摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRYADVISORY BOARD*Chairman: J. M. Ottaway {Glasgow, U.K.)'L. S. Bark (Salford, U.K.)E. Bishop (Exeter, U.K.)W. L. Budde (U.S.A.)D. T. Burns (Belfast, U.K.)L. R. P. Butler (South Africa)H. J. Cluley (Wembley, U.K.)E. A. M. F. Dahmen (The Netherlands)L. de Galan (The Netherlands)A. C. Docherty (Billingharn, U.K.',D. Dyrssen (Sweden)G. Ghersini (Italy)J . Hoste (Belgium)A. Hulanicki (Poland)'6. W. Kirby (Glasgow, U.K.)W. S . Lyon (U.S.A.)H. V. Malmstadt (U.S.A.)G. W. C. Milner (Harwell, U.K.)*G. J. Dickes (Bristol, U.K.)'A. C. Moffat (Aldermaston, U.K.)E. J. Newman (Poole, U.K.)H. W. Nurnberg (West Germany)*T. B. Pierce (Harwell, U.K.)E. Pungor (Hungary)f . H.Scholes (Middlesbrough. U.K.)D. Simpson (Thorpe-le-Soken, U.K.)"J. M. Skinner (Billingharn, U.K.1'J. D. R. Thomas (Cardiff, U.K.)"A. M. Ure (Aberdeen, U.K.)K. C. Thompson (Shetfield, U,K.)A. Walsh, K.B. (Australia)G. Werner (German Democratic Hepublic)T. S. West (Aberdeen, U.K.)"P. C. Weston (London, U.K.)"J, Whitehead (Stockton-on-Tees, U.K.)J . D. Winefordner (U.S.A.)P. Zuman {U.S.A.)'Members of the Board serving on the Analytical Editorial BoardEditor: P. C. WestonSenior Assistant Editor: R. A. YoungAssistant Editors: Mrs. J. Brew, Miss D. ChevinR EG I0 NAL ADVlSO RY ED IT0 RSDr. J . Aggett, Department of Chemistry, University of Aucklanrf, Private Bag, Auckland, NEW ZEALAND.Professor L. Gierst, Universiti! Libre de Bruxelles Facult6 des Sciences, Avenue F.-D.Hoosevelt 50,Professor H . M. N. H . Irving, Department of Theoreticai Chemistry, University of Cape Town, Ronde-Professor W . A. E. McBryde, Faculty of Science, University of Waterloo, Waterloo, Ontario, CANADA.Dr. 0. Qsibanjo, Department of Chemistry, Universitv of ibadan, ibadan, NIGERIA.Dr. G. Rossi, Chemistry Division, Spectroscopv Sector, CEC Joint Research Centre, EURATOM, ispraDr. I. Rubeska, Geological Survey of Czechoslovakia, Malostranski, 19, 118 21 Prague 1. CZECHO-Professor J. Ruzicka, Chemistry Department A, Technical University of Denmark, 2800 Lyngby,Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN.Professor L. E. Smythe, Department of Chemistry, University of New South Wales, P.O.Box 1,Kensington, N.S.W. 2033, AUSTRALIA.Professor P. C. Uden, Department of Chemistry, University of Massachusetts, Amherst, MA 04 003,U.S.A.Bruxelles, BELGIUM.bosch 7700, SOUTH AFRICA.Establishment, 21 020 lspra (Varese). ITALY,SLOVAK I A.DEN MARK.Editorial: Editor, The Analyst, The Royal Society of Chemistry, Burlington House,Piccadilly, London, WIV OBN. Telephone 01 -734 9864. Telex No. 268001Advertisements: Advertisement Department, The Royal Society of Chemistry, Burlington House,Piccadilly, London, WIV OBN. Telephone 01 -734 9864. Telex No. 268001The Analyst (ISSN 0003-2654) is published monthly by The Royal Society of Chemistry, BurlingtonHouse, London W I V OBN, England. All orders accompanied with payment should be sent directly toThe Royal Society of Chemistry, The Distribution Centre, Blackhorse Road, Letchworth, Herts. SG6 1 HN,England. 1983 Annual subscription rate U K f 93.50, Rest of World €99.00, USA $201 .OO. Purchased withAnalytical Abstracts U K f226.50, Rest of World f238.50, USA $487.00. Purchased with AnalyticalAbstracts plus Analytical Proceedings UK €251 .OO, Rest of World f265.00, USA $539.00. Purchasedwith Analytical Proceedings UK f l l 7 . 5 0 , Rest of World f 124.50, USA $253.00. Air freight and mailingin the USA by Publications Expediting lnc., 200 Meacham Avenue, Elmont, NY 11003.USA Postmaster: Send address changes to: The Analyst, Publications Expediting inc., 200 MeachamAvenue, Elmont, NY 11003. Second class postage paid at Jamaica, NY 11431. 411 .stherdespatches outside the UK by Bulk Airmail within Europe, Accelerated Surface Post outside Europe,PRINTED IN THE UK.Volume 108 No 1282 0 The Royal Society of Chemistry 1983 January 198

ISSN:0003-2654    

DOI:10.1039/AN98308FX001

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

ANALAO 108 (1 282) 1-1 36 (1 983) January 1983THE ANALYST117334353586471768’1859299106109116119122126THE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRYCONTENTSAn Automated Viscometer Based on High-precision Flow Injection Analysis. Part I.Apparatus for High-precision Flow Injection Analysis-D. Betteridge, W. C. Cheng,E. L. Dagless, P. David, T. B. Goad, D. R. Deans, D. A. Newton and T. B. PiercePart II,Measurement o f Viscosity and Diffusion Coefficients-D. Betteridge, W. C. Cheng.E. L. Dagless, P. David, T. B. Goad, D. R. Deans, D. A. Newton and T. B. PierceGas-chromatographic Procedure f o r the Determination o f Environmental Organo-chlorine Residues in Avian Tissues w i t h Confirmation o f Identities by ChemicalDerivatisation-Andrew J.Trim, Peter M. Brown, Peter J. Bunyan, Edward M. Odam andPeter I. StanleyDetermination o f Sulphide Produced by Desulfovibrio Species o f Sulphate-reducingBacteria-I. K. Al-Hitti, G. J. Moody and J. D. R. ThomasAir-segmented Continuous-flow Visible Spectrophotometric Determination o f Cephalo-sporins in Drug Formulations by Alkaline Degradation t o Hydrogen Sulphide andFormation o f Methylene Blue and Determination o f Sulphide-producing ImpuritiesIncluding Cephalosporins i n Penicillin Samples-Mohamed A. Abdalla, Arnold G. Fogg,John G. Baber and Christopher BurgessAtomic-absorption Determination of Mercury i n Geological Materials by Flame andCarbon-rod Atomisation After Solvent Extraction and Using Co-extracted Silveras a Matrix Modifier-Richard F.Sanzolone and T. T. ChaoNovel Static Cold Vapour Atomic-absorption Method for the Determination ofMercury-Ping-Kay Hon, Oi-Wah Lau and Man-Chaun WongDifferential Electrolytic Potentiometry. Part XXVI. Direct Polarisation in Acid -Base Titrimetry in Toluene - Methanol Mixtures-Abdalla M. S. Abdennabi andEdmund BishopVoltammetric Determination o f Hydrazine and Hydroxylamine-Francis X. Dias andBruno JaselskisEffects o f Filter-paper Adsorption and Desorption on the Spectrophotometric andRadiochemical Measurement o f Phosphate in Filtered Solutions-Jeffrey C. Hughesand William S. G. MacpheeApplication o f a Modified Catalytic Effect: Determination o f Nanogram Amounts o fZinc i n Milk Samples Using a Kinetic - Fluorimetric Method-A.Moreno, M. Silva,D. Perez Bendito and M. ValcarcelSpectrofluorimetric and Spectrophotometric Determination o f Aluminium with theSalicyloylhydrazones o f Pyridine-2-aldehyde and Pyridoxal-M. Gallego, M.Valcarcel and M. Garcia-VargasSpectrophotometric and Complexometric Determination o f Iron Using the Iron( Ill) -Arsenazo Ill System-F. Bosch Reig, J. Martinez Calatayud, M. C. Garcia Aivarez-Coqueand M. C. Pascual MartiAn Automated Viscometer Based on High-precision Flow Injection Analysis.REPORTS BY THE ANALYTICAL METHODS COMMITTEEDetermination of lpronidazole in Animal FeedsDetermination of Tin in Organic Matter by Atomic-absorption SpectrometrySHORT PAPERSDetermination o f Nitrovin i n Animal Feedingstuffs by High-performance LiquidChromatography-Michael J.Gliddon, Christopher Cordon and Geoffrey M. ParnhamSpectrophotometric Determination o f Sulphide and Sulphite with Cacotheline-Nemani Krishna Murty and Narasimhadevara Venkata Srinivasa RaoDetermination o f lnosine by Polarography-Yoshio Kato and Akira HatanoBOOK REVIEWSSummaries of Papers in this Issue-Pages iii, iv, v, vi, vir, viii, ix, xPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class a t New Yark. USA, Post OfficReading _.Second- - handCopies -RSC members have the advantage thatthey may subscribe to this journal a t amost attractive discount price.The convenience of having yourpersonal copy, rather than borrowinga library copy, is obvious but there aremany other advantages of member-ship.Details of membership will besent if you write 200 in one of theboxes on the Reader Enquiry Servicepage*BUREAU OF ANALYSEOSAMPLES LTDannounce the availability ofNEWCERTIFIED REFERENCEMATERIALSfor twelve trace elementsinNICKEL BASE ALLOYSBCS/SS 345 for normal trace levelsBCS 346/SS 346A for enhancedtrace levelsFor full details write, telephoneor telex to:BAS Ltd., Newham Hall, Newby,Middlesbrough, Cleveland, TS8 9EATelephone: Middlesbrough 31 721 6Telex: 587765 BASRIDA200 for further information. See page XIV A202 for further information. See page xivReprint of an Important Analyfiwl Chemistry ReviewIn an attempt to ensure that major developments in chemistry reach as wide anaudience as possible the RSC is to make available a reprint of an important reviewwhich was published in Chemical Society Reviews Vol 10, No 1, pp 11 3-1 58entitled :Modern Analytical Methods for EnvironmentalPolycyclic Aromatic Compoundsby K.D. Bartle, M. L. Lee, and S. A. WisePolycyciic aromatic compounds are major pollutants of the environment,originating from many sources.available for identification and analysis and provides the reader with acomprehensive and authoritative source of information on the subject. The paperis divided into the following sections:I NTRO D UCTl ON; SAM PLE PREPARATION; C H ROMATOG RAPH I C M ETH ODs;MASS SPECTROMETRY; SPECTROSCOPIC METHODSThis review, which contains more than 400 references, will be of interest toenvironmental, petroleum and analytical chemists.f2.50 ($5.00) by sending a stamped addressed envelope measuring 6" x 9"minimum, together with your remittance, to:This paper reviews the techniques that areA copy may be purchased forDr, A. Kabi, The Royal Society of Chemistry,The University, NOTTINGHAM, NG7 2RDEnglan

ISSN:0003-2654    

DOI:10.1039/AN98308BX003

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

January, 1983 SUMMARIES OF PAPERS I N THIS ISSUESpectrophotornetric and Complexometric Determination of IronUsing the Iron(II1) - Arsenazo I11 SystemA spectrophotometric study of iron(II1) - Arsenazo I11 aqueous solutions isdescribed. At various reactant concentrations Arsenazo I11 forms 2 : 1, 1 : 1and 1 : 2 complexes with iron(II1). Beer’s law is obeyed up to iron con-centrations of 0.7 p.p.ni. A method for the cornplexometric determinationof iron(II1) with EDTA, using Arsenazo I11 as a metallochromic indicator, isproposed. The method allows the determination of iron in the range 0.5-60 mg.Keywords : Iron (111) - A rsenazo I I 3 complexes ; spectrophotometric deter-mination ; covnplexometvic titrations ; A rsenazo 111 metallochromic indicatorF. BOSCH REIG, J.MARTINEZ CAL-ATAYUD, M. C. GARCIA ALVAREZ-COQUE and M. C. PASCUAL MART1Departamento de Quimica hnalitica, Facultad de Quimicas, Universidad de Valencia,Valencia, Spain.Analyst, 1983, 108, 99-105.Determination of Ipronidazole in Animal FeedsReport Prepared by the Medicinal Additives in Animal FeedsSub-committee (A)Keywords : Ipronidazole determination ; animal feeds ; gas - liquid chromato-P P h YANALYTICAL METHODS COMMITTEEThe Royal Society of Chemistry, Burlington House, Piccadilly, London, W1V OBN.Analyst, 1983, 108, 106-108.Determination of Tin in Organic Matter byAtomic - absorption SpectrometryReport Prepared by the Metallic Impurities in Organic MatterSub-committeeKeywords : Tin determination ; atomic-absorption spectrometvy ; canned foods\ 1xANALYTICAL METHODS COMMITTEEThe Royal Society of Chemistry, Burlington House, Piccadiily, London, WlV OBNAnalyst, 1983, 108, 109-1 15x SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Nitrovin in Animal Feedingstuffs by High-performance Liquid ChromatographyJanuary, 1983Short PaperKeywords : Nitvovin determination ; growth-Promoting drugs ; high-performanceliquid chromatographyMICHAEL J.GLIDDON, CHRISTOPHER CORDON and GEOFFREY M.PARNHAMDivisional Product Research and Development Laboratories, Cyanamid of GreatBritain Ltd., Fareham Road, Gosport, Hampshire, PO1 3 0AS.Analyst, 1983, 108, 116-1 19.Spectrophotometric Determination of Sulphide and Sulphitewith CacothelineShort PaperKeywords : Sulphide determination ; sulphite deternzination ; spectrophoto-metvy ; cacothelineNEMANI KRISHNA MURTY and NARASIMHADEVARA VENKATASRINIVASA RAODepartment of Engineering Chemistry, Andhra University, Waltair, India.Analyst, 1983, 108, 119-122.Determination of Inosine by PolarographyShort PaperKeywords : Inosine determination ; polarographyYOSHIO KATOGifu College of Pharmacy, Mitahora, Higashi, 6-1, Gifu, Japan 502.and AKIRA HATANOMitsui Petrochemical Industries, Waki-Cho, Kuga-Gun, Yamaguchi-Ken, Japan 740.Analyst, 1983, 108, 122-125Janaary , 1983 THE ANALYST \ xiAnalytical Sciences MonographsNo.4 ElectrothermalAtomisation for AtomicAbsorption Spectrometryby C. W. FullerSince the introduction of atomic absorptionspectrometry as an analytical technique, by Walsh,in 1953, the use of alternative atomization sourcesto the flame has been explored.At the present timethe two most successful alternatives appear to bethe electrothermal atomiser and the inductively-coupled plasma. In this book an attempt has beenmade to provide the author’s views on the historicaldevelopment, commercial design features, theory,practical considerations, analytical parameters of theelements, and areas of application of the first ofthese two techniques, electrothermal atomisation.Hardcover 135pp 0 851 86 777 4f18.00 ($34.00) RSC Members f13.50No. 5 Dithizoneby H. M. N. H. IrvingThe author of this monograph, who has beenclosely associated with the development ofanalytical techniques using this reagent for manyyears, and who has made extensive investigationsinto the properties of its complexes, has gatheredtogether a body of historical and technical data thatwill be of interest to many practising analyticalchemists.Hardcover 11 2pp 0 851 86 787 1f12.50 ($24.00) RSC Members f9.50No.6 lsoenzyme AnalysisEdited by D. W. MossThis monograph attempts to draw together the mostimportant experimental techniques which haveresulted from the modern recognition that enzymesfrequently exist in multiple molecular forms. Thismonograph also indicates the advantages andlimitations in isoenzyme studies of these modernexperiments.Brief Contents:Multiple Forms of Enzymes; Separation of MultipleForms of Enzymes; Selective Inactivation of MultipleForms of Enzymes; lmmunochemistry of MultipleForms of Enzymes; Catalytic Differences betweenMultiple Forms of Enzymes, Methods of ObtainingStructural Information, Selection of Methods ofAn a l ysis.Hardcover 171 pp 0 85186 800 2fl2.00 ($23.00) RSC Members f9.00No.7 Analysis of AirbornePollutants in WorkingAtmospheresThe Welding and SurfaceCoatings Industriesby J. Moreton and N. A. R. FallaThis Monograph covers the following:Part I The Welding Industry: Airborne Pollutantsin Welding; Sampling of Welding WorkshopAtmospheres; Analysis of Welding Fumes andPollutant Gases.Part II The Surface Coatings Industry: Origin ofAirborne Pollutants in the Surface CoatingsIndustry; Collection and Analysis of GaseousAtmospheric Pollutants in the Surface CoatingsIndustry; Collection and Analysis of ParticulateAtmospheric Pollutants in the Surface CoatingsIndustry; Future Trends Relating to Sampling andAnalysis in the Welding and Surface CoatingsIndustries.Hardcover 192pp 0 851 86 860 6f15.00 ($29.00) RSC Members f12.00No.8 The Sampling of BulkMaterialsby R. Smith and G. V. JamesThe literature of analytical chemistryexhaustively covers the many techniques nowavailable to the analyst.feature common t o all analyses, is in contrastonly sparsely documented. Comparatively feworiginal papers on this subject have beenpublished in the last fifty years; there are veryfew reviews available, and perhaps as a resultsampling is badly neglected in most instructionalcourses in analytical chemistry.ThisMonograph will go some way towards filling agap in the literature and should stimulateinterest in the development of sampling as afield of study.B r i e f ContentsIntroduction; Glossary of Terms; Establishment of aSampling Scheme, Sampling Theories; Apparatusfor Sampling; Sampling Methods; Appendices 1-4.Hardcover 200pp 0 851 86 81 0 Xf16.50 ($32.00) RSC Members f10.75Orders: .RSC Members should send their orders to:The Membership Officer, The Royal Society of Chemistry30 Russell Square, London WC1 B 5DTAll other orders should be sent to:The Royal Society of Chemistry, Distribution Centre,Blackhorse Road, Letchworth, Herts. SG6 1 H NSampling, the oneThe Royal Society ofChemistrxii THE ANALYST Janwary, 1983The Royal Society of ChemistrySpecialist Periodical ReportsEnvironmentalChemistry Vol, 2Senior Reporter: H.J. M. BowenThe first volume of this series was published in1975 and emphasized environmental organicchemistry whereas this second volume is deliber-ately slanted towards inorganic chemicals,covering the broad fields of the atmosphere andthe hydrosphere, soils, and human diets.Reviewers of all these subjects agree that far toolittle information is available on the chemicalforms of the elements in environmental reservoirs,thus laying down a challenge to analyticalchemists. A broad review of mycotoxins is how-ever included partly to redress the balance ofinorganic topics.Brief Contents :Inorganic Particulate Matter in the Atmos-phere :Methods of Sampling and Analysis; GeneralPhysical and Chemical Composition of Particu-lates; Characteristics of Emissions from SpecificSources; Atmospheric Transport and Dispersionof Particulates; Removal of Particulates from theAtmosphere; Effects of Airborne and DepositedParticulates; Future Research Needs and Con-clusions;The Elemental Content of Human Diets andExcreta :Outline of Ingestion, Absorption, Excretion;Methodological Problems, Inputs, Outputs,Deficient Concentrations, and Oral Toxicities ofthe Elements;The Elemental Constituents of Soils:The Alkali Metals: Lithium, Sodium, Potassium,Rubidium, and Caesium; The Alkaline EarthElements: Beryllium, Magnesium, Calcium,Strontium and Barium; Titanium, Zirconium, andHafnium; Vanadium, Niobium, and Tantalum;The Lanthanides or Rare Earth Elements, andYttrium and Scandium; Molybdenum andTungsten; Chromium, Manganese, Iron, Cobalt,and Nickel; Copper, Zinc, and Cadmium; TheNoble Metals; Mercury; Boron, Aluminium,Gallium, Indium, and Thallium; Carbon, Silicon,Germanium, Tin, and Lead; Nitrogen, Phos-phorus, and Sulphur; Hydrogen and Oxygen;The Halogens: F, CI, Br, and I; Arsenic, Selenium,Antimony, and Bismuth; Thorium and Uranium;Radionuclides, Organic Soils;Mycotoxins:Biogenesis of Mycotoxins; The Importance ofMycotoxins in the Environment; Analysis ofMycotoxins; Occurrence in Food and AnimalFeed; Metabolism and Mode of Action of Myco-toxins; Control of Mycotoxins in the Food Chain;Occurrence, Distribution, and ChemicalSpeciation of some Minor DissolvedConstituents in Ocean Waters :Individual Elements; Additional Aspects ofChemical Speciation;Hardcover 301 pp 0 851 86 765 0Price f33.00 ($63.00) RSC Members f19.00Still available:Volume 1.Hardcover 212pp 0 85186 755 3Price f15.50 ($30.00) RSC Members f7.50Special Package Price (Vols 1 €t 2)Non-RSC Members only f39.00 ($75.00)Miscellaneous PublicationsThe Periodic Table ofthe ElementsThe Royal Society of Chemistry has produced acolourful wall chart measuring 125cm x 75cmcovering the first 105 elements as they existtoday.Each group is pictured against the same tintedbackground and each element, where possiblephotographed in colour and discussed withregard to its position in the hierarchy of matter.Additional information for each element includeschemical symbol, atomic number, atomic weightand orbits of electrons.This chart is particularly useful for both teachersand students and would make a worthwhileaddition to any establishment.Price f2.20 ($4.00) RSC Members f1.00Teacher Members f4.60 for 10Prices for The Periodic Table subject to VAT in the UKRSC members should send their orders to: The Royal Society of Chemistry, The Membership Officer, 30 RussellSquare, London WCl B 5DT.Non-RSC members should send their orders to: The Royal Society of Chemistry,Distribution Centre, Blackhorse Road, Letchworth, Herts SG6 1 HN. ;p<:' VThe Royal Society of ChemistryBurlington HouseLondon W1V OBJanuary, 1983 THE ANALYST xiiiReprint of an Important Analytical Chemistry ReviewStandardised Thin-Layer ChromatographicSystems for the Identincation ofDrugs and Poisonsby A.H. Stead, R. Gill, T. Wright, J. P. Gibbs, A. C. MoffatThe October '82 issue of The Analyst featured a review entitled Standardised Thin-LayerChromatographic Systems for the Identification of Drugs and Poisons. The wide use of TLCfor the analysis of drugs and poisons in biological fluids and pharmaceutical preparationssuggests that this article will be of great interest to many analysts working in the field.Consequently, The Royal Society of Chemistry has decided to make separate reprints available.COVERAGEThis review gives criteria for good systems and applies them to the selection of the eight mosteffective. The selected systems are standardised by the use of standard running conditions andthe use of reference compounds.Rf X 100 values are given for 594 basic, 48 neutral and 152acidic drugs on the selected systems both in alphabetical order and ascending order of Ri foreach system to aid the identification of unknown drugs. Further identification is enhanced bythe inclusion of various locating procedures.Price f5.75 ($11.50)-including p b- p to all U.K. and European destinations, andSurface mail outside Europe. $13.00 for Airmail outside Europe.To order, send payment to Dr. A. Kabi at the address below:The Royal Society of ChemistryThe UniversityNottingham, NG7 2RD, EnglandNotice to SubscribersSubscriptions for The Analyst, Analytical Abstracts andAnalytical Proceedings should be sent to:The Royal Society of Chemistry, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HN, EnglandRates for 1983 (including indexes)u K1EireThe Analyst alone ... ...... ... f93.50The Analyst and Analytical Proceedings ... f 1 1 7.50The Analyst and Analytical Abstracts ... f226The Analyst, Analytical Proceedings andAnalytical Proceedings alone ... ... f44Analytical Abstracts ... ... ... f251Analytical Abstracts a Ion e ... ... ... f165Rest ofUSA World$201 f99$95 f46$253 f124.50$487 f238.50$539 f265$355 f17TUCK IN UNDER FLAP ATHE ANALYST January, I983READER ENQUIRY SERVICEFor further information about any of the products featured in the advertise- $ iments in this issue, please write the appropriate A number in one of the :boxes below. 0 :Postage paid if posted in the British lsles but overseas readers must affix & ia stamp.- n .(Please use BLOCK CAPITALS)NAME ......................................................................................................................................................................................... iOCCUPATION ................................................................................................................................................................. iADDRESS ............................................................................................................................................................................. iSECOND FOLDPostagewill bePaid byLicenseeDo not affix Postage Stamps if posted inGt. Britain, Channel Islands or N. Irelandi 0 :: c : I : -I; IBUSINESS REPLY SERVICELicence No. W.D. 106I IReader Enquiry ServiceThe AnalystThe Royal Society of Chemistry: m :: 0 :: 0 :: m iBurlington HousePiccadilly London W I E 6WFENGLANDTHIRD FOLD2 :3 :- n :y ip :0 ;CI A D

ISSN:0003-2654    

DOI:10.1039/AN98308BP007

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst January 1983 Vol. 108 $9. 17-32 17 An Automated Viscometer Based on High-precision Flow Injection Analysis Part 11." Measurement of Viscosity and Diffusion Coefficients D. Betteridge,f. W. C. Cheng E. L. Dagless,; P. David and T. B. Goad D. R. Deans University College Swansea Swansea S A 2 8PP Petrochemicals and Plastics Division Imperial Chemical Industries Ltd. Wilton Clevelaaad TS6 8 J E D. A. Newton and T. B. Pierce Instrumentation and Applied Physics Division A E R E Harwell Didcot Oxfordshire OX1 1 ORA It is shown that viscosities in the range of 0.2-190 CP can be determined on 6O-pl samples in an analysis time of about 1 min with reasonable accuracy and a precision of 0.2-0.30/;. The method is basecl on timing a sample injected into a flowing carrier stream over a fixed distance.The factors affecting the dispersion of the sample within the carrier stream have been critically investi-gated. It is shown that chemical interactions especially hydrogen bonding, between the sample slug and the carrier stream can crucially affect the dispersion pattern so that correct choice of carrier is essential for the success of the method. I t can also be used for the determination of molecular diffusion coefficients and it is possible that it could be adapted to cover a wider range of viscosities. The implications of the study to flow injection analysis are discussed. Keywords Viscosity ; viscometer ; diffusion coeficient ; computer-controlled analytical instrumentation ; flow injection analysis In Part I an apparatus was described which was capable of first injecting a slug of liquid of 10-60~1 into a tube through which a carrier liquid is flowing and then subsequently of measuring the time taken for it to pass a known distance along the tube and determining its pattern of dispersion.In this part the use of the apparatus as a viscometer and for the measurement of diffusion coefficients is described and evaluated. It is shown that it is possible to determine the viscosity of a range of samples rapidly and precisely. Care is needed in selecting the appropriate carrier liquid as for some carrier - sample combinations chemical interactions such as hydrogen bonding occur and influence both the flow-rate and the sample dispersion pattern. These findings are of some relevance to the practice of flow injection analysis where such carrier - sample interactions may be encountered and peak height measurements are commonly employed.Its value for the determination of diffusion coefficients is harder to demonstrate. Experimental The details of the apparatus are given in Part I. Apparatus Essentially it consists of a straight length of tubing 0.086 cm i.d. along which under a constant head of pressure the flowing sample is injected by a chromatography slide valve into a carrier stream. The timing of the passage of the sample slug and its dispersion are measured via two photometric or con-ductimetric detectors spaced along the flow tube after the point of sample injection. The temperatures of the carrier and sample are carefully controlled and monitored throughout the system.The apparatus is controlled by a microcomputer that also collects and processes * For Part I of this series see Analyst 1983 108 1. t Present address BP Research Centre Sunburv-on-Thames Middlesex TW16 7LN. f Present address Department of Electrical Engineering University of Bristol Senate House Bristol, BS8 1TH 18 BETTERIDGE et d. AUTOMATED VISCOMETER BASED ON Analyst VOl 108 the signals for the detectors. The values of the most important parameters adopted for most of the experiments described here are sample volume 60 pl; two detectors positioned at 20 and 60 cm respectively from the injection valve; and flow-rate as indicated in the legends to the figures. Reagents Before use they were de-gassed in an ultrasonic bath and thermostated at the temperature of the measure-ment.The viscosity of each solution was determined with an Ostwald viscometer just before use. After the measurement of the most viscous samples the solution was flushed out with acetone to prevent any sample carry-over. The organic carrier stream was usually cyclohexane and the aqueous one was usually a 1 M solution of sodium chloride. These were coloured with a dilute solution of a suitable indicator when photometric cells were employed. The precise experimental details are given in the text. Commercially available chemicals were used without further purification. Theory There are two different ways of determining the viscosity of a sample with the above One is to view it as a modified Ostwald viscometer in which the time taken for Allowance has to be made The other approach is to determine the dispersion of the The apparatus.the sample to pass over a fixed distance is a measure of viscosity. for the effect of the carrier stream. sample and use that as a measure of viscosity and diffusion coefficient of the sample. relevant theories are briefly reviewed. Modified Ostwald Viscometer Approach The Hagen - Poiseuille relationship112 gives the flow-rate I; for a liquid of viscosity q, flowing through a straight horizontal tube of length 1 and radius a under constant pressure, P and isothermal conditions such that where t is the time taken for the liquid to traverse the distance L. If the liquid is discharged into air there is a need to correct the pressure for the kinetic energy of the fluid.3 This has been taken to be negligible in this study because of the length of pipe after the last celi.Let the liquid flowing through the tube be taken as the carrier and be subscripted 1 whilst that injected is the sample subscripted 2. The flow-rate is measured by the time taken for the carrier t, or sample t, to pass along a given length of the tube 1. The change in the time taken for the sample to pass along the same length of tube relative to the carrier is clearly dependent upon the difference in viscosities and the relative amounts of sample I,, and carrier I (Il + I = L ) . It may be deduced that the following relationship holds: which yields if one uses a fixed length of tubing and sample size and a constant pressure, the simple linear relationship .. ' * (3) At = k6 A7 . where It may also be shown that January 1983 HIGH-PRECISION where These are the key equations which which via kb provide a means for FLOW INJECTION ANALYSIS. PART I1 19 8L a2P k == -enable reasonable system parameters to be selected and testing the validity of the approach. It is to be noted that provided the viscosity of one of the ll’quids is known that offhe other can be evaluated. Hence whether determining the viscosity of the injected sample relative to the standard carrier or the viscosity of the carrier relative to the standard sample the same equations hold. The Diffusion Coefficient Approach The dispersion of a sample plug introduced into a fluid stream flowing under laminar-flow conditions has been studied by several workers prior to the advent of flow injection analysis (FIA).5-9 Taylor in a paper which has proved the starting point of most subsequent discussion argued that there were two principle modes of di~persion.~ The first which he termed “convection,” is the parabolic plug flow of classical physics.The second is mole-cular diffusion with radial diffusion being more important than longitudinal. The effect of molecular diffusion is to give a more symmetric dispersion of the sample which results in Gaussian peaks being observed under appropriate conditions. Under conditions of pure convection the sample front proceeds down the tube a t twice the mean velocity u whereas the fraction of sample at the wall of the tube has zero velocity and hence gives rise to a large tail on an asymmetric sample peak.If however radial molecular diffusion is taken into account molecules at the head of the sample diffuse away from the centre towards the wall and experience a drop in velocity whereas those at the wall diffuse towards the centre and gain velocity. He calculated that for molecular diffusion to be dominant the following condition must be fulfilled : L a2 ->- U (3.8)20 where L is the length of the tube over which the sample is dispersed U is the maximum velocity and D is the molecular diffusion coefficient. (This is sometimes called “Taylorian conditions’’ but in view of the number of cases Taylor considered and delineated its use is confusing.) It follows that for convection to be predominant the time of flow must be less than a2/(3.8)20 and this is the condition assumed in the modified Ostwald viscometer approach.For most of the experimental results given below the boundary condition is a flow-rate of 0.5 cm3 min-l. Assuming that molecular diffusion predominates then the concentration distribution of the sample of mass M is given by Mexp [ - (x - X)2/4KTt] C = . . 2a2 (r3KTt)0.5 where C is the concentration t is the time of the measurement Z is the point of maximum concentration and x is the distance from it K = (a2u2)/48D and is known as the Taylor diffusion coefficient. This is a form of the normal error graph and it follows that where or 1 au * * (7) D=-(-)t 24 (T where (T is half the peak width at 0.607 of its height 20 diffusion coefficient.BETTERIDGE et aZ. AUTOMATED VISCOMETER BASED ON AnaZyst VoZ. 106 Hence the peak height and the peak width are proportional to the square root of the The Stokes - Einstein relationship islo . . . . . . . . RT 67~77~~ N Dz-where N is Avogadro’s Number R is the gas constant and yS is the radius of the solute particle which is assumed to be a sphere. If this holds then * * (9) C,,,. = const. ( t / 7 7 ) O a 5 . . Although Taylor’s description is substantially correct and it provides an excellent starting point it has been subject to a number of criticisms and modifications have been proposed. Taylor himself proposed a more stringent set of condit ions.ll Ark7 and Ananthakrishnan et aZ.6 took the effects of longitudinal molecular diffusion into account and demonstrated that they were not completely negligible.The effect of their studies and that of Philip,g was to give more complex equations. Indeed Ananthakrishnan et aZ. preferred to use a dimension-less unit in order to achieve a solution of the equations. They defined the valid boundary conditions for the use of the equations of Taylor and Aris themselves as well as those for pure convection and diffusion. We have used these criteria to ensure that in the diffusion studies Taylor’s equations are valid. Taylor’s treatment also assumes that all of the sample is injected as an infinitely thin plug, an assumption which must be questioned in view of the dimensions of sample length and tube length commonly employed in FIA. This point along with many other aspects of the diffusion model has been dealt with by Vanderslice et aL.12 They concluded that if the sample size was less than 20% of the total volume it had little effect on the dispersion pattern.Bate et a1.l3 have pointed out that the pure convection model cannot be true otherwise some of the sample would remain stuck to the wall of the tube. Apart from these theoretical considerations experimental arrangements have appreciable effect on the dispersion pattern as was shown in Part I. disruptions to the flow patterns by injection valves detector cells etc.,l7$ l8 all contribute to the dispersion. Nevertheless the fact remains that in principle the total dispersion is the sum of the parts17 : Coiled tubes,14-16 turbulent where 20 is the peak width at 0.607 of the peak height.Hence under diffusion controlled dispersion the graphs are Gaussian and if all of the interfering factors are constant they may be used to obtain diffusion coefficients with appropriate calibration. A further limitation is that the Einstein - Stokes relationship is not valid for all systems, even for some with spherical parti~1es.l~ Finally and most importantly all of these theories and others like the tank-in-series theory,l7T2O which have been used to explain sample dispersion assume that there is no chemical interaction between the sample and the carrier. As will be seen below the validity of this assumption depends crucially on the nature of the solvents employed. Results and Discussion Determination of the Effect of Physical Parameters fully discussed in Part I.problems encountered whilst dealing with solutions of a wide range of viscosities. seen that in some instances chemical effects have to be taken into account. The effects of the physical parameters on the measurement of peak time and height were Those discussed here relate to testing of the theories discussed above and the practical It will be Tzcbe radizcs Problems arising from the flow tube becoming blocked were encountered when samples of high viscosity i.e. 70-100~0 aqueous glycerol solutions and 60% sucrose solutions wer January 1953 HIGH-PRECISION FLOW INJECTION ANALYSIS. PART I1 21 injected into a carrier flowing through a 0.086 cm i.d. tubing. In these circumstances larger internal bore flow tubes were used for example 0.10 and 0.125 cm i.d.However these were unsatisfactory when determining the velocity of low viscosity samples as they reduced the experimental precision obtained with the system (Table I). TABLE I EFFECT OF TUBE RADIUS ON PRECISION The conditions were as follows carrier cyclohexane coloured with 0.04% PAR; flow-rate 2.15 cm3 min-l (6.17 cm s-l) ; detector photometric two cells 20 and 60 cm respectively from the injection valve; and sample volume 30 p1. Coefficient of variation in t,* yo f 1 Flow tube material . . . . PTFE Kel-F 7-Internal tube radius/cm . . 0.043 0.05 0.0625 Sample concentration yo Glycerol. . . . 5 60 70 80 90 100 Sucrose . . . . 20 40 60 0.21 0.52 0.497 0.647 1.20t 2.oot 0.20 0.28 o.9ot 0.22 0.35 0.50 0.22 0.50 0.28 0.56 0.33 0.81 0.31 1.507 0.50 0.25 0.45 0.29 0.35 0.50 0.31 * The coefficient of variation in t was based on four experi-7 Experimental runs not necessarily successive owing to occas-mental values.sional pipe blockage. Sample volame The effect of sample volume experienced by Vanderslice et a1.12 in which the dispersion of the sample plug was affected by sample volumes greater than 20% of the total volume of the system was investigated to determine whether the relationship between time of flow and sample viscosity [equation (4)] was affected in a similar manner. Sample volumes between 5 and 100 pl were used the maximum volume being set by the appearance of a flat apex to the peaks resulting from an insufficient length of PTFE between the electrodes of the annular conductivity cell.The 5-p1 volume was obtained directly from the slide in the chromato-graphy valve the bore of the sample chamber being identical with the flow tube and there-fore not disturbing the flow. However larger volumes than 5 p1 would have required sample chambers with substantially larger bores and these would have disturbed the flow pattern. Therefore sample volumes greater than 5 pl were obtained via the sample loop. A linear graph of sample plug length vcysus t / q was obtained (Fig. l) which demonstrates that there is no deviation from linearity as a result of the sample dispersion being affected by the sample volume. For a length of 60 cm between the injection valve and the second cell, sample volumes larger than 60 p1 were greater than 20% of the volume of the flow tube and were therefore above the limits set by Vanderslice et al.Flow-rate Two systems were studied in detail to determine the effect of flow-rate on the relationships t = f(7) and peak height = f [ ( t / y ) o - 5 ] although similar trends were observed in other systems. The first which was aqueous based involved two annular conductimetric cells positioned 20 and 70 cm from the injection valve respectively whilst the second involving organic solvents employed two photometric cells positioned 20 and 60 cm from the injection valve. For the aqueous system plots of time for the sample to flow between two points v e r ~ ~ s sample viscosity produced a linear graph for the fast flow-rate of 3.56 cm3 m F whilst plot 22 9 1 -BETTERIDGE et al.AUTOMATED VISCOMETER BASED ON Analyst VoZ. 108 ::// 0 10 20 30 40 50 60 Glycerol % 100 80 .?? 60 F 40 20 E" 0 E CT 2000 1500 .-a II a II p 1000 .- w m -$ 500 w 5 2000 .-a r m 0. > Y 1500 a 1000 a 500 CT .- c -1 1 I I 1 I 10 20 30 40 50 60 Glycerol % 1 3 5 7 9 ViscosityIcP Fig. 2. For glycerol solutions injected into 1 M sodium chloride solution effect of flow-rate on (I) percentage composition and viscosity V e n u s time and (11) percentage composition and ( t / q ) O . . " uwsus peak height. Flow-rate A 3.56 c1n3 min-' (10.21 cm s-l), B 1.16 cm3 min-l (3.33 cm s-l) and C 0.52 cm3 min-l (1.49 cm s-l) ; detector annular conductivity; carrier 1 M sodium chloride solution; sample volume 30 pl; flow tube between cells 50 cm; temperature 22.3 O C ; and tube internal radius 0.043 cm January 1983 HIGH-PRECISION FLOW INJECTION ANALYSIS.PART 11 23 In the organic system aqueous glycol solutions coloured with O.OS~o 4-(2-pyridylazo)-resorcinol (PAR) were injected into a toluene carrier. Linear graphs were obtained for time veysus viscosity and peak height ueysus ( t / 7 7 ) O e 5 for all flow-rates (Fig. 3). This indicates that the curvature in the calibration graphs for the aqueous system was a result of chemical interaction between the sample and carrier and not a consequence of physical anomalies in the system. 30 v) 3 20 E .-I-10 I 10 20 30 40 50 60 Glycerol YO v) . g 20 .-I-10 ViscosityIcP 2 2500 m 9) 1 Y m .-2000 > m al .- w -a 1 500 10 20 30 40 50 60 Glycerol % I 1.7 2.2 2.7 3.2 3.7 4.2 ( t / q ) 0 .5 Fig. 3. (I) For aqueous solutions of glycol injected into toluene effect of flow-rate on (a) percentage composition and viscosity veysus time and (b) percentage composition and ( t / ~ ) ~ - ~ veysus peak height. Flow-rate A 3.40 cm3 min-' (9.75 cm s-l) B 1.20 cm3 min-l (3.44 cm s-l) and C 0.61 cm3 min-l (1.75 cm s-l); detector photometric (sample coloured with PAR) ; carrier toluene; sample volume 30 p1; flow tube between cells 40 cm; temperature 20.1 "C; and tube internal radius 0.043 cm. (11) Peak height VeYsus (a) composition and (b) ( t / v ) O s 5 . Flow-rate A 3.40 cm3 min-l (9.75 cm s-l) B 1.20 cm3 min-l (3.44 cms-l) and C 0.61 cm3 min-l (1.75 cm s-l) ; detector photometric; carrier toluene; sample volume 30 p1; flow tube between cells 40 cm; temperature 30.1 "C; and tube internal radius 0.043 cm.Tube length A photometric system consisting of five cells positioned 20 40 60 80 and 100 cm respec-tively from the injection valve was employed to determine the effect of the tube length on the sample velocity. Two different carriers whose viscosities were similar cyclohexane and water were used to show the effect of tube length on sample - carrier interaction. Aliquots of 30% ethanol coloured with 0.58% xylenol orange injected into a carrier of cyclohexane produced a linear graph for the time the sample plug took to travel between two cells versus distance travelled (Fig. 4). However when the cyclohexane carrier was replaced by water, and the same flow-rate produced a curved graph was obtained for identical flow conditions [Fig.3 ( b ) j . These results corroborate those described above in which non-linear graphs of time of flow veysus sample viscosity were obtained for an aqueous carrier. The viscosity of ethanol increases as the percentage composition increases until a maximum viscosity i 24 BETTERIDGE et al. AUTOMATED VISCOMETER BASED ON Analyst VoZ. 108 Length of tubeicm Fig. 4. Effect of tube length on sample residence time. Flow-rate 2.40 cm3 min-I (6.89 cm s-l) ; sample 30"/0 ethanol + 0.05% xylenol orange; volume 30 pl; carrier (a) cyclo-hexane and (b) water; detector photometric five cells 20, 40 60 80 and 100 cm from injection valve; flow tube, PTFE; temperature 22 "C; and tube internal radius, 0.043 cm.reached at 48% by mass after which the viscosity decreases. The curvature in Fig. 4 may be explained by hydrogen bonding increasing the viscosity of the sample to approach this maximum value. If there was no chemical interaction the graph expected can be calculated by application of equation (3) and this is shown in Fig. 4. I t is seen that the theoretical graph parallels the experimental one for the cyclohexane carrier whereas for the aqueous system the deviation is first in the direction of increasing viscosity and then towards decreasing viscosity as expected with greater dilution of sample. Investigation into Sample - Carrier Interaction and Its Effect on the Measurement of Viscosity Initially observed anomalies to the theoretical relationships were attributed to sample -carrier interaction mainly through hydrogen bonding; the geometry of the detector being 26 23 5 20 P .E 1 1 - -a, c 17 % 14 n a, c 8 5 J CHC13 Pro panol L I I 2 4 6 8 10 Vi sco st i yic P Fig.5. Relationship between time and viscosity for samples injected into toluene. 0 First experi-mental run; and 0 average of four experimental runs. Carrier toluene + 0.0259h PAN; sample volume 60 p1; flow-rate 2.50c1n3 min-l (7.17cm s-l) ; detector photometric two cells 20 and 60 cm from injection valve; temperature 20.8 "C; and tube internal radius 0.043 cm. 17 l9 t 2 15 -a, c '3 3 % 1 1 .E 9 a, n a, F 7 5 Octanol / [k G&o h ex a n e Met ha no1 \ 1 biethyl ether 2 4 6 - 8 10 Viscosi tykP Fig.6. Relationship between time and viscosity for samples injected into cyclohexane. Carrier, cyclohexane + 0.02596 PAS; sample volume 60 pl; flow-rate 2.15 c1n3 inin-' (6.17 cm s-l); detector, photometric two cells 20 and 60-cni from inection valve; temperature 20.8 "C; and tube internal radius 0.043 cm January 1983 HIGH-PRECISION FLOW INJECTION ANALYSIS. PART 11 26 such as to alter the velocity profile of the sample; and impurities in the sample causing the solution to have a different viscosity to that listed in the literature. A re-designed apparatus and improved experimental procedure removed the last two of these possibilities and led to the conclusion that anomalies such as those described above could only be due to interaction between the sample and carrier.The effect of hydrogen bonding occurring in certain systems was therefore studied in order that the systems in which it occurs and therefore deviate from theory could be predicted. In these systems it would be necessary to obtain a calibration graph using standards of known viscosity and similar composition to the sample under investigation in order to account for the distortion in the experimental results as a result of chemical interaction. All the experiments were performed using the automatic chromatography sample valve with a sample loop of volume of 60 pl. A single length of PTFE tubing was used between the injection valve and waste with two photometric cells positioned along it 20 and 60 cm, respectively from the valve.Two organic systems were studied Figs. 5 and 6 where there was no possibility of hydrogen bonding occurring between the sample and carrier. Both cyclohexane and toluene produced approximately linear graphs for sample travel time veyszts viscosity with a variety of samples, each value being the mean of three experimental determinations. When the organic carriers were replaced with ethanol and water (Figs. 7 and 8) deviations from linearity were observed for a series of alcohols although samples which did not contain available hydroxy groups for bonding did fall on a linear plot. Hydrogen bonding results in an increase in the molecular mass of the sample thus the sample plug takes longer to travel down the flow tube. This can be seen in the peak profiles shown in Fig.9 where a sample of acetone was injected into four different carriers. In systems where ethanol or water was used as a carrier the peak width was wider and the peak height shorter than in the toluene and cyclohexane systems. 26 23 v) - 2 20 - al 17 4- 3 E 'F 11 2 14 8 5 CHCI? 1 I 2 4 6 8 10 Viscosi tyicP Relationship between time and viscosity €or samples injected into ethanol. Carrier ethanol + 0.05% methyl orange; sample volume 60 p1; flow-rate 2.10 cm3 min-l (6.03 cm s-l) ; detector, photometric two cells 20 and 60 cm from injection valve; temperature 20.8 "C; and tube internal radius 0.043 cm. Fig. 7. 23 20 v) v) 1 -8 17 p 14 C W Y a 13 .- ; l1 I-8 5 Diethyl rther / Cyclo hexane 2 4 6 8 10 Fig.8. Relationship between time and viscosity for samples injected into water. Carrier water + 0.05% methyl orange; sample volume 60 p1; flow-rate 2.20 cm3 min-l (6.31 cm s-I) ; detector photometric two cells 20 and 60 cm from injection valve; temperature, 20.8 "C; and tube internal radius 0.043 cm. Vi scosityicP The slopes of the linear graphs in Figs. 5-8 were. determined to yield values for kk from equation (3). The parameter k6 is a constant for the system provided that all the experi-ments are performed under identical flow conditions. Table I1 compares the experimentally obtained values for kk for the different carriers with those derived from the relationship k6 - tlZ,/Lql. The largest relative errors between the experimental and theoretical values for kb were obtained in the ethanol and water system 26 BETTERIDGE et al.AUTOMATED VISCOMETER BASED ON Analyst VoZ. 108 Fig. 9. Peak shapes resulting from injection of acetone into various carriers. Sample acetone ; volume 60 p1; carrier (A) toluene + 0.05% PAN (B) ethanol + 0.05% PAR (C) water + 0.05% PAR and (D) cyclohexane + 0.05% PAN; detector photometric, two cells 20 and 60 cm from injection valve; flow-rate, (A) 2.50 cm3 min-' (7.17 cm s-l) (B) 2.10 cm3 min-' (6.17 cm s-l) (C) 2.20 cm3 min-l (6.31 cm s-l) and (D) 2.15 cm3 min-l (6.03 cm s-l). where the samples that lay on the linear portion of the graph were used to determine A;. These were mainly inert samples without hydroxy groups to take part in hydrogen bonding, with the exception of propan-2-01 and the sample of ethanol in the ethanol carrier.The large relative error may be a consequence of insufficient samples being used to yield the value for k6. Table I11 lists viscosity values for the samples investigated obtained by substituting the experimentally obtained values for k6 into equation (3). The cyclohexane and toluene carriers yielded results similar to those obtained via the Ostwald viscometer the majority of values having a small negative deviation which may be attributable to experi-mental error. Large positive variations in the viscosity values were obtained for samples of alcohols acetone and diethyl ether injected into water or ethanol carriers. TABLE I1 EXPERIMENTAL AND THEORETICAL VALUES FOR k6 Values for k'D Experimental from Derived from A I \ Carrier slope of linear graph equation (3) Cyclohexane .. 1.62 1.53 Ethanol. . 1.90 1.41 Toluene 2.45 2.60 Water . . 1.96 1.64 The effect of sample - carrier interaction on the relationship of peak height veYsus ( t / 9 ) 0 - 5 was observed for a slow flow-rate of 0.92 cm3 min-l in two systems involving toluene and ethanol as carriers. In each system a series of alcohols coloured with 0.075% PAR were injected into the respective carriers and the variation in the absorbance measured down stream. A linear graph for peak height veysus ( t / q ) O s 5 was obtained with toluene as the carrier. However a curved graph was obtained when the toluene was replaced by ethanol as the carrier (Fig. 10). The peak maxima and 0 for the peaks resulting from the ethanol system are, respectively shorter and larger than their counterparts in the toluene system (Table I V ) .Thus a greater degree of sample dispersion is caused in the ethanol system owing to the hydrogen bonding increasing the molecular mass of the sample. Measurement of the Viscosity of the Carrier by Injecting Samples of Known Viscosity A possible use for the FIA based viscometer is in production line situations where the liquid being manufactured is tapped from a production line and used directly as a carrier fo January 1983 HIGH-PRECISION FLOW INJECTION ANALYSIS. PART 11 27 the viscometer. A sample of known viscosity is injected and the viscosity of the unknown carrier determined via a rearranged form of equation (3) i.e., Sample Toluene .. Cyclohexane . . Tetrachloromethane Chloroform . . Methanol . . Ethanol. . . . Propan-1-01 . . Butan-1-01 . . Pentan-1-01 . . Octanol . . . . Propan-2-01 . . Diethyl ether . . Acetone. . . . Water . . Dimethylformamide . . . . . . . . . . . . . . . . . . . . . . . . . . TABLE I11 EXPERIMENTAL VISCOSITY VALUES Viscosity/cP Carrier* . . . . . . . . . . . . . . . . . . . . . . . . Cy clohexane 0.60 (+0.05) --0.85 ( - 0.10) 0.66 (+ 0.09) 0.73 (+0.13) 1.16 2.29 (-0.01) 3.04 (+0.02) 3.73 (0) 9.16 (0) 2.29 (-0.11) 0.35 ( + 0.13) 0.29 (- 0.04) (- 0.07) --0.91 (-0.07) Toluene --0.97 0.88 0.51 0.55 1.05 2.13 2.93 3.63 9.09 (- 0.07) 2.38 0.22 0.47 (-0.13) (- 0.07) (- 0.06) (-0.05) ( - 0.15) (-0.17) (-0.09) (-0.10) (- 0.02) ( 0) (+ 0.11) --1.55 (+ 0.57) Ethanol 0.78 1.09 1.04 0.73 (+0.23) (- 0.01) (+0.09) ( + 0.16) ---_.6.41 7.52 8.41 11.67 2.46 2.15 3.30 4.52 3.78 (+4.11) ( + 4.50) (+ 4.68) (+2.51) (+0.06) (+ 1.93) (+ 2.94) (+3.52) (+ 2.80) Water' 0.64 (+0.09) 1.10 (0) 1.00 ( f 0 . 0 5 ) 0.64 (+ 0.07) 1.36 (+ 0.76) 1.72 3.21 5.15 5.62 9.28 2.44 1.62 1.26 ( 1-0.52) (+0.91) ( + 2.03) (+ 1.89) ( + 0.12) ( f 0 . 0 4 ) (+0.90) (+ 1.40) --1.36 ( + 0.38) Ostwald experimental values 0.55 1.10 0.95 0.57 0.60 1.20 2.30 3.02 3.73 9.16 2.40 0.22 0.36 1.00 0.98 Literature valu es5 16 0.587 1.06 0.969 0.566 0.5945 1.194 2.26 2.98 3.68 9.13 2.43 0.242 0.358 1.002 0.85 * Values in parentheses are deviations from values obtained via Ostwald viscometer.A high accuracy in the viscosity values might not be required; however precision is important because one wants to be sure that any fluctuations in the measurement between successive runs are due to varying composition of the manufactured liquid and not a consequence of the viscometer. I I 1 I 1 I 1 2 3 4 5 6 7 3 4 5 6 7 (tiq)O Fig. 10. Effect of carrier on relationship between peak height and ( t / 7 7 ) 0 . 5 . Carrier (a) toluene and ( b ) ethanol; sample volume 60 p1; flow-rate 0.92 cm3 min-l (2.64 cm s-l) ; detector photometric two cells 20 and 60 cm from injection valve; temperature 20.8 "C; and tube internal radius 0.043 cm 28 BETTERIDGE et al.AUTOMATED VISCOMETER RASED ON Analyst Vol. 108 TABLE IV PEAK CHARACTERISTICS OF ALCOHOLS INJECTED INTO TOLUENE AND ETHANOL The conditions were as follows flow-rate 0.92 cm3 min-1 (2.64 cm s-1); sample a series of alcohols coloured with 0.075% PAR; carrier toluene or ethanol; sample volume 60 p1; detector photometric two cells 20 and 60 cm, respectively from the injection valve. Carrier r Toluene Sample Methanol . . 692 Ethanol. . . . . . 800 Butan-1-01 . . . . 1275 Pentan-1-01 . . . . 1380 Octan-1-01 . . . . 1498 Propan-1-01 . . . . 1010 \ Ethanol 7 7-5.05 489 7.30 5.80 -6.45 730 8.55 7.20 855 8.80 7.80 1023 9.20 9.10 1249 10.50 a*/s Peak height o*/s -* Half peak width at 0.61 peak height.An experiment was set up to simulate a production line situation in which a carrier of 50% glycerol was gradually diluted throughout a series of experiments. Aliquots of 30 pl of cyclohexane were injected as the sample. A graph of the time taken for a sample to travel between two cells positioned 40 cm apart wmus the number of the run was plotted which also permitted the level of the carrier to be realigned with a mark on the side of the constant-head. As can be seen from Fig. 11 run 7 was performed before the whole flow system had properly equilibrated at the new concentration hence the time for the sample to flow between the two cells was high for that concentration.The viscosity of the carrier for each concentra-tion was determined via equation ( 1 1 ) . The composition of the carrier was determined by interpolation between the relevant points on a calibration graph of viscosity zwsus percentage composition21 and found to be in agreement with that actually employed. C e P .- ; G 7 ' n I-43% Glycerol L a - + 30% Glycerol 3 6 9 12 61 Number of run Fig. 11. Determination of the viscosity of an unknown carrier. Sample cyclo-hexane ; carrier glycerol solutions diluted through experiments ; detector photometric, two cells 20 and 60 cm from injection valve; and temperature 20 "C. Viscosity Measurements above Room Temperature The ability of the flow system incorporating the thermostated water-jacket to maintain a constant temperature above room temperature and the effect of temperature on the viscosity of several samples was investigated.Two systems were studied both of which used toluene as a carrier; the first involved various concentrations of aqueous glycerol solution whilst the second was concerned with samples of acetone butanol and propanol. All the samples were coloured with 0.050/ PAR. The flow tube consisted of PTFE tubing, 0.086 cm i.d. with two photometric cells situated 20 and 60 cm respectively from th January 1983 HIGH-PRECISION FLOW IN JECTIOK AN.4LYSIS. PART I1 29 injection valve. A flow-rate of 1.8 cn13 min-l was maintained throughout the entire series of experiments. Temperatures above 60 "C were not studied in these systems consisting of photometric cells owing to a noticeable deterioration in the response from the cells which resulted from a breakdown of the optical components.The ability of the water-jacket to maintain the temperature gradients being set up was examined up to 68 "C and it was found to be a very effective method for temperature control. Hence temperatures between 20 and 50 "C were investigated for the two chemical systems. Fig. 12(a) shows the relationships between the sample residence time between two points and the viscositv of the sample determined via equation (3) and the Ostwald viscometer and literature ;alues21*22 for* various temperatures between 20 and 50 "C. glycerol system logarithmic graphs were employed. 1 2 3 Glycerol O/O ViscositykP I In the aqueous 1 TemperaturePC Tem per at u re/"C Fig.12. Viscosity measurements for various solutions a t different temperatures (I) glycerol - water mixtures and (11) acetone propanol and butanol. Conditions 0 experimental values from FIA system ; 0 experimental values from Ostwald viscorneter ; and A literature value.21 The experimental values are in close agreement with those taken from the literature; the graphs of viscosity zleysuus temperature [Fig. l2(b)] follow the expected trend i.e. as the temperature increases so the viscosity of the sample decreases. Each value is based on the mean of three experimental runs with the exception of 90% glycerol in which only a single run was undertaken owing to the length of time (10 min) required for the sample to travel the 40 cm.The largest relative errors observed between the experimental and literature values for viscosity were observed in the high percentage glycerol solutions. For example, the experimental viscosity of 90% glycerol was determined as 190 cI' at 20 "C whilst the literature7 cited it as 219 cP. No viscosity measurements above 35 CP were made with the 0s twald viscomet ers. For a greater accuracy in the viscosity values of highly viscous samples wider bore tubing should be used as the flow tube. However this was impossible in this set of experiments as the wider i.d. tubing would not have been satisfactory for the lower viscosity samples. The highest accuracy obtained between the experimental values obtained via the FIA systems and those yielded from an Ostwald viscometer was for the sample of acetone at 40 "C.Generally the accuracy varied between 0.3 and 10% with the exception of 10% glycerol at 50 "C where a larger error of 23% was obtained 30 BETTERIDGE et al. AUTOMATED VISCOMETER BASED ON Analyst YoZ. 108 Measurement of Diffusion Coefficients Huber and van Vught have used Taylor's method for determining diffusion coefficients in binary systems by measuring the elution curves resulting from both gas and liquid chroma-t o g r a p h ~ . ~ ~ They state that the accuracy of the solution depends on how closely the experi-mental peak approximates to a Gaussian shape and how accurately the other parameters in equation (7) are determined. In addition the effect of the tube not being straight and the influence of the uncertainty in the measurement of the tube radius do not appear at low flow-rates provided the tube radius is not too large.Therefore the accuracy with which D can be determined depends mainly on the accuracy of measuring 0 which can only increase if the time of analysis is increased. However we have obtained anomalous results when flow times of 5-10min have been used. The sample components have separated to some degree making the accurate measurement of the peak width and diffusion coefficient impossible. 24 Diffusion coefficients for potassium permanganate and PAR injected into various carriers, were measured by the following method. Two photometric cells were situated along a single flow tube 20 and 120 cm respectively from the injection valve.To ensure that the dispersion caused by the injection valve to the sample plug was eliminated the standard deviation of the peak from the first cell was subtracted from that at the second cell to yield the spread of the graph 0 which resulted from the transport of the sample plug along the flow tube. TABLE V EXPERIMENTAL VALUES FOR DIFFUSION COEFFICIENTS OF 5 x M POTASSIUM PERMANGANATE SOLUTION INJECTED INTO WATER AT DIFFERENT FLOW-RATES The conditions were as follows carrier water; sample 5 x M KMnO solution; volume, 30 p1; detector photometric two cells 20 and 120 cm respectively from the injection valve; temperature 20.1 "C. Experimental values of u* and D t f Run 1 Flow-rate/ p---bh-7 cm3 min-1 u/cm D x lo5 0.3 13.8 3.5 0.4 16.2 3.4 0.6 18.7 3.6 0.8 21.3 3.8 1.0 23.8 3.9 Run 2 Run 3 z: &&u/cm D x 105 o/cm D x 105 o/cm D x lo5 13.8 3.5 13.8 3.5 13.8 3.5 15.9 3.5 15.7 3.6 15.9 f 0.3 3.5 f 0.1 19.2 3.4 17.9 3.9 18.6 i 0.7 3.6 f 0.3 22.9 3.3 21.9 3.6 22.0 & 0.9 3.6 & 0.3 22.4 4.4 25.9 3.3 24.0 f 1.9 3.8 & 0.6 * 0.5 Peak width at 0.61 peak height; the mean value is based on three runs.t Diffusion coefficient (cm2 s-l). Values f the range for experimental values under specific flow-rate. Measurements that were taken under flow-rates satisfying Taylor's stricter conditionsll yielded the most reproducible results. In this instance the sample took over 2 min to flow through the system which necessitated the time interval over which each data point was collected to be increased to 500ms ensuring that the whole peak was collected.Inter-polation between the relevant data points yielded the value of CJ to within 0.1 s. Difficulty was experienced in obtaining diffusion coefficients from the literature for the potassium permanganate and PAR system studied. Furth and Ullmann are quoted by Taylor5 as studying a concentration range of 0-0.06 M potassium permanganate at 18 "C and obtaining the range 0.435 x for the diffusion coefficient measurements. The values in Table VI are greater than the range they give but the temperature at which these were measured was greater than in Furth and Ullmann's system. Bird quotes several diffusion coefficients for liquid binary systems in the region of 1 x This factor was obtained in all the experimental diffusion coefficient measure-Owing to the lack of data with which to make sound comparisons it is impossible to 1.5 x cm2 s-1.25 ments January 1983 HIGH-PRECISION FLOW INJECTION ANALYSIS.PART 11 31 to decide on the accuracy of the method.* I t looks feasible but in view of all of the possible experimental sources of contribution to the peak broadening it is preferable to use the method on a comparative basis. This however requires reliable standards for calibration. TABLE VI DIFFUSION COEFFICIENTS FOR POTASSIUM PERMANGANATE SOLUTION INJECTED INTO WATER ETHANOL AND ACETONE The conditions were as follows flow-rate 0.3 cm3 min-l; detector photometric two cells, 20 and 120 cm respectively from the injection valve; temperature 20.1 "C; sample volume 30 pl.Carrier Water Ethanol Acetone I A 7 Concentration of KMnO (-*-- r - - ~ (-L-, 5 x 10-4 . . . . . . 13.8 3.5 15.6 2.7 14.5 3.2 5 x lo-' . . . . . . 14.9 3.0 16.0 2.6 14.0 3.4 5 x 10-5 . . ,. . . 15.4 2.8 16.2 2.5 14.6 3.1 5 x 10-5 . . . . 15.7 2.7 16.8 2.3 15.0 2.9 injected/M u*/cm Dt x lo5 u*/cm D t x lo5 u*/cm D t x lo6 * 0.5 Peak width a t 0.61 peak height; mean value based on three runs. t Diffusion coefficient (ema s-l). Conclusions It has been shown that it is possible by the proposed method to measure the viscosity of a range of samples rapidly reliably and with a precision of 0.2-0.3%. The successful opera-tion depends on close control of instrumental and experimental operating conditions. Of the former it is desirable to work with short tube lengths and high flow-rates which requires good data handling and processing capabilities.I t is essential to maintain close control of the temperature of the system and to carry out timings via two detectors using the inlet valve only as a source for a sample. The common practice of using one detector and the inlet valve as a time-controlling device results in lack of precision. Experience has shown that it is important to match the sample and carrier both physically, i.e. similar viscosities and chemically i.e. prevent chemical interaction. Hence an ideal system consists of a carrier of similar viscosity to the sample under investigation whilst acting purely as an inert transporter for the sample. Generally precisions between 0.2 and 0.3% were obtained with the instrument although in systems where viscous samples were under investigation higher coefficients of variation in the results were obtained typically between 1.0 and 2.0%.For highly viscous samples improved precisions were obtained by increasing the internal bore of the flow tube which also decreased the likelihood of tube blockage. Provided that no hydrogen bonding occurs in a system the relationships expressed in equations (1)-(9) are generally valid. Other forms of intermolecular interaction are known to cause deviations from general relationships relating viscosity to physical parameters26 and these are minor relative to hydrogen bonding but should be borne in mind. In situations where accuracy is of importance use should be made of calibration graphs of viscosity veyszts percentage composition (based on samples of known viscosity and of similar composition to the unknown sample) in order to determine the latter's viscosity.The calibration graphs will take account of errors in the experimental data resulting from hydrogen bonding preferential diffusion of the indicator and dispersion of the sample plug resulting from the cell geometry etc. A viscosity range of 0.2-190 CP was investigated by this method but viscosities above this range may be determined if tubing with the appropriate sized internal diameters is selected. * Recently we have read the manuscript of a paper by Gerhardt and Adams (Anal. Chern. December 1982, in the press). They report the use of FIA for the measurement of diffusion coefficients based on the equations of Vanderslice et.aZ.,12 and draw the conclusion that the method is accurate and potentially more useful than existing methods for the measurement of diffusion coefficients 32 BETTERIDGE et aE. In extending these conclusions to FIA systems in general two caveats need to be entered, Firstly in FIA it is common to use coiled tubes and as the flow-rate increases secondary flow predominates.l5,l6 This helps the mixing process but is disadvantageous in viscosity measurements as it alters the parabolic velocity profile resulting from convection. Secondly, many of the effects noted would be smoothed out by use of a peristaltic pump operating at a constant flow-rate rather than at a constant pressure. However precision higher than 2-3y0 which is typical for most FIA systems is possible if the procedures adopted in this study are followed.The use of improved data acquisition procedures and processing of data via a microcomputer compared with using chart recorders and manually reading off the pertinent data better constructed flow cells that do not disturb the A ow and close control of temperature throughout the system are especially important in achieving high precision. Close temperature control is of particular importance as many reactions are temperature dependent and temperature variations in the laboratory through the day will introduce experimental error. The use of a two-detector system that enables the flow-rate to be monitored frequently and also takes away from the injection valve its role as a timing device will also improve the precision.It appears to us that these pro-cedures could be incorporated on H A procedures with advantage especially gradient analysis titrations and “stop-flow” kinetics. The authors gratefully acknowledge the award of an SAC Studentship to P.D. and the award of an SRC CASE award to T.B.G. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. References Poiseuille J. L. C. R. Acad. Sci. 1840 11 961 and 1041. Hagen G. Ann. Phys. Chewi. 1839 46 423. \‘an Wazer J. R. Lyons J. W. Kim K. Y . and Colwell R. E. “Viscosity and Flow Measurement,” Interscicnce New York 1963 Chapter 4. Goad T. B. PhD Thesis University of Wales 1982. Taylor G. Proc. R. SOC. London Ser. A 1953 219 186. Ananthakrishnan Y. Gill W. N. and Barduhn A. J . AIChE. J . 1965 11 1063. Aris R. Pvoc. R. Soc. London Ser. A 1956 235 67. Griffiths A Proc. Phys. Soc. London 1911 23 190. Philip J. R. A u s t . J . Phys. 1963 16 287. Glasstone S. “Textbook of Physical Chemistry,” Second Edition Rfacmillan London 1947 p. 261. Taylor G. Proc. K. Soc. London Ser. A. 1954 225 473. Vanderslice J . T. Stewart K. K. Rosenfeld A. G. and Higgs D. J. Talanta 1981 28 11. Bate €I. Rowlands S. Sirs J . A and Thomas H. W. J . Phys. D 1969 Ser. 2.2 1447. Taylor G. Proc. R. Soc. London Ser. A 1954 223 446. Tijssen R. A n a l . Chim. Acta 1980 114 71. Tijssen R. PhD Thesis Technische Hogeschool Delft 1979. Levenspiel O. “Chemical Reaction Engineering,” Second Edition John Wiley New York 1972, Reijn J . Rl. Van der Linden W. E. and Poppe H. A n a l . Chim. Acta 1980 114 105. Barrow G. M. “Physical Chemistry,’’ Third Edition McGraw-Hill New York 1973 p. 744. RfiZiEka J . and Hansen E. H. A n a l . Chim. Acta 1978 99 37. Segur J . K. and Oberstar H. F. I n d . Eng. Chem. 1951 43 2118. “Handbook of Chemistry and Physics,” Sixtieth Edition CKC Press Cleveland OH 1981. Huber J . F. I<. and \‘an Vught G. Ber. Bztnsenges. Phys. Chem. 1965 69 821. David P. PhD Thrsis University of Wales 1978. Bird R. B. Stewart TV. E. and Lightfoot E. N. “Transport Phenomena,” John Wiley New York, Partington J . R. “An ;Idvanced Treatise on Physical Chemistry,” Volume 2 Longmans London, Chapter 9. 1963 p. 504. 1951 pp. 71-127. Received August 16th 1982 Accepted September 14th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800017

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst January 1983 Vol. 108 pp. 33-42 Gas-chromatographic Procedure for the Determination of Environmental Organochlorine Residues in Avian Tissues with Confirmation of 33 Identities by Chemical Derivatisation Andrew J. Trim Peter M. Brown Peter J. Bunyan Edward M. Odam and Peter I. Stanley Tolworth Laboratory Agricultural Science Service AWinistry of Aguicultuve FishrrLes and Food Hook R Z S ~ South Tolworth Surbiton Surrey IiT6 7NF A procedure is described for the determination of low levels of organochlorine residues in avian tissues collected for environmental monitoring. The analytical scheme combines published and modified methods for the identifi-cation confirmation and quantitation of these residues. A hexane extract is cleaned u p using chromatography on alumina and silica gel followed by reaction with chromiuni(V1) oxide before measurement of residues using electron-capture gas chromatography.Identities of a range of compounds are confirmed by chemical derivatisation. Keywords Gas chromatography ; organochlorine pesticide yesidues ; chemical derivatisation ; environmental monitoring There is much published literature on organochlorine residue analysis and it is difficult to select the best method for a particular application. We have examined many of these methods and have combined the most suitable with our own modifications and additions to produce a comprehensive scheme for the determination and confirmation of small residues of organochlorines in the presence of interferents found in the tissues of predatory birds.VC7e hope that the procedure described will assist analytical chemists engaged in measuring organochlorine pesticide residues in environmental samples to confirm the identification of small residues without the need for a highly sensitive mass spectrometer. Primarily as a result of agricultural pest control the environment has become contanii-nated with low levels of organochlorine insecticides.l Many of these are persistent2 and have undesirable long-term eff e c k 3 The Laboratory has measured residues in the tawny owl (Strix aluco) and the barn owl (Tyto a h ) to monitor the withdrawal of the organo-chlorines used in agric~lture.~ Owls were chosen because they are at the top of a food chain are widely distributed and are sedentary. They are vulnerable to collision with road vehicles providing a source of carcases for post-mortem examination and chemical analysi~.~v~ A reliable method was required to determine accurately the range of organochlorine residues present in owl liver and muscle.It is particularly important to be able to measure and confirm the identity of residues that occur at levels below 0.1 mg kg-l. We now describe a procedure for the measurement of alpha-HCH gamma-HCH and aldrin (HHDN) and the measurement and confirmation of liexachlorobenzene heptachlor-2,3-epoxide dieldrin (HEOD) p@’-DDT $p’-DDE $P’-TDE endrin7 and PCB (polychlori-nated biphenyls as Aroclor 1254 equivalent). The owl sample is extracted with hexane, the extract cleaned up and separated into two fractions by liquid chromatography (using alumina and then silica gel) and finally a portion is cleaned up further using cliromium(V1) oxide.Measurement and tentative identification are made by reference to standard solutions using gas chromatography with electron-capture detection. This detector is not specific and coincident retention times are common so it is particularly important to confirm residue identities. Mass spectrometry is unsuitable for this because of limited sample size and the presence of interfering compounds. We therefore followed the approach of Hamence et nl.* by chemically converting the compounds of interest into products having different charac-teristic retention times giving confirmation of identity. Crown Copyright 34 TRIM et al. GC PROCEDURE FOR ORGANOCHLORINE RESIDUES Analyst Vol.108 Method Materials Analytical standards. Acetone. Hexane. Diethyl ether. Methanol. Acetic acid. Pyridine. Propanol. Chromium( V I ) oxide reagent. Solids >99Tl pure from commercial sources. Rathburn (Walkerburn) Ltd. glass distilled. Rathburn (Walkerburn) Ltd. glass distilled. BDH Chemicals Ltd. AnalaR. James Burrough Ltd. analytical-reagent grade. Dry glacial BDH Chemicals Ltd. AnalaR. Pierce and Warriner (UK) Ltd. silylation grade. Hopkin and Williams Ltd. laboratory-reagent grade re-distilled. Prepare a saturated solution of chromium(V1) oxide (BDH Chemicals Ltd. technical grade) in acetic acid as follows. Dry the chromium(V1) oxide at 100 "C for 2 h and then put 5 g of this into a 150-ml round-bottomed flask. Add 50 ml of acetic acid whilst swirling attach a micro-Snyder column and heat in a boiling water-bath for 15 min.Stopper the flask and shake well then loosen the stopper slightly and leave over-night at ca. 18 "C in a safe place. Caution-This mixture may ignite on handling and may freeze below ca. 15 "C. Potassium hydroxide. Hydrochloric acid concentrated 36% H Z i V . Antimony( V ) chloride. BDH Chemicals Ltd. technical grade. Store over anhydrous sodium sulphate. Sand. BDH Chemicals Ltd. laboratory-reagent grade fine sifted and acid washed. Heat at 400 "C overnight. Sodium sulphate anhydrous. Heat at 400 "C overnight and store in a glass jar. Anti-bumping granules. Hopkin and Williams Ltd. Pre-extract with hexane but let the solvent evaporate completely before use.Alwnina. ICN Pharmaceuticals GmbH W200 neutral super I. Deactivate with water (10% Vim) and check its activity with organochlorine standards and by fat ~ a p a c i t y . ~ A 500-600-mg amount of owl fat should be retained per 22-g alumina column. Silica gel. Wash with water then with diethyl ether and leave to dry overnight in a shallow trough. Activate by heating a 3-mm layer at 130 "C for 7 h. Adjust the activity with water (ca. 5-6% V/m) until the desired separation of organochlorines is obtained.1° BDH Chemicals Ltd. AnalaR. Rinse the pellets with hexane. Hopkin and Williams Ltd. UltraR. BDH Chemicals Ltd. AnalaR. 70-230 mesh No. 60 E. Merck AG "extra pure." Apparatus Glassware should be borosilicate with ungreased ground joints taps and stoppers and a solvent wash should be tested by gas chromatography to demonstrate absence of contamina-tion.Plastics and rubbers should be avoided. 150-mm i.d. all-glass. Mortars and pestles. Soxhlet extraction apparatus. These should contain an 80 x 28 mm i.d. cellulose thimble Pre-(a) 300 x 18 mm i.d. columns with tap and porosity 1 (b) 150 x 8 mm i.d. columns (Whatman Ltd. single thickness) and be fitted with a 150-ml round-bottomed flask. extract this apparatus with methanol for 3 h and then overnight with hexane. sinter. with quartz-wool plugs. Graduated test-tubes. High-pressure reaction vessels. (UK) Ltd. Silli-vials]. Syringes. Gas chromatograph. Liquid chro,matography columns. Pack these with a slurry of 22 g of alumina in hexane. Dry pack these with the silica gel (2 g).10 ml with two-valve micro-Snyder columns to fit. 2 ml with PTFE-lined septum cap [Pierce and Warriner 2 ml all-glass with Luer-fit needle for injecting acid into Silli-vials. With an electron-capture detector (Pye Unicam Ltd. GCV). Extraction the liver or muscle and weigh out ca. 5 g wet-mass ( M ) for extraction. Weigh a round-bottomed flask containing some anti-bumping granules. Dice a sample of Mix this with 5 January 1983 I N AVIAN TISSUES WITH CHEMICAL DERIVATISATION 35 of sand and with 20 g of sodium sulphate and grind the mixture in a mortar. Extract the dry powder for 3 h with hexane in a Soxhlet apparatus using the weighed flask as a receiver and a solvent exchange time of 1-2 min. Determination of dry-mass and fat-mass Weigh out another 2 g of the diced sample and determine its dry-mass ratio (in grams per gram wet-mass) by drying to a constant mass at 65 "C so that the results may be corrected for the water lost from tissues during storage.Re-weigh the round-bottomed flask and so find the mass of the extract by subtraction. Take a weighed aliquot (ca. 8 ml) and calculate the proportion P of the extract that this represents and so find the wet-mass of ex-tracted tissue in the aliquot by multiplying P and M Evaporate this aliquot with nitrogen to constant mass at ca. 18 "C and calculate the fat to tissue ratio (in grams per gram wet-mass) so that the results may be corrected for variations in the lipid contents of individual birds. Clean-up Removal of lipid alumina column Rotary evaporate the sample to 0.5 ml and apply it whilst still warm to a 22-g alumina column.Elute sequentially with three separate 3-ml hexane flask washes and then with hexane, collecting a total of 150ml. Rotary evaporate this volume and make up with hexane for preliminary gas chromatography (Fig. 1). Removal of interferents silica gel column on silica gel (see Materials) using the method of Holden and Marsdenlo as follows. Clean up the remaining extract by the method of Telling et U Z . ~ as follows. Separate the electron-capturing compounds into two fractions by column chromatography Evaporate 200 150 E E p 100 C 0 Q Q) Cr 50 0 5 X I 4 and X I I 48 36 24 12 0 Time/min Fig. 1. Gas chromatogram of an alumina-cleaned muscle extract from a typical tawny owl 1 hexachlorobenzene ; 2 alpha-HCH; 3 gamma-HCH; 4 hepta-chlor-2,3-epoxide; 5 pp'-DDE; 6 dieldrin (HEOD); 7 pp'-TDE'; X unknowns, thought to be polychlorinated biphenyls (PCB); and Y unknown thought not to be PCB 36 TRIM et al.GC PROCEDURE FOR ORGANOCHLORINE RESIDUES Analyst Vol. 208 a 10-ml aliquot of the 25-ml extract with nitrogen to 0.5 ml and quantitatively apply it to a silica gel column. Elute with 10 ml of hexane (fraction 1 ) then with 10 ml of 10% VjV diethyl ether in hexane (fraction 2). Fraction 1 is suitable for gas chromatography without dilution. Dilute a 1-ml aliquot of fraction 2 ten-fold with hexane before determining the organochlorine concentrations (see Determination by Gas Chromatography).If necessary, clean-up the remainder of the undiluted fraction 2 by treatment with chromium(V1) oxide. Removal of interferents treatment with chromium( V I ) oxide Add 8ml of fraction 2 to a 150-ml round-bottomed flask and evaporate with nitrogen, just to dryness at ca. 18 “C. Allow the sample to spread thinly over the bottom of the flask. Add 10 ml of the chromium(V1) oxide reagent swirl well and leave at 18 & 3 “C for 25 min. Add 10ml of hexane and shake well. Transfer the mixture into a 25-ml separating funnel. Retain the hexane layer and wash this three times with 10% VjV methanol in water (10 7 and 5 ml respectively) prior to gas chromatography (Fig. 3). Swirl the sample add 10 ml of 10% VjV methanol in water and swirl again. Determination by Gas Chromatography After alumina clean-up subject the sample to preliminary gas chromatography (Fig.1) so that appropriate standards may be prepared and contamination or serious losses can be recognised later should they occur. Measure hexachlorobenzene aldrin $$‘-DDE $$’-DDT and PCB in fraction 1 (Fig. 2). Aldrin is converted into dieldrin in vivo and is unlikely to be found. Measurement of PCB will only be approximate unless the peak retention times and height ratios are similar to those of the standards. Use the sum of the products of peak height and retention time as the basis of calibration for PCB. Measure alpha-HCH gamma-HCH heptachlor-2,3-epoxide dieldrin endrin and $$’-TDE in fraction 2 ; this fraction generally contains more co-extracted lipid than fraction 1 and should be diluted for gas chromatography.If the sensitivity of the gas chromatograph is insufficient to allow dilution and especially if interferents causing negative peaks occur measure heptachlor-2,3-epoxide, 150 100 . s 0 Q fn CT 50 60 48 36 24 12 0 Tirneimin Fig. 2. Gas chromatogram of the silica gel “fraction 1” from the alumina-cleaned extract shown in Fig. 1 1 hexachloro-benzene; 5 pp‘-DDE7; X unknowm, thought to be polychlorinated biphenyl January 1983 I N AVIAN TISSUES WITH CHEMICAL DERIVATISATION 40 20 0 E E 40 s g 20 0 --1 -a ln 40 20 -0 --37 36 24 12 0 Timeimin Fig. 3. Gas chromatogram of the silica gel “fraction2 ” from the alumina-cleaned extract shown in Fig. 1 and the effect of clcan-up by chromium(V1) oxide reagent (a) before treatment with chromium(V1) oxide; (b) after treat-ment (note the rcmoval of negative peaks) ; (G) mixed standard for comparison with (a) and ( b ) .1 Hexachlorobenzcne 1 pg; 2 alpha-HCH 2 pg; 3 gamnia-HCH 2 pg; 4 heptachlor-2,3-epoxide 6 pg; 5 pp’-DDE G pg; 6 tlicldrin (HIWII) 9 pg; 7 pp’-TDE 15 pg; and 8 pp’-DDT,’ 45 pg. dieldrin and p$’-TDE after the chromium(1’1) oxide clean-up of undiluted fraction 2 (Fig. 3). There should be little p$’-DDT here but if necessary this may be measured and added to the result obtained from fraction 1. Measurement of endrin may be hampered if an aged gas-chromatographic column is used as this can cause endrin to isomerise (see Results and Discussion). Apparatus and conditions Fit the gas chromatograph with a 1.5 m x 4 mm i.d.glass colunin filled with 1 part of 3% m/m SP-2100 methylsilicone on 100-120-mesh Supelcoport (Supelco Znc.) at the detector end and 4 parts of 376 m/m SP-2401 trifluoropropylmethylsiliconc on 100-120-mesh Supel-coport (Supelco Inc.) at the injection end. Condition the column at ca. 190 “C with nitrogen overnight before connecting it to the detector. Set the temperature of the injection port to 170 O C the detector to 300 “C and the colunin to 160 “C and use oxygen-free nitrogen as carrier gas. Using 24-pg injections of $p’-DDE, adjust the column head pressure to ca. 2.1 bar so that a retention time of 10-15 niin is obtained and adjust the attenuator and current settings to give a peak height of SOYo full scale on a base line with acceptable noise.Measwement and calculation The organochlorine concentrations are found by comparing the sample response with those of a calibration range of mixed standards injected daily. Use 1-p1 injections throughout. Select one of the standards from the calibration range and use this repeatedly between samples to monitor drifts in sensitivity. Table I gives the retention times obtained the limits of measurement and the recoveries. Leave the top 70 nim unpacked. Calculate the residue R in mg kg-l wet-mass from C 25 M - P M m R 38 TRIM et al. GC PROCEDURE FOR ORGANOCHLORINE RESIDUES Analyst VoZ 108 where C is the concentration (ngml-l) in the 25-ml extract M is the mass (g) of sample extracted and P is the proportion removed for determining the fat to tissue ratio.Where residues have been cleaned up by the chromium(V1) oxide reagent correct the results for the change in volume. Divide by the dry-mass ratio or by the fat to tissue ratio to express the residues on a dry-mass or fat - mass basis respectively. TABLE I IDENTIFICATION AND RECOVERY OF THE ORGANOCHLORINES INVESTIGATED Compound Hexachlorobenzene . . . . alpha-HCH . . . . . . gamma-HCH . . . . pp’-DDE . . . . pp’-TDE . . . . . . PP’-DDT . . . . Aldrin . . . . . . Heptachlor-2,3-epoxide . . Dieldrin. . . . * . Endrin . . . . PCB . . . . Relative retention 0.17 0.23 0.31 0.42 0.82 1.00 1.32 1.55 1.80 2.04 Multiple peaks (pjb’-DDE = 1) Lower limit of measurement r A 1 Approximate Concentration in wet-mass/ solutionlng ml-l mg kg-l 0.5 0.003 10 0.05 10 0.05 n.d.t n.d.t 3 0.02 3 0.02 4.5 0.02 n.d.t n.d.7.5 0.04 22.5 0.1 120 0.6 Recovery*, % 74 74 74 -85 74 100 74 -85 74 85 n.d.t * These figures are based upon a t least three determinations except for those of aldrin and t n.d. = not determined. endrin which are determined from a single experiment. Confirmation In each of the following confirmatory procedures the sample is evaporated with nitrogen just to dryness. In each instance a derivative is prepared the reaction mixture is cooled 10 ml of water are added and the product is partitioned into 2 ml of hexane. The hexane layer is treated then as described in the relevant section. The hexane solution is then examined by gas chromato-graphy primarily for changes in retention time.In some instances the identities of the products are unknown but Table I1 shows the changes in retention times observed the lower limits of confirmation and “response factors” (i.e. response from the derivative divided by the corresponding mass of the parent compound). If interferences occur which prevent confirmation by these procedures then a second silica gel separation is conducted on the products. PCB elutes in fraction 1 and PP’-dichloro-benzophenone (the derivative of $$’-DDE) in fraction 2. The derivatives of dieldrin and of heptachlor-2,3-epoxide are eluted with 10 ml of diethyl ether in a third fraction. Alterna-tively the peaks of interest may separate from this interference on a different gas-chromato-graphic column (e.g.3% SP-2100 alone or 3% DEGS). Except where indicated this is done in a graduated test-tube. pp’-DDE Evaporate 5 ml of fraction 1 add 5 ml of the chromium(V1) oxide reagent and shake well. Extract into Quantitative formation of pp’-dichloro-Keep at 30 “C for 15 min. hexane and wash the hexane layer with water. benzophenone confirms @’-DDE.I1 Add 3 ml of water and keep at ca. 18 “C for 3 d. PCB remains largely unaffected and may interfere. pp’-DDT To the hexane layer produced in the $p’-DDE confirmation add three pellets (about 2.8 g) of the potassium hydroxide and then about 1.5 ml of methanol. Shake vigorously for 30 s, attach a micro-Snyder column and heat at 100 “C until no more vapour is leaving through the valves.Extract into hexane and wash the hexane layer with water. Continue heating for 5 min more. Quantitative formation of @’-DDE confirms @’-DDT. January 1983 I N AVIAN TI S S U E S WITH C H E 31 I C A L I) E R I VAT I SAT1 0 N 39 PCB Perchlorination Completely evaporate the hexane layer produced in either of the jh$'-DDE or $$'-DDT confirmation procedures in a high-pressure reaction vessel (Silli-vial). Add 0.1 g of sodium sulphate and then 0.2 ml of antimony(V) chloride. Seal with a septum cap and heat at 170 "C for 3 h then cool. Pierce the cap with a spare syringe needle to release the pressure, then carefully inject 1 ml of 20y0 V/V concentrated hydrochloric acid in water from a 2-ml syringe to destroy excess of reagent. Extract into liexane and wash the hexane layer with 15% m/V potassium hydroxide in water.Gas chromatograph at 210 "C measuring the peak heights (N and H,) and the retention times ( T and T2) of the two peaks obtained. The response is then taken to be (HITl) + (N,T,). Quantitative perchlorination to a mixture of decachlorobiphenyl and an unknown compound (possibly bromononacliloro-biphenyl)12 confirms PCB.12 For calibration perchlorinate standard solutions of Aroclor 1254. Hexachloro benzene m/V potassium hydroxide in propanol. column. chlorophenyl propyl ether confirms hexach10robenzene.l~ procedure on hexachlorobenzene standards. Evaporate a second 5 ml of fraction 1 dissolve in 0.5 ml of pyridine and add 1 ml of 10%) Heat at 100 "C for 10 min under a micro-Snyder Formation of penta-For calibration carry out this Extract into hexane and wash the hexane layer with water.End rin Evaporate 4 ml of the chromium(V1) oxide cleaned fraction 2 add 3 ml of 50% V/V con-centrated hydrochloric acid in water shake and then heat at 100 "C for 15 min under a micro-Snyder column. Extract the product identified by Chau and Cochrane14 as the half cage &Ketone into hexane and wash the hexane layer with water. Dieldrin heptachlor-2,3-epoxide p p'- T D E Evaporate 4 ml of the chromium(V1) oxide cleaned fraction 2 dissolve into 1 ml of acetic acid and add 2 ml of concentrated hydrochloric acid. Heat at 100 "C for 15 min under a micro-Snyder column. Add another 1 ml of the hydrochloric acid and continue heating for a further 5 min. Extract into 2 ml of hexane and evaporate a 1.7-ml aliquot of this just to dryness.Add 5ml of the chromium(V1) oxide reagent and keep at 40°C for 15min. Extract again into hexane and wash the hexane layer with water. For calibration carry out this procedure on appropriate standards. Results and Discussion Extraction A procedure was required for the complete extraction of lipid-soluble organochlorines f rom owl tissues. Maceration of tissue in acetone followed by filtration and partition from aqueous acetone into hexane,15 gave low and variable recoveries of organochlorines from owl tissues. As an alternative to maceration ultrasonic disintegration was found to be too slow and centrifugation offered no advantage over filtration. Complete extraction was achieved by the Soxhlet procedure described but this also extracted large amounts of lipid.I t was decided to adopt this procedure and to investigate clean-up methods for removing the extracted lipid. Clean-up Removal of lipid The partition method of Maunder et aZ.16 gave low recoveries of organochlorines added to owl tissue extracts and also produced emulsions and introduced interferents. The removal of lipid with oleum - sulphuric acid mixture as described by Schechter et a1.I' gave problems with emulsions and interferents and destroyed dieldrin and lieptachlor-2,3-epoxide and was therefore unsuitable. Satisfactory clean-up was obtained both with the alumina column method of Greve and Grevenstukls and that of Telling et ~ 1 . ~ The latter method was slowe 40 TRIM et d. GC PROCEDURE FOR ORGANOCHLORINE RESIDUES Analyst VOZ 108 than the former yet had a greater fat capacity and was procedurally more robust.Re-coveries with the Telling et aZ. method were generally greater than 90% and more than 90% of the co-extracted lipid was eliminated. I t was therefore selected as part of the analytical scheme. Removal of interferents Gas chromatography of tlie alumina cleaned extract was complicated by the presence of PCB (Fig. 1). Several procedures have been devised to separate PCB from organochlorine pesticide^.^^^^^^^^-^^ The method described by Holden and J4arsden1° but using silica gel of the nearest particle size (as that quotedlO was no longer available) was found to be both rapid and reproducible with recoveries normally greater than 80%. Each of the organochlorine compounds other than $$’-DDT was eluted in either fraction 1 or fraction 2 .More than 80% of jb$’-DDT was eluted in fraction 1 the remainder in fraction 2. The method cannot separate $p’-DDT $$’-DDE or hexachloro-benzene from PCR as these all elute in fraction 1 (Fig. 2) but application of the confirmation techniques described denionstrated that in the owl samples investigated interference between PCR and these three compounds was not a problem and so this method proved adequate for our needs. Accurate low-level measurement of heptachlor-2,3-epoxide and dieldrin had previously been very uncertain because interference from compounds causing negative peaks prevented the accurate measurement of residues in fraction 2 (Fig. 3). This interference was removed by tlie chromium(V1) oxide procedure with complete recovery of heptachlor-2,3-epoxide dieldrin and pp’-DDT.If applied to fraction 1 the same procedure gives good recoveries of PCR while eliminating $p’-DDE. The conditions employed are milder than those described by Szelewski et n1.,25 which caused significant losscs of PCB. A secondary clean-up was therefore required. Measurement The lower limits of measurement of residues by gas chromatography using the analytical scheme described are shown in Table I. These represent amounts giving at least 2.50/ full-scale deflection on the chart recorder and are expressed as mg kg-l without correction for losses during the procedure. The low limits of measurement particularly for hexachloro-benzene liept achlor-2,3-epoxide PP’-DDE and dieldrin are particularly valuable for moni-toring environmental residues.The recovery results (Table I) were obtained using 5-g samples of muscle or liver spiked with 100 pl (five punctures from a 250-pl syringe) of hexane containing a mixed standard to give ca. five times the limit of measurement in tissues 2 h at room temperature prior to extraction. The recoveries were consistent between samples and for these small residues the recoveries are considered to be satisfactory. Confirmation of Residues Gas-chromatographic retention time alone is not accepted as identification of a residue even if tlie compound of interest has already been selected by two stages of adsorption-liquid cliromatograpli ? and by treatment with chromium (17) oxide reagent.Therefore identification was based upon characteristic changes in gas-chromatographic retention times following appropriate chemical treatments. Both qualitative and quantitative aspects of the cferivatisations were examined ; a combination of retention time retention time change on derivatisation and response factor (z.(?. response of the derivative divided by the corre-sponding mass of the parent compound) provides proof of identity. For the confirmation of P$‘-DDE the nietliod of Collins et aZ.ll using chromium(V1) oxide was modified to increase the yield of $P’-dichlorobenzoplienone. Conversion was 75% within 5 min of adding water but was complete after 3 d at room temperature. For $$’-DDT the method of Hamence et aZ.8 was found to give incomplete conversion to $$’-DDE but by increasing the temperature and concentration of the alkali better recovery was obtained.The method adopted for confirmation of PCB is a combination of the methods of Berg et aZ.20 and of Crist and 1Ioseman.l2 To minimise interferences and to reduce the risk of explosion, as much of the hexane as possible is removed before perchlorination and the chloroform is omitted. Florisil column clean-up12 after the perchlorination was found to be unnecessary but a potassium hydroxide wash was required before gas chromatography and this replaces the wash by sodium hydrogen carbonate.12 A recovery of 92:; was obtained from cleaned u Janzmry 1983 IN AVIAN TISSUES WITH CHEMICAL DERIVATISATION 41 owl-liver extract spiked to 380 ng ml-1 with PCR. The method of Collins ct al.13 was found to be suitable for confirming hexachlorobenzene but pentacliloroplienol gives the same product.Wienke and Burkez6 reacted endrin with a mixture of hydrochloric acid and zinc chloride. In our procedure the zinc chloride was dispensed with as it caused too much interference and appeared to be unnecessary. A product having the same retention time as the ketone describedz6 was obtained using sulphuric acidI4 or an aged gas-chromatographic column. Hydrochloric acid was chosen as an alternative to sulpliuric acid as it was to be used for other confirmatory tests. Dieldrin and heptachlor-2,3-epoxide react with hydrochloric acid probably giving chlorohydrins.26 These were obtained in greater and purer yields (as determined by gas chromatography) than the hromohydrins or bromoacetate~~~ obtained by Hamence et al.* Dissolving the sample in acetic acid was found to increase the apparent yield of chlorohydrin.Chromium(V1) oxide was tried as a clean-up for the products of the hydrochloric acid reaction. I t was found that the retention times changed and the peaks became larger and sharper which improved the confirmation. The clean-up was then modified to enhance this effect. The identities of the products are unknown but may be similar to the ketone reported by Cochrane,28 although acetylation has not been ruled out. @$’-TDE also forms an unknown product when reacted with chromium(1’1) oxide reagent. In this instance the initial reaction with hydrochloric acid is unnecessary but is convenient if heptachlor-2,3-epoxide dieldrin and $P’-TDE are to be confirmed together.Keactioii of j!@’-TDE with chromium(V1) oxide reagent is temperature dependent and moisture sensitive. Fig. 4 illustrates some of the confirmatory procedures described. In this maniple PCB did not occur. The silica gel separation has been omitted in order to show P)’-DDE being confirmed along with dieldrin and $$’-TDE. With the exception of PCB (where the limit is 0.5 mg kg-1) and of alpha-HCH and gamma-HCH for which a suitable method has not been found the chemical derivatisations allow the confirmation of residues below 0.1 nig kg-l (Table 11). E E --v) P 60-a 120r 11 9 90 I ’L I I 1 I I I I I 60 48 36 24 12 0 60 48 36 24 12 0 Ti me/m in Fig. 4. Gas chromatogram of an alumina-cleaned muscle extract from a typical barn owl showing some of the chan;;es in retention used in residue confirmation.(a) After reaction with hydrochloric acid - acetic acid followed by chromiuni(\.’l) oxide in acetic acid 9 derivative of pp’-DDE ; 10 derivative of p$’-TDE ; and 11 derivative of dieldrin. ( b ) Before reaction 5 pp’-DDE; 6 dieldrin (HEOD): and 7 pp’-TDE.’ Conclusions The analysis of environmental background levels of organochlorine pesticides in avian predator tissue samples is made difficult by the presence of lipid and other interferents in the extracts. The analytical scheme described is suitable for the analysis of ten individual organochlorine compounds together with polychlorinated biphenyl mixtures at low levels in avian tissue samples. The procedure is robust and self-checking and only solutions whic 42 TRIM BROWN BUNYAN ODAM AND STANLEY TABLE I1 CONFIRMATION OF RESIDUE IDEXTITIES Lowest confirmable level in cleaned-up owl sample f--?- Relative retention Concentration Approximate Approximate of derivative in solution/ wet-mass/ “response factor”*/ Compound (pp’-DDE = 1) ng ml-l mg kg-l mm pg-l Hexachlorobenzene .. . . 0.36 1.5 0.007 24 alpha-HCH . . . . . . gamma-HCH . . . . . . pp’-DDE . . . . . . Aldrin . . . . . . . . Heptachlor-2,3-epoxide . . Dieldrin. . . . . . . . Endrin . . . . . . . . pp’-TDE . . . . . pp’-DDT . . . . . . PCB . . . . . . . . ---1.10 0.82 2.87 5.50 - 7 -- -- - I 5 0.02 13 0.07 11 0.06 2.38 19 0.09 1.00 11 0.06 n.d.t 100 0.5 Not fully investigated 0.48 1.2 0.68 Not fully investigated 0.25 1.8 n.d.tf * This is the peak height given by the derivative per equivalent unit mass of parent compound.t n.d. = not determined. The response is calculated from two peaks. are very similar to standards free of co-extractives are compared with such standards for the purposes of measurement. The scope of this procedure together with its sensitivity offers considerable advantages over existing analytical methods for organochlorine compounds. The analytical scheme has been successfully applied to owls collected in the UK during the 1970s and the results of this environmental monitoring are being published elsewhere. 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. References “European Colloquium Problems Raised by the Contamination of Man and his Environment by Persistent Pesticides and Organo-halogenated Compounds (EUR 5196),” Commission of the European Communities Luxembourg 1975. Robinson J . Annu. Rev. Pharmacol. 1970 10 353. Ratcliffe D. A. J . Appl. Ecol. 1970 7 67. Stanley P. I. and Elliott G. R. Agro-Ecosystems 1976 2 223. Glue D. E. Ornis Scand. 1971 4 97. Hardy A. R. Hirons G. J. M. and Stanley P. I. in “Recent Advances in the Study of Raptor Worthing C. R. Editor “The Pesticide Manual,” Sixth Edition British Crop Protection Council, Hamence J . H. Hall P. S. and Caverly D. J. Analyst 1965 90 649. Telling G. AT. Sissons D.J . and Brinkman H. W. J . Chromatogr. 1977 137 405. Holden A. V. and Marsden K. J. J . Chromatogr. 1969 44 481. Collins G. B. Holmes D. C. and Jackson F. J. J . Chromatogr. 1972 71 443. Crist H. L. and Moseman R. F. J . Assoc. Off. Anal. Chem. 1977 60 1277. Collins G. B. Holmes D. C. and Wallen M. J . Chromatogr. 1972 69 198. Chau A. S. Y. and Cochrane W. P. J . Assoc. Off. Anal. Chem. 1969 52 1220. Taylor A. Rea R. E. and Kirby D. R. Analyst 1964 89 ,497. Maunder M. J. de F. Egan H. Godly E. W. Hammond E. W. Roburn J. and Thomson J., Schechtcr ILI. S. Milton A. and Haller H. L. I n d . Eng. Chem. Anal. Ed. 1947 19 51. Creve P. A, and Grevenstuk W. B. F. Meded Rzjksfac. Landbouwwet. Gent 1975 40 1115. Bacaloni A Goretti G. Lagana A. and Petronio B M. J . Chromatogv. 1979 175 169. Berg 0. W. Diosaday P. L. and Rees G. A. V. Bull. Envzron. Contam. Toxicol. 1912 7 338. Cooke XI. Nickless G. and Roberts D. J . J . Chromatogr. 1980 187 47. Kveseth N. J . and Brevik E. M. Bull. Envzron. Contam. Toxicol. 1979 21 213. Picer M. and Ahel M I J . Chromatogr. 1978 150 119. Teichman J. Bevenue A. and Hylin J. W. J . Chromatogr. 1978 151 155. Szelewski M. J . Hill D. R. Spiegel S. J . and Tifft E. C. Jr. Anal. Chem. 1979 51 2405. Wiencke W. W. and Burke J . A, J . Assoc. Off. Anal. Chem. 1969 52 1277. Maybury K. B. and Cochrane W. P. J . Assoc. Off. Anal. Chem. 1973 56 36. Cochrane W. P. .J. Assoc. 08. Anal. Chem. 1969 52 1100. Musial C. J. Peach 31. E. and Stiles D. A. Bull. Environ. Contam. Toxzcol. 1976 16 98. Diseases,” Chiron Publications Keighley 1981 p. 159. Croydon 1979. Analyst 1964 89 168. Received M a y 26th 1982 Accepted August 19th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800033

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst January 1983 Vol. 108 $9. 43-52 43 Determination of Sulphide Produced by Desdfovibrio Species of Sulphate-reducing Bacteria 1. K. Al-Hitti G. J. Moody and J. D. R. Thomas Applied Chemistry Department Redwood Building U WIST Cardifl CF1 3N U Orion Model 94-16A ion-selective electrodes have been used in indirect and direct modes for determining the sulphide produced by various Desulfovi-brio species of sulphate-reducing bacteria namely D. desulfuricans D. gigas and D. vulgaris. The sulphide determined by both modes matched the amounts expected from the nutrient sulphate present in the culture media and also matched the sulphide determined gravimetrically (lead sulphide) for the indirect mode. All three species of bacteria also thrived on sulphite thiosulphate and metabisulphite metabolic intermediates but would not grow in the stable dithionate or in the absence of inorganic sulphur species.Indirect monitoring with the sulphide ion-selective electrode and with daily renewal of the trapping 1 M sodium hydroxide and 2% ascorbic acid solutions in the monitoring flask permits the resolution of lag growth and stationary and death phases of the various bacteria. Keywords Sulphate-reducing bacteria ; Desulfovibrio bacteria ; sulphide ion-selective electrode Anaerobic sulphate-reducing bacteria of the Desulfovibrio species are widely distributed and cause many corrosion problems.ls2 The major product of their action is hydrogen sulphide. The problems created by the bacteria require methods of monitoring their presence and growth .Desulfovibrio species derive their energy from the anaerobic reduction of sulphate accom-panied by simultaneous oxidation of organic substances. The general equation of such metabolic reductions of sulphate may be represented as follows3 : n“C” + MSO + H20 + M(Ac) + C02 + H2S . . where “C” is the organic substrate and M a metal. Therefore the liberation of hydrogen sulphide is an indication of the growth of these organisms. A test of the American Society for Testing and Materials (ASTM)4 is frequently used to monitor the growth of sulphate-reducing bacteria. The sulphide ions produced react with the iron(I1) salt of the culture to give black iron(I1) sulphide. A referee method of the ASTM test uses an iodimetric determination of hydrogen sulphide in the absence of any black precipitate.The test considers sulphate-reducing bacteria to be absent if the con-centration of hydrogen sulphide is less than 50 p.p.m. M sulphide) are well within the range of a sulphide ion-selective electrode Earlier studiess have shown that the sulphide ion-selective electrode gives an earlier indication of the growth of sulphate-reducing bacteria than the ASTM method. Also Nedwell and Baratg have used an Orion 94-16A sulphide ion-selective electrode to measure sulphide in slurries removed from estuarine marshland with sterilised syringes. The object of this investigation was to extend the applicability of the sulphide electrode for detecting the growth of sulphate-reducing bacteria. Three species of Desulfovibrio have been studied using various inorganic sulphur sources as growth energy suppliers.Hydrogen sulphide levels of 50 p.p.m. (1.5 x Experimental Refrigerated freeze-dried cultures obtained from the National Collect ion of Industrial Bacteria Aberdeen (NCIB) were used namely Desulfovibrio vulgaris Strain No. NCIB 8457, 1975; Desulfovibrio desulfuricans Strain No. NCIB 8307 1979; and Desulfovibrio gigas Strain No. NCIB 9332 1979 44 AL-HITTI et aj. DETERMINATION OF SULPHIDE FROM Analyst VOZ. 108 Preparation of the Starter Culture of Desulfovibrio Species Cultures were revivedlO in a slightly modified version of Postgate's medium for sulphate reducers,l1,l2 with the following composition K,HYO, 0.25 g ; NH4C1 0.50 g ; yeast extract, 0.50 g ; MgS0,.7H20 1.00 g ; Na,S04 (anhydrous) 0.50 g; CaC1,.2H20 0.05 g ; sodium lactate (70% m/m solution) 2.50 cm3; and doubly de-ionised water 500 cm3.The pH of the medium was adjusted to 7.40 with 5 hi sodium hydroxide solution and autoclaved at 125 "C for 30 min. After cooling the preparation was completed by adding 0.50 g of sodium thioglycollate and 0.50 g of sodium ascorbate in order to lower the oxidation - reduction potential of the medium and thereby initiate growth of the sulphate-reducing bacteria.13 The pH was finally adjusted to 7.80 and autoclaved at 125 "C for 30 min. The medium was then rapidly cooled and anaerobicity was assured by vigorous bubbling of white-spot nitrogen for at least 30 min prior to inoculation. The basal medium was then inoculated with freeze-dried Desulfovibrio species. Air above the medium was flushed out with nitrogen.The sealed-container was incubated at 30 "C and bacterial growth became evident after 2-3 d. This medium called the starter culture was used as the source for sub-cultures. An aliquot of the starter culture (inoculum) was aseptically added to a fresh basal medium of appro-priate composition allowing for variations in inorganic sulphur species (Tables I and 11). In these instances the magnesium salt used was magnesium chloride (0.83 g) except for the runs with magnesium sulphate when no further magnesium salt was added. Experimental Assemblies for Following Growth of Desulfovibrio Bacteria prior to inoculating with the starter culture. Indirect method Hydrogen sulphide produced by the bacterial action was swept from the culture flask A, (Fig.1) with white-spot nitrogen into a monitoring flask B fitted with a sulphide and a double-junction reference electrode. Two modes were employed and the sub-culture flask was autoclaved in each instance Sterile cotton-wool Plasticine and 250 cm3 of NaOH solution (1 M) a t 30 "C A +5 g of ascorbic acid B Fig. 1. Indirect method of monitoring sulphide production by Desulfovibrio bacteria. Direct method The culture after sterilisation and pre-saturation with white-spot nitrogen was sealed in with a sulphide electrode double-junction reference electrode and combination-glass p January 1983 DESULFOVIBRIO SPECIES O F SULPHATE-REDUCING BACTERIA Plasticine and Parafilm seals Combination Double junction 500 cm3 of medium _-<- reference - - - - - eiectrode - - _ _ - at 30 "C - - _ _ Magnetic - _ stirrer bar 45 Fig.2. Apparatus for the direct method of monitoring sulphide production by Deulfovibrio bacteria. electrode (Fig. 2) and the medium was gently stirred. monitored as well as e.m.f. of the sulphide ion-selective electrode cell. During growth studies pH was Potentiometric Measurements A calibrated Orion Model 94-16A sulphide ion-selective electrode was used with a Corning, Model 476002 double-junction reference electrode with an outer 10% potassium nitrate filler solution. The pH measurements were made with a combination EGEN 4 pH (glass) electrode. The e.m.f. and pH measurements were recorded with a Beckman Model 4500, digital pH - millivolt meter in conjunction with a Servoscribe Model RE541 potentiometric recorder.Turbidimetric Measurements A Hilger and Watts Uvichem spectrophotometer was used at 350nm for turbidimetric monitoring of bacterial growth. Aliquots (2 cm3) of the culture medium were taken periodically and the transmittance was measured against a blank of pre-inoculated medium. Three variations were studied continuous bubbling of nitrogen through the medium and blank; gentle stirring of the culture and the blank; and no stirring of the culture and blank. Gravimetric Determination of Sulphide Produced by Cultures the monitoring flasks by indirect-mode type of experiments. flasks were charged with 0.5 M lead nitrate. at 110 "C and weighed. The sulphide produced was determined gravimetrically by precipitation of the sulphide in For these the monitoring The filtered lead sulphide precipitate was dried Results and Discussion Indirect Method Sztlphate in the basal medium Sweeping off hydrogen sulphide by a continuous stream of white-spot nitrogen into the monitoring flask B (Fig.1) was found to be effective as sulphide ions are hydrolysed to hydrogen sulphide at the pH of the growth medium i.e., and . . * * (2) S2- + H,O + HS- + OH- . . . . HS- + H20 + H,S + OH- . . . . ' * (3) Although the initial pH of the culture medium was 7.0-7.5 the pH eventually increased to approximately 8.5 as a result of removing acidic hydrogen sulphide and carbon dioxide, produced by the following mechanism 46 AL-HITTI et al. DETERMINATION OF SULPHIDE FROM 2CH,CHOHCOONa + Na,SO -+ 2CHJOONa + CO + Na2C0 + H2S + H,O 2CH,CHOHCOONa + MgSO -+ 2CH,COONa + CO + MgCO + H,S + H,O Analyst VoZ.108 (4) (5) The bacterial growth was monitored continuously by measuring the e.m.f. using a sulphide ion-selective electrode against the double-junction reference electrode. Sharp increasingly negative potential changes were recorded in the early stages after inoculation which was attributable to hydrogen sulphide being brought in by the starter culture. There was then little change until the e.m.f. again changed in a negative deviation. The e.m.f. values were converted into concentrations of sulphide from calibrations' of the sulphide electrode and concentrations of sulphide collected in the monitoring flask were plotted against time (Fig. 3). Fig. 3 (graph B) depicts the usual type of pattern that reflects the steps of bacterial growth.However for very fresh starter cultures the pause (step) period was not observed (Fig. 3 graph A). Time/h Fig. 3. Production of sulphide by D. vulgaris (indirect method). B, 0.564 g of Na,SO as inorganic sulphur nutrient and starter culture aged 73 d ; A 0.564 g of Na,SO as inorganic sulphur nutrient and starter culture aged 10 d ; and C no inorganic sulphur nutrient for starter culture aged 49 d. The growth of bacteria was signified by the increasingly negative e.m.f. readings which started after 2-3 d following the pause (step) period (depending on the age of starter culture). The reduction of sulphate to sulphide by sulphate-reducing bacteria was found to be stoicheiometrically complete [equations (4) and ( 5 ) ] and no sulphate turbidity was observed when the medium after the end of the growth was mixed with acidified barium chloride solution.The total amount of sulphide produced by D. desuZfuricans D. gigas and D. vulgaris detected potentiometrically by the sulphide electrode for various sulphur species in the medium is shown in Table I and compared with gravimetric determinations. Allowing for the difficult experimental conditions there is good agreement between the expected values and the sulphide levels as indicated by the sulphide ion-selective electrode and gravi-metric recoveries (Table I). To confirm that sulphide collected in the alkaline medium-monitoring flask originated from sulphate bacterial growth was examined in the absence of inorganic sulphur in the basal medium.The sharp increase in sulphide concentration in the early stages due to sulphide brought in by the inoculum was followed by much lower final sulphide levels (Fig. 3 graph C) than was the case in the presence of a sulphur source in the basal medium. Therefore, Desulfovibrio species do not prosper without an inorganic sulphur source. Inorganic Sulphur other than Sulphate in the Basal Medium organisms have been intensively investigated. Mechanisms of the metabolism of inorganic sulphur compounds by sulphate-reducing The first step shown in Scheme 1 involve January 1983 DESULFOVIBRIO SPECIES OF SULPHATE-REDUCING BACTERIA 47 the activation of sulphate by its reaction with adenosine triphosphate (ATP) in the presence of the enzyme ATP-sulphurylase to form adenosine phosphosulphate (APS) and pyro-p h o ~ p 1 i a t e .l ~ ~ ~ ~ The second iniportant step is the reduction of APS in the presence of APS-Enters cell so42- / bS042- + ATP APS + PPi -bpi I 2e __.+ SO3'- + AMP I i"' reductase where the products are sulphite and adenosine monophosphate (AMP). Sulphite is reduced to sulphide in the presence of the enzyme sulphite-redu~tase.~~,~~ This can be by a cyclic mechanism.ls Therefore sulphite dehydrates to metabisulphite which is reduced via an intermediate (dithionite ions S,O,,-) to give trithionate S30G2-. Trithionate is reduced to thiosulphate and regenerates some sulphite. Finally thiosulphate is reduced to give sulphide and regenerates more sulphite. The cyclic reaction was severely criticised by Chambers and Trudingers17 who used labelled 35S in their experiments and concluded that trithionate and thiosulphate are not intermediates in sulphite reduction and that reduction to sulphide takes place directly.They considered trithionate and thiosulphate to be by-products. To follow the above this study has centred on sulphite thiosulphate metabisulphite and dithionate as electron-acceptor species instead of sulphate because they can be metabolic intermediates in the reduction of sulphate. The masses of inorganic sulphur substrates were selected in order that the sulphide collected in the monitoring flask B (Fig. 1) would always be the same for comparable reductions. The organisms used were grown properly and the product of reduction of these inorganic sulphur species was sulphide (Table I and Fig.4). The reactions were found to be stoicheiometric and consistent with the following equations : 2CH3CHOHCOONa + Na,SO + 2CH,COONa + CO + Na,CO + H,S + H2 (6) 2CH3CHOHCOONa + Na,S,03 -f 2CH3COONa + CO + Na,CO + 2H,S (7) 3CH3CHOHCOONa + Na,S,O + 3CH3COONa + 2C0 + Xa,CO + 2H2S + H20 (8) Total sulphide production by D. gigas D. deszti@ricans and D. vdgaris in various media are summarised in Table I and the sulphide ion-selective electrode data are compared with gravimetric data 48 AL-HITTI et al. DETERMINATION OF SULPHIDE FROM TABLE I INDIRECT METHOD DATA OF TOTAL SULPHIDE PRODLJCTION BY DESCILFOVIUXIO Analyst VoZ. 108 SPECIES IN MEDIA CONTAINING VARIOUS INORGANIC SULPHUR (si) SOURCES 9- in monitoring flask/g -A- - Age of Sulphur source D.desulfuricans data-0.500 g Na,SO . . 0.100 g MgS0,.7H,O. . No inorganic . . . . 0.564 e Na.SO. . . 0.564 g Na,SO . . 0.500 g Na,SO . . 0.100 g MgSO4.TH,O D . gigas data-No inorganic . . . . 0.564 g Na,SO . . 0.500 g Na,SO, 0.500 g Na,S,0,.5H,O ‘ 0.377 g Na,S,O, 0.479 g Na,S,O,.SH,O’ 0.564 g Na,SO . . 0.564 g Na,SO . . 0.500 g Na,SO . . 0.100 g MgS04.7H,0 D . uulgaris data-No inorganic . . . . 0.564 g Na,SO, 0.500 g Na,S,O,.BH,O’ 0.500 g Na,SO. . . 0.377 Na;S,O, 0 . 4 ~ g N~,s,o,.~H,O’ 0.564 g of Na,SO . . 0.564 g Na,SO . . 0.564 g Na,SO . . Si* in medium/g . . 0.113 . . 0.01:3 . . 0.127 . . 0.127 . . 0.129 . . 0.12i . . 0.127 . . 0.127 . . 0.127 -.. 0.113 . . 0.013 . . 0.127 . . 0.127 . . 0.129 . . 0.127 . . 0.127 . . 0.127 . . 0.127 -. . 0.113 . . 0.013 . . 0.12i . . 0.129 . . 0.127 . . 0.12i . . 0.12i . . 0.12T . . 0.127 . . 0.127 -Sulphide electrode 0.108 0.012 4 x lo-& 0.128 0.133 0.152 0.112 0.003 --0.106 0.011 0.136 0.128 0.133 0.128 0.001 3 x 10-5 --0.112 0.012 0.006 0.128 0.160 0.124 0.124 ::E} --0.128 inoculuml Gravimetric d 0.105 43 0.016 11 None 22 0.131 11 0.127 3 9 0.113 121 0.152 143 None 1 S 6 0.118 4 0.137 35 0.108 24 0.011 150 None 161 0.131 28 0.141 25 0.142 29 0.145 24 None 45 0.124 133 0.111 14 10 0.011 42 -None 49 0.135 73 0.182 66 0.124 201 0.137 105 None u.119 133 0.115 2s 0.118 4 Remarks -Traces only of black precipitate in gravimetry --Traces only of black precipitate in gravimetry Sulpliidc collected in 2 M NaOH and precipitated as Sulphide collected as PbS in Pb(NO,), PbS with acidified reagent -Traces only of black precipitate in gravimetry --Traces only of black precipitate in gravimetry Sulphide collected in 2 M NaOH and precipitated as Sulphide collected as PbS in Pb(NO,), PbS with acidified reagent -Traces only of black precipitate i n gravimetry -Traces only of black precipitate in gravimetry Different ages of starter culture used Sulphide collected in 2 M NaOH and precipitated as PbS with acidified reagent Sulphide collected as PbS in Pb(NO,), Sodium thioglycollate omitted * Sulphur initially present in the medium In contrast to other inorganic sulphur sources all the Deszilfozdwio species failed to grow with dithionate in the basal medium and it is recalled that the high stability of this com-pound resists the actions of most reducing and oxidising agents.1g~20 The absence of a dithionatase enzyme may also contribute to the metabolic stability of ditliionate.Fig. 4 shows the production of sulphide by D . deszd&ricam with the various inorganic sulphur sources and emphasises the lack of growth by the total sulphide in the nionitoring flask remaining at approximately lo-* hi compared with that brought in by the inoculum. Bacterial-growth Monitoring by Turbidimetry They showed a steady decline in transmittance during the early stages (40-120 h) of growth.Thereafter, the transmittances remained steady at about l0-20%. However inadequate stirring or low nitrogen bubbling rates led to an increase in transmittance values to the 30-4076 range owing to clumping and settling. Clumping has long been regarded as a hindrance in bacterial growth studies21 but this and other disadvantages of the turbidimetric method i.e. the tediousness of frequent sampling are not encountered with sulphide ion-selective electrode monitoring. Furthermore the ion-selective electrode method described has the advantage of being in a closed system. Turbidinietric measurements were carried out for comparison purposes. Growth Phases of Desulfovibrio Species by Indirect Monitoring The sulphide ion-selective electrode was also used to determine the growth phases of the three Deszdjovibrio species by the indirect method over growth periods of 1 month.In this instance the concentration of sulphide collected was determined daily and for this th January 1983 DESULFOVIBRIO SPECIES OF SULPHATE-REDUCING BACTERIA 49 0 50 100 150 200 250 Time/h Fig. 4. Production of sulphide by D. desulfuricuns (indirect method). Sulphur nutrients used A sulphate; B sulphite; C thiosulphate; D meta-bisulphite; and E dithionate. solution of 1 M sodium hydroxide and 2Yq ascorbic acid in the monitoring flask was renewed daily. Ry way of illustration Fig. 5 depicts the growth phases of D. gigas resolved by these potentiometric studies with the sulphide electrode. The shape of the graph reflects the intervals of the growth phases Between runs there were some variations in the dimensions of the phases (probably due to different ages of the starter culture) but the profile shows the sequence of bacterial growth.1 0 - 2 z . -10-4 10-5 r Ih -10-6 10-7 Fig. 5 . Growth phases of D. gigas (indirect method) with different ages (circles 87 d ; and triangles 153 d) inocula. A Lag (pause) phase ; B logarithmic growth phase ; C stationary phase; and D, death phase. The first increase in sulphide concentration was due to sulphide brought in by the starter culture. The response proceeded either at a relatively constant level or there was a drop (as shown in Fig. 5) until the organisms entered their logarithmic phase. The period of the initial or lag phase (the period required for the initiation of extensive activity) was found to depend on the age of the starter culture.The lag phase lasted up to 4 d with D. gigas when the age of the starter culture was 87 d but was 8 d when the age was approximately doubled (153 d). Topely and Wilson22 also found that the lag phase of bacterial multiplication tends to be longer when the inoculum is supplied from old cultures. The same trend was obtained with D. desuZ&ricans and D. vdgaris 50 AL-HITTI et al. DETERMINATION OF SULPHIDE FROM Analyst VoZ. 108 By the time the micro-organisms have reached their maximum metabolic activity at the end of the lag phase they commence their multiplications with simultaneous production of an immense amount of hydrogen sulphide (logarithmic phase in Fig.5 ) . The logarithmic phase ends when the culture medium becomes so altered that it no longer provides the con-ditions necessary to maintain uniform rate of growth and the organisms enter the stationary phase. The rate of production of sulphide is now steady because the organisms cease to multiply as a result of depletion of sulphate which is the main energy supplier. As the sulphate becomes exhausted the organisms gradually reduce in number and finally enter their death phase until the medium becomes sterile. This conclusion is compatible with that of Alico and Liegey13 who found that the depletion of sulphate concentration is the major significant factor contributing to population decline. Hence the indirect potentiometric method offers an effective alternative reliable way for following the growth phases and could usefully replace the turbidimetric method.Direct Monitoring of Generated Sulphide The growth pattern of Desulfovibrio species monitored by this method is illustrated in Fig. 6 for D. vulgaris. The e.m.f. of the sulphide ion-selective electrode (with respect to the reference electrode) was approximately -300 mV prior to starter culture inoculation. This altered to -600 mV on inoculating the starter culture owing to the sulphide being carried over in the inoculum. The subsequent change of e.m.f. to about -470mV is due to its hydrolysis. The bacteria enter their logarithmic growth phase when the e.m.f. values become more negative (region K to L in Fig. 6). As the e.m.f. values become steady (L to M Fig.6) the total sulphur source has been reduced to sulphide. The steady state (L to M in Fig. 6) represents the stationary and death phases as sulphide concentration accuinulates during bacterial growth. The J to K response in Fig. 6 represents the lag phase discussed above. > -650 E . -0 9 a -550 8 0 I + a -500 -600 w 9) -C 2 7 n !?? 6 m al t -450 I I I 0 50 100 150 200 250 2 -4001 ui Timeih Fig. 6. method). Sulphide production and pH curves of D. vulgaris (direct The major difficulty of this monitoring method is the hydrolysis of sulphide in the pH range of the culture which is about 7.0 according to equations (2) and (3). The sulphide ion-selective electrode responds only to the activity of free sulphide ions.Therefore a glass (pH) electrode was immersed in the medium for experiments based on this direct method of monitoring in order to determine the total sulphide ion c~ncentration.~ When the first hydrolysis stage [equation (2)] is considered in mildly alkaline solution, sulphide activity as2- may be calculated from the relation a - (9) . . as2- = fs2- [S2-IT/ ( 1 + - ‘El) . . . . where K is the ionisation constant of HS- andf,z- the activity coefficient of the sulphide ion January 1983 DESULFOVIBRIO SPECIES OF SULPHATE-REDUCING BACTERIA 51 At lower pH most of the sulphide is present as H,S and its ionisation constant (K,) should also be considered in order to assess [s2-]T. From the resulting relation~hip,~~ which takes into consideration the effect of K and K, it is possible to calculate the total sulphide ion concentration at the steady state of e.m.f.measurements using the following relation : [ P I T = Fas2- . . . . . . (10) where F is a pH-dependent factor for converting sulphide ion activity into total sulphide concentration [S2-IT. In this work F corresponds to 30 “C and ionic strength of culture media. The main sulphur source used in the direct monitoring method was sulphate in which all three strains of bacteria thrived (Table 11). Several runs in the absence of an inorganic sulphur source confirmed the inability of the organisms to grow without this energy supplier (Table 11). The bacteria grow in other sulphur sources; for example D. desuyuricans were successfully grown in thiosulphate and D. vztlgaris in sulphite and metabisulphite (Table 11).TABLE I1 TOTAL SULPHIDE PRODUCTION BY DESULFOVIBRIO SPECIES FROM VARIOUS INORGANIC SULPHUR Si SOURCES (DIRECT METHOD) Sj* in 0.564 g Na,SO . . . . . . 0.127 0.100 g MgS0,.7H10 . . . . 0.013 No inorganic sulphur . . . . -Sulphur source medium/g 0.564 g Na,SO . . . . . . 0.127 0.100 g MgS0,.7H,O . . . . 0.013 No inorganic sulphur . . . . -0.564 g Na,SO . . . . . 0.127 0.100 g MgS0,.7H20 . . . . 0.013 0.500 g NalS,0,.5H,0 . . . . 0.129 0.500 g Na,SO . . . . . . 0.127 0.377 g NalSIO . . . . . . 0.127 * Sulphur initially present in the medium. No inorganic sulphur . . . . -S1- by Sz-electrode/g 0.123 0.013 0.000 I 0.134 0.010 0.002 0.136 0.010 0.002 0.122 0.123 0.122 Desulfovihrio species D.vulgaris D. vulgarbs D. vulgaris L). desulfuricans 11. desulfuricans D. desulfuricans I). gigas D . gigas D. gigas D . desulfuricans D. vulgaris D. vulgaris Age of inoculum/d Remarks 93 -42 49 Culture contains 0.50 g of sodium 11 -11 32 Culture contains 0.50 g of sodium 28 -152 -172 Culture contains 0.50 g of sodium 121 -254 -265 --thioglycollate. N o growth -thioglycollate. No growth thioglycollate. No growth Conclusion The sulphide ion-selective electrode provides a convenient way of determining and moni-toring sulphide produced by sulphate-reducing bacteria illustrated here by studies on three Desulfovibrio species namely D. desulfuricans D. gigas and D. vulgaris. The bacteria also thrive on certain other metabolic intermediates such as sulphite thiosulphate and meta-bisulphite.This wider activity rules out possible use of the bacteria as practical selective sensors for sulphate. The various studies show that it is possible to directly monitor growth and sulphide pro-duced by the bacteria with a sulphide ion-selective electrode. However in order to dis-tinguish between the “stationary” and “death phases’’ of the bacterial life-span it is necessary to have an indirect method whereby the sulphide-trapping liquid medium in the secondary or monitoring chamber is periodically renewed. Finally each of the sulphide ion-selective electrode methods described here is more quanti-tative than both the ASTM tests and turbidimetric methods of determining growth of sulphate-reducing bacteria.The turbidimetric method is especially prone to errors because of the clumping of bacterial colonies. The authors thank the University of Salah Al-Deen Iraq for paid leave of absence granted to 1.K.Al-H. References 1. 2. 3. 4. Roberts G. A. H. BY. Corros. J. 1969 4 318. Crombie D. J. Moody G. J. and Thomas J . D. R. Chew,. I n d . (London) 1980 500. Salle H. J. “Fundamental Principles of Bacteriology,” McGraw-Hill New York 1967 p. 329. American Society for Testing and Matcrials “Standard Methods of Test for Sulphate-reducing Bacteria in Industrial Water and Water-formed Deposits,” ASTM Philadelphia 1965 D993-58 52 AL-HITTI MOODY AND THOMAS Hseu T. M. and Rechnitz G. h. A n a l . Chem. 1968 40 1054. Light T. S. and Swartz J .L. A n a l . Lett. 1968 1 826. Crombie TI. J. Moody G. J. and Thomas J . I). Ti. Anal. Chiin. i l c t a 1975 80 1. Crombie D. J. Moody G. J. and Thomas J . D. K. Lab. Pract. 1980 29 259. Nedwell D. B. and Barat I. M. Microbiol. Ecol. 1981 7 305. Mackenzie A. R. “National Collection of Industrial Bacteria,” Aberdeen 1976 personal communi-Postgate J . R. in Schlegcl H. G. and Kroger E. Editors “Anreichungskulfur und Mutanteri-Pankhurst E. S. in Shapton D. A . and Board R. G. Editors “Isolation Anaerobes,” Academic Alico R. K. and Liegey F. W. J . Bacteriol. 1966 91 112. Peck H. D. Proc. Natl. Acad. Sci. USA 1959 45 701. Ishirnoto M. J . Biochem. (Tokyo) 1959 46 205. Siegel L. M. in Singer T. P. and Ondarzo I<. XI. Editors “Mechanism of Oxidizing Enzymes,’‘ Chambers L. A, and Trudingers P. A . J . Hacteriol. 1975 123 36. LeGall J. and Postgate J . I<. A d v . hficrobiol. Physiol. 1973 10 81. Wells A. F. “Structural Inorganic Chemistry,” Oxford University Press London 1975 p. 594. Schmidt M. and Siebert W. “Comprehensive lnorgariic Chemistry,” Volume 2 Pergamon Press, Jennison A t . W. .I. Bacteriol. 1937 33 461. Topely W. \V. C. and Wilson G. S. “Principles of Bacteriology and Immunity,” Fourth Edition, “Sulphide Ion Electrode Silver Ion Electrode (Model 94-16) ,” Instruction Manual Orion Research cation. auslese,” Fischer Stuttgart 1965 p. 190. Press London 1971 p. 223. Elsevier New York 1978 p. 201. Oxford 1973 p. 878. Arnold London 1955. Inc. Cambridge MA 1970. Keceived August 5th 1982 Accepted September 17th 1982 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23

ISSN:0003-2654    

DOI:10.1039/AN9830800043

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst Janzmry 1983 Vol. 108 $9. 53-57 Air-segmented Continuous-flow Visible S pect ro p hot o met r i c Deter m i nation of Cephalosporins in Drug Formulations by Alkaline Degradation to Hydrogen Sulphide and Formation of Methylene Blue and Determination of Sulphide-producing Impurities Including Cephalosporins in Penicillin Samples 63 Mohamed A. Abdalla and Arnold G. Fogg John G. Baber and Christopher Burgess Chemistify Department Loz.tghbovough University of Technology Lozighbovough Leicestevshive LE11 3T U Glaxo Opevations ( U K ) Ltd. Bamard Castle Co. Durham DL12 8DT A manual visible spectrophotometric method for the determination of cephalo-sporins by alkaline degradation to sulphide and formation of methylene blue has been adapted for use with an air-segmented AutoAnalyzer I system.The system has been tested for the determination of twelve ceplialosporins; rectilinear calibration graphs were obtained with good precision in the general range 8-80 pg ml-l of cephalosporin. The automated procedure was tested as a method of determining trace amounts of cephalosporins and other sulphide-producing impurities in penicillin G and V samples. The detection limit was calculated t o be 1-2 p g g-l of cephalosporin in penicillin samples. Keywords /3-Lactains ; alkaline degradation ; hydrogel.t sulphide ; wiethylene blue ; autonzatic analysis A manual visible spectrophotometric method for the determination of cephalosporins based on alkaline hydrolysis to hydrogen sulphide and formation of methylene blue has been reported previous1y.l ?2 All of the cephalosporins studied gave a reproducible yield of sulphide on degradation in 0.5 M sodium hydroxide solution in a boiling water-bath.Recom-mended hydrolysis times varied from 30 min to 1 h. The yield of sulphide depended on the cephalosporin used and varied from 14% for cefuroxime to S4.4Y0 for cephalexin. The alkaline solution of the degraded cephalosporin was treated with zinc acetate solution NN-dimethyl-p-phenylenediamine solution and iron( 111) solution to produce methylene blue from the sulphide by a cyclisation reaction; the methylene blue was then measured spectro-photometrically at 667 nm. Penicillins did not interfere when present at similar levels to the cephalosporins as they do not give hydrogen sulphide under these conditions.This method has now been automated using an air-segmented continuous-flow system and details of this system are reported here together with an extension of the method to the determination of cephalosporins in penicillin samples. The use of zinc acetate was found to be unnecessary and it was excluded in this automated procedure. Experimental Apparatus Air-segmented continuous-flow analysis was carried out with an AAI AutoAnalyzer mani-fold using a two-speed AutoAnalyzer peristaltic pump and a Newton Instruments automatic sampler. Visible spectrophotometric measurements were made with a Pye Unicam SP600 spectro-photometer fitted with a Hellma flow cell (80-11.1 capacity) and connected to a Pye Unicam AR55 recorder 54 ABDALLA et al. CONTINUOUS-FLOW SPECTROPHOTOMETRY OF Analyst VOZ.108 Reagents NN-Dimethyl-p-phenylenediamine sulphate solution 0.005 M. Dissolve 0.93 g of NN-di-methyl-p-phenylenediamine sulphate in 750 ml of distilled water. Add 186 ml of concentra-ted sulphuric acid cool and dilute to 1 1 with water. Ammonium iron(II1) sulphate solution 0.25 M. Dissolve 60.8 g of ammonium iron(II1) sulphate dodecahydrate in 0.5 M sulphuric acid and dilute to 500 ml with 0.5 M sulphuric acid. Dissolve 40 g of sodium hydroxide in water add 100 ml of Triton X-100 and dilute to 1 1 with water. Prepare as above but using only 20 g of sodium hydroxide. Sodium hydroxide solution 1 M with detergent. Sodium hydroxide solution 0.5 M with detergent. Air (0.8) 23T 7T Re-sample A U n Procedure At the high temperatures used it was found necessary to re-sample after the hydrolysis step (at de-bubbler 1) and to include a glass bulb as a pulse suppressor in the air line before the water-bath in order to counteract the pulsing that occurred in the system when used at 80 "C the temperature finally adopted.Cephalosporins were dissolved in water with the exception of 7-aminodeacetoxycephalo-sporanic acid (7-ADCA) and 7-aminocephalosporanic acid (7-ACA) which were dissolved in 2 rnl of 0.5 M sodium hydroxide solution and then diluted rapidly before extensive hydrolysis could occur. Stock standard solutions of cephaloglycin cefuroxime cephoxazole cephalon-ium cephaloridine cephalothin cefaclor and cefazolin were prepared at 0.02% m/V those of 7-ADCA and 7-ACA at 0.01% m/V and those of cephalexin and cephradine at 0.008~o m/V concentrations.Calibration graphs were obtained by presenting to the analyser solutions prepared by diluting 5 10 15 and 20 ml of these stock solutions to 50 ml. To obtain calibration graphs in the presence of high concentrations of penicillin G and penicillin V (sodium salts) 5-g amounts of the penicillin sample were dissolved in water, aliquots of standard cephalosporin solution were added and the solutions were diluted to 25 ml in calibrated flasks. Triton X-100 (10% m/V) was incorporated in the 1 M sodium hydroxide solution and the 0.5 M sodium hydroxide wash solution. A schematic diagram of the analysis manifold finally adopted is shown in Fig. 1. These solutions were presented to the AutoAnalyzer system.(0.8) I " -AFS (0.32) Flow cell (1.6) Waste Waste 80 PI 667 nm Fig. 1. Recommended AutoAnalyzer manifold. Sample time 36 s ; wash time 72 s. Numbers in parentheses are the flow-rates in millilitres per minute January 1983 CEPHALOSPORINS BY DEGRADATION AND METHYLENE BLUE FORMATION 55 Results An AutoAnalyzer I system for carrying out the colour reaction on pre-hydrolysed cepha-lexin solutions was shown to give good rectilinear calibration graphs. When the delay coil contained in a water-bath was incorporated in the system in order to effect the hydrolysis step on-line the optimum temperature of the water-bath was found to be 80 "C although unsatisfactory pulsing was observed in the system when temperatures above 70 "C were used. This problem was overcome by incorporating a small glass bulb (to act as a pulse suppressor) in the air line segmenting the sample solutions before the hydrolysis stage and by de-bubbling and re-sampling the hydrolysed sample solution automatically after it had passed through the water-bath.Data obtained for calibration graphs for cephalexin at various water-bath temperatures are given in Table I. The calibration graphs show good rectilinearity. Problems of repro-ducibility were experienced at 95 "C and it was necessary to replace water lost by evaporation from the water-bath. The latter problem could have been overcome by using silicone oil in the bath but for these reasons 80 "C was adopted for routine use. Samples remain in the water-bath for about 30 min. Previously several cephalosporins were shown to give their maximum yield of sulphide after 30 min in a boiling water-bath but others require hydro-lysis for up to 50 min at this temperature.Clearly maximum formation of sulphide is not achieved in the present AutoAnalyzer system. The minimum wash time was determined by the usual method of observing the time to reach a steady state on changing from the standard solution of lowest concentration to that of highest concentration and vice versa. From this it was clear that over 60 samples per hour could be run even without curve regeneration but all of the present work was carried out at a rate of 30 samples per hour. The wash time was twice the sample time. Rectilinearity was achieved when water was used as the wash liquid but a steadier base line was achieved by using 0.5 M sodium hydroxide solution as wash liquid without loss of rectilinearity and this was adopted.The system was cleaned after use by passing a 10% V/V solution of Triton X-100 through all the liquid channels. Typical signals obtained with cephalexin solutions are shown in Fig. 2. The recommended system is shown in Fig. 1. TABLE I EFFECT OF TEMPERATURE OF HYDROLYSIS WATER-BATH ON THE CALIBRATIOX GRAPH FOR CEPHALEXIN Concentration of cephalexin in Absorbance presented solution/ r- A \ pg ml-I 50 "C 60 "C 70 "C 80 "C 90 "C 10 0.075 0.147 0.155 0.199 0.199 20 0.147 0.259 0.304 0.366 0.366 30 0.208 0.387 0.420 0.509 0.509 40 0.299 0.497 0.588 0.699 0.700 Data for calibration graphs for the determination of 12 cephalosporins are given in Table 11 together with results obtained on passing a standard sulphide solution through the system.The calibration graphs show good rectilinearity with low coefficients of variation. Detection limits (twice the standard deviation above the blank) were in the range 0.3-0.9 pg ml-I of cephalosporin in the solutions presented to the analyser. Results for the determination of three cephalosporins in penicillin G and V samples are given in Table 111. Firstly penicillin G and V at the high concentrations used (0.2 mg ml-l in the solutions presented) give appreciable blanks (equivalent to about 0.25 pg ml-l of cephalosporin) owing to the slight colour of the degrada-tion products. The size of this blank was readily determined by replacing the NN-dimethyl-p-phenylenediamine sulphate solution with a sulphuric acid solution of the same acidity.Maintaining the same acidity is important as the colour decreases with increasing acidity. Secondly rectilinear calibration graphs with low coefficients of variation were obtained and detection limits for the determination of the cephalosporins in samples of penicillin were calculated to be in the range 1.25-2 ,ug g-l. Comparison of the slopes of the calibration graphs for particular cephalosporins in Tables I1 and I11 clearly indicates that the high con-centration of penicillin affects the yield of sulphide obtained from the cephalosporins. Several points emerged from this study 56 ABDALLA et al. CONTINUOUS-FLOW SPECTROPHOTOMETRY OF Analyst Vol. 108 60 I 100 I +- Time Fig.2. Signals obtained for calibration with standard cephalexin solutions. Cephalexin con-centration of presented solutions A 8 ; B 16; C 24; D 32 pg ml-l. Results with standard sulphide solutions indicate a slight decrease (48% instead of 520,;) in the yield of methylene blue from sulphide in these solutions. Appreciably increased over-all yields of methylene blue are obtained from cephaloridine and cephalothin in the presence of high concentrations of penicillin G or V whereas for cephalexin there is a decreased over-all yield in the presence of penicillin G but not penicillin V. When a 350 pg ml-l solution of ampicillin was presented no methylene blue was formed. Insufficient ampicillin was available to study the determination of cephalosporins in ampi-cillin but there should be no difficulty in doing this.TABLE I1 CALIBRATION GRAPHS FOR DETERMIXATION ON CEPHALOSPORINS IN STANDARD SOLUTIONS Concentration range Slope of presented in sample calibration graph/ Coefficient of Detection limit/ Cephalosporin cups/pg ml-1 ml pg-l x variation,* pg nil-’ Sodium sulphide . . . . Cephalexin . . . . Cephradine . . . . 7-ACA . . . . . . 7-ADCA . . . . . . Cephaloglycin . . . . Cefuroxime . . . . Cephoxazole . . . . Cephaloridine . . . . Cephalothin . . . . Cephalonium . . . . Cefaclor . . . . . . Cefazolin . . . . . . 8-32 8-32 8-32 10-40 10-40 20-80 20-80 20-80 20-80 20-80 20-80 20-80 20-80 10.0 7.0 6.8 2.1 3.5 3.6 2.0 3.8 1.1 2.3 0.89 3.0 7.0 1.01 0.41 0.73 0.62 0.66 0.66 0.72 0.71 0.63 0.81 0.42 0.45 0.51 0.3 0.3 0.3 0.5 0.4 0.5 0.8 0.8 0.9 0.7 0.5 0.4 0.6 * Based on calibration graph of four points plus blank.Discussion The recommended automated procedure is highly satisfactory for the determination of cephalosporins. The method is not stability indicating as several cephalosporin degradation products are known to degrade further to sulpl~ide.~ In this respect the method is inferior to the imidazole method4 for cephalosporins and penicillins which is stability indicating owing to the direct reaction of the irnidazole with the /I-lactam ring. The present method, however is selective for cephalosporins in the presence of penicillins uses the visible rat her than the ultraviolet spectrum and does not require the use of niercury(I1).In dilute aqueous solutions both the manual and automatic procedures have been show Janzmy 1983 CEPHALOSPORINS BY DEGRADATION ANT) METHYLENE BLUE FORMATIOX 57 TABLE I11 CALIBRATION GRAPHS FOR DETERMINATIOK OF CEPHALOSPORINS IN PENICILLIX G AND v SAMPLES Equivalent con-centration range in penicillin/ Sample Cephalosporin Pg K1 I Penicillin G . . . . Sodium sulphide Cephalexin Cephaloridine Cephalot hin Cephalexin Cephaloridine Cephalothin Penicillin V . . . . Sodium sulphide 0.5-10 0.5-10 1-20 1-20 0.5-10 0.5-10 1-20 1-20 Slope of calibration graph*/ 9.2 3.6 3.4 3.1 9.2 7.0 4.0 5.1 111 pg-1 x 10-Coefficient of -3 variation,t :h 1.02 0.67 0.70 0.68 1.03 0.52 0.70 0.68 Detection limit/ 1.25 1.25 1.5 1.5 1.25 1.25 2.0 2.0 PLg g-l * 'L'he absorbance of the blank was 0.0130 (penicillin G) and 0.0132 (penicillin V).7 Based on calibration graph of four points plus blank. to be highly reproducible despite the fact that the molar yield of sulphide from a cephalo-sporin will be between 14 and 65:/ (manual procedure) and that the yield of methylcne blue from sulphide is only 5270.192 A reduction in this yield of methylene blue from sulphide from 52% to 48% was observed in the automatic procedure in the presence of high concentra-tions of penicillins. On the other hand increased over-all yields of methylene blue were observed in the presence of high concentrations of penicillins for cephaloridine and cepha-lothin.The yield of hydrogen sulphide from cephalexin is known to be highly dependent on pH395y6 and more significantly degradation routes depend markedly on other matrix effects such as buffer constituents and degredand c~ncentration.~,~ Thus the observation of matrix effects in the recommended procedure is not unexpected. The precision and day-to-day reproducibility however have been shown to be excellent for the determination of cephalo-sporins in penicillin samples. Clearly a standard additions method must be used in develop-ing any limit test and the blank due to the colour of penicillin degradation products must be determined and subtracted The method will not distinguish between cephalosporins and other sulphide-producing degradation products or impurities in the penicillin samples.I t should be possible to distinguish trace amounts of inorganic sulphide from cephalo-sporins and other sulphide-producing impurities if this were felt to be necessary by flushing out a weakly acidic solution of the penicillin sample with nitrogen before presenting the solution to the AutoAnalyzer system. The procedure and system used here are capable of providing a limit test for 1-2 pg g-l of cephalosporin and other sulphide-producing impurities in penicillin samples and should be particularly useful in detecting cross-contamination of penicillin samples. By using an even more concentrated solution of the penicillin this detection limit can be reduced slightly. The use of a modified system incorporating a double-beam spectrophotorneter would allow the automatic subtraction of the blank due to the colour of the penicillin degradation products. The detection limit might well be reduced by the use of a more modern spectrophotometer. A sample of cephradine was kindly provided by E. R. Squibb and Sons Ltd. a sample of ampicillin by Beecham Pharmaceuticals samples of cephaloglycin cefaclor and cefazolin by Lilly Research Centre Ltd. and samples of the other cephalosporins and penicillins by Glaxo Operations (UK) Ltd. One of us (MAA) thanks The British Council for financial support and the University of Khartoum for leave of absence. Assistance with the initial develop-ment of the system from G. R. Lloyd is gratefully acknowledged. References 1 . 2. 3. 4. 5. 6. Abdalla M. A. Fogg A. G. and Burgess C. Analyst 1982 107 213. Fogg A. G. Abdalla M. A. and Henriques H. P. Analyst 1982 107 449. Fogg A. G. Fayad N. M. and Burgess C. Anal. Chim. Acta 1979 110 107. Bundgaard H. J . Pharm. Pharmacol. 1972 24 790. Bundgaard H. Arch. Pharm. Chem. Sci. Ed. 1976 4 25. Bundgaard H. Arch. Pharm. Chem. Sci. Ed. 1977 5 149. Received June 18th 1982 Accepted August 20th 188

ISSN:0003-2654    

DOI:10.1039/AN9830800053

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

58 Analyst January 1983 Vol. 108 $9. 58-63 Atom ic-a bso r pti on Deter m i nation of M ercu ry i n Geological Materials by Flame and Carbon-rod Atomisation After Solvent Extraction and Using Co-extracted Silver as a Matrix Modifier Richard F. Sanzolone and T. T. Chao United States Department of the Interior Geological Survey Box 25046 Denver CO 80225 USA Based on modifications and expansion of the original Tindall’s solvent extraction flame atomic-absorption procedure an atomic-absorption spectro-photometric method has been developed for the determination of mercury in geological materials. The sample is digested with nitric and hydrochloric acids in a boiling water-bath. The solution is made ammoniacal and potassium iodide and silver nitrate are added. The mercury is extracted into isobutyl methyl ketone as the tetraiodomercurate( 11).hclded silver is co-extracted with mercury and serves as a matrix modifier in the carbon-rod atomiser. The mercury in the isobutyl methyl ketone extract may be determined by either the flame- or the carbon-rod atomisation method depending on the con-centration level. The limits of determination are 0.05-10 p.p.m. of mercury for the carbon-rod atomisation and 1-200 p.p.tn. of mercury for the flame atomisation. Mercury values for reference samples obtained by replicate analyses are in good agreement with those reported by other workers with relative standard deviations ranging from 2.3 to 6.90/. Recoveries of mercury spiked a t two levels were 93-1060/,. Major and trace elements commonly found in geo-logical materials do not interfere.Keywovds IWevcuvy determination ; geochemical analysis ; carbon-vod atomic-absorption spectvoplaotovnetvy ; matrix modijier ; solvent extraction Primary and secondary dispersion of mercury that results in leakage halos may have great potential for the discovery of deeply buried deposits. Geochemical prospecting based on the analysis of near-surface soils and rocks for trace amounts of mercury may reveal an aureole adjoining or overlying buried mineral occurrences.1,2 The close association of mercury with precious metals and with the elements of volcanogenic deposits makes it a useful pathfinder element for gold silver antimony and massive sulphide deposits especially those containing lead and zinc.3 Numerous methods have been reported in the literature for the determination of mercury.Kokot4 and Ure5 have comprehensively reviewed and discussed various procedures for mercury determination with applications in geochemical analysis. The popular cold-vapour atomic-absorption methods are in principal simple and sensitive but in practice the accurate determination of mercury in natural samples is diffic~lt.~ Such methods may either entail elaborate procedures to separate and trap the mercury from the sam.ple suffer from inter-ferences caused by the diverse geological matrices especially substances that absorb ultra-violet light,2 or the methods are applicable only to a specific narrow range of concentrations of mercury. Therefore a rapid sensitive and interference-free method capable of deter-mining mercury throughout the wide range of concentrations found in geological materials would be a valuable tool for geochemical prospecting.The method proposed here is based on the original Tindall’s solvent extraction flame atomic-absorption procedure,6 but with significant modifications and expansion. The changes con-sist of precise defining of experimental conditions and interferences; reduction of the amount of some reagents used; extension to carbon-rod atomisation ; and use of the co-extracted silver as a matrix modifier. These changes have resulted in an atomic-absorption method using flame and/or carbon-rod atomisation that is capable of determining mercury from the crustal abundance level to the level of ore-grade concentrates SANZOLONE AND CHAO Experimental* 59 Apparatus A Varian AA-6 atomic-absorption spectropliotometer equipped with a Model 63 carbon-rod atomiser a simultaneous background corrector a mercury hollow-cathode lamp a Model 53 automatic sample dispenser and an automatic gas control unit was used.The instrument settings were as follows wavelength 253.7 nm; lamp 2.5 mA; and slit width 0.5 nm. A fuel lean air - acetylene flame was used with the oxidant flow meter set at 4.5 and the fuel flow meter set at 0.2. Conditions for the carbon-rod atomisation were as follows drying voltage, 6.5 (about 100 "C) for 25 s ; ashing voltage 3.0 (about 150 "C) for 5 s; ramp atomise voltage, 5.0 (about 1200 "C) at rate 3.0 (voltages are arbitrary settings not volts). The nitrogen flow meter was set at 8.0 and cooling time A was used for the automatic sample dispenser.All absorbance measurements were made using the peak-height mode. Reagents to 100ml. acid and store in an amber-glass container. before use, Potassium iodide solution 20y0 m/V. Silver nitrate stock solution 1 .O M. Silver nitrate solution 0.01 M. Concentrated hydrochloric acid. Concedrated nitric acid. Concentrated ammonia solution sp.gr. 0.88. Isobutyl inethyl ketone. Mercury- free solid material. Dissolve 20 g of potassium iodide in water and dilute Dissolve 8.49 g of silver nitrate in 50 ml of 1 yo V / V nitric Dilute the silver nitrate stock solution with water immediately Medium-textured soil or sediment ground to the same fineness as the sample and heated at 800 "C for 4 h in a muffle furnace to drive off any trace amounts of mercury.Dissolve 1.353 g of mercury(I1) chloride in 1000 ml of 1% VjV hydrochloric acid. Dilute the 1000 pg ml-l mercury stock solution with 1% V/V hydrochloric acid. Mercury stock solution 1000 pg ml-1. Mercury standard solutions 100 and 10 pg ml-l. Store the solution in a Pyrex glass bottle. Procedure Weigh 1.00 g of the rock soil or stream-sediment sample (less than 100 mesh) into a 25 x 200 mm culture-tube. Add 2 ml of concentrated nitric acid and 5 ml of concentrated hydro-chloric acid to the tube and place it in a boiling water-bath (about 92 "C) under a fume hood for 20 min agitating occasionally. Remove the tube from the bath and adjust the volume to 20 ml with water. Allow the contents of the tube to cool add 10 ml of concentrated am-monia solution and mix.Cool add 1.0 ml of potassium iodide solution and mix; then add 0.2 ml of silver nitrate solution and mix. Add 2 ml of isobutyl methyl ketone stopper shake for 1 min and centrifuge to separate the organic layer. Determine the mercury in the organic solution by aspiration into the flame or by injecting 5 or 10 pl of the organic solution with the automatic sample dispenser for carbon-rod atomisation. Standards are prepared by weighing 1 .OO-g portions of the mercury-free material into 25 x 200 mm tubes and adding appropriate volumes (using precision Eppendorf pipettes) of the 10 or 100 pg ml-I mercury standard solutions to contain 0 0.05,0.1,0.5 1,3 5 and 10 pg of mercury respectively for the carbon-rod determinations and to contain 0 1 3 5 10 50, 100 and 200 pg of mercury respectively for the flame determinations.The standards are then carried through the same procedures as the samples and absorbance readings are com-pared to determine the mercury values. The calibration proved to be linear and reproducible. For this work a geochemical exploration reference material (GXR-6 in Table I) was roasted for 4 h at 800 "C in a muffle furnace to prepare the mercury-free material used for standard preparation. Although there may be a build-up of silver in the carbon rod the calibration graphs run on different days using the same rod remain reproducible. * IJse of trade names in this paper is for descriptive purposes only and does not constitute endorsement by the US Geological Survey 60 SANZOLONE AND CHAO AAS OF HG IS GEOLOGICAL M ~TERIALS A~zalyst VOZ.108 Results and Discussion Sample Decomposition The nitric acid - hydrochloric acid digestion used in this procedure is the decoinposition technique used by Tindall.6 The digestion is Iiighl~7 effective in dissolving mercury from the samples. A time study sliowecl no difference in tlie amount of mercury released wlien the digestion time was ~ a r i e d from 5 to 75 min. Prolonged heating did not result in a loss of mercury as long as the solution w a s not heated to dryness. Samples containing significant amounts of organic material ( i . c . marine sediment samples MESS-1 and BCSS-1 listed in Table I with approximately 6.0 and 4.40,; organic matter, respectively) require that the organic matter he oxidised completely.If organic material remains after the digestion an interference o(-curs as a result of smoke forination during the atomisation step. For such samples add 4 in1 of concentrated nitric acid and heat the samples in a hot water-bath until the brown fumes of nitrogen dioxide cease to evolve (about 30 min) ; at this point add 5 ml of concentrated liydrochloric acid and continue tlie digestion for 20 min as specified in the procedure. A soil sample spiked wit11 cinnabar was digested in the same manner as the organic-rich samples above and also with potassium cliloratc - concen tratcd hydrochloric acid which has been reported to decompose cinnabar co~iipletely.~ The mercury values obtairied by the two decomposition techniques agreed with each other within experimental error.Solvent Extraction of Mercury Mercury is extracted quantitatively from an arnrnoniacal solution into isobutyl methyl ketone as tlie tetraioclomcrcurate(~1). The tlistribution ratio ( D ) was determined to be approximately 1000. Mercury in the isobutyl methyl ketone extract remains stable for at least 48 11. A large excess of potassium iodide or ammonia lias no effect on the efficiency of the extraction. Tlitrefore tlie amounts of tliese two reagents have been reduced relatilye to the original Tindall's rnctliod. As suggested by Tindal16 and confirmed i n this investigation, only tliat portion of isobutyl methyl ketone that i i separated from tlie aqueous phase contains mercury. The portion of isobutyl methyl ketone that remains in the aqueous phase and that associates with the residue from the digestion and the precipitated hydroxides is free from mercury.Therefore ensuring that the volume of isobutyl metliyl ketone separated is equal for the standards and samples is of paramount importance. To meet this condition standards with the same mass of mercury-free material a5 the samples are carried through the procedure so that any loss of isobutyl tnetliyl ketone clue to physical trapping or surface adsorption by the solid materials may be cclualised. I:quilibration of the iwbutj.1 nietliyl ketone with an aqueous solution similar to that used iir tlrc sm~ple preparation as tried prior to its u.;e in the extraction ; however unequal volumei of tlrc organic 1 a j - t ~ still rebulted from untlissolved solid niaterial and precipitated iron which tend to hold back some iwhutyl methyl ketone in tlrc aqucous layer.Tjilution of the isobutyl riietliyl ketone layer to volume would r e d t in a more cumbersome procedure and a loss of srniitivity. For determinations that require extreme accuracy and for wliicli the amount of sample niatcrial is adequate a roayted TABLE I CAIIRON-ROD ATOMIC ABSORPTION AND COMPARISON WITH KEPORTEI) VALUES REPLICATE DETERMINATIONS O F illERCLTK\- I N VARIOUS SXXIPL1i)S BY FLAME AKD IZclative hIear1 s taiitinrd I<(-portcd valucs p.p,rn. Dctc.riniiratinn ( n =- a) IZaiigc de 0 1 1 ------ ----A- ~ - --- 7 Sample Iicfcrciicc 1.1" I<t,fcrcricc IS* Rcf(,rcnce 1:lt 3.91) :i.W) 2 . i .-) 2 . i i GXR-3 (Fe - \ I l l d<y"islt) 0.:;; 0.3:;- li.:lti o.:;.; GSIZ-1 (Cu rrrill iicaci) .0 . 1 ; 0.12-O.l:; l i . 1 :: GSR-5 (Soil) . . . . 0 . 1 1 o . l \ - - l l l > 0.1:i GYK-6 (Soil) . . . . Carlioti rotl 0.0s I ).()$-I I . ( I!) rl,llS UCSS-1 (Btariiic scdiiuciit) C'arboir rotl ti.1:; l l . l ? - ~ l . l 1 6 !N i 0.129 0.01 z MESS-1 (hlariiic. scdiiriciit) Carboil rod 0.1s 0 . 1 i - O l ! i 0 . 1 i l + 0.1ll.I * Cold vnpour atomic ahsorption. t Cold vapour atomic absorption ;uid isotope dilution solid - sourcc mass sprctronlctry J~Z~ZILLWY 1983 AFTER SOLVENT EXTRACTION WITH L ~ G XS MATRIX AIODIFIER 61 mercurjr-free portion of the sample may be uscd in the <taritlards to ~iisiire equal retention of isobutyl methj-1 ketone by the standards and s;imple\. This tinic-consuniiiig procedure requires calibration graphs for each sample type and i y considercd to be unnecessary for geochemical exploration lvork in the light of the precision and accuracy data obtained in this work using the same roasted geochemical reference material for standarcls (Table5 I and 11).A whitish precipitate is deposited on the inside of the tube around the isobutyl methyl ketone layer owing to the presence of silver but this precipitate does not seem to affect the analytical results. TABLE I1 RECOVERY OF KNOWN AMOUNTS OF MERCURY ADDED TO VAliIOUS S.4llPLES ICach rei;ult is the avcragc of duplicate malyscs GXR-1 GXR - 1 GXR-2 GXR-2 GXR-3 GXR-4 GXR-5 GXR-6 S a 111 p 1 e (Jasperoicl) . . . . . . (Jasperoid) . . . . . . (Soil) . .. . . . (Soil) . . . . . . . . (Fe - Mn deposit) . . . . (Cu mill head) . . . . (Soil) . . . . . . . . (Soil) . . . . . . . . > , AIESS-1 (Marine sediment) . . BCSS-1 (Marine sediment) . . Flanic Carbon rod Flame Carbon rod Carbon rod Carbon rod Carbon rod Carbon rod 4.00 4.00 7.84 3.0(i s 00 1 1.4!1 4.01 1.00 5.0s 4.01 2 . 0 0 5.9!) 2.55 2.00 4.55 2.55 4.00 6.17 2.54 0.50 2.H4 2.54 1.00 3.30 0.35 0.25 0.63 0.35 0.50 0.90 0.13 0.10 0.22 0.13 0.20 0.33 0.14 0.10 0.25 0.14 0.20 0.35 0.08 0.05 0.13 97.!) !15 7 l O l . 4 !1!).7 It l o . 0 !M.1 !)3.4 !)3.2 105.0 105.9 95.7 100.0 104.2 102.9 100.0 0.08 0.10 0 . 1 9 105.6 Carbon rod 0.18 0.20 0.40 105.3 0.18 0.40 0.58 100.0 Carbon rod 0.13 0.10 0.22 95.7 0.13 0.20 0.33 1 00.1 Matrix Modification These modifiers reduce the loss of mercury during the ashing stage and increase the efficiency of atomisation. Matrix modification techniques currently employed require a separate addition of the modifier to the furnace along with the analyte. In the proposed procedure silver is chosen as the matrix modifier; the silver is co-extracted with mercury uiider the experimental conditions thus eliminating the need for a separate addition. Fig. 1 shows tlie extraction curve for silver using the same extraction conditions as proposed for mercury. ilbsorhance data were obtained from the organic layer by flame atomic absorption. To eliminate possible Various matrix modifiers have been used for furnace determination of mercury.s-ll 0 200 400 600 800 1000 Silver in aqueous solutionivg Fig.1. Extraction of silver from solution into isobutyl methyl ketone 62 SANZOLONE AND CHAO AAS OF HG IN GEOLOGICAL MATERIALS AnaZyst VoZ. 108 factors that may have limited the absorbance to 0.600 as shown in Fig. 1 readings were also taken for residual silver in the aqueous layer. No absorbance values were seen until approxi-mately 200 pg of silver had been added to the system; from this point residual silver in the aqueous layer was directly proportional to the silver added i.e. the organic layer was saturated with silver. Based on this result 0.2 ml of 0.01 M silver nitrate solution was added to the aqueous solution to give the maximum silver content in the isobutyl methyl ketone (210 pg).Fig. 2 is a comparison of calibration graphs obtained with A the co-extracted silver as a matrix modifier; B silver added separately; C ammonium sulphide; and D without a matrix modifier. Clearly use of the co-extracted silver as a matrix modifier gives the highest absorbance values over the entire range of mercury. Our experimentation has shown by using separate additions that amounts from about 0.1 to 20 pg of silver show no variable effect on the peak height of mercury. Kirkbright et aZ.ll used from 30 to 120 pg of silver with no change seen in the mercury absorbance. The amount of silver in A also apparently has a wide useful range in that silver accumulates at about 1 pg per injection and no effect in absorbance can be seen over the useful life of the tube (approximately 200 firings).Fig. 3 illustrates the effect of ashing temperature on the absorbance of mercury with and without various matrix modifiers; again the enhancing effect of the co-extracted silver is shown. 0.5 a 0.4 u C 0.3 ft 2 0.2 a 0.1 0 1 2 3 4 5 Mercury in aqueous solution/pg 0.5 I + 6 0.3 0.1 Fig. 2. Calibration graphs for mercury by carbon-rod atomisation with various matrix modifiers. A lo-$ extract of the isobutyl methyl ketone and ramp atomisa-tion set a t rate 3.0 and voltage 5.0 (about 1200 "C) were used for all curves. A Co-extracted silver; drying 100 O C for 25 s ; ashing 160 "C for 5 s. B 5 pl of 0.01 M AgNO solution; drying 100 "C for 40 s ; ashing 150 "C for 5 s. C 5 pl of 5% m/V (NH4)2S solution; drying 150 "C for 40 s ; ashing 200 "C for 5 s.D Without matrix modifier; drying 100 "C for 25 s ; ashing, 150 "C for 5 s. 0 100 200 300 400 500 Ash i ng tern peratu re /"C Fig. 3. Effect of various matrix modifiers on the carbon-rod atomisation of mercury a t different temperatures. A 10-p1 extract of isobutyl methyl ketone from the aqueous solution con-taining 5 pg of mercury and optimum settings for each modifier were used. A Co-extracted silver; B 5 pl of 0.01 M AgNO solution; C 5 pl of 57" m/V (NH,),S solution; and D without matrix modifier. Sensitivity The sensitivity (1% absorption) of the carbon-rod atomisation for mercury has been determined to be 2.2 x 10-10 g A linear relationship exists between absorbance values and varying amounts of mercury (0.05-10 pg corresponding to 0.05-10 p.p.m.of mercury in the sample) using either 5 or 10 pl of the isobutyl methyl ketone extract for atomisation. Owing to the low atomisation temperature (about 1200 "C) the carbon rod can be fired 200 times before any deterioration expressed as a loss of sensitivity can be observed. For levels of mercury between 1 and 200 pg in the isobutyl methyl ketone extract (corre-sponding to 1-200 p.p.m. of mercury in the sample) the flame technique may be used. The sensitivity (1% absorption) of the flame technique is 0.73 pg ml-l. For concentrations of mercury greater than 200 p.p.m. in the sample a smaller sample mass should be used. The co-extracted silver has no effect on the flame determination of mercury and should be added to all samples so that if the mercury concentration is too low to be determined by the flame technique an aliquot of the extract may be used for electrothermal determination January 1983 AFTER SOLVENT EXTRACTION WITH AG AS A MATRIX 31OT)IFIER Interferences There are no interferences in the mercury determination from major and trace elements commonly present in geological materials.The flame technique for mercury determination was evaluated for interferences at two levels of mercury (20 and 500 pg) and the carbon-rod technique was also evaluated for interferences at two levels of mercury (0.25 and 2.0 pg). Mercury determination was found to be interference free from as much as 40(;( of iron 25c:/o of aluminium calcium or manganese 200/ of magnesium potassium or sodium 5000 p.p.iii.of antimony arsenic bismuth cadmium. cobalt copper lead niolybdenum nickel tin tellurium or zinc and 100 p.p.m. of gold in the sample. Various combinations of the abovt elements also showed no interference effect. Background absorption caused some noise for the blank and for low levels of mercury; the use of the background corrector can eliminate this problem for the carbon-rod work and can reduce the noise level for the flame work. (’are sliould tx taken with samples containing a high amount of organic material. 1;ailur-e to oxitlise corn-pletely the organic material may result in an interference from the smoke that is generated during the atomisation stage. 63 Results for Geological Samples The proposed method was applied to six US Geological Survey geoclieniical reference samples (GXR 1-6)l2 and to two marine sediment reference materials (AIJSSS-1 and BCSS-1) .I 3 The results are in good agreement with those reported by other investigators (Table I ) . Replicate determinations of mercury in two geochemical samples by flame atomic absorption gave relative standard deviations of 6.4 and 6.7% (Table I). Keplicate analyses of eight geo-chemical samples by carbon-rod atomisation gave relative standard deviations ranging from 2.3 to 6.9% (Table I). Two samples were spiked with two levels of mercury prior to sample digestion and recoveries by flame atomic absorption ranged from 94.1 to 100.Oq/o. Spiking of eight geochemical samples with two levels of mercury and subsequent determination by carbon-rod atomisation gave recoveries that ranged from 93.2 to 105.90/ (Table 11).Conclusions The method for the determination of mercury by flame and carbon-rod atomic absorption proposed in this paper is rapid sensitive and interference free; the method is capable of measur-ing mercury from the crustal abundance level up to ore-grade concentrates and is suitable for a variety of geological materials including sulphides and organic-rich samples. The use of the co-extracted silver as a matrix modifier is a unique feature of the method that not only en-hances the absorbance signal of mercury but also simplifies the measurement operation. About 50 samples can be analysed for mercury per 8-h day. The method should fulfil the need for a satisfactory procedure for the determination of mercury in geological materials.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References Levinson A. A “Introduction t o Exploration Geochemistry,” Appliwl I ’ublishing Calg:q, McCarthy J . H. J r . J . Geochem. Explov. 1972 1 143. Rose A. W. Hawkes H. E. and Webb J . S “Geochemistry i n Ylincral l ~ ~ x p l o r ~ ~ t i o ~ t ” Academic Kokot M. L. Minev. Sci. Eng. 1974 6 236. Ure A. &I. Anal. Chim. Acta 1975 76 1. Tindall I. M. At. Absorpt. Newsl. 1967 6 104. Chao T. T. and Sanzolone R. F. U.S. GeoZ. Suvv. J . l i e s . 1977 5 409. Tssaq H. J . and Zielinski W. L. Jr. Anal. Chew. 1974 46 1436. Ediger R. D. A t . Absovpt. Newsl. 1975 14 127. Alder J . F. and Heckman D. A. Anal. Chem. 1977 49 336. Kirkbright G. F. Shan Hsiao-Chuan and Snook I.<. D. A t . Spectvosc. 1980 1. 8.5. Allcott G. H. and Lakin H. W. in Elliott I . L. and Fletcher IV. K. Edituvs “Geochemicsl Rerman S. “Marine Sediment Keference Materials,” Divisjon of Chernistry Sational Research Gladney E. S. Perrin D. R. Owens J . W. and Knab D. rZnaZ. Chenz. 1979 51 1567. 130 C. L. (Minerals Studies Laboratory University of Texas at Austin) personal communication, Received June 15th 1982 Accepted A ugztst 1 Oth 1982 Alberta Canada 1974 p. 332. Press New York 1979 p. 561. Exploration 1974,” Elsevier Amsterdam 1975 p. 659. Council Ottawa Canada 1981. 1982

ISSN:0003-2654    

DOI:10.1039/AN9830800058

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

64 Analyst Jawary 1983 Vol. 108 $$. 64-70 Novel Static Cold Vapour Atomic-absorption Method for the Determination of Mercury Ping-Kay Hon Oi-Wah Lau and Man-Chaun Wong Depavtwient of Chemistvy Chincse University of Ho~zg l<orig Shatiti iY.T. Hong l<ong A static cold vapour procedure for the determination of mercury by atoniic-absorption spectrometry using a very simple and inexpensive apparatus is described. The riiercur~~-laden air is forced into the absorption cell b\- water displacement. Alaximuni sensitivity is obtained. wIie11 the volume of the displaced air is equal to the internal volume of the absorption cell ant1 thc mercury solution is 9 hi in sulphuric acid. 'lhe peak absorbance exhibited a marked tlcclinc for Iij-tlrochloric acid concentrations above 1.5 M and for nitric acid concentrations above 3 1.1.'I'lic calibration graph obtained for mercury(I1) in 9 AI sulpliuric acid is linear from 0 to 17 ng m-l and tlie sensitivity is 0.08 tig nil-1. ,2 wiIidowless absorption cell can also be uscd i\Tith a narrower linear calibration range. The detection limit \{-as found to be 0.08 ng nil-1 for both cells with and without end 1vincloiL s. Keywovds .l.lei.cuvy detevigizization ; cold ~lapouv ~ t o ~ g ~ ~ c - ~ ~ b s o v ~ t i o g ~ sfwctvo-metry ; static procedure The cold vapour atomic-absorption (CVAA) method is commonly used for the determination of trace amounts of total mercury in solution. The principle used is that described by Poluektov tt d.,l whereby mercury(I1) is reduced to mercury(0) with tin(I1) ions.The resultant eleniental mercury is swept out of solution with a carrier gas and directed to an absorption cell where tlie atomic-absorption signal of mercury is measured. Hatch and Ott2 used a closed system in which the air containing the released mercury vapour was continu-ously recirculated. Many improvements to the method3 have reduced the detection limit to as low as 0.001 parts per lo9 (p.p.b.) of mercury. Cold vapour atomic fluorescence spectro-metry has also been employed to determine mercury with low limits of detection (c.g., 0.02 ng m-l) and a wide linear calibration range (e.g. from 0.2 to 1000 ng)' The reduction - aeration methods described above suffer from a reduction in sensitivity because the mercury vapour is diluted by the carrier gas and the use of cell windows requires the incorporation of moisture traps to prevent fogging.These traps however introduce undesirable memory effects. Moisture traps are not essential if the absorption cell is heated, but then a heating device such as a light-bulb needs to be present above the cell. To over-come the above problems a number of methods that do not involve the use of a carrier gas have been reported. Staintonj developed a syringe procedure in which tlie mercury was partitioned between the sample solution and air in the syringe used to inject the mercury-laden air into tlie absorption cell. Clinton6 described a static vapour apparatus where the mercury vapour is simply transferred to the absorption cell by displacing it from the reaction vessel with tap water and the mercury vapour is subsequently removed by suction.Ahother simple apparatus that uses water displacement to transfer the equilibrated mercury vapour was devised bj' Chapman and Dale.' The water was added from a separating funnel and the absorption cell had no end windows in order to avoid condensation problems. very simple stationary cold vapour method8 has been proposed in which a stoppered 4-cni CJ' cell is used for the reduction of rnercury(II) the partition of the reduced mercury between the liquid and gas phases and the subsequent atonric-absorption measurement. Ll background corrector is necessarjr to correct for non-atomic absorption. A stationary cold vapour atomic-absorption attachment has also been de\-elopeclg and applied to the direct deter-mination of total mercury in undigested biological samples.Howel-er the partition equilibrium times were 12 and 6 min respectivelj. for the 80- and 45-mni absorption cells. Further background correction was again necessary and the mercury vapour in the absorption cell was removed by flushing the cell wit11 air from an air pump. X similar single unit stationary cold vapour generator was describedlO for the determination of mercury in bio-logical tissues by elect ro t hernial at omic-absorpt ion spectrometry HON LAU AND WON(; ti5 The purpose of this work was to devise a procedure using a very simple and inexpensive static-vapour apparatus for the determination of iiiercurj7( I I) in solution tlic C V U method and using water displacement to transfer tlie equilibrated inercurj- to tlie absorption cell.The construction and operation of the apparatus are described and the optimum conditions for tlie determination and the sensitivity precision and accuracy o f the proposed method are reported. Experimental Apparatus The instrunient used was a Perkin-Elmer Model 360 atomic-absorption 5pcctronieter with the burner removed. A Varian-Techtron iiiercury hollow-cathode lanip was uscd as the light source and the wavelength and band width were set at 253.7 and 0.2 11111 respectively. The lamp current was 4.5 m4. For easy construction the absorption cell was made of rectaIigular-scctioii Plexiglass, having dimensions 17 cm x 3.3 mm x 5 mm (the 5 nim section being mounted vertically) (see Fig. 1). The cell had an internal volume of 2.8 in1 and two quartz windows were attached to its ends with epoxy cement.The cell was fixed on a metal support whicli in turn was fitted to the burner support hole. The inlet and outlet ports ( 2 inn1 i.(l.) were about 2 inni from each end of the absorption cell. A windowless cell with similar diniensioiis was also made but the inlet port was in the middle of the cell. The reduction vessel was a glass vial (3.6 x 1.5 cni i d . ) with a screw-cap whose liner was removed and replaced with a layer of silicone-rubber. Two pieces o f Tcflon tuhi rig p:~sscrl through the screw-cap; the longer one (25 cm x 2 niiii 0.~1.) conncctecl the reaction vial to the inlet port of the absorption cell and the shorter one (11 cni x 1 ~iiiii o.d.) had about 2.6 cm of its length inside the vial and was used for the injection of reductant and distilled water with syringes into the vial.t A 1'1 J I -B H c t - 10 mercurv Fig. 1. Scheniatic diagraiii of the 5tatic vapour apparatus. A A\bsorption cell LZ nietal .;upport, C PTFE tubing; D reduction vessel; 13 silicone rubber, I; magnetic bar; G magnetic stirrcr; H PTFE tubing; I, incident beaniintensity ; I transmitted besin intensity; and J exhaust. Reagents A stock solution containing 1000 p.p.m. of mercury was prepared by dissolving 1.3535 g of inercury(I1) chloride in 50 1x1 of concentrated hydrochloric acid and diluting to 1 1 with water. l'roni this stock solution, working standard solutions were freshly prepared by appropriate dilution with a solution containing 1% V/V of nitric acid and 0.0020/ w/V of potassium permanganate.This solution need not be purified by blowing nitrogen tlirough it to remove trace amounts of mercury as we found no difference in the value of the blank absorbance whether we purged the solution for 30 rnin with nitrogen before use or not. All reagents used were of analytical-reagent grade. The reductant was a 1% m/V tin(T1) chloride solution in 1 31 liydrocliloric acid 66 Analyst Vol. 10s Procedure The absorption cell was aligned properly to allow maximum intensity of the light beam to pass through. Sample standard or blank solution (1 ml) was pipetted into the reduction vessel containing a small magnetic bar and the vessel was immediately capped. The reduc-tant [0.2 ml of the 1% tin(I1) chloride solution] was injected with a 1-ml syringe into the reduction vessel through the shorter Teflon tubing.The 1-ml syringe was immediately replaced with a 10-ml plastic syringe containing 2.7 ml of distilled water. After the solution in the reduction vessel had been stirred for 1.5 min with a magnetic stirrer the water in the syringe was injected quickly and completely into the vessel to force the mercury vapour into the absorption cell through the longer Teflon tubing. The maximum absorbance at 253.7 nm was immediately read from the meter or recorded with a recorder. For each measurement at least three readings for three 1-ml aliquots of the solution were recorded and then averaged. The reduction vessel was temporarily replaced with a dry vial. Then air was injected from an empty 10-ml syringe to force the mercury vapour out of the apparatus.Zero absorbance was observed nearly instantaneously. The reduction vessel was emptied and rinsed three times with distilled water before the procedure was repeated. About 1 g of orchard leaves 0.1 g of cyclohexanebutyric acid mercury(I1) salt and 0.5 g of the mercury standard in oil were dissolved separately by an acid digestion procedure for organic samples.ll The sample solutions were diluted with distilled water and the concentra-tions of mercury in these solutions were deduced from a calibration graph obtained from solutions containing 0-17 ng ml-l of mercury in 9 M sulphuric acid. HON et al. STATIC COLD VAPOUR AAS Results and Discussion The method proposed here is similar to the methods proposed by Clinton6 and Chapman and Dale,' as water displacement is used to transfer the mercury-laden air to the absorption cell.However the present procedure is simpler; no suction is necessary to remove water and mercury vapour no separating funnel is used as the displacement apparatus and nitrogen is not needed to purge the residual mercury from the apparatus. Further when the experi-mental conditions are optimised the sensitivity achieved by the proposed method is better than both of these methods. Absorption Cell and Reduction Vessel The atomic-absorption signal of mercury is expected to increase with increasing length of the absorption cell but to decrease with increasing diameter of the cell as is true for the reduction - aeration m e t h ~ d . ~ With our atomic-absorption spectrophotometer however, the maximum cell length that can be used is 17 cm and we have also made the cross-sectional area of the cell small.Further the absorption cell is connected to the reduction vessel with a short piece of Teflon tubing with very small diameter (2 mm 0.d.). The effect of the volume of the displaced air on the peak absorbance for a constant absolute mass (25 ng) of mercury(I1) was studied by adding 0.2 ml of the 1% tin(I1) chloride solution to four vials each containing 1.0 ml of a mercury(I1) standard (25 ng ml-l). These vials had the same diameter but different heights. For each vial distilled water was injected just up to the neck of the vial and the volume of the displaced air was equal to the volume of water injected.The results shown in Fig. 2 indicate that the maximum absorbance was obtained when the volume of the displaced air was 2.7 ml which was approximately equal to the volume of the absorption cell (2.8 ml). This finding is similar to that observed by Clinton.6 It is also interesting that the absorbance of mercury decreases more rapidly for smaller volumes of the displaced air compared with the result for larger volumes of the displaced air (see Fig. 2). I t appears that when the volume of the displaced air is small it is not as efficient to extract the reduced mercury vapour into the gas phase. When the volume of the displaced air is greater than 2.8 ml however part of the mercury-laden air will be driven out of the absorption cell so that the signal will also decrease.Effect of Acids and Tin(I1) Chloride on Absorption Koirtyohann and Khalil12 found that sulphuric acid but not nitric or hydrochloric acid has significant effect on the partition constant K defined as the ratio of the concentration of mercury in air to that in liquid. Tong* observed however that the peak absorbance readin January 1983 METHOD FOR DETERMINING MERCURY 67 I I I 2 3 4 5 6 7 0.4 Volume of displaced airiml Fig. 2. Effect of the volume of displaced air on the absorbance at 253.7 nm of 25 ng of mercury(I1). for each measurement is strongly dependent on the acidic medium. The effect of acids on the peak absorbance for a constant mass (15 ng) of mercury was studied and the results are shown in Fig. 3. Although our results are in general similar to those previously reported by Tongs there are two important differences.Firstly the peak absorbance exhibited a marked decrease for hydrochloric acid concentrations above 1.5 M and for nitric acid concentrations above 3 M and in both instances the peak absorbance remained virtually constant after the marked decrease. Secondly the maximum absorbance observed in sulphuric acid was about 1.4 times the corresponding values in either nitric acid or hydro-chloric acid whereas Tongs reported the maximum absorbance in all three acids to be around 0.25. The decrease in absorbance in nitric acid may be ascribed to its oxidising properties, which will make the reduction of mcrcury(I1) by tin(I1) chloride more difficult. On the other hand the high concentration of chloride ion will make the reduction of mercury(1) more difficult as evidenced by the much smaller standard reduction potential of the couple Hg,Cl,(s)/Hg (0.268 V) compared with that of the couple Hg,SO,(s)/Hg (0.615 V).The comparatively high absorbance at higher sulphuric acid concentration may be attributed to the comparatively greater ease of reduction of the mercury ions superimposed on a greater partition constant in this acid.12 I A I 0 2 4 6 8 10 Concentration of acid/M Fig. 3. Effect of acids on the absorbance of 15ng of mercury(I1) A sulphuric acid; B nitric acid; and C hydro-chloric acid. I t is also noted from Fig. 3 that when the highest sensitivity of the method is not desired, it is much simpler and more economical not to add any acid other than those already present in the mercury standard and tin(I1) chloride solutions 68 HON et al.STATIC COLD VAPOUR AAS Analyst VoL. 108 The effect of the concentration of tin(I1) chloride solution (0.2 nil) on the absorbance of 15 ng of mercury(I1) in 1 ml of 9 M sulphuric acid was studied and the results are shown in Fig. 4. The absorbance was found to remain essentially constant when the concentration of tin(I1) chloride was greater than 0.70/,. \Ve preferred to use 1% tin(I1) chloride because it would yield a slightly higher absorbance of mercury. 0.2 ' I I I 0.5 1 .o 1.5 2.0 Concentration of tin(ll) chloride o/o Fig. 4. Effect of the concentration of tin(I1) chloride ( 0 . 2 ml) 011 the absorbance of 15 ng of mercury(I1) in 1 rril of 9 hi sulphuric acid. The effect of the mixing time is shown in Fig.5 which indicates that the maximum absorbance was obtained when the mixing time was in the range 1.5-5 min and we chose to use a mixing time of 1.5 min for all subsequent experiments. In all measurements the speed of the magnetic stirring bar was fixed at about 1200 rev niin-1 and the volume of the liquid phase was also fixed at 1.2 nil as these two parameters can also affect the absorbance signal of mercury. 0.2 ' 1 1 2 3 4 5 Mixing ti meirn i n Fig. 5 . Effect of mixing time on the absorbance of 15 ng of mercury(I1) in 1 ml of 9 M sulphuric acid reduced by 0.2 ml of 1% tin(I1) chloride solution. Calibration Graphs Sensitivities and Precision Chapman and Dale7 used an absorption cell without windows to eliminate fogging problems.We have therefore prepared calibration graphs (Fig. 6) using absorption cells both with and without quartz end windows. The calibration graph for cell B (without end windows) was linear in the range 0-12 ng ml-l of mercury and gave a correlation coefficient of 0.9996 and a slope of 0.059 p.p.b.-l. On the other hand the calibration graph for cell ,4 (with end windows) was linear in the range 0-17 ng m-l of mercury and gave a correlation coefficient of 0.9999 and a slope of 0.053 p.p.b.-l. Hence it is more desirable to use the cell with end windows as a wider linear calibration range can be used despite the fact that the slope of the calibration graph is slightly lower and the construction more complicated than the windowless cell. Further the absorbances for the blank using cell ,4 and cell B are 0.030 and 0.033, respectikrely indicating clearly that the fogging problems are insignificant when the cell with windows is used.iZ possible explanation is that the air originally present in the absorption cell is driven out faster when windows are absent so that tlic mercurj~-laden air is less severely diluted. The sensitivity or cliaracteristic concentration is 0.08 ng nil-' of niercur). in 9 31 sulphuric acid for 17; light absorption for the cell with end windows and that for the windowless cell The slightly higher sensitivity associated with cell €3 is unexpected J a w a r y 1 983 METHOD FOR DETERMINING MERCURT ( i9 is 0.07(5) ng nil-'. VjV nitric acicl (a1)oitI 0.16 M) for the cell with end wiiiclow~. The detection limit defined as the concentration that yields an absorbance twice that of the standard deviation of the absorbance of a blank was found to be 0.08 ng ml-l for tlic cell with end windows and that for the windowless cell was also 0.08 ng ml-I.The sensitivity is 0.13 ng ml-1 of mercilry iii I Concentration of rnercuryhg mi-' Fig. 6. Calibration graphs for the determination of mercury in 1 in1 of 9 M sulphuric acid using absorption cell A with quartz windows and cell B without wiiiclow. The absorbance increases rapidly to a maximum and remains constant for about 6 s for the cell with end windows and for a shorter time for the windowless cell. However in both instances it is possible to read the absorbance directly from the meter without difficulty. I t is important to inject water into the reduction vessel as quickly as possible so that the mercury-laden air is forced into the absorption cell with minimum mixing.With a little practice reproducible results can be obtained. The relative standard deviation of 10 replicate determinations of a mercury standard solution (4 ng ml-l) was found to be 2.5O/, for the cell with end windows and 1.1% for the windowless cell. Comparison with Other Static Cold Vapour Methods The proposed method is much simpler in operation than other static cold vapour niethods. Throughout the experiment the reduction vessel is attached to the absorption cell which is fixed in position. We need no background corrector or any carrier gas or vacuum pump to remove the residual mercury vapour from the system.After optimisation the sensitivity or characteristic concentration (0.08 ng ml-l) obtained by the proposed method is better than those obtained by the methods proposed by Clinton6 (0.12 ng ml-l using a 10-ml sample), Chapman and Dale' (1 ng as characteristic mass) Stainton5 (0.2 ng ml-1) or Tong8 (0.17 ng ml-l). The sensitivity is also better than that obtained by the reduction - aeration procedure of Hawley and Ingle3 (0.20 ng ml-l) in which a 20-cm cell was used. Determination of Mercury in Standard Reference Materials The accuracy of the proposed method was checked by determining the mercury content in orchard leaves (NBS Standard Reference Material 1571) and in two Merck standards for mercury. The orchard leaves were certified to contain 0.155 sfr 0.015 pgg-l of mercury.The Merck standards were cyclohexanebutyric acid mercury( 11) salt certified to contain 38.9y0 of mercury (oil-soluble standard for AAS E. Merck Art. 4431) and 1 g kg-l of mercury (mercury standard dissolved in oil E. Merck Art. 15063) respectively. The results are given in Table I. The mercury contents in these samples obtained by the present method agree well with the certified values demonstrating the accuracy of this method 70 HON LAU AND WONG TABLE I MERCURY CONTENT IN STANDARD REFERENCE MATERIALS Value obtained by Sample present method Certified value Cyclohexanebutyric acid mercury(I1) salt . . . . 38.9 37.4 37.1y0 38.97; Mercury standard dissolved in oil . . . . . . 0.099 0.098 0.097o;b 0.1 yo NBS orchard leaves . . . . . .. . . . 0.170 0.161 0.157 pg g-l 0.155 2 0.015 pg g-1 Conclusion The proposed static cold vapour atomic-absorption procedure for the determination of mercury in solution requires no special apparatus or carrier gas and the procedure is simple. Precise and accurate results can be obtained. The analysis time is 3.5 min per solution and this can be further reduced to 3 min with slightly lower sensitivity if the mixing time used is 1 min instead of 1.5 min. The detection limit may be further lowered and the linear cali-bration range extended if cold vapour atomic-fluorescence spectrometry is used. Further, the method can be modified to determine elements that form volatile hydrides. Research along these lines is now in progress. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Poluektov N. S. Vitkun K. A. and Zelyukova Yu. V. Z h . Anal. Khim. 1964 19 937. Hatch W. K. and Ott W. L. Anal. Chem. 1968 40 2085. Hawley J . E. and Ingle J . D. Jr. Anal. Chem. 1975 47 719. Ebdon L. Wilkinson J . R. and Jackson K. M‘. Anal. Chinz. A d a 1981 128 45 and references Stainton ILI. P. Anal. Chem. 1971 43 625. Clinton 0. E. Lab. Pract. 1974 705. Chapman J . F. and Dale L. S. Anal. Chim. Acta 1978 101 203. Tong S. L. Anal. Chem. 1978 50 412. Tong S. L. and Leow W. K. Anal. Chem. 1980 52 581. Bourcier I). K. and Sharma R. P. J . Anal. Toxzcol. 1981 5 65. Van Loon J C, “Analytical Atomic Absorption Spectroscopy-Selected Methods,” Acadernic Koirtyohann S. K. and Khalil M. Anal. Chem. 1976 48 136. cited therein. Press New York 1980 p. 160. Received June 2nd 1982 Accepted July 27th 198

ISSN:0003-2654    

DOI:10.1039/AN9830800064

出版商:RSC    

年代:1983

数据来源: RSC

摘要:

Analyst, January, 1983, Vol. 108, $9. 71-75 71 Differential Electrolytic Potentiometry Part XXVI.* Direct Polarisation in Acid - Base Titrimetry in Toluene - Methanol Mixtures Abdalla M: S. Abdennabit and Edmund Bishop Unzversaty of Exeter, Chrmistvy Department, Stockev Road, Exeter, E X 4 4QD The application of direct current differential electrolytic potentiometry to acid - base titrations in toluene - methanol mixed solvent has been investi- gated. Antimony electrodes have been examined in titrations of diverse acids with sodium methoxide. Normal differentiation was observed only in the titration of weak acids. Malfunction of the anode in titration of stronger acids was reflected in the differential curves, a n d was accompanied h y ;I decrease in mass of the electrode. The zero-current antimony electrode exhibited similar behaviour with the strongest acids.The effect has been traced to a chemical reaction between methanol and the antimony oxide in the presence of stronger acids. Keywords : Diffevential electvolytic potentiometvy ; antimony electvodes ; inethanol - toluene mixed solvent ; non-aqueous acid - base titvimetvy Direct current differential electrolytic potentiometry (d.c. DEP) has been applied to all types of titrimetric reactions in aqueous media,l-s and to non-aqueous acid - base reactions in acetic anhydride - acetic acid mixed ~ o l v e n t . ~ The best results in acid - base titrations have been achieved with antimony ele~trodes,~-59~ desirably restricted to the pH range 2-10, and not exposed to ligan(1s that would combine with antimony.1° Antimony has been used for ze- o- _urreiit potentiometric titrations of both weak and strong acids in non-aqueous solvents such as ethylenediamine,'l~12 butylamine13 and benzene - methanol mixtures.14 Warner and Haskell13 reported slow equilibration of antimony electrode potentials in the butylamine solvent.The DEP technique accelerates the antimony electrode response in non-aqueous solventsg and requires no reference electrode or salt bridge, which poses diffi- culties in non-aqueous media. This paper reports the results of a study of the appiicability of DEP to acid - base titra- tions in toluene - methanol mixtures and of the behaviour of polarised and zero-current antimony electrodes in this medium, the titrations being simultaneously monitored wit 11 a glass electrode.Experimental The apparatus and titration assembly have been described previo~sly.~9~~ The glass electrode was an EIL GHS23. Reagents Chemicals were of AnalaR grade. S o d i z m methoxide solzztion. Freshly cut pieces of sodium metal (4.Og) were washed in methanol and added cautiously to 100 ml of methanol contained in a 1-1 flask kept at 0 "C; atmospheric carbon dioxide was excluded by continuously purging the vessel with nitrogen. Toluene and methanol were finally added in amounts to give a volume ratio of 3 : 2 and the solution was standardised against weighed amounts of benzoic acid using th ymol blue as the indicator. Dehydration was effected by adding 10 nil of pure xylene to 2.Og of the solid acid to form an azeotropic mixture with the water and removing the solvent in a rotary evaporator at 100 "C, with final desolvation at 80 "C under reduced pressure.The solid was dissolved in the appropriate amount of 3 + 2 toluene - methanol. ToZuc~ze-p-sztZi3honic acid solution, 0.1 moll-I. * For details of Part XXl' of this series, see reference list, p. 75. f Present address : University of Petroleum and Minerals, 1726 UI'lI, P.0. Box 144, Dhahran Inter- national ,Airport, Dhahran, Saudi Arabia.72 ABDENNABI AND BISHOP: DIFFERENTIAL A n a l y s t , Vol. 108 Hydrogen chloridc ill tolue.12 e - mcthnnol. Hydrogen chloride gas, evolved by dropwise addition of sulphuric acid to hydrochloric acid contained in a closed vessel, was passed by means of a delivery tube into a calibrated flask containing methanol. I'lihen the required amount had been transferred to give a final concentration of approximately 0.1 moll-1 of hydrogen chloride, the solution was niade up with toluene and methanol in amounts to give the 3 : 2 volume ratio, aiid tlie solution was standardised against sodium niefhoxide using thymol blue as the indicator.Benxoic, p-nitvo btvzxoic, $yvzwic, salicylic nizd tvichlovoacet ic acid solut ioya, 0.1 in01 1-l. These were prepared by direct weighing of the acids and dissolution in 3 + 2 toluene - methanol. This was prepared by direct weighing and dissolution in tlie solvent mixture; the solubility limits tlie attainable concentration. The required amount of anhydrous litliiuni chloride was dissolved in nzethanol and toluene was added to give the required volume ratio.Ammoizizw flcroxydisztl~hnte solzitioiz, 0.003 niol 1-I. L i t h i u m chloi.ide .soZzitio?z, 0.05 mol 1-'. Procedure The cell was cliargecl with 40 nil of 0.05 niol lkl lithium chloride supporting electrolyte and 10 nil of 0.1 niol1--1 titrand solution. Three antimony electrodes and the glass electrode were fitted in the cell lid. A lion-aqueous double salt filled witli 0.05 rnol 1--1 lithium chloride solution, connected the cell solution to the SCE. Two of the antimony electrodes were polarised a t a current density of 1 p 4 c r r 2 b y the constant-current source, the third being a zero-current indicator electrode along with the glass electrode, so that DEP and zero-current potentionietry could be performed simultaneously for strict comparison.The microburette containing tlie sodium methoxide titrant was protected from carbon dioxide by a soda-lime guard tube. Oxygen to sustain the oxide film on tlie antimony electrodes was passed tlirougli tlie external inlet to the cell, and stirring was effected magnetically. ,4fter each addition of titrant, the potentials of the anode (X), catliode (C), zero-current (Z) and glass electrode (G) with respect to the SCE, together with the differential potential (EA = A - C), were serially measured on a Corning EEL, Xodel 12, pH nieter. Readings were recorded when the potential drift became less tlian 1 in\' niiii-l. The end-point region was traversed in increments of titrant of 0.01 nil. Results and Discussion Solvent Tlie weakly acidic properties of alcohols as solvents give less sharp titration of ~7eak acids than do less acidic solvents such as acetonitrilc ; conzrnonly, an aprotic solvent such as benzene is added to the alco1ioll4 to sharpen the end-point for weak acid titration.The low dielectric constant of tlie aprotic sol\rent suppresses ionisation so that the acidity or basicity of the solute is not enhanced, and solvolysis of reaction products is also suppressed, so decreasing the buffering of the titration curve and giving sharper breaks. Toluene is here substituted for the toxic benzene, in a volume ratio of 3: 2 toluene - Inetliaiiol. Methanol in this ratio proved necessary ; less methanol caused precipitation of sodium methoxide and fluctuations in potential indicating increased solution resist ance.Titration of Acids The normal pattern for a basic titrant3-j is obtained, arid iy illustrated for benzoic acid in Fig. 1 and for salicylic acid in 1;ig. 2. The catliode is reducing dissolved oxygen, and so may exhibit a less negati1.e potential than tlie zero-current electrode. Tlie weakness of these acids produces zero-current curves for aiitiniony and glas5 electrodes of poor h r p n e s s . The differential curves exhibit small, but sliarp and liiglilj- reproducible peaks, permitting precise location o€ end-points coincident with the equi\dence points. For certain acids, abnormal behaviour is displayed bj. the anode, as shown in Fig. 3 for anirnoniuni p c r o ~ ~ ~ d i s u l ~ ~ l i ~ t e and in Fig. 4 for pj-ruvic acid. Instead of tlie normal pro- gression to more negatiire potentials, the anode iiio\-es to le5s ncgative potentials, reverting to nornial beliaviour after the end-point, and affecting the diffcrcntial curve.,llthougli the equivalence point is encompassed by the double peak of tlie cliff erential curve, its preciseJnnunvy, 1983 ELECTROLYTIC POTESTIOMETR\7. PXRT XSVI 73 €*ImV 4 80 I 0.5 1.0 ml I I O' Volume of titrant E,imV 1 *O L' -4001 I O a o m Volume of titrant Fig. 1. Titratiou of 10 nil of Fig. 2. Titration of 10 nil of 0.1 moll-1 benzoic acid solution 0.1 niol 1-1 salicylic acid solu- with 0.1 rnol 1-1 sodium meth- tion with 0.1 mol 1-1 sodium oxide solution. Medium, 0.05 methoxide solution ; conditions moll-' of lithium chloride plus toluene - methanol, 3 + 2 V / V . Current density 1 pA c i r 2 .Curves: A, anode-S.C.E.; C, cathode- S.C.E. ; 2, zero-current antimony electrode- S.C.E. ; G, glass electrode-S.C.E.: and EA, differential potential A-C. and symbols as in Fig. 1. Volume of titrant €;ig. 3. Titration of 10 nil o t 0.003 mol 1 -l ammonium peroxydisulpha te solutioii xvith 0.01 in01 I-' of sodium methoxide solution, condi- tions and symbols as in Fig. 1. location becomes uncertain without the guidance of the zero-current curves. This behaviour becomes more pronounced as the acid becomes stronger, as illustrated in Fig. 5 for p-nitro- benzoic acid and in Fig. 6 for trichloroacetic acid and in the latter the zero-current antimony electrode shows distortion. Even greater malfunction arises with strong acids, as in Fig. 7 for hydrogen chloride and in Fig.8 for toluene-p-sulphonic acid, both of which show pro- found aberration of the zero-current antimony electrode curve. The malfunction was proved not to be due to instability of the 741 amplifiers in the current source or to migration of copper from the soldered leads to the operating face of the antimony electrodes by reconstruction with 40 J amplifiers and by preparation of GN antimony slug electrodes contacted to brass by prcssure16 ; the phenoniena remained reproducible, and -200 > E 3 z v) -300 I? s b UI - 400 E,lrnV -180 \ \ '. l- G 0 0.5 i.oml I i Volume of titrant Fig. 4. Titration of 10 nil of 0.1 niol 1-l pyruvic acid solution with 0.1 rriol 1-1 sodium niethoxide solution ; condi tioris and symbols as in Fig. 1. E\'mV 4120 I 4 t G I : -200 v, Y! -300 A hk I \ c 2 i \ L -400 1 \ \ 1 0 0.5 " :.Om - Volume of titrant Fig.5. Titration of 10 ml of 0.1 mol 1-1 P-nitrobenzoic acid solution with 0.1 in01 I-1 sodium nicthoxide solution ; condition5 anti symbols as i n Fig. 1.74 ABDENNABI AND BISHOP: DIFFERENTIAL Analyst, Vol. 108 reminiscent of oxygen starvationlo despite saturation of the solution with oxygen. Di- antimony pentoxide formation is unlikely : indeed, the affected electrodes were blackened by a film that was easily wiped off with a tissue. Electroreduction at the anode being dis- counted, chemical reaction was explored. The literature contains no information on the subject. > E $ 2 -200 x 2 L' -300 b - 400 EA/ m V '- G D 0.5 1.oml - I Volume of titrant > E c! 2 @ 'Gi 100 p 200 3 Lu 300 E,hV .--.Volume of titrant 0.5 i.oml I O- Volume of titrant Fig. 6. Titration of 10 ml Fig. 7. Titration of 10 ml Fig. 8. Titration of 10 ml of 0.1 mol 1-1 trichloroacetic of 0.1 mol 1-1 hydrogen of 0.1 moll-' toluene-p-sul- acid with 0.1 moll-' sodium chloride with 0.1 mol 1-1 phonic acid with 0.1 in01 1-1 methoxide solution; condi- sodium methoxide solution ; sodium riiethoxide solution ; tions and symbols as in conditions and symbols as in conditions and symbols as Fig. 1. Fig. 1. in Fig. 1. Chemical Reaction at the Electrode Surface No blackening occurred with benzoic or salicylic acid, implicating a stronger acid. Anodisation in the lithium chloride solution produced simply an oxide film. Such a coated slug electrode was weighed, left in the mixed solvent solution of a stronger acid for a period and re-weighed.For each of the acids in Figs. 5-8 a decrease in mass occurred. A slug left in a toluene solution of trichloroacetic acid suffered no decrease in mass. The combination of methanol and a stronger acid appears necessary, and the reaction is chemical rather than electrochemical. The black deposit was carefully removed, dissolved in 6 mol 1-1 hydrochloric acid, tested" and identified as antimony metal. An oxide coated slug was weighed, immersed for a time in the trichloroacetic acid solution in the mixed solvent and re-weighed, the decrease being calculated as oxide. The black deposit was dissolved in hydrochloric acid and titrated with bromate to determine the amount of antimony. The mass of antimony was less than the mass of oxygen on the basis of diantimony trioxide, but it is difficult to remove completely the powdery deposit with safety.,4 search was made for possible oxidation products of methanol, particularly formaldehyde, but no chemical tests of adequate sensitivity were found. These observations accord with the dissociative chemisorption of methanol18 : the moiety being oxidised to carbon dioxide with no other productslg; other moieties such as COOH may also be less strongly ad~orbed.l~-~l The adsorption occurs on open circuit electrodes,lg but appears to be potential dependent,22 which would explain the observed dependence on the strength of the acid. The active adsorbates supplied by dominant mass transfer of methanol reduce the oxide film on the electrode faster than it can be replaced by reaction with dissolved oxygen or by anodic polarisation.Janwxry , 1983 ELECTROLYTIC POTENTIOMETRY.PART XXVI 75 Conclusions Although weak acids in toluene - methanol mixed solvent can be titrated satisfactorily by the DEP technique when zero-current potentiometry is unsatisfactory, stronger acids engender chemical reaction between the electrode surface and methanol leaving a metallic instead of an oxide surface, destroying the metal - metal oxide function and changing the nature of the electrode process.1° While glass electrodes are evidently suitable in the titration of stronger acids in the mixed solvent, neither polarised nor zero-current antimony - antimony oxide electrodes can be used with any confidence.References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Bishop, E., Mikrochim. A d a , 1956, 619. Bishop, E., Analyst, 1958, 83, 212. Bishop, E., and Short, G. D., Analyst, 1962, 87, 467. Bishop, E., and Short, G. D., Analyst, 1964, 89, 415. Bishop, E., and Short, C. D., Analyst, 1964, 89, 587. Bishop, E., and Dhaneshwar, R. G., Analyst, 1962, 87, 207. Bishop, E., and Dhaneshwar, R. G., Analyst, 1962, 87, 845. Bishop, E., and Dhaneshwar, R. G., Anal. Chew., 1964, 36, 726. Abdennabi, A. M. S., and Bishop, E., Analyst, 1982, 107, 1032. Bishop, E., and Short, G. D., Tulanta, 1964, 11, 393. Moss, M. L., Elliot, J . H., and Hall, R. T., Anal. Chem., 1948, 20, 784. Martin, A. J . , Anal. Chem., 1957, 29, 79. Warner, B. R., and Haskell, W. W., Anal. Chew., 1954, 26, 770. Fritz, J . S., and Lisicki, N. M., Anal. Chew., 1951, 23, 589. Bishop, E., and Hartshorn, L. G., Analyst, 1971, 96, 885. Bishop, E., Cofrk, P., East, G. A., and Thorne, F. C., to be published. Burns, L). T., Townshend, A,, and Carter, A. H., “Inorganic Reaction Chemistry,” Volume 2, Part Podlovchenko, B. I., and Gorgonova, E. P., Dokl. Akad. Nauk SSSR, 1964, 156, 673. Podlovchenko, B. I., Frutnkin, A. N., and Stcnin, V. F., Elektrokhimiya, 1967, 4, 4. Breiter, M. TV., J . Electroanal. Chem., 1967, 14, 407. Breiter, M. W., J . Electroanal. Chem., 1967, 15, 221. Podlovchenko, B. I., and Stenin, V. F., Elektrokhimiya, 1966, 3, 649. A, Ellis Horwood, Chichester, 1981, p. 40. Non-References 2, 3, 4, 5 , 6, 7 and 9 are to Parts 11, V1, XIV, XV, V, V I l I and XXV, respectively, Keceived June 7th, 1982 Accepted J u l y 30th, 1982 of this series.

ISSN:0003-2654    

DOI:10.1039/AN9830800071

出版商:RSC    

年代:1983

数据来源: RSC