDP complex is ca.65 mV. For the studies involving ZDDP (alkyl = isobutyl) and copper(I1) oleate dissolved in ASTM solvent, this difference is only ca. 35 mV [Fig. ll(A-C)]. The peak separation is related to the free energy of adsorption of the complex, that is, the greater the free energy of adsorption, the stronger the adsorption of the reactant at the mercury surface. More energy is required in the form of electrical energy to effect reduction, resulting in a peak potential separation that increases with increasing extent of adsorption.= It therefore follows from this study that the adsorption of CuDDP (alkyl = isopropyl) is stronger than that of CuDDP (alkyl = isobutyl) at the mercury electrode. Copper(II) Perchlorate - ZDDP (Alkyl = Isopropyl) This interaction is predictably similar to that of copper(I1) oleate and ZDDP (alkyl = isopropyl) with the same type of differential-pulse polarographic and linear sweep voltam- metric waves being evident in both instances (Figs.15 and 16). Notably, it is again apparent that the use of ZDDP (alkyl = isopropyl) as a complexing agent results in adsorbed species contributing significantly to the electrode processes taking place in solution. I I 1 1 I 0 -0.2 -0.4 -0.6 -0.8 -1.0 PotentialN Fig. 15. Differential-pulse polarograms of 0.79 m M zinc diisopropyl- dithiophosphate A) and 0.14 mM copper(I1) erchlorate (in ethanol) in the resence o z 0.79 mM zinc diisopropylditLophosphate (B) at the S M D ~ Scan rate: 10 mV s-1 -0.2 -0.4 -0.6 -0.8 PotentialN Fig.16. Linear sweep voltammograms of 0.14 mM copper(I1) perchlorate (in ethanol) in the resence of 0.79 mM zinc disopropyl- dithio hosphate at the HMDE farea = 0.020 cm2) at scan rates of 20, 50 anal00 mV s-1, respectively, for A, B and C Copper(II) Perchlorate - m D P (Alkyl = Isopropyl) The addition of copper(I1) perchlorate (0.1 mM) to NH4DDP (alkyl = isopropyl) (0.42 mM) results in the formation of a CuDDP complex and this yields a differential-pulse polaro- graphic wave I11 in the range previously encountered for the formation of CuDDP involving ZDDP, namely, between -0.45 and -0.50 V (Fig. 17). Again, adsorption post-peaks manifest themselves when the stationary electrode voltam- metric behaviour of the interaction is investigated (Fig. 18). Oxidation State of Copper With regard to the oxidation state of copper following the various interactions and undergoing reduction, the electron spin resonance signals for the in situ formation of CuDDP (Fig.19) closely resembles that reported by Shopov and Yordanov31 who studied a range of CuIIDDPs. The formation of a CuIDDP complex would be expected to produce a "splitting" of the polarographic waves reminiscent of the polarographic studies32933 where stabilisation of theANALYST, MAY 1987, VOL. 112 607 1 I I I I I 0.2 0 -0.2 -0.4 -0.6 -0.8 -1 PotentialN Fig. 17. Differential-pulse plarogram of 0.10 mM copper(I1) perchlorate (in ethanol) in t e presence of 0.42 mM ammonium diisopropyldithiophosphate at the SMDE. Scan rate: 10 mV s-l llla C A 1 -0.1 -0.3 -0.5 -0.7 -0.9 PotentialN Fig.18. Linear swee voltammograms of 0.1 mM copper(I1) perchlorate (in ethanol7 in the presence of 0.42 mM ammonium diisopropyldithiophos hate at the HMDE (area 0.020 cm2) at delay times of 30,180 and If00 s, respectively, for A, B and C. Scan rate: 50 mV s-1 copper(1) oxidation state occurs. No corresponding behaviour has been encountered in the CuDDP interactions reported here, suggesting that under the conditions imposed the existence of CuIDDP in quantifiable amounts is unlikely. Conclusion Electrode Processes and Wave Assignments The voltammetry of CuIIDDP complexes formed in situ can be represented by the following scheme: Cu2+ s CullDDPp -t 2e CuO + 2DDP- (wave Ill) (wave II) (wave I) 2Hgo + GDDP- e 2Hgl1DDP3- + 4e- 2Hgl1DDP3- + HgO e 3HgllDDP2 + 2e- In addition to the above three waves, there is a further reduction process when ZDDP is present, due to the reduction of zinc to zinc metal (wave IV).Adsorption waves for CuIIDDP and HgIIDDP are also observed when either ZDDP (alkyl = isopropyl) or NI&DDP (alkyl = isopropyl) are present and these waves have been designated IIIa and IIa, respectively. The CuDDP (alkyl = isobutyl) complexes do not exhibit adsorption characteristics except when ASTM solvent is used. H- Fig. 19. Electron spin resonance spectrum of 0.98 mM copper(I1) perchlorate (in ethanol) in the presence of 1.2 m~ zinc diisopropyl- dithiophosphate Table 5. Differential-pulse polarographic peak potentials for CuIIDDP complex formed in situ Wave assignment I I1 IIa I11 IIIa IV Efl .. +0.16 -0.14 -0.30 -0.45 -0.50 -1.00 The observed differential-pulse peak potentials (Ep) for the interactions discussed are summarised in Table 5 for con- venience. Stability Constant Determination The data contained in Table 3 and Subrahmanya’sM equation enable the stability constant of CuIIDDP to be determined This approach yields a value of 17.3 for log(stabi1ity constant), which agrees with the literature35 value of 17.4. However, Subrahmanya’s equation is rigorously applicable only when (W1a remains constant for the irreversible reduction of the complexed metal ion. A decrease in ana with increasing concentration of the CuDDP complex has been encountered in these studies; hence, the value of 17.3 should be seen in this light. The Science and Engineering Research Council is thanked for a research studentship (to M.J. H.) within its CASE scheme in association with Esso Chemicals, Abingdon, Oxon. Dr. Lynne Griffiths, Dr. T. Colclough and Dr. J. Marsh of Esso Chemicals are thanked for their enthusiastic encouragement and many helpful suggestions. Also, Professor J. E. Simao of the Univeridade do Minho, Portugal, is thanked for discus- sions on voltammetric assignments made possible by NATO Grant No. 84/0069. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Ford, J. F., J. Znst. Petrol., London, 1968,54,535. Kennedy, G . W., and Patterson, W. L., Znd. Eng. Chem., 1956,48,1917. Burn, A. J., Tetrahedron, 1966,22,2153. Bum, A. J . , in Mayor, F. R., Editor, “Oxidation of Organic Compounds,” Volume I, American Chemical Society, Wash- ington, DC, 1968, p.323. Colclough, T., and Cunneen, J. I., J. Chem. Soc., 1964,4799. Bridgewater, A. J., Dever, J. R., and Sexton, M. D., J. Chem. SOC., Perkin Trans. 2, 1980, 1006. Kendall, P. F., and Rimmer, A., Chem. Znd. (London), 1962, 43, 1864. Lewkowitsch, P. R. E., Chem. Znd. (London), 1962,27,1214. Jenkins, G . I., and Humphreys, C. M. A., J. Znst. Petrol., London, 1965,51,493.608 ANALYST, MAY 1987, VOL. 112 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Fodor, G. E., and Newman, F. M., ASLE Trans., 1977, 22, 389. Perry, S. G., J. Gas Chromatogr., 1964,2,93. Legate, C . E., and Burnham, H. D., Anal, Chem., 1960, 32, 1042. Butlin, A. G., and Lynes, A., in Hodges, D. R., Editor, “Recent Analytical Developments in the Petroleum Industry,” Applied Science, Barking, 1974, p. 283. Coates, J. P., J. Inst. Petrol., London, 1971,57,209. Killer, F. C. A., and Amos, R., J. Inst. Petrol., London, 1966, 52, 515. Brook, A. J. W., Davies, J. E., and King, B. M. J., in Hodges, D. R., Editor, “Recent Analytical Developments in the Petroleum Industry,” Applied Science, Barking, 1974, p. 97. Lamotte, A., and Auvray, J., J. Chromatogr., 1974,97, 213. Leighton, D., Moody, G. J., and Thomas, J. D. R., Analyst, 1974,99,442. Jamson, B., and Hillman, D. E., J. Chromatogr., 1978, 150, 499. Paaza, S., Analyst, 1984, 109, 1313. Shafiqul Alam, A. M., Martin, J. M., and Kapsa, Ph., Anal. Chim. Acta, 1979, 107, 391. Sexton, M. D., J. Chem. SOC., Perkin Trans. 2, 1984, 1771. Yordanov, N. D., Alexiev, V., Macicek, J., Glowiak, T., and Russell, D. R., Transition Met. Chem., 1983, 8, 257. Bond, A. M., Casey, A. T., and Thackeray, J. R., J. Electrochem. SOC., 1973, 120, 1502. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. Rudzinski, W., and Fernando, Q., Anal. Chem., 1978,50,472. Bard, A. J., and Faulkner, L. R., “Electrochemical Methods,’’ Wiley, New York, 1980, p. 219. Nicholson, R. S., and Shain, I., Anal. Chem., 1964,36, 706. Wopschall, R., and Shain, I., Anal. Chem., 1967,39, 1515. Webber, A., Shah, M., and Osteryoung, J., Anal. Chim. Acta, 1984, 154, 105. Bond, A. M., “Modern Polarographic Methods in Analytical Chemistry,” Marcel Dekker, New York, 1980, p. 193. Shopov, D., and Yordanov, N. D., Inorg. Chem., 1970, 9, 1943. von Stackelberg, M., and von Freyhold, H., 2. Electrochem., 1940,46, 120. Clark, G. C. F., Moody, G. J., and Thomas, J. D. R., Anal. Chim. Acta, 1978, 98, 215. Subrahmanya, R. S., “Advances in Polarography,” Pergamon Press, Oxford, 1960, p. 674. Toropova, V. F., Cherkasov, R. A., Savel’eva, N. I., Gorsh- kova, V. N., and Pudovik, A. N., Zh. Obshch. Khim., 1971, 41, 1469. Paper A61400 Received October 22nd, 1986 Accepted November 26th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9871200601

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ANALYST, MAY 1987, VOL. 112 609 Voltammetry of Copper(1) 0,O-Di( I -methylethyl)phosphorodithioate Miles J. Hutchings, G. J. Moody and J. D. R. Thomas* Department of Applied Chemistry, Redwood Building, UWIST, P.O. Box 13, Cardiff CFl3XF, UK The reduction of copper(1) - diisopropyldithiophosphate [dialkyldithiophosphate (DDP), alkyl = isopropyl] in an ethanolic medium gives rise to well defined polarographic waves. In the differential mode the polarographic wave divides when the concentration of CulDDP (alkyl = isopropyl) is between 0.3 and 0.4 mM. Limiting current data obtained in the d.c. polarographic mode suggest that the oxidation state of the copper species undergoing reduction is common before and after the division occurs and ESR studies confirm that it is a copper(l1) species that is present in solution.Stationary electrode voltammetric observations show that the electrode process occurring is electrochemically irreversible and diffusion controlled. In addition, at long delay times (>60 s), an adsorbed species manifests itself and this can be attributed to the adsorption of a mercury(l1) - DDP complex formed from the interaction of mercury metal with the dialkyldithiophosphate ligand. From an analytical standpoint, the differential polarographic currents are linearly related to the CulDDP (alkyl = isopropyl) concentration between 0.07 and 1.0 mM with an inflection to higher slope occurring at 0.4 mM. Keywords : Voltam me try; diisop rop yldith iop h ospha te; copper(1) dialkyldithiop hospha tes; 0,O ' -dialk yl- phosphorodithioates The use of metal 0, 0'-dialkylphosphorodithioates, com- monly known as dialkyldithiophosphates (MDDPs) , as antioxidants and the reasons for adopting an investigative voltammetric approach have been discussed in a project' detailing the in situ formation of copper dialkyldithiophos- phate complexes. The aim of this paper is to present and discuss voltammetric observations for the direct analysis of copper 0, 0'-di( 1-methylethyl)phosphorodithioate, that is, copper(I) diisopropyldithiophosphate [ CuIDDP, alkyl = iso- propyl] in an ethanolic medium.Experimental Reagents and Solutions Solutions of copper( I) di( 1 -me thyle thy1)dithiophosphate [CuIDDP, alkyl = isopropyl] (a gift from Esso Chemicals, Abingdon) were prepared using absolute ethanol (James Bunough analytical-reagent grade) as were supporting elec- trolyte solutions of either lithium nitrate (Fluka, purum) or sodium perchlorate (BDH Chemicals, AnalaR grade).The PAR 174A polarographic and associated instrumenta- tion and procedures used are as described previously1 with a drop time of 0.5 s and a modulation amplitude of 50 mV. Results and Discussion Differential-pulse polarography , d.c. polarography and cyclic voltammetry were used to investigate CuIDDP (alkyl = isopropyl) in solution at the mercury electrode. From an electrochemical and speciation perspective, it is the reduction of copper in the complex to copper metal that is of interest, and it is this process that is featured in the following discussion. The electrochemistry of the DDP ligand is briefly mentioned but this has been discussed more extensively elsewhere1 in studies involving interactions between DDPs and copper salts.Differential-pulse Polarography Fig. 1 shows differential-pulse polarograms of CuIDDP (alkyl = isopropyl) in a supporting electrolyte medium of 0.1 M * To whom correspondence should be addressed. -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8 -0.9 Potentia IN Fig. 1. Differential-pulse polarograms of copper(1) diisopropyldi- thiophosphate at the static mercu drop electrode (SMDE), showing effect of concentration. [CuIDDP = 0.054,0.11,0.38,0.54 and 0.65 mM, respectively, for A, B, C, 'is and E. Scan rate, 10 mV s-1; supporting electrolyte, 0.1 M lithium nitrate in ethanol Table 1. Differential-pulse polarographic data for CuIDDP (alkyl = isopropyl) in ethanol at the SMDE Concentration of CuIDDPImM -E,*N I,t/pA -E,tN I,tlpA 0.054 0.512 0.175 0.110 0.526 0.385 0.160 0.540 0.635 0.220 0.555 0.780 - 0.270 0.565 0.950 - 0.380 0.585 1.130 0.525 1.110 0.430 0.575 1.410 0.545 1.430 0.540 - S$ 0.560 1.930 0.650 0.575 2.400 0.750 0.592 3.130 0.860 0.599 3.680 0.970 0.608 4.550 SS s'c * Profiles observed at low concentrations (original).t Profiles observed at high concentrations (new). $ S = shoulder. WJmV 92 96 106 - - - - - 108 108 105 104610 ANALYST, MAY 1987, VOL. 112 "I .I 0 1 2 3 4 5 6 7 8 9 10 [Cu'DDP1/10-4 M Fig. 2. Plot of differential-pulse polaro raphic peak current versus concentration of copper(1) disopropylditfiophosphate. Scan rate, 10 mV s-l; supfmrting electrolyte, 0.1 M lithium nitrate in ethanol.0, Before splittmg; W, after splitting 0.5 CCA I A -0.2 -0.4 -0.6 -0.8 PotentialN Fig. 3. D.c. polarograms of copper(1) diisopropyldithiophosphate of concentrations 0.05,0.11,0.16 and 0.22 mM, respectively, for A, B, C and D. Scan rate, 10 mV s-1; supporting electrolyte, 0.1 M sodium perchlorate in ethanol lithium nitrate dissolved in absolute ethanol. At low concen- trations only a single reduction process is observed. However, on increasing the concentration of the complex, a second reduction process becomes evident and the original reduction wave disappears. These results are tabulated in Table 1 and a graph of peak height (Ip) versus concentration is shown in Fig. 2. It can be seen from Fig. 2 that the polarographic waves give different concentration - current slopes with the change in the wave profile occurring when the CuIDDP (alkyl = isopropyl) concentration is between 0.3 and 0.4 mM.Where possible, values of peak widths at half-peak height (Wi) have been included in Table 1. D.c. Polarography The d.c. polarograms of the CuIDDP (alkyl = isopropyl) system show a characteristic increase in limiting currents (IL) with increasing concentration (C) of analyte (Fig. 3 and Table 2). Values of I L K are essentially constant (Table 2), suggest- ing that the electroactive species undergoing reduction exists in a single oxidation state throughout the concentration range investigated (0.05-0.97 mM). This range includes the region where a change in the differential-pulse polarographic profile was observed (0.3-0.4 mM) leading to the conclusion that copper possesses a common oxidation state before and after the change.The inference that a single oxidation state participates in the reduction process is supported by the observation that W* values are not significantly different Table 2. D.c. polarographic data for CuIDDP (alkyl = isopropyl) in ethanol at the SMDE [ CuIDDP]/ I4 - E*l/ ILICSIPA mM (= C) -EJV I=/@ mV aa* a,? lmmol-1 0.054 0.532 0.37 98 0.56 0.53 6.9 0.110 0.546 0.68 74 0.65 0.70 6.2 0.160 0.558 1.35 76 0.70 0.68 8.4 0.220 0.563 1.70 74 0.68 0.70 7.7 0.650 0.591 4.05 90 0.66 0.57 6.2 0.750 0.593 4.65 72 0.70 0.72 6.2 0.860 0.607 5.75 70 0.72 0.74 6.7 0.970 0.617 7.20 70 0.77 0.74 7.4 * om,, determined from log (Id - Z)/I VS. E data of Fig. 3. ? ma, determined from IE+ - E+I = 51.7/ana mV.4 ZL/C: f = 7.0; s = 0.81; CV = 11.6%. -0.3 -0.5 -0.7 -0.9 PotentialN Fig. 4. Linear sweep voltammograms of 0.11 mM copper(1) diiso- propyldithiophosphate at the HMDE (area 0.020 cm2) at scan rates of 50, 100 and 200 mV s-1 for A, B and C, respectively. Supporting electrolyte: 0.1 M sodium perchlorate in ethanol before and after the change (Table l), as Wi depends on the oxidation state of the electroactive species.2 Possible explana- tions for the difference in peak current - concentration profiles at low and high concentrations of CuIDDP (alkyl = isopropyl) (Fig. 2) are a change in electrochemical reversi- bility of the reaction in the differential mode, or differing diffusion coefficients of the electroactive species at the low and high concentrations.Stationary Electrode Voltammetry Typical linear sweep voltammograms for solutions of CuI- DDP (alkyl = isopropyl) undergoing reduction at the hanging mercury drop electrode (HMDE) are shown in Fig. 4. A shift of the peak potential to more negative values with increasing scan rate is indicative of an electrochemically irreversible electrode process3 (Table 3). The irreversibility of the process is also borne out in cyclic voltammetric observations where values for the difference between anodic and cathodic peak potentials (Epa - Epc) are significantly larger than those expected for a reversible process4 (Fig. 5 and Table 4). Also, ratios of anodic to cathodic peak currents (Zpa/Zpc) differ from unity, again suggesting that the electrode process being investigated is irreversible .3 Values of IJVt are essen- tially constant at concentrations of 0.11 and 0.55 mM CuI- DDP (alkyl = isopropyl) (Table 3), suggesting that the reduction processes are diffusion controlled.In contrast, large variations in the values of ZJV preclude adsorption processesANALYST, MAY 1987, VOL. 112 611 ~~~ ~ Table 3. Linear sweep voltammetric data for 0.11 and 0.55 mM CuIDDP (alkyl = isopropyl) at the SMDE [CuIDDP]/ Scan rate/ mM mVs-l(=V) 0.11 50 100 200 500 0.55 50 100 200 500 IE, -Epnll - E P Idpi mV M a * 0.603 0.50 70 0.68 0.616 0.77 68 0.70 0.630 1.09 65 0.73 0.655 1.45 67 0.71 0.622 1.40 50 0.95 0.652 2.00 65 0.73 0.665 2.65 66 0.72 0.690 3.05 72 0.66 IdW pA v-1 St 2.24 2.44 2.44 2.05 6.26 6.32 5.93 4.31 IdVl pl v-1 s 10.00 7.70 5.45 2.90 28.00 20.00 13.25 6.10 * ana, determined from IE, - Ep,l = 47.7/ana.t ZdV (0.11 mM): f = 2.29; s = 0.187; CV = 8.16%. ZdVt (0.55 IIIM): f = 5.71; s = 0.946; CV = 16.60%. Table 4. Cyclic voltammetric data for 0.11 and 0.55 mM CuIDDP (alkyl = isopropyl) at the SMDE [CuIDPP]/ Scan rate/ mM mVs-l(=V) 0.11 50 100 200 500* 0.55 50 100 200 500 -EpaN 0.520 0.505 0.470 0.605 0.585 0.515 0.532 - * Anodic data difficult to interpret owing to broad profile. - EpcN ZpJCLA 0.603 0.22 0.616 0.32 0.630 0.44 0.622 0.60 0.652 0.90 0.665 1.10 0.690 1.45 0.655 - ZPJ@ 0.50 0.77 1.09 1.45 1.40 2.00 2.65 3.05 IEpa - EpcV IpalZpc mV 0.44 83 0.41 111 0.40 160 0.43 17 0.45 67 0.42 150 0.47 158 - - t 2 w 3 V I , -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8 PotentialN Fig.5. Cyclic voltammograms at the HMDE (area = 0.020 cm2) of (a) 0.11 mM copper(1) diisopropyldithiophosphate; (b) 0.11 mM copper(1) diisopropyldithiophosphate at delay times of 8,60,167 and 227 s for A, B, C and D, respectively; and (c) 0.04 mM ammonium diisopropyldithiophosphate. Scan rates, 100 mV s-1; supporting electrolyte, 0.1 M sodium perchlorate in ethanol as major contributors to the electrochemistry of CuIDDP (alkyl = isopropyl) at the HMDE. The cyclic voltammogram shows an interesting develop- ment in the form of a pre-peak when the delay time is increased [Fig. 5(b), A]. (The delay time is the time interval between the formation of the mercury drop and the commen- cement of the voltammetric scan.) For comparison, the cyclic voltammogram of ammonium DDP (alkyl = isopropyl) is Table 5.Determination of n from linear sweep voltammetric data at the SMDE [CuIDPP]/ Scan rate/ mM mVs-l(=V) 0.11 50 100 200 500 0.55 50 100 200 500 m a 0.68 0.70 0.73 0.71 0.95 0.73 0.72 0.66 Id@ 0.50 0.77 1.09 1.45 1 .# 2.00 2.65 3.05 *n determined from Z, = 2.99 x 10% ana+ACoDoW. n* 1.91 2.05 2.01 1.70 0.90 1.04 1.02 0.75 shown in Fig. 5(c). This time-dependent pre-wave may be attributed to the formation of an adsorbed mercury(I1) - DDP lspecies as characterised by Bond et aZ.5 Also, Hutchings et aZ. 1 have discussed a pre-wave of this type encountered in studies of CuDDP where the interaction of mercury metal with DDP produced a pre-wave. Determination of the Number of Electrons (n) Involved in the Reduction of Cu'DDP (Alkyl = Isopropyl) in Solution The reduction of CuDDP (alkyl = isopropyl) at the mercury electrode can be represented as follows: Cu(DDP), + ne + Hg Cu(Hg) 3- nDDP- (1) Voltammetry Application of electrochemical theory for the determination of n from an analysis of the polarographic - voltammetric waveform is based on a knowledge of diffusion coefficients for the participating electroactive species.The average value of IJC from d.c. polarographic data is 7.0 x 10-3 pA 1 mol-1 (Table 2). Applying a modified form of the Cottrell equation,6 which takes into account electrode sphericity, enables diffu- sion coefficients to be evaluated with the premise that either a one- or two-electron change is taking place. The respective diffusion coefficients are 3.7 x 10-5 and 1.1 X 10-5 cm2 s-1.612 ANALYST, MAY 1987, VOL.112 Table 6. Coulometric results obtained with copper(I1) perchlorate, CuIDDP and ZDDP (alkyl = isopropyl). Background current = 200 FA Mass taken/ Oxidation - reduction Expected Observed Final Copper(I1) perchlorate . . 1.0022 + 0.2/- 0.3 2 3.0426 4.6729 300 CuIDDP . . . . . . . . 0.3415 -0.8* 1 0.5185 1.0820 976 ZDDP . . . . . . . . 0,3655 +0.1/- 1.1 2 1.0787 3.8470 2476 Standard mg potentiaW n charge/C charge/C current/pA 2 1.0370 * Reduction only. I 1 H- Fig. 6. Electron spin resonance spectrum of copper(1) diisopropyldi- thiophosphate in ethanol after de-gassing An estimate of the diffusion coefficient (Do) by application of the Stokes - Einstein relation is possible’: at 25 “C, where q is the viscosity of the solvent, d is the density of the pure substance and M is its relative molecular mass.Using the Stokes - Einstein approach yields a value of 4.27 x 10-6 cm2 s-1 for Do. On comparing this with the values calculated from the limiting currents, it is clear that the experimental value of 1.1 x 10-5 cm2 s-1 is large for a two-electron change, and the even larger value of 3.7 x 10-5 cm2 s-1 for n = 1 is unlikely to describe the reduction observed here for what started out as a copper(1) species. Earlier discussions have indicated that the electrode process occurring at the HMDE is irreversible and diffusion con- trolled. By applying the relevant equation for such a process8 and using the value of 1.1 x 10-5 cm2 s-1 as the diffusion coefficient enables a value for n to be determined under linear sweep voltammetric conditions.Table 5 lists the calculated n values at 0.11 and 0.55 mM of CuIDDP (alkyl = isopropyl) in solution. It can be seen from these results that the experimen- tally determined diffusion coefficient adequately describes a two-electron process at a concentration of 0.11 mM, but it does not account for the relatively low peak currents, and hence small n values, obtained at 0.55 mM of CuIDDP (alkyl = isopropyl). At this higher concentration a value of ca. 3 X 10-6 cm2 s-1 for the diffusion coefficient would be required to describe a two-electron reduction process. Controlled Potential Coulometry (CPC) It is clear from the above discussion that a technique that does not rely on the determination of diffusion data is desirable for providing information that is unambiguous in respect to the number of electrons involved in a particular process.CPC is a useful method for studying electrode reactions and for determining the n value of a process without prior knowledge of electrode area of diffusion coefficient. In all the experiments conducted in ethanolic media, the final current was greater than the background current attributable to the supporting electrolyte alone (Table 6). In each instance, this had the effect of indicating a greater charge than that expected for complete electrolysis. Hence, a credible value for n could not be obtained. This type of behaviour is encountered when a reaction of the electrolysis product regenerates starting material or another electroactive substance.However, with copper(I1) perchlorate as a primary standard it is difficult to ascribe the observations to the above explanation. Blanke- spoor9 noticed this type of behaviour when looking at the oxidation of zinc (DDP) (alkyl = isopropyl) in a non-aqueous medium. Electron Spin Resonance (ESR) Although elemental analyses and magnetic moment studies irrefutably show that CuIDDP (alkyl = isopropyl) is present in the solid state, dissolution in ethanol produces a strong copper(I1) ESR signal indicating transformation to CuIIDDP (alkyl = isopropyl) (Fig. 6). This is accordance with observa- tions of Yordanov et aZ.,lo who stated that CuIDDP (alkyl = isopropyl) produces CuIIDDP (alkyl = isopropyl) in solution via an inner self-redox reaction. Conclusion The voltammetric behaviour of copper(1) diisopropyldithio- phosphate in an ethanolic medium is complex.Reduction processes observed at high and low concentrations of the complex are diffusion controlled and involve a common oxidation n u m b for the copper species undergoing reduction at a mercury electrode, but the reduction process is compli- cated by adsorption of an electroreducible species on to the surface of the mercury drop. Nevertheless, graphs of differen- tial-pulse polarographic currents versus concentration of the CuIDDP (alkyl = isopropyl) are linear from 0.07 to 1 mM, with an inflection to higher slope occurring at 0.4 mM. However, with regard to determination in specific systems, it can be envisaged that interactions with other components will call for preliminary treament to assure analytical usefulness of vol- tammetry for this type of compound.The Science and Engineering Research Council is thanked for a research studentship (to M. J. H.) within its CASE scheme in association with Esso Chemicals, Abingdon, Oxon. Dr. Lynne Griffiths, Dr. T. Colclough and Dr. J. Marsh of Esso Chemicals are thanked for their inspiring encouragement and many helpful suggestions. Also, Professor J. E. Simao, Universidade do Minho, Portugal is thanked for discussions made possible by NATO Grant No. 84/0069. References 1. Hutchings, M. J., Moody, G. J., and Thomas, J. D. R., Analyst, 1987, 112, 601. 2. Baird, A. J., and Faulkner, L. R., “Electrochemical Methods,” Wiley, New York, 1980, p. 195. 3. Baird, A. J., and Faulkner, L. R., “Electrochemical Methods,” Wiley, New York, 1980, p. 219.ANALYST, MAY 1987, VOL. 112 613 4. Matsuda, H., and Ayabe, Y., 2. Elektrochem., 1955,59,494. 5 . Bond, A. M., Casey, A. T., and Thackeray, J. R., 10. Yordanov, N. D., Alexiev, V., Macicek, J., Glowiak, T., and J . Electrochem. SOC., 1973,120, 1502. 6. Baird, A. J., and Faulkner, L. R., “Electrochemical Methods,” Wiley, New York, 1980, p. 145. 7. Meites, L., “Polarographic Techniques,” Wiley, New York, Paper A61399 1 9 6 5 , ~ . 144. Received October 22nd, 1986 8. Nicholson, R. S., and Shain, I., Anal. Chem., 1964, 36, 706. Accepted November 26th, 1986 9. Blankespoor, R. L., Znorg. Chem., 1985, 24, 1126. Russell, D. R., Transition Met. Chem., 1983, 8, 257.

ISSN:0003-2654    

DOI:10.1039/AN9871200609

出版商:RSC    

年代:1987

数据来源: RSC