摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRYADVISORY BOARD"Chairman: J. M. Ottaway (Glasgow, U.K.)R. Belcher (Birmingham, U.K.)E. Bishop (Exeter, U.K.)D. T. Burns (Belfast, U.K.)L. R. P. Butler (South Africa)E. A. M. F. Dahmen (The Netherlands)L. de Galan (The Netherlands)A. C. Docherty (Billingham, U.K.)D. Dyrssen (Sweden)G. Ghersini (Italy)R. Herrmann (West Germany)J. Hoste (Belgium)M. T. Kelley (U.S.A.)"G. E. Penketh (Wilton, U.K.)"T. B. Pierce (Harwell, U.K.)E. Pungor (Hungary)D. I. Rees (London, U.K.)P. H. Scholes (Middlesbrough, U.K.)S. Siggia (U.S.A.)*J. M. Skinner (Billingham, U.K.)A. A. Smales, O.B.E. (Thornaby, U.K.)"J. D. R . Thomas (Cardiff, U.K.)K. C. Thompson (Sheffield, U.K.)A. Walsh, K.B.(Australia)G. Werner (German Democratic Republic)T. S. West (Aberdeen, U.K.)'P. C. Weston (London, U.K.)'J. Whitehead (Stockton-on-Tees, U.K.)J. D. Winefordner (U.S.A.)P. Zuman (U.S.A.)'G. J. Dickes (Bristol, U.K.)'G. W. Kirby (Glasgow, U.K.)*J. N. Miller (Loughborough, U.K.)G W. C. Milner (Harwell, U.K.)E. J. Newman (Poole, U.K.)H. W. Nurnberg (West Germany)'Members of the Board serving on the Analytical Editorial BoardEditor: P. C. WestonAssistant Editors: Mrs. J. Brew, Mrs. P. A. Fellows, R. A. YoungREGIONAL ADVISORY EDITORSDr. J. Aggett. Department of Chemistry, University of Auckland, Private Bag, Auckland, NEW ZEALAND,Professor L. Gierst, Universit6 Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50,Professor H.M. N. H. Irving, Department of Theoretical Chemistry, University of Cape Town, Ronde-Professor W. A. E. McBryde. Faculty of Science, University of Waterloo, Waterloo, Ontario, CANADA.Dr. 0. Osibanjo. Department of Chemistry, University of Ibadan, Ibadan, NIGERIA.Dr. G. Rossi, Chemistry Division, Spectroscopy Sector, CEC Joint Research Centre, EURATOM, lspraDr. I. Rubeika. Geological Survey of Czechoslovakia, Malostranskb 19, 118 21 Prague 1, CZECHO-Professor J. Rhiicka, 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,Bruxelles, BELGIUM.bosch 7700, SOUTH AFRICA.Establishment, 21020 lspra (Varese), ITALY.SLOVAKIA.DENMARK.Kensington, N.S.W.2033, AUSTRALIA.Professor P. C. Uden, Department of Chemistry, University of Massachusetts, Amherst, Mass. 01003,U.S.A.Editorial: Editor, The Analyst, The Royal Society of Chemistry, Burlington House,Piccadilly, London, W l V OBN. Telephone 01 -734 9864. Telex No. 268001Advertisements: Advertisement Department, The Royal Society of Chemistry, Burlington House,Piccadilly, London, W1 V OBN. Telephone 01 -734 9864. Telex No. 268001The Analyst is published monthly by the Royal Society of Chemistry, Burlington House, London W1V OBN,England. The Analyst cannot be purchased on its own as a current year subscription. 1981 AnnualSubscription price purchased with Analytical Abstracts: UK f168.00, Rest of World f177.00, US$41 6.00, including air speeded delivery. Purchased with Analytical Abstracts plus Analytical Proceedings:UK fl90.00, Rest of World €200.50, US $471.50, including air speeded delivery. Application to mail atsecond class postage rate is pending at Jamaica, N.Y. 11 431. Change of address and orders with paymentin advance to The Royal Society of Chemistry, The Distribution Centre, Blackhorse Road, Letchworth,Herts. SG6 1 HN, England. Air Freight and mailing in the US by Publications Expediting Inc., 200 MeachamAvenue, Elmont, N.Y. 1 1 003. All other despatches outside the UK by Bulk Airmail, and Accelerated SurfacePost outside Europe. PRINTED IN THE UK.Volume 106 No 1267 0 The Royal Society of Chemistry 1981 October 198

ISSN:0003-2654    

DOI:10.1039/AN98106FX037

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

ANALAO 106 (1 267) 1 025-1 136 (1 981 ) October 1981THE ANALYSTTHE ANALYTICAL JOURNAL OF THE ROYAL SOCIETY OF CHEMISTRY1025103610421057107110761082108810961103110911191122112511301136CONTENTSDevelopment and Evaluation o f Selected Assays for Drugs and Drug Meta-bolites in Biological Materials-A. P. De Leenheer and H. J. C. F. NelisMonitoring Exposure t o Toxic Gases in Workplace Atmospheres-D. T. CokerAnalysis of Additives and Process Residues in Plastics Materials-D. C. M.Various Applications o f Functional Group Analysis-G. D. B. van HouwelingenAspects o f the Analysis o f Drugs and Drug Metabolites by High-performanceLiquid Chromatography-G. G. SkellernDetermination o f Mercury Vapour i n Air Using a Passive Gold Wire Sampler-J.E. Scott and J. M. OttawayField Method for the Determination of Aromatic Primary Amines i n Air. PartI. Generation o f Standard Atmospheres o f Amines-D. W. Meddle andA. F. SmithField Method for the Determination o f Aromatic Primary Amines in Air. PartII. Development o f a Sensitive Field Test-D. W. Meddle and A. F. SmithDetermination o f Gold in Tissue and Faeces by Atomic-absorption Spectro-photometry Using Carbon Rod Atomisation-R. M. Turkall and J. R. BianchineDetermination o f Riboflavin and Flavin Mononucleotide in Foodstuffs UsingHigh-performance Liquid Chromatography and a Column-enrichmentTechnique-I. D. Lumley and R. A. WigginsPartVII. Potentiometric Precipitation Titrations-Gunnar Gran, Axel Johanssonand Sten JohanssonSquirrel1Automatic Titration by Stepwise Addition o f Equal Volumes o f Titrant.SHORT PAPERSHigh-performance Liquid Chromatographic Determination o f Chlorpromazineand Thioridazine Hydrochlorides in Pharmaceutical Formulations-A. C.MehtaStudy o f 3-Propyl-5-hydroxy-5-D-arabinotetrahydroxybutyl-3-thiazolidine-2-thione as a Reagent for the Spectrophotometric Determination o fThallium(1)-J.AznBrez, J. R. Castillo and C. LuesmaCOMMUNICATIONFlow Injection Sample Introduction for Atomic-absorption Spectrometry:Applications of a Simplified Model for Dispersion-J. F. Tyson and A. B .ldrisBOOK REVIEWSErratumSummaries o f Papers in this Issue-Pages iv, v, vi, vii, viiiPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class at New York USA.Post OfficAnalytical Sciences MonographsNo. 6 lsoenzyme AnalysisBy D. W. MossThis monograph attempts to drawtogether the most important experimentaltechniques which have resulted fromthe modern recognition that enzymesfrequently exist in multiple molecularforms. This monograph also indicatesthe advantages and limitations inisoenzyme studies of these modernanalytical techniques.Brief Contents :Multiple Forms of Enzymes; Separationof Multiple Forms of Enzymes; SelectiveInactivation of Multiple Forms ofEnzymes; lmmunochemistry of MultipleForms of Enzymes; Catalytic Differencesbetween Multiple Forms of Enzymes;Methods of Obtaining StructuralInformation; Selection of Methods ofAnalysis.Hardcover 171 pp 8%” x 5Q” 0 851 86 800 2 f 12.00 ($32.50)Orders to: The Royal Society of Chemistry DistributionCentre, Blackhorse Road, Letchworth, Herts. SG6 1 HNNotice to SubscribersSubscriptions for The Analyst, Analytical Abstracts and Analytical Proceedingsshould be sent to:The Royal Society of Chemistry, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HN, EnglandRates for 1981 (including indexes)U K/ Rest ofEire USA WorldThe Analyst, and Analytical Abstracts .. .. .. .. f168 $416 f177The Analyst, Analytical Abstracts, and Analytical Proceedings . , f 190 $471.50 f200.50Analytical Proceedings alone* . . .. .. .. .. f30 $70.50 f31.50Analytical Abstracts alone . . .. . . .. .. , . f129.50 $322 f137Subscriptions are not accepted for The Analyst alone*NEW FOR 198

ISSN:0003-2654    

DOI:10.1039/AN98106BX039

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

October, 1981 THEANALYTICAL SCIENCES MONOGRAPH No. 3Pyrolysis -Gas Chromatographyby R. W. May, E. F. Pearsonand D . ScothernMany papers have been published, particu-larly over the past decade, o n aspects o fpyrolysis-gas chromatography. A largenumber o f different types of apparatus havebeen used, o n a wide range o f samples. Thismonograph attempts t o present the availableknowledge in a form useful t o the practisinganalyst, helping in the choice of an appro-priate method and in the avoidance of themore common pitfalls in this, perhapsdeceptively, simple technique.Hardcover 117pp 8t” ~~~ 6” 0 85186 767 7f12.50 (RSC Members f9.50)Orders t o :THE ROYAL SOCIETYOF CHEMISTRY,Distribution Centre,Blackhorse Road, Letchworth,Herts., SG6 1 HNBUREAU OF ANALYSEDSAMPLES LTD.Newham Hall, Newby,Middlesbrough, Cleveland, TS8 SEAadvise newCertified Reference Materialsnow availableBCS 18314 Leaded GunmetalES 579-1 60% Nb Ferro-NiobiumES 580-1 Low C Ferro-ChromiumFor further details, and a copy ofthe new BAS catalogue, telephoneMiddlesbrough 31 721 6or telex 587765 BASRIDQALYST iiiTechniques in Visibleand UltravioletSpectrometryThe titles in this series arise from working partiesset up by the Ultraviolet Spectrometry Group,whose members are drawn both frommanufacturers and from specialist users in industryand universities.Standards inAbsorption SpectrometryEdited by C Burgess and A Knowles for theUltraviolet Spectrometry GroupThis title reviews current practices and ideas inabsorption spectrometry, and makesrecommendations on the choice and use o fstandards, methods o f calibration.and the designand construction of spectrometers and cells.It offers a practical and authoritative guide foranalytical chemists and biochemists and for thoseconcerned with the development o f instruments.January I98 I 240 x I59 mm I68 pagesHb 0412224704 L9.50Standards inFluorescence SpectrometryEdited by J N Miller for the UltravioletSpectrometry GroupFluorescence spectroscopy is a technique of greatsensitivity and selectivity with enormous potentialin chemistry and biochemistry. This book shouldhelp t o promote i t s use by suggesting standardprocedures and materials which will allow users tomake reliable comparisons of their results.November 1981 240 x 159 mm I28 pagesHb 041222500X f8.50Particle Size MeasurementThird EditionT Allen, Postgraduate School o f PowderTechnology, University o f BradfordThis title has established itself over the years as thereference book on particle measurement. Therange o f applications for the technology describedhere is enormous and includes chemicalengineering.metallurgy, pharmaceuticals and foodscience. This new edition, like i t s predecessors,should be of particular interest to the industrialmarket. The readership will include postgraduatestudents and research workers in powdertechnology and engineers in any industry dealingwith products in powder form.Powder Technology SeriesJuly I98 I 538 pages IllustratedHardback 0412 154102 (24.50CHAPMAN AND HALLI I New Fetter Lane, London EC4P 4EE733 Third Avenue.New York N Y I00 I 7A201 for further information. See page x A202 for further information. See page iv SUMMARIES OF PAPERS IN THIS ISSUE October, 1981Summaries of Papers in the IssueDevelopment and Evaluation of Selected Assays for Drugsand Drug Metabolites in Biological MaterialsAn over-all strategy for developing and evaluating assays for drugs in biologicalmaterials is described. Each stage, i.e., starting points, nature of the bio-logical sample, characterisation of the compound, choice of the analyticalmethod, selection of an internal standard, isolation procedure, calibration,evaluation and quality control, are discussed separately. Chromatographictechniques, which are usually preferable to approaches not involving a separa-tion step, such as competitive binding assays, are emphasised.Some contro-versial points concerning the use of internal standards are discussed andspecial approaches for calibration, including an outline of assay criteria to beevaluated, are presented. The general principles are illustrated with examplesfrom the authors’ own experience.graphy; assay evaluationKeywords ; Drug assay; biological materials; assay strategy ; chromato-A. P. DE LEENHEER and H. J. C. F. NELISLaboratoria voor Medische Biochemie en voor Klinische Analyse, R. U.G., AkademischZiekenhuis, 135 De Pintelaan, B-9000 Ghent, Belgium.Analyst, 1981, 106, 1025-1035.Monitoring Exposure to Toxic Gases in Workplace AtmospheresThe basis, objectives and shortcomings of personal monitoring as a means ofboth estimating the level of exposure and preventing the occurrence of overexposure to airborne toxic substances are discussed.Techniques that areavailable, and the circumstances in which they are suitable are reviewed alongwith recent new developments and possible future alternative approaches.Keywords : Toxic gas monitoring; workplace atmospheres ; health and safetyD. T. COKEREsso Europe Inc., Esso Research Centre, Abingdon, Oxfordshire, OX13 0AE.Analyst, 1981, 106, 1030-1041.Analysis of Additives and Process Residues in Plastics MaterialsAnalytical problems, compounded by the increasing number and variety ofadditives used by the plastics industry and the necessity to reduce processresidues and contaminants to the minimum, are described.An outline isgiven of general analysis schemes in use for the examination of PVC com-pounds, polyolefin compounds and acrylics. This is followed by some specificexamples chosen particularly to illustrate the integrated approach, utilisingchemical, physical and complex instrumental methods, that is required tosolve some of the more difficult analytical problems. Methods used includesolvent extraction, titration, ultraviolet, visible and infrared spectrophoto-metry, X-ray fluorescence, gas chromatography, liquid chromatography, thin-layer chromatography and mass spectrometry. The methods are often usedin combination.Keywords ; Plastics analysis ; additives ; process residuesD.C. M. SQUIRRELLImperial Chemical Industries Limited, Petrochemicals and Plastics Division, P.O.Box 8, Welwyn Garden City, Hertfordshire, AL7 1HD.Analyst, 1981, 106, 1042-1056October, 1981 SUMMARIES OF PAPERS IN THIS ISSUEVarious Applications of Functional Group AnalysisThe importance of functional group analysis is demonstrated with a numberof examples in various fields. Special attention is given to the determinationof end groups in several kinds of high and low relative molecular massmaterials.Methods for the determination of the following groups are described:hydroxyl groups (polymers, esters), carboxyl groups (polymers, ester-inter-change products), unsaturated groups (vinyl ester end groups in polymers),anhydride groups (polymers), amino groups (polymers, derivatised silicamaterials), epoxide groups (resins, derivatised silica materials) and quaternaryammonium compounds (surfactants).The methods include derivatisation procedures and several techniques forquantitative determination (spectrophotometry, X-ray fluorescence, coulo-metry and potentiometric and photometric titration),Keywords : Functional group analysis ; polymers ; derivatisationG. D.B. VAN HOUWELINGENAkzo Research, Corporate Research Department Arnhem, P.O. Box 60, Arnhem,The Netherlands.Analyst, 1981, 106, 1057-1070.Aspects of the Analysis of Drugs and Drug Metabolites byHigh-performance Liquid ChromatographyThe biotransformation of a drug may result in a diversity of metabolites withwidely differing physico-chemical properties.With the aid of suitableexamples the applicability of adsorption, partition and paired ion high-performance liquid chromatography to the measurement of drugs and theirmetabolites is discussed in relation to their lipophilicity and polarity.Keywords : Drug analysis ; drug metabolite analysis ; high-performance liquidchromatographyG. G. SKELLERNDrug Metabolism Research Unit, Department of Pharmaceutical Chemistry,University of Strathclyde, Glasgow, G1 1XW.Analyst, 1981, 106, 1071-1075.Determination of Mercury Vapour in Air Using a PassiveGold Wire SamplerThe determination of atomic mercury vapour in air in the range 10-120 pgis described.Mercury vapour was collected by exposing a 1-cm length ofgold wire to the air for 5 min, then thermally desorbed from the wire formeasurement by atomic-fluorescence spectrometry. The results obtainedshowed good agreement with those obtained from an acid permanganate wetsampling system.Keywords : Mercury determination ; air analysis ; gold wire ; atomic-fluorescencespectrometryJ. E. SCOTTAnalytical Laboratory, Philips Hamilton, Wellhall Road, Hamilton, Scotland,ML3 9BZ.J. M. OTTAWAYDepartment of Pure and Applied Chemistry, University of Strathclyde, CathedralStreet, Glasgow, G1 1XL.Analyst, 1981, 106, 1076-1081.vi SUMMARIES OF PAPERS IN THIS ISSUEField Method for the Determination of Aromatic Primary Aminesin Air.Part I. Generation of Standard Atmospheres of AminesA generator for the production of standard atmospheres of eleven aromaticamines, a t very low concentrations, is described. The concentration of theamine produced is determined by gas - liquid chromatography of the hepta-fluorobutyrate derivative using electron-capture detection.October, 1981Keywords : Aromatic primary amine determination ; a i r ; standard atmosphere;gas - liquid chromatographyD. W. MEDDLE and A. F. SMITHLaboratory of the Government Chemist, Cornwall House, Stamford Street, London,SE1 9NQ.Analyst, 1981, 106, 1082-1087.Field Method for the Determination of Aromatic Primary Aminesin Air. Part 11. Development of a Sensitive Field TestA simple field test for primary aromatic amines in air is described in whichthe sample is drawn through an impregnated paper to form a coloured stainproduced by the reaction of aromatic primary amine groups with 4-(dimethyl-amin0)cinnamaldehyde.The test is sensitive to about long of amine andhas been found to be applicable to eleven different amines. It does not dis-tinguish individual compounds and can be affected by airborne alkaline oracidic compounds.Keywords : Aromatic primary amine determination ; a i r ; field methodD. W. MEDDLE and A. F. SMITHLaboratory of the Government Chemist, Cornwall House, Stamford Street, London,SE1 9SQ.Analyst, 1981, 106, 1088-1095.Determination of Gold in Tissue and Faeces by Atomic-absorptionSpectrophotometry Using Carbon Rod Atomisation-4 simple and accurate atomic-absorption spectrophotometric method isdescribed for the measurement of gold in tissue and faeces following chryso-therapy. Samples and standards are disrupted with a quaternary ammoniumhydroxide solubiliser in toluene, diluted with isobutyl methyl ketone, andanalysed directly using a carbon rod atomiser. The detection limit for goldis 50 & 25 pg and the calibration graph is linear up to approximately 3 ng.Concentrations of gold in tissue determined by this method correlate wellwith results obtained by flame atomic-absorption spectrophotometry and byneutron-activation analysis.Keywords : Gold determination; carbon rod atomic-absorption spectrophoto-metry; tissue analysis; faeces analysisR. M. TURKALL and J. R. BIANCHINEDepartment of Pharmacology, College of Medicine, Ohio State University, Columbus,Ohio 43210, USA.Analyst, 1981, 106, 1096-1102

ISSN:0003-2654    

DOI:10.1039/AN98106FP125

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

October, 1981 SUMMARIES OF PAPERS IN THIS ISSUEDetermination of Riboflavin and Flavin Mononucleotide inFoodstuffs Using High-performance Liquid Chromatography and aColumn-enrichment TechniqueA high-performance liquid chromatographic method has been developed forthe determination of riboflavin and flavin mononucleotide (FMN) using aC,, reversed-phase column packing material. Mean recoveries of 100%(&5y0) for riboflavin and .102yo (+2y0) for FMN were obtained. Multipledeterminations of the vitamin in two samples indicated a precision of & 5%.The results agree well with those obtained by microbiological assay. Ribo-flavin and FMN are eluted within 9 min, and the total time for the analysisof most samples is 3.5 h. The lifetime of the reversed-phase column isfar greater than that of silica columns.After 500 sample loadings nodeterioration of column efficiency occurred. The method is suitable for thedetermination of riboflavin down to the 0.01 mg per 100 g level. For sampleswith a lower vitamin content trace enrichment can be used, which involvesonly a few minutes’ extra analysis time.Ribojavin determination ; flauin mononucleotide determination ;high-performance liquid chromatography ; microbiological assay; traceenrichmentKeywords1. D. LUMLEY and R. A. WIGGINSLaboratory of the Government Chemist, Cornwall House, Stamford Street, London,SE1 9NQ.Analyst, 1981, 106, 1103-1108.viiAutomatic Titration by Stepwise Addition of Equal Volumes ofTitrant. Part VII. Potentiometric Precipitation TitrationsThe method used throughout this series to add the titrant stepwise with equalvolumes each time is well suited to precipitation titrations. The advantagethat one has to wait for equilibrium to be reached a t only a few points is ofspecial importance.Two methods for calculation of the equivalence volume in precipitationtitrations have been evaluated.One is based on the solution of a set oflinear equations and the other is an extended version of the Gran I method.The methods are characterised by the following facts: (1) side-reactions canbe accounted for; (2) the values of the solubility product and the stabilityconstants of complexes may be unknown; (3) the solutions may be so dilutethat a complete precipitation is not obtained; and (4) no accurate calibrationof the electrode couple is required.The calculation methods have been tested on the determination of chlorideby titration with silver nitrate solution.The two methods agree closely andthe errors in the calculations are negligible. The over-all errors in the deter-minations are 0.1-0.2~0 a t chloride concentrations down to lo-* M. Atlower concentrations the relative errors are greater.The methods have also been tested on the determination of sodium fluoridewith lanthanum nitrate in ( a ) neutral and unbuffered solution, ( b ) acidicsolution, (c) solution buffered with acetic acid - acetate and ( d ) solutionbuffered with formic acid - formate. The calculations were performed onexperimental curves according to Lingane.For ( a ) and ( b ) good agreementbetween calculated and expected values were obtained. In the acetate-buffered solution too high a value was obtained in spite of the fact that’acetateand mixed acetate - fluoride complexes were considered. For ( d ) the buffercapacity was too small to give satisfactory results.Keywords : Gran I method; potentiometric precipitation titrations ; sodiumfluoride determination ; chloride determinationGUNNAR GRAN, AXEL JOHANSSON and STEN JOHANSSONDepartment of Analytical Chemistry, The Royal Institute of Technology, S- 100 44Stockholm, Sweden.Analyst, 1981, 106, 1109-1118viii SUMMARIES OF PAPERS IN THIS ISSUEHigh-performance Liquid Chromatographic Determination ofChlorpromazine and Thioridazine Hydrochlorides inPharmaceutical FormulationsShort PaperOctober, 1981Keywords : Chlorpromazine hydrochloride determination ; thioridazine hydro-chloride determination ; pharmaceutical formulations ; high-performanceliquid chromatographyA.C. MEHTAPharmacy Department, The General Infirmary, Leeds, LS1 3EX.Analyst, 1981, 106, 1119-1122.Study of 3- Propyl-5- hydroxy- 5- D- arabinotetrahydroxybutyl-3-thiazolidine-2-thione as a Reagent for the SpectrophotometricDetermination of Thallium( I)Short PaperKeywords : 3-Propyl-5-hydroxy-5-~-arabinotetrahydroxybutyl-3-thiazol-idine-2-thione reagent ; spectrophotometry ; thallium (I) determinationJ. AZNAREZ, J. R. CASTILLO and C. LUESMADepartment of Analytical Chemistry, Science Faculty, University of Zaragoza,Zaragoza, Spain.Analyst, 1981, 106, 1122-1125.Flow Injection Sample Introduction for Atomic-absorptionSpectrometry: Applications of a Simplified Model for DispersionCommunicationKeywords : A tomic-absorption spectrometry ; flow injection ; sample intro-duction ; dispersion modelJ.F. TYSON and A. B. IDRISDepartment of Chemistry, Loughborough University of Technology, Loughborough,Leicestershire, LEI 1 3TU.Analyst, 1981, 106, 1125-1129THE ANALYST October, 1981READER ENQUIRY SERVICEFor further information about any of the products featured in the advertise- gments in this issue, please write the appropriate A number in one of the r:Postage paid if posted in the British Isles but overseas readers must affixa stamp.boxes below. nPostagewill bePaid byLicenseeH i l lDo not affix Postage Stamps if posted inGt. Britain. Channel Islands or N. Ireland(Please use BLOCK CAPITALS)NAME .........................................................................................................................................................................................OCCUPATION .........ADDRESS ....................... .........................................................................................................................................SECOND FOLDI ~~~ ~ I BUSINESS REPLY SERVICELicence No. W.D. 106Reader Enquiry ServiceThe AnalystThe Royal Society of ChemistryBurlington HousePiccadilly London W1 E 6WFENGLAND21THIRD FOL

ISSN:0003-2654    

DOI:10.1039/AN98106BP129

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

OCTOBER 1981 The Analyst Vol. 106 No. 1267 Development and Evaluation of Selected Assays for Drugs and Drug Metabolites in Biological Materials* A. P. De Leenheert and H. J. C. F. Nelis Laboiatoria vooi Medische Biochemie en VOOY Klinische A nalyse, R. U.G., A kademisch Ziekenhuis, 135 De Pintelaan, 3-9000 Ghent, Belgium An over-all strategy for developing and evaluating assays for drugs in biological materials is described. Each stage, z.e., starting points, nature of the bio- logical sample, characterisation of the compound, choice of the analytical method, selection of an internal standard, isolation procedure, calibration, evaluation and quality control, are discussed separately. Chromatographic techniques, which are usually preferable to approaches not involving a separa- tion step, such as competitive binding assays, are emphasised.Some contro- versial points concerning the use of internal standards are discussed and special approaches for calibration, including an outline of assay criteria to be evaluated, are presented. The general principles are illustrated with examples from the authors’ own experience. Keywords : Drug assay ; biological materials ; assay strategy ; chromato- graphy ; assay evaluation There is a continuing increase in the number of papers being published in analytical chemistry.l However, analytical work, especially that of a clinical nature, is still the subject of mis- understandings and prejudices. These are attributable to a striking ignorance by many scientists of the basic difficulties involved in the analysis of drugs and endogenous compounds in biological materials, and especially the difficulties imposed by the complexity of biological matrices.Measurements can rarely be applied directly to blood or urine and extensive sample pre- treatment and/or chromatographic separation are usually required. Unlike purely “chemical assays,” the drug might not be directly available but tightly bound to proteins or other macromolecules. Losses inevitably occur during isolation and require compensation by use of appropriate internal standards. In clinical assays more thorough evaluation of assay criteria is needed than with “chemical asays” because the results are crucial in the inter- pretation of pathological states and might be used in the diagnosis and treatment of patients. It is therefore important that each assay should be capable of measuring therapeutic levels of a drug in a realistic matrix, e.g., serum or urine, and not simply for artificially high con- centrations in an unreal medium such as aqueous solutions.In this paper, we discuss our basic strategy for the development of new assays for drugs and possible metabolites in biological materials in terms of the different steps in the analytical approach, illustrated with selected examples from our experience. In this regard it should be remembered that a drug assay should always be dynamic in nature and subject to con- tinouous modification and improvement during its application. Development of a Drug Assay Starting Point A thorough definition of the need to set up a new drug assay should always precede experi- mental work.Requests from physicians or drug companies should be extensively docu- * Plenary Lecture presented a t the Joint NL - UK Symposium on Quantitative Organic Analysis, Noordwijkerhout, The Netherlands, April 22-24, 1981. t To whom correspondence should be addressed. 10251026 Analyst, Vd. 106 mented by an exhaustive computerised literature search. Assays for proposed new drugs are essential with a view to establishing protocols for registration. Procedures for deter- mining levels of the compound in serum, urine and tissues are required in order to evaluate its fate in the organism, in terms of pharmacokinetics and metabolism. Therapeutic drug monitoring is a field that is steadily gaining interest.Uncontrollable irregularities in the absorption and distribution of drugs often result in poor bioavailability and hence unsatisfactory therapeutic response. In such cases, monitoring of plasma levels in treated patients is necessary to optimise dosage schedules. Drugs that require thera- peutic drug monitoring include compounds with a narrow therapeutic index, e.g., cyto- statics, and substances that are apt to produce serious side-effects when their plasma levels reach a toxic range, e.g., antiarrhythmics. Certain pathological conditions such as renal diseases enhance the above risks because of drug accumulation and warrant close monitoring of patients. Apart from clinical considerations, the originality of the analytical problem itself should also be defined, as existing methods are often unsatisfactory and need major improvements to sample size , sensitivity, precision, specificity and analysis time.The most logical approach in the developmental stage of an assay involves opposite chronology to the final protocol. Although isolation obviously precedes the actual measure- ment of a compound, it can be more convenient to develop the analytical technique first. This is followed by selection of the internal standard and development of the isolation pro- cedure. DE LEENHEER AND NELIS: SELECTED ASSAYS FOR DRUGS Nature of the Biological Sample Several analytical problems and interferences from endogenous compounds can be antici- pated and dealt with if the normal constitution and the peculiarities of the biological matrix are known and understood.However, dramatic changes in the matrix may occur in patho- logical conditions. As an example, samples from uraemic patients, an important group for which drug monitoring is often essential, contain numerous unidentified substances that are absent in normal humans. In addition to potential interferences, attention should be paid to the possible binding of drugs by macromolecules. With the exception of saliva and urine, most biological materials contain at least two forms of a drug, i.e., free and protein bound. Occasionally, binding to other macromolecules or other interactions, such as chelation, may occur. Concern with the phenomenon lies in the fact that only “free” drugs are pharmacologically active. How- ever, nearly all assays determine “total” drug contents and to achieve this, special measures are often needed to release the compound quantitatively from its binding sites.Removal of proteins and other substances such as lipids, salts and pigments (urine) is also obligatory in order to protect equipment, e.g. , columns, from deterioration. Finally, particular samples including tissues and faeces pose specific problems with respect to hetero- geneity and interferences. Characterisation of the Compound to be Assayed The selection of a particular analytical technique and isolation method is dependent on the physico-chemical characteristics of the drug to be assayed. These include pK values, lipophilicity, e.g., expressed as partition coefficient for octanol - water at different pHs, solubility in aqueous and organic media, polarity (presence of functional groups), and stability towards air, light , heat and pH conditions. Spectral characteristics [ultraviolet (UV) maxima, molar absorptivity], are also important.As mentioned above, biological factors such as binding are significant for sample prepara- tion and a knowledge of pharmacological properties can be extremely valuable. Thera- peutic levels in blood and urine are also useful data for establishing the sensitivity require- ments of the procedure. Finally, information about other drugs that may be co-administered will contribute to predicting expected interferences. Choice of Analytical Method For each drug assay, the correct choice of analytical methodology is a challenge as, in general, few methods have equivalent performance.Although many techniques are available, only a few are adequate from the standpoint of sensitivity and specificity.October, 1981 AND DRUG METABOLITES IN BIOLOGICAL MATERIALS 1027 Chemical techniqzles without a separation step Chemical methods not involving a separation step include spectrophotometry and fluori- metry but, although convenient, both approaches have poor selectivity. Endogenous interferences are likely to occur and it is often impossible to distinguish parent compounds and metabolites. In addition, spectrophotometry, unlike fluorimetry, usually lacks the sensitivity required to determine realistic drug levels. Competitive binding assays The Yalow and Berson principle of competitive binding2 was initially applied only to the determination of biopolymers, such as proteins. Organic molecules of low relative mole- cular mass, e.g., drugs, have to be covalently linked to proteins to generate immunogenic properties.According to the nature of the protein that couples with the drug-hapten in the competitive reaction, those assays can be classified into immunoassays, using drug-specific antibodies, and competitive binding assays (CPB), using a variety of other proteins. Radioimmunoassays (RIA) employ radioactively labelled ligands and involve separation of antibody-bound and free labelled drug-hapten. Enzyme immunoassays (EIA) are based on labelling drug-haptens with specific enzymes and can be further subdivided into so-called heterogeneous (ELISA) and homogeneous (EMIT) procedures.Other immunoassays, eg., fluorescence immunoassay (FIA), are also being developed. In other competitive protein binding assays, competition occurs between unlabelled drug and tritium-labelled tracer for drug target proteins. More and more drugs are now analysed on a routine basis using the above techniques, partly as a result of tremendous commercial pressure and publicity. The widely available ready-to-use kits allow rapid processing, often without any sample pre-treatment, of large numbers of samples with relatively high sensitivity and can be handled by technicians without special training. However, the analytical specificity and hence accuracy of many com- mercially promoted assays is sometimes questionable. Structurally related and even other compounds can interfere significantly, e g ., the cross-reactivity of diphenylhydant~in~ and diazepam* in the protein-binding assay of thyroxine (T,) . The stereochemically structural similarity between those drugs and T, was discovered following X-ray crystallographic examination. In addition, although external quality control surveys show that within-laboratory vari- abilities of immunoassays are controllable, between-laboratory precision is usually extremely poor. Variations are ascribed to the unpredictable quality of antisera, the properties of which are difficult to define. Combination of competitive binding assays with high-resolution separation techniques, e.g., high-performance liquid chromatography (HPLC) , can yield significant improvements in analytical performance. The chromatographic step enhances the specificity, while the sensitivity of the immunoassay is maintained, as seen for the RIA determination of certain pep tide^.^ Chromatographic methods Recent developments in column chromatography have led to the present popularity of HPLC.Microparticulate packing materials and the development of stable bonded phases have improved the efficiency of column separations tremendously. Reversed-phase materials show great potential for the analysis of drugs in biological materials, as polar compounds are generally not retained on this type of stationary phase and will therefore not interfere with the analysis of more lipophilic drugs. The sensitivity of the currently available detectors (UV absorbance, fluorescence, electro- chemical) is insufficient for ultratrace analysis, but the use of new detectors (such as laser fluorescence) may eventually overcome this drawback.Detectors based upon direct coupling of liquid chromatography with mass spectrometry (LC - MS) are being tested but so far they do not show the same promise as their gas chromatography - mass spectrometry (GC - MS) counterparts. There is frequently ignorance on other ways of enhancing the intrinsic sensitivity of an HPLC system. As the detectability of a compound is a function of the peak height, which itself is directly related to the degree of dilution of the compound in the column, the use of shorter columns with smaller internal diameters than most commercial columns generally1028 DE LEENHEER AND NELIS: SELECTED ASSAYS FOR DRUGS Analyst, VoZ.106 results in improved sensitivity. The theoretical concepts underlying this effect have been thoroughly described by Karger et a1.' Contrary to some beliefs: columns with reduced dimensions can be packed easily. In this respect, we found a 15 x 0.32 cm i.d. column to be a satisfactory compromise in terms of sensitivity, efficiency and loadability. Finally, HPLC is ideally suited for the analysis of unstable compounds : no derivatisation is necessary and separations can be carried out at room temperature, thus preventing possible thermal degradations. Despite the growing importance of HPLC, gas chromatography (GC) continues to be of value.9 Its inherent disadvantages, i.e., obligatory derivatisation of polar compounds lacking sufficient vapour pressure and potential thermal instability of certain compounds, are largely compensated for by sensitivity and specificity.Capillary GC offers excellent resolution, unprecedented in chromatography, and is suitable for quantitative work. However, its performance is strongly affected by technological factors, specifically the configuration of the injection device1* and the quality of the columns used. Instruments specially designed for use with capillary columns are essential for high- performance work and have only recently appeared on the market. Nitrogen - phosphorus and electron-capture detectors provide greatly improved sensitivity and selectivity for particular compounds. The sophisticated GC - MS approach is capable of further advancing the frontiers in this respect and quantitative GC - MS has become an important tool in drug analysis.l1 Selection of an Internal Standard Most errors in drug analysis occur during extraction, or during subsequent evaporation of organic solvents and derivatisation.12 Suitable internal standards should minimise these errors by compensation, provided that they are added at the earliest possible stage. A basic requirement is close structural similarity with the drug to be assayed. Too often, however, internal standards do not bear any relationship to the drug of interest, and this practice should be strongly discouraged. In GC - MS, where selected ions of the drug and an internal standard are monitored, choice of the latter is critical. Labelled analogues are often preferable (isotope dilution MS), but there is controversy as to the superiority of this kind of internal standard.13 Although it behaves chemically like the unlabelled compound, the position of the label may affect some of its properties, e.g., pK values.Also, the MS fragmentation pattern may undergo changes due to an isotope effect. It is common practice to add a labelled analogue as a carrier to prevent losses of low amounts of unlabelled compound, e.g., by ad~orpti0n.l~ This effect is well documented in radiochemistry,14s15 but its relevance in isotope dilution MS is question- able at present. In no way should the use of a carrier make up for the neglect of other measures to bring an assay under control. Further, labelled analogues are usually contami- nated by some unlabelled substance, which has a systematic contributory effect.For all of these reasons, some workers claim homologues to be superior, even for GC - MS work. Frequently, a homologue will yield fragment ions with the same m/z value as the drug of interest, thus allowing single-ion monitoring. Usually, they are more readily avail- able or cheaper to prepare than labelled analogues. In general, the use of an internal standard belonging to the same chemical class as the drug of interest often presents more problems than is commonly assumed. Even when there are close structural similarities, such as with chemical homologues, pronounced differences in physico-chemical behaviour may arise, as exemplified later. Isolation Procedurels Prior to analysis, the drug has to be isolated from its biological matrix by methods depen- ding on the nature of the sample and the analytical method chosen. Sample pre-treatment is also usually needed to remove interfering substances and to protect equipment from deterioration.Concentrating the extract to a small end-volume may be helpful in increasing sensitivity . Conventional approaches for liberating a drug from binding proteins include precipitation with acids or organic solvents, dialysis and extraction of the drug in combination with salt addition and/or pH adjustment. Even a simple dilution step may be sufficient to dissociate complexes.October, 1981 AND DRUG METABOLITES IN BIOLOGICAL MATERIALS 1029 The commonest way of isolating drugs from a biological medium is by double-phase solvent extraction.Selection of a suitable solvent depends on extraction efficiency and selectivity, the latter with respect to co-extraction of interferents. Partition of a drug into an organic solvent is favoured by lipophilicity, which is at its maximum in the un-ionised form and is pH dependent. Compounds of high polarity are difficult to extract and require strongly polar, and hence non-selective, solvents. Sometimes, salting-out may increase the efficiency, but a special approach to extract charged species consists of adding a counter ion and extracting the lipophilic ion pair formed. An alternative to solvent extraction for polar drugs is the use of a column of alumina or ion exchanger. Interfering materials are washed out prior to elution of the compounds preferentially with an organic solvent.The technique is usually characterised by nearly quantitative recoveries and excellent reproducibility. A less frequently used but less reliable variation to this approach is adsorption on to solids, e.g., charcoal, directly suspended in the sample. Calibration Samples of drug-free material (serum, urine), containing known drug concentrations, should always be employed for calibration. Alternative “external” approaches based on assaying or even just injecting (in the case of chromatography) pure standard solutions are unacceptable because they ignore the specific peculiarities of the biological matrix. However, even drug- supplemented samples are subject to criticism, as an externally added drug might indeed not be present in the same physico-chemical state as an in vivo delivered drug. In chromatography, peak-height or peak-area ratios (compound to internal standard) are usually plotted against concentration.The calibration graphs are calculated by linear regression analysis but in order to avoid inaccuracies near the origin weighted regression analysis should be performed. This follows because carrying out linear regression of y upon x implies, among other constraints, that there is no variance in x and a constant variance in Some workers propose special approaches to extend the linear range artificially,l* but much greater accuracy is obtained by calculating the actual non-linear calibration graph by using weighted polynomial regression a n a l y s i ~ . ~ ~ ~ ~ ~ A computer program calculates the different poly- nomials by means of matrix mathematics and the residual variances around the curves are used in an F-test to select the best fitting polynomial.This approach has been applied in our laboratories to isotope dilution MS in general and to radio-ligand assays of thyroid hormones and CPB of 25-hydroxycholecalciferol.20 y.17 In isotope dilution MS, non-linear calibration graphs are frequently obtained. Evaluation Several criteria have to be evaluated in order to check the reliability and the over-all performance of an assay. Problems involved in linearity and calibration have been discussed above. The extraction recoveries of the drug and internal standard provide useful information and can be calculated either by running radioisotopically labelled compounds through the procedure or by assaying drug-supplemented samples by an “external” calibration method.Assessment of accuracy is a more fundamental problem because the “true” value can never be known with absolute certainty. Various parameters such as linearity and recovery indirectly provide some information about the accuracy. Comparison of the assay with other existing methods, e.g., a reference method, is also very relevant. Precision or reproducibility of an assay is defined as the coefficient of variation (relative standard deviation) of the results at a certain drug concentration. Both within-day and between-day reproducibilities can be evaluated. Sensitivity is usually expressed as the detection limit obtained under the conditions used. The issue of analytical specificity has already been alluded to.Drugs that are thera- peutically combined with the drug of interest should be checked for possible interferences or cross-reactivity. Blank samples should be included in each run in order to detect possible endogenous interfering substances. The term analytical specificity also covers more funda- mental aspects. For example, in GC - MS, careful selection of ions to be monitored may1030 DE LEENHEER AND NELIS: SELECTED ASSAYS FOR DRUGS Analyst, vol. 106 affect the specificity considerably. It has been shown that the relative molecular masses of known compounds are not randomly distributed but that the frequency of occurrence is maximal between values of 200 and 350. Moreover, most compounds possess even rather than odd relative molecular masses (ratio 4 : 1).Millard21 states that, as a result, maximum specificity will be obtained by selecting ions whose masses exceed roughly 350. Also, fragment ions with even instead of more common odd m/z values should be preferentially monitored.21 Two comments can be made in this respect. Quality Control Routine drug assays should be regularly checked for reliability by using internal quality control schemes. This involves the preparation of serum pools to which known amounts of drug are added, but it is essential that the samples thus prepared can be stored for long periods under conditions of stability. With each determination on patients’ samples a control sample is also analysed and the result is interpreted in terms of established limits of variation.Inter-laboratory surveys on a similar basis are currently organised by different institutions and are often imposed by law. Control samples are distributed to the laboratories partici- pating in such an external quality control programme and the results are statistically evaluated. The status of quality control in the UK has recently been reviewed.22 Applications The above basic strategy has been used for many years in our laboratories for the develop- ment of new specific assays for drugs and metabolites in biological materials. This is illustra- ted below by three practical examples, viz., the determination of bromhexine, doxycycline and 5-fluorouracil, 5-fluorodeoxyuridine and 5-fluorouridine. Each step is described in terms of the general principles discussed above.Determination of Bromhexine in Plasma Starting points ment of respiratory infections. circulating levels of the drug in plasma. hitherto, mainly tsGicause of the lack of suitable analytical methods. recommended dosage schedule could be considered as unreliable. technique was required to deal with this analytical challenge. Bromhexine is a mucolytic drug that is frequently combined with antibiotics in the treat- Rapid metabolism probably accounts for the extremely low No pharmacokinetic parameters were available Consequently, the Hence, a highly sensitive Characterisation of the drug substituents and a primary amino group in its structure. philic and soluble in non-polar organic solvents. Chemically, bromhexine is a hydrochloride of a basic substance, containing two bromine The free base is reasonably lipo- Analytical techniques A chromatographic method was desirable to separate the parent drug from metabolites and interfering endogenous compounds.HPLC with UV detection was dismissed in view of the relatively poor chromophore of the molecule and the extremely low concentrations to be detected. The low polarity of the molecule makes it very suitable for GC but, although it can be chromatographed without derivatisation, trifluoroacetyl derivatives were prepared for reasons of ~electivity.~3 As the compound contains electronegative groups, an electron- capture detector (ECD) (nickel-63) was chosen to ensure sensitivity. Details of the pro- cedure can be found elsewhere.24 Concentrations below the detection limit could still be measured by GC- MS with selected ion monitoring.25 The same technique allowed the detection of different metabolites of the drug in urine. Structures were assigned for these on the basis of the recorded mass spectra of their trifluoroacetyl derivatives.The choice of GC as the analytical technique was straightforward for several reasons.October, 1981 AND DRUG METABOLITES IN BIOLOGICAL MATERIALS 1031 Internal standards A previously described GC method for determining the drug in milk and tissues of treated animals employed a structurally unrelated insecticide as an internal standard.2* However, we synthesised various homologues of br~mhexine~~ and selected the N-propyl derivative as an internal standard on the basis of its favourable chromatographic properties.The GC - MS method25 is based on the isotope dilution approach. Trideuterated brom- hexcne was synthesised for use as the inimakstandzrd. - Ions at mJz--2% and 308 were - chosen for monitoring the drug and the internal standard, respectively. Non-specific inter- ferences from co-extracted endogenous substances prevented the use of ions at lower m1.z values (112 and 115, respectively). Isolation method Bromhexine can be extracted from plasma with a non-polar organic solvent but alkalinisa- tion is helpful as the free base is more lipophilic. Extraction was therefore carried out with hexane in sodium hydroxide - triethanolamine medium, the triethanolamine being necessary to avoid a low recovery of the internal standard.The same isolation method was applicable in the GC - MS technique. Evaluation The GC assay with electron-capture detection (ECD) was found to be linear over the two concentration intervals studied, i.e., 1.6-12.7 and 7-115.8 ng ~ m - ~ . These cover the thera- peutic range for oral administration of the drug. An extraction recovery of 90.1 r f 5.7% (n = 12) was calculated for bromhexine in the 6.3-53.1 ng ~ m - ~ range. The within-day coefficient of variation was 6% (2 = 4.3ng~rn-~; n = 8) whereas the between-day value averaged 8.6% (2 = 11 ng cm-3; n = 13). However, a substantial difference in the between- day reproducibility of both methods (GC - ECD and GC - MS) was found for levels below 10 ng ~ m - ~ . The average coefficient of variation for the GC - ECD assay, as determined at five different concentrations, was 16% compared with 3.5% for the GC - MS method.The GC-MS assay was also clearly superior to the GC-ECD assay from the standpoint of sensitivity. The detection limit of the latter was 1 ng cm-3 but a more realistic sensitivity limit was 3 ng ~ m - ~ . The GC - MS method permitted quantitation of down to 0.5 ng ~ m - ~ in plasma. Applications and conclusions In order to obtain realistic pharmacokinetic data following the empirical dosage schedule, plasma levels as a function of time were studied in human volunteers24 and dogs.27 The results confirmed the low bioavailability of bromhexine. The GC - MS method was used to study the fate of the drug in horses% with similar conclusions on low bioavailability.As the existence of several metabolites was demonstrated, biotransformation probably causes the rapid disappearance of bromhexine from plasma. It was concluded that, provided that the metabolites are pharmacologically inactive, the recommended dosage regimen is sub- therapeutic and should be readjusted. Determination of Doxycycline in Serum, Urine, Tissues and Faeces Starting points Considerable doubts existed about the metabolism of doxycycline, a semi-synthetic tetra- cycline antibiotic. Some biological observations suggested the possibility of biotransforma- tion but metabolites had not been isolated or analytically demonstrated. Hence, improved methodology for the analysis of this drug was desirable because existing procedures (fluori- metric, microbiological) lacked specificity.Characterisation of the compounds Unf avourable physico-chemical properties seriously complicate the analysis of this drug in biological materials. Tetracyclines are amphoteric, very soluble in water and almost insoluble in most organic solvents. Their polyfunctional structure accounts for a high degree of polarity, although doxycycline is more lipophilic than other members of the group. In1032 DE LEENHEER AND NELIS: SELECTED ASSAYS FOR DRUGS Analyst, VoE. 106 addition, these compounds are unstable at different pHs (alkaline, weakly acidic). More promising from the analytical point of view is the presence of an extended chromophore (6 = 140001mol-1cm-l at 350nm). In serum at therapeutic concentrations of a few micrograms per cubic centimetre, over 80% of the drug is reversibly bound to proteins.Analytical technipe Instability in basic media and the low chemical reactivity of some substituents prevent appropriate derivatisa- tion. On the other hand, their favourable spectrometric characteristics led us to investigate HPLC with detection at 350 nm. Although tetracyclines display a more intense maximum at 270 nm, the former wavelength was chosen for reasons of selectivity. Efficient liquid chromatography of doxycycline turned out to be possible only under rigorous conditions. Poor HPLC properties (tailing, low efficiency) are a direct consequence of the peculiar physico-chemical nature (amphotericity, polarity, low diffusion coefficient) of the compound. However, one particular reversed-phase material, viz., an octyl-bonded phase, was superior to all other packing materials tested.% Classical solvent mixtures, such as water - methanol or water - acetonitrile, could not be used and the mobile phase had to be modified with acids for satisfactory efficiency.Maximum retention occurred at pH 2.1, where doxycycline exists in its mono-cationic form. Therefore, retention mechanisms can mainly be rationalised in terms of ionic, rather than hydrophobic, interactions.30 Columns with reduced dimensions (10 x 0.2 ern i.d.) were used to optimise the sensitivity. GC of the polysubstituted tetracyclines is difficult and inconvenient. Internal standard Three commercially available tetracyclines were tested for use as internal standards. They all eluted from the column more or less close to doxycycline but initially none of these derivatives proved satisfactory.Metacycline was not extractable by the adopted procedure (see below). 6-efli-Doxycycline, which most closely resembles doxycycline in structure, yielded approximately 20-30y0 lower recoveries; this was due in part to adsorption on to glass. Demeclocycline epimerised during evaporation of serum extracts in ethyl acetate. However, ascorbic acid inhibits this reaction and with this additive demeclocycline is acceptable as an internal standard. Isolation method31 The polar and amphoteric nature of tetracyclines leads to substantial difficulties in their isolation by double-phase solvent extraction. Only moderately polar solvents, e.g., ethyl acetate, were found to be applicable.Serum samples were buffered at pH 6 near the isoelectric point of doxycycline, where it displays minimal aqueous solubility. Only one buffer, an uncommon phosphate - sulphite system, yielded satisfactory, clean extracts. Sulphite is assumed to form conjugates with serum pigments and keep them in the aqueous phase. The high concentration of salts (2 M for both species) ensured dissociation of the drug - protein adducts and probably also contributed to the efficiency by a salting-out effect. Evaluation Within- day and between-day coefficients of variation were 1.8% (Z = 2.5 pg ~ r n - ~ ; n = 10) and 4.8% (Z = 2.6 pg cm-3; n = 9), respectively. The extraction recovery averaged 87.7 -& 4.3% (n = 12). was obtained, which is far beyond the range of therapeutic levels.Linear relationships were observed over the therapeutic range (0-6 pg ~ m - ~ ) . A detection limit of 50 ng Application to other biological materials a homogenisation step had to be included in the analysis. Hydrochloric acid is excellent for this purpose. Apart from giving a pH where doxycycline is stable, it has the additional advantage of dissociating chelates, especially as chelate formation between doxycycline and divalent cations may reduce the extractability of the drug from faeces. A pre-extraction stage with diethyl ether removes interfering lipids. Doxycycline was also assayed in a In order to quantify doxycycline in heterogeneous materials such as tissues32 andOctober, 1981 AND DRUG METABOLITES IN BIOLOGICAL MATERIALS 1033 variety of tissues obtained from patients who had received the drug parenterally before surgery.No modifications of the procedure were necessary to analyse urine samples for doxycycline and metabolites were not detected.= “Free” doxycycline concentrations were measured in saliva. Although samples of this fluid can be injected directly on to the column, the doxycycline levels were too low to be detected this way and the normal extraction and concentration procedure was therefore followed. A similar procedure, based on HPLC, was developed for the determination of minocycline, another semi-synthetic tetracycline.@ Unlike doxycycline, this compound was found to be metabolised in humans, as several polar by-products were found in urine.= Determination of 5-Fluorouracil, 5-Fluorodeoxyuridine and 5-Fluorouridine in Serum and Urine Starting points 5-Fluorouracil (FU) and 5-fluorodeoxyuridine (FUdR) are useful cancer chemotherapeutic agents.Until recently, existing analytical methods lacked the sensitivity for studying the disposition of these drugs over an extended time period following administration. They are rapidly distributed and excreted, resulting in low circulating blood levels. Dosage regimens remained largely empirical and were not supported by actual pharmacokinetic data. Characterisation of the compounds The most striking physico-chemical property of FU and FUdR is their high polarity and the compounds act as weak acids. Protein binding is non-existent and therapeutic levels are often in the ng ~ m - ~ range.Analytical techniques Several GC methods had been described for the determination of FU, but they all had serious disadvantages. They either lacked sensitivity, employed unsuitable internal standards or involved silylation of the compound, a reaction which is difficult to control. Few assay methods existed for FUdR. Several major improvements in the GC of these drugs were introduced. Stable butyl (FU)35 and peralkyl ( FUdR)36 derivatives, showing excellent GC properties, were prepared. Sensitivity of the assays was greatly enhanced by use of a multiple ion detector (FU)37 or a nitrogen - phosphorus detector (FUdR).= Another con- siderable gain in sensitivity and specificity was obtained in the FUdR analysis by replacing packed colums~ with glass capillary systems.3s Ultimate limits in this respect were reached in a new GC - MS method, based on chemical ionisatioH of the compound with ammonia gasq0 The use of an epimeric form of FUdR as internal standard permitted single-ion monitoring, which further contributed to the excellent sensitivity of the procedure. Internal standard Close structural analogues of FU, viz., 5-chlorouracil or ls0-labelled FU (isotope dilution MS), were chosen as internal standards.37 Similarly, 5-chlorodeoxyuridine was used for the FUdR detex~nination,~8*~9 whereas the 3’-epimer served as an internal standard in the CI - MS procedure.40 Isolation Problems concerning the isolation of these compounds from aqueous media are obviously associated with their pronounced polarity.They are not readily extracted by double-phase extraction techniques unless very polar non-selective solvents are used.Column extraction overcomes these diffic~lties.~7+ At pH 10, the compounds occur almost exclusively as mono-anions and are therefore retained on an anion-exchange resin. Subsequent elution was carried out with methanolic acetic acid. This approach was applied successfully to plasma samples. The presence of an excess of salts was deleterious to both column performance and derivatisation and pigments also interfered The analysis of urine presented additional problems.1034 DE LEENHEER AND NELIS: SELECTED ASSAYS FOR DRUGS AnahySt, Vd. 106 with the chromatography. Two further clean-up steps were therefore introduced for the determination of FUdR in urine.39 A barium salt precipitation served to remove inorganic sulphates and phosphates.After the anion-exchange chromatography, Sephadex LH-20 chromatography permitted further purification to remove some pigments. A metabolite of FU and FUdR, 5-fluorouridine (FURD) (a ribonucleoside), was also relevant for the determination in urine.*l cis-Diol-containing substances are known to be selectively retained on boric acid gels, and in a mildly alkaline medium a stable complex is formed between FURD and immobilised phenylboronic acid. Subsequent acidification dissociated this complex for elution of the compound to give sufficiently clean residues for use on the ion-exchange column. E valzcation The column extraction technique resulted in excellent recovery of the compounds, e.g., 96.8 & 2.4% (FU) and 93.2 & 2.1% (FUdR).Even with additional clean-up steps the recoveries were still acceptable, e.g., 70 5 4% (FUdR in urine) and 68 If: 6% (FURD in urine). Within-day coefficients of variation for the FU and FUdR assays averaged 7.2% (3 = 50 ng cm-3; n = 5) and 4% (5 = 297 ng ~ m - - ~ ; n = 12), respectively. Corresponding between-day values were 8.4% (5 = 50 ng ~ m - ~ ; n = 5) for FU and 6.2% (Z = 295 ng cm-3; n = 20) for FUdR. The detection limit of the GC - MS approach for FU was estimated to be 2 ng cm-3, whereas a value of 50 ng ~ m - ~ was obtained for the determination of FUdR in plasma using a nitrogen-selective detector. Serum FUdR concentrations down to 1 ng cm-3 could be measured by the CI - MS procedure. Conclusion : Future Prospects Limitations of analytical methods for the determination of drugs in biological materials mainly concern specificity, accuracy and sensitivity.Specificity can be improved either by more complex sample pre-treatment, with consequent additional sources of error, or by using more sophisticated equipment. The ultimate goal should be analysis without sample preparation. Much progress has been made in enhancing sensitivity and detection limits of assays are steadily improving.42 Whereas the detectability of drugs in biological media was confined to the nanogram range some years ago, current ultra-trace analysis can be applied to amounts as low as a few picograms or occasionally even femtograms (10-15 g). However, as pointed out elsewhere,42 with the current rate of progress it would take at least another 20 years to reach the 10--23g level.Whether or not the ideal of detecting one single molecule will be accomplished within reasonable time is still speculative. This work was supported in part by the National Medical Research Foundation (F.G.W.O.) through grants 20007, 20452 and 3.0004.76. The authors acknowledge the cooperation of C. F. Gelijkens, L. M. R. Vandecasteele-Thienpont, J. A. A. Jonckheere and M.-C1. Cosyns- Duyck who carried out the experimental work described in the Applications section and gave permission to quote from their papers. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Braun, T., Bujdoso, E., and Lyon, W. S., Anal. Chem., 1980, 52, 617A. Yalow, R. S., and Berson, S. A., J . Clin.Invest., 1960, 39, 1157. Wolff, J . , Standaert, M. E., and Rall, J. E., J . Clin. Invest., 1961, 40, 1373. Schussler, G. C., J . Pharmacol. Exp. Ther., 1971, 178, 204. Loeber, J , G., Verhoef, J., Burbach, J. P. H., and Witter, A., Biochem. Biophys. Res. Commun., Arpino, P. J., and Guiochon, G., Anal. Chem., 1979, 51, 683A. Karger, B. L., Martin, M., and Guiochon, G., Anal. Chem., 1974, 46, 1640. Parris, N. A., Editor, “Instrumental Liquid Chromatography,” Elsevier, Amsterdam, 1976, p. 32. Powers, J. L., in SadtSe, W., and Beelen, G. C. M., Editors, “Drug Level Monitoring,” John Wiley, New York, 1980, p. 51. Verzele, M., Redant, G., Qureshi, S., and Sandra, P., J. Chromatogr., 1980, 199, 105. Caprioli, R. M., Liehr, J. G., and Seifert, W. E., in Waller, G. R., and Dermer, 0.C., Editors, “Biochemical Applications of Mass Spectrometry,” First Supplementary Volume, John Wiley, New York, 1980, p. 1055. Millard, B. J., GC-MS News (Japan), 1978, 6, 64. 1979, 86, 1288.October, 1981 AND DRUG METABOLITES IN BIOLOGICAL MATERIALS 1036 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. Millard, B. J., GC-MS News (Japan), 1978, 6, 80. Thomson, S. J., in Lenihan, J . M. A., and Thomson, S. J., Editors, “Activation Analysis-Principles De Soete, D., Gijbels, R., and Hoste, J., Editors, “Neutron Activation Analysis,” Wiley-Interscience, Reid, E., Analyst, 1976, 101, 1. Schoeller, D. A., Biomed. Mass Spectrom., 1976, 3, 266. Millard, B. J ., Editor, “Quantitative Mass Spectrometry,” Heyden, London, 1978, p.75. Jonckheere, J . A., and De Leenheer, A. P., Abstract of paper to be presented a t the 29th Annual Conference on Mass Spectrometry and Allied Topics, Minneapolis, Minn., May 24-29th, 1981. Jonckheere, J . A., Steyaert, H., and De Leenheer, A. P., Abstract of paper to be presented a t the 25th Annual Convention of the Canadian Society of Clinical Chemists, Edmonton, Alberta, July Millard, B. J., Editor, “Quantitative Mass Spectrometry,” Heyden, London, 1978, p. 127. Whitehead, T. P., Analyst, 1980, 105, 1009. De Leenheer, A. P., and Vandecasteele-Thienpont, L. M. R., J . Chromatogr., 1979, 175, 301. De Leenheer, A. P., and Vandecasteele-Thienpont, L. M. R., J . Pharm. Sci., 1980, 69, 99. Jonckheere, J . A. A,, Thienpont, L. M. R., De Leenheer, A. P., De Backer, P., Debackere, M., and Belpaire, F. M., Biomed. Mass Spectrom., 1980, 7, 582. Eichler, D., and Kreuzer, H., Arzneim.-Forsch., 1975, 25, 615. Vandecasteele-Thienpont, L. M. R., Belpaire, F. M., Braeckman, R. A., and De Leenheer, A. P., Arzneim.-Forsch., 1980, 30, 1643. De Backer, P., Vandecasteele-Thienpont, L. M. R., Jonckheere, J. A. A., Belpaire, F. M., Debackere, M., and De Leenheer, A. P., Zentralbl. Veterinaermed., 1980, 27, 740. De Leenheer, A, P., and Nelis, H. J. C. F., J . Chromatogr., 1977, 140, 293. Nelis, H. J . C. F., and De Leenheer, A. P., J . Chromatogr., 1980, 195, 35. De Leenheer, A. P., and Nelis, H. J. C. F., J . Pharm. Sci., 1979, 68, 999. Nelis, H. J . C. F., and De Leenheer, A. P., Clin. Chim. Acta, 1980, 103, 209. Nelis, H. J. C. F., and De Leenheer, A. P., J . Pharm. Sci., 1981, 70, 226. De Leenheer, A. P., and Nelis, H. J. C. F., J . Pharm. Sci., 1979, 68, 1527. De Leenheer, A. P., and Cosyns-Duyck, M.-Cl., J . Chromatogr., 1979, 174, 325. De Leenheer, A. P., and Gelijkens, C. F., Anal. Chem., 1976, 48, 2203. Cosyns-Duyck, M.-Cl., Cruyl, A. A. M., De Leenheer, A. P., De Schryver, A., Huys, J. V., and De Leenheer, A. P., and Gelijkens, C. F., J . Pharm. Sci., 1978, 67, 417. Gelijkens, C. F., and De Leenheer, A. P., Anal. Biochem., 1980, 105, 106. Gelijkens, C. F., and De Leenheer, A. P., and Sandra, P., Biomed. Mass Spectrom., 1980, 7 , 572. De Leenheer, A. P., and Gelijkens, C. F., J . Chromatogr. Sci., 1978, 16, 552. Ahuja, S., Chem. Tech., 1980, 702. and Applications,” Academic Press, London, 1965, p. 73. London, 1972, p. 348. 5-8th, 1981. Belpaire, F. M., Biomed. Mass Spectrom., 1980, 7 , 61.

ISSN:0003-2654    

DOI:10.1039/AN9810601025

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

1036 Analyst, October, 1981, Vol. 106, pp. 1036-1041 Monitoring Exposure to Toxic Gases in Workplace Atmospheres* D. T. Coker Esso Europe Inc., Esso Research Centre, Abingdon, Oxfordshire, OX13 6AE The basis, objectives and shortcomings of personal monitoring as a means of both estimating the level of exposure and preventing the occurrence of over exposure to airborne toxic substances are discussed. Techniques that are available, and the circumstances in which they are suitable are reviewed along with recent new developments and possible future alternative approaches. Keywords : Toxic gas monitoring; workplace atmospheres ; health and safely Whenever volatile substances are being manufactured, processed, transported or handled in any way in industry there is a possibility that employees may be exposed to vapours, and if this exposure becomes excessive then toxic effects may occur among those over-exposed.Most industrialised nations have safety and health legislation covering employees at work, which is designed to prevent damage to their health by imposing a statutory obligation on their employer to provide a healthy and safe workplace. This paper will consider only how employers can fulfil this obligation with respect to preventing damage to their employees’ health arising from over-exposure to toxic gases in the workplace. Toxic Damage It is important to note that all substances are toxic and their potential to damage the human body is a matter of their inherent toxicity and the dose the person receives. The dose is the amount of the substance that gets into the body and depends on four main factors : (i) the concentration of the substance in the inhaled air; (ii) the respiration rate of the person, which can vary by as much as five times between sedentary tasks and very strenuous tasks; (iii) the exposure time; (iv) the absorption factor, i.e., the fraction of the total amount of a substance inhaled that is actually absorbed into the body.This varies considerably, water-soluble substances being more easily absorbed than insoluble ones. The pattern of exposure will also affect the absorption factor, as the amount absorbed from a continuous steady concentration will differ from that absorbed from a rapidly fluctuating con- centration even though the average concentrations are the same.Toxic damage may occur at one or more locations in the body and the degree of damage will be related to the concentration of the toxic substance or more commonly one of its toxic metatjolites formed in the body, in the “at risk” or “target” organs. However, the degree of damage is dependent not only on the dose of the substance but also on individual suscepti- bilities, which may arise from many factors: men and women may well be affected differently; age may have an effect, often the young are more at risk; the health of the person, their general fitness and such factors as fatness may have an influence ; pregnant women and the foetus are a special case; individual variations in metabolism and body chemistry may influence the degree of toxic damage, particularly if the damage is caused by a toxic metabolite rather than the original substance; the degree of toxic damage may be compounded by smoking, drinking alcohol, taking drugs or by exposure to other toxic substances.* Plenary Lecture presented at the Joint NL - UK Symposium on Quantitative Organic Analysis, Noordwijkerhout, The Netherlands, April 22-24, 1981.COKER 1037 Protection of Employees Ideally, the protection of employees’ health against damage from toxic substances would best be achieved by monitoring the level of the substance or its metabolites in the target organs, but the problems of “invasive” sampling on a routine basis make this approach impractical. As a next-best option, the monitoring of body fluids, mainly urine and blood is used in certain instances, e.g., blood analysis for lead and urinalysis for metabolites of tri- chloroethylene and benzene. There has also been a resurgence of interest in exhaled breath analysis for organic substances now that sampling and analytical techniques of sufficient sensitivity are available.In practice, the most widely used monitoring technique is to measure the concentration of the inhaled air, sampled close to the mouth. This concentration is called the “exposure” and should not be confused with “dose.” It is evident from the previous discussion that exposure will not give a very accurate assessment of the concentration of the toxic substance in the target organ as it will not account for variations in respiration volume and rate, absorption factors, exposure by other routes (dermal, oral) and non-occupational exposure.However, this is the system in widespread use and the rest of this discussion will review the principles and techniques available for personal exposure monitoring. Air-quality Standards Standards for acceptable levels of exposure to organic substances in the air breathed have been established and are usually referred to as Threshold Limit Values (TLVs),l which can be defined in several ways. All listed substances have an 8-h time-weighted average (TWA) concentration, to cover a typical workshift, but there are also 15-min averaged short-term exposure limits (STEL) and ceiling values, which should not be exceeded at any time, quoted for several substances. These different TLVs are applied to cover the different types of hazardous situation that may arise, and in general the type of TLV applied for a substance depends usually on the type of toxic effect that occurs at the lowest concentration.For instance, benzene will produce narcosis at very high concentrations, whereas at lower con- centrations it does not cause narcosis but is irritating to the respiratory system; these are both acute or short-term effects. However, prolonged and repeated exposure over months or years to even lower levels can lead to damage to the blood and blood-forming organs, and in order to prevent this chronic toxic effect the TWA TLV would be applied. Alternatively, with substances such as hydrogen sulphide there is no chronic effect from long-term exposure to low levels, only eye and throat irritation.At higher concentrations severe acute toxic effects ranging from headaches to coma and death can occur. The TWA measurement is therefore of little use and a limit on the immediate concentration should be applied. They are personal exposure limits for healthy adults in the workplace but do not apply to levels in the home or outside environment or to the very old, children, sick, etc. They refer to an airborne concentration or level and can be applied to gases, vapours, dusts, mists, noise, radiation, heat and micro-organisms. In most countries (outside the USA) they are not statutory limits except in a few specific cases. In the UK they are published by the Health and Safety Executive and are qualified by four notes: (i) TLVs are not sharp dividing lines between “safe” and “dangerous” concentrations; (ii) the best working practice is to reduce the concentration of all airborne contaminants as far below the TLV as is reasonably practicable, whether or not they are known to present a hazard; (iii) the absence of a substance from the list does not necessarily indicate that it is safe; (zv) the application of any TLV in a particular situation should be interpreted by a trained occupational hygienist.The scope and applicability of TLVs are clearly defined but frequently exceeded. Monitoring Objectives The prime purpose of legislation is the “protection of employed persons at work” to ensure that their health, well-being and working efficiency are not impaired. Any monitoring programme or system should be capable of protecting the employee, and this can be achieved in several ways.For protection against substances giving acute toxic effects, the monitoring1038 COKER : MONITORING EXPOSURE TO TOXIC Analyst, Vol. 106 instrument should be capable of warning the wearer before a hazardous level or the dose limit is reached. For protection against substances giving chronic toxic effects, the retro- spective measurement of shift average exposures offers no direct protection from over- exposure but indicates the necessity for remedial actions such as engineering control or operational procedure improvements, or the need for personal protection. Protection of the employee as a monitoring objective is directly connected with protecting the employer, as failure to demonstrate that workplace conditions are satisfactory can result in action by the Health and Safety Inspectorate, either in improvement or closure notices or in legal prosecution.Legal action can also be taken directly by employees against the employer and in extreme cases these can be for criminal negligence. Plant control is a secondary result of monitoring as increased exposure measurements may indicate leakages, incorrect operational procedures, etc. Keeping exposure records is also very important, both as a record of working conditions should some unexpected health effect arise years after exposure and to provide data for epidemiology where medical statistics are correlated with past exposure, which is ultimately the best test of the correct setting of a TLV. Monitoring Location TLVs are personal exposure limits, and when a person is moving around the workplace where there are typically considerable variations in airborne contaminant levels, then measurements on the person are necessary to obtain an accurate estimate of exposure.Monitoring at fixed locations within the workplace seldom gives a good correlation with personal monitoring unless the person does not move about or when contaminant levels in the workroom are reasonably uniform and constant. Fixed location monitoring is allowable as a means of estimating exposure, but there is then an onus to show that, firstly, some correlation exists between personal and fixed location exposure estimates and, secondly, that within the limits of that correlation the personal exposure limit has not been exceeded.In practice this means that to demonstrate compliance with a TLV from fixed location exposure estimates these need to be lower than personal exposure estimates to allow for the uncertainty in the correlation. In general, therefore, personal exposure monitoring is to be preferred and fixed location monitoring is best suited to plant leakage control or as an early warning alarm for large emissions. Monitoring Techniques Different techniques are necessary to account for the toxic properties of different substances and the particular objectives of the monitoring. Biological monitoring has many advantages, as previously discussed, but unfortunately also many drawbacks. Routine blood sampling presents obvious problems and urinalysis is complicated by the need to collect the samples at set times following the end of a workshift, which can be awkward.Further, there may also be some interferences from naturally occurring metabolites. However, exhaled breath can be quickly and easily sampled at the beginning of the next workshift and, as the contaminant and not its metabolites are measured, the occurrence of interferences is much less. Spot measurements using detector tubes or direct reading m,eters can be useful for prelimi- nary assessments of the type and approximate level of any airborne substances present in the workplace, which is necessary prior to designing a comprehensive monitoring programme. Other useful devices are continuous recording meters, which can be used to identify the areas or operations in the workplace that contribute most to the over-all exposure.These include portable flame ionisation or photoionisation detectors and the paper-tape monitors that comprise a cassette of paper tape impregnated with selective reagents that is wound slowly past an air sampling port. Any contaminant in the air flowing through the port produces a stain on the paper, which is quantified using reflectance spectrometry. For acutely toxic substances where effects are related to the immediate concentration but no chronic effects occur, then personal hazardous level warning devices offer the best protec- tion and shift average exposure measurement is of little use. There are many such instru- ments available commercially, which can be worn in an overall pocket and have an audio - visual alarm that can be set to give warning when an appropriate level is reached. TheseOctober, 1981 GASES IN WORKPLACE ATMOSPHERES 1039 usually have rapid response electrochemical cells as sensors and are used typically for substances such as hydrogen sulphide, nitrogen dioxide, ozone, oxygen deficiency, chlorine and carbonyl chloride (phosgene).A similar type is also used for substances that are acutely toxic but where the dose is more important than the immediate concentration. These integrate the concentration with time and again trigger when a pre-set “dose” is reached, They are called dosemeters, which, strictly, is a misnomer as they measure the integrated concentration with time, not the dose received by the body. Typical applications for this type are for carbon monoxide and nitrogen oxide.Monitoring for substances with mainly chronic toxic effects, which includes most organic substances, is by measurement of the average concentration over the workshift. This can be achieved by collecting the air directly into a gas bag or evacuated container but this approach has many drawbacks and so the most widely used technique is to remove the contaminant from the air using either a liquid bubbler or a solid adsorbent. Bubblers have many disadvantages: they are clumsy to wear; there is a large dilution factor in the absorbent solution; air sampling rates need to be high so large pumps are required; and the sampling rate must be controlled to achieve complete absorption and avoid spray and evaporative losses.Because of these disadvantages, bubblers are usually only used for very high boiling, reactive or polar substances that cannot be quantitatively recovered from solid adsorbents. Vapour adsorption tzcbes The most commonly used technique for monitoring organic vapours is the vapour adsorption (VA) tube, which is simply a bed of adsorbent in a tube or container through which air is drawn. Any contaminant in the air is trapped on the adsorbent and held for subsequent analysis. The adsorbed contaminant is removed for analysis either by solvent extraction or by heating the adsorbent and purging off the desorbed vapour in an inert gas stream. Originally these were large, clumsy, glass tubes packed with silica gel, charcoal, etc., and they also required large air sampling pumps to attain acceptable detection limits owing to the large desorption solvent volume needed.In the USA, NIOSH developed the miniature charcoal- in-glass VA tube which quickly gained acceptance as it is small and unobtrusive and needs only a small, low flow-rate personal sampling pump. The tubes are analysed by breaking and removing the two charcoal beds (one for sampling, one as back-up) into small vials. Desorption solvent (usually carbon disulphide), is added and the extract analysed, usually by gas chromatography. This method also has disadvantages, including : the desorption step is time consuming and uses a toxic, flammable solvent; the sensitivity is limited by a dilution of the sample with the solvent of at least 100-fold; the solvent and any impurities interfere with the analysis; desorption is often incomplete ; and the sampling tube is not re-useable.These disadvantages have led to the increasing popularity of heat desorption VA tube systems over the last few years. They use a metal tube with a single adsorbent bed of either charcoal (for permanent gases) or porous polymer bead adsorbents (e.g., Porapak, Tenax, Chromosorb) for vapours from liquids. These are analysed using a heat desorber, which is attached to the analyser (gas chromatograph, infrared spectrometer, atomic-absorption spectrometer, etc.) . For gas- chromatographic (GC) analysis, the desorber is connected into the carrier-gas line just before the column. The desorption procedure is to clamp the VA tube into the desorber, divert the carrier gas flow through it then flash-heat the tube in a heated metal block; the trapped vapour is rapidly released on heating and carried directly on to the GC column as a discreet slug in the carrier gas.Conditions can be easily arranged so that desorption can be made on to an isothermal column giving a very rapid analysis, eliminating sample preparation, the use of and interference from desorption solvents and giving a very sensitive determination, as sample dilution is also eliminated. An alternative procedure is to desorb the vapour into a reservoir from which aliquots can be withdrawn, allowing replicate analyses or the use of more than one type of analyser. The VA tubes are clean and ready for re-use after desorption. Passive adsorption systems This is a vapour adsorption system that does not have air pumped through an adsorbent bed but has Another recent development is the so-called “dose badge’’ or passive sampler.1040 COKER : MONITORING EXPOSURE TO TOXIC Analyst, Vol.106 a layer of adsorbent covered by a stagnant air layer across which contaminants diffuse and are trapped by the adsorbent at a rate proportional to their concentration in the air, elimi- nating the need for a pump. Analysis of these is similar to that of the pumped tubes, both solvent and heat desorption types being available. These showed promise initially in laboratory studies but field evaluations have generally been less promising, as their accuracy is rather poor in field conditions probably owing to the rapid fluctuations in concentrations that commonly occur in the workplace.Exhded breath analysis As previously mentioned, this technique is attracting much interest as an easier and more logical means of assessing workplace air quality than personal (inhaled air) monitoring but at present it is mainly in the experimental stage. Its potential advantages over personal monitoring include both technical advantages, viz., (i) exhaled breath (EB) sampling takes only 1-2 min of the subject’s time and eliminates the necessity for wearing pumps and samplers throughout the workshift with their attendant problems of encumbering the subject, intrinsic safety and reliability, (ii) the subject can tamper with a personal sampler but not with EB sampling and (iii) personal samplers can become contaminated with water or liquid sprays, etc., in the workplace; and also the opportunity for the measurement of body burden, viz., the effect of a toxic substance on the body is related to the dose the body receives but personal monitoring measures only the average airborne concentration.There is a considerable amount of literature on the subject and techniques are reasonably established but the main problem is that standards are set in terms of airborne concentrations and EB analysis has yet to prove that it is a more logical and effective means of assessing the acceptability of workplace air quality. Future Trends Irrespective of whatever techniques may be used in the future, it is evident that the require- ment to monitor the workplace atmosphere will become a routine part of plant management.This will probably be carried out more comprehensively than at present and will be common to most industrialised countries. Although both personal and fixed location monitoring are likely to be used, it is probable that personal monitoring will be the definitive approach, with multi-point fixed location plant monitors being used for plant leakage control and as a “watchdog” on the plant atmosphere. The ultimate personal monitor could be envisaged as a highly selective, highly sensitive and stable detector cell incorporated into a small personal monitor comprising a power supply and voltage stabiliser and circuitry to detect when a hazardous level is reached and sound a bleeper and/or a flashing light (and perhaps also to transmit this warning to the plant control room via radio).Another circuit would record the measured levels continuously on solid- state memory and also integrate the level with time; these values would be examined at the end of the shift by connection to an external read-out instrument. The level of contaminant would also be displayed on a meter built into the monitor. The electronic technology for such an instrument already exists but it is doubtful if detector cell development will reach the level of sophistication needed for monitoring one organic substance in the presence of many others. Personal monitoring using vapour adsorption systems is still likely to be the most commonly used technique in the forseeable future. It remains to be seen how much impact passive samplers will have, but with the increasing evidence of their inaccuracy under field con- ditions it seems unlikely that they will displace the more reliable and accurate pumped tube method to a significant extent. It is often said that good accuracy is not a prime necessity of personal monitoring, but with the trend toward lower TLVs and a tighter control over workplace air quality the requirement for accuracy is becoming more apparent. Further, the total real cost of monitoring in terms of manpower and resources is high and the cost saving in eliminating the pump is only a small fraction of the total. It is a shortsighted policy to diminish the value of monitoring data, measured in terms of the ability of an employer to demonstrate his compliance with standards, his competence in controlling work- place conditions and their possible use as a defence against legal action by an employeeOctober, 1981 GASES I N WORKPLACE ATMOSPHERES 1041 claiming that a health problem arose from his employers negligence by using monitoring methods of debatable accuracy and reliability. It is to be hoped that the advantages offered by exhaled breath sampling will be explored and adopted in the future. It is unlikely however, that this will happen very rapidly as it involves a complete change in direction that may take many years to achieve. Reference 1. American Conference of Governmental Industrial Hygienists, “Threshhold Limit Values for Chemical Substances and Physical Agents in the Workroom Environment,” American Conference of Governmental Industrial Hygienists, Cincinnati, 1978.

ISSN:0003-2654    

DOI:10.1039/AN9810601036

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

1042 Analyst, October, 1981, Vol. 106, fie. 1042-1056 Analysis of Additives and Process Residues in Plastics Materials* D. C. M. Squirrel1 Imperial Chemical Industries Limited, Petrochemicals a%d Plastics Division, P.O. Box 6, Welwyn Garden City, Hertfordshire, AL7 1HD Analytical problems, compounded by the increasing number and variety of additives used by the plastics industry and the necessity to reduce process residues and contaminants to the minimum, are described. An outline is given of general analysis schemes in use for the examination of PVC com- pounds, polyolefin compounds and acrylics. This is followed by some specific examples chosen particularly to illustrate the integrated approach, utilising chemical, physical and complex instrumental methods, that is required to solve some of the more difficult analytical problems.Methods used include solvent extraction, titration, ultraviolet, visible and infrared spectrophoto- metry, X-ray fluorescence, gas chromatography, liquid chromatography, thin- layer chromatography and mass spectrometry. The methods are often used in combination. Keywords : Plastics analysis ; additives ; process residues The title of this paper introduces a very large subject, even if one considers only the more common plastics. The increasing number and variety of additives used by the industry and the necessity to reduce process contaminants to the absolute minimum for quality and/or safety requirements add daily to the analytical problems. Therefore, no attempt will be made to cover the whole field, but first an outline of the general analysis schemes in current use in the author’s laboratory for PVC compounds, polyolefin compounds and acrylics will be given. This is followed by some specific examples chosen particularly to show the inte- grated approach utilising chemical, physical and complex instrumental methods that is required to solve some of the more difficult analytical problems. It would be wrong to say that expensive and complex instrumentation is essential for many of the analyses requested, but its availability often speeds the solution of the problems and provides additional important inf onnation.General Schemes In the analysis of an “unknown” plastic, characterisation into a broad group is usually relatively simple, taking into account the origin of the sample, its use, appearance, elemental composition and, most important, its infrared spectrum.Much of the detail of this and the analytical schemes given in Figs. 1-3 is outside the scope of this paper, but further information is given in reference 1. Where identification, particularly of minor organic components, is required, then some separation from the plastic compound is often necessary. Special care and specialised techniques are required when dealing with laminates and surface-coated films. For major components, the separation is made quantitatively and the analysis is completed by gravimetric, titrimetric or spectrophotometric methods. For volatile components, separation, identification and quantification can often be carried out in one analytical process.Some additives and contaminants, having strong spectral characteristics, can be determined directly in a pre- pared sample without any separation stage, for example, the infrared analysis of film samples. X-ray fluorescence is of particular value and is widely used when the composition contains characteristic elements of atomic number greater than 12. Direct analysis of samples in granular form is possible and Tables I and I1 show that even for light elements, e g . , for In the schematics shown in Figs. 1-3, the major points to note are as follows. * Plenary Lecture presented a t the Joint NL - UK Symposium on Quantitative Organic Analysis, Noordwij kerhout, The Netherlands, April 22-24, 1981.SQUIRRELL 1043 organic phosphorus compounds in nylon, precision and sensitivity are reasonable when a thin polypropylene window is used to support the granules in the presentation container.With a variation in granule size from 0.026 to 0.050 g per granule, little change in the co- efficients of variation was noted. PVC COMPOSITION Extract with diethyl ether I I 1 SOLUBLE INSOLUBLE I 1 I Plasticisers, most lubricants, most stabilisers, UV absorbers PVC polymer: filler, pigment, stabiliser, modifier, processing aid, residual plasticiser, emulsifier. some lubricants Extract with methanol f I SOLUBLE Some stabilisers, emulsifiers, residual blowing agent INSOLUBLE PVC polymer, filler, pigment, stabiliser, modifier, processing aid, residual plasticiser, some lubricants Extract with tetra hydrof uran and centrifuge 3000 rev min" 1 SOLUBLE PVC polymer, processing aid, residual plasticiser, modifier (in suspension), some lubricants I INSOLUBLE Filler, pigment, stabiliser, some modifiers I c 1 SOLUBLE PVC polymer, processing aid, residual plasticiser, some lubricants I Add methanol INSOLUBLE Modifier, some filler SOLUBLE Residual plasticiser, some lubricants I INSOLUBLE PVC polymer, processing aid Fig.1. Analysis of PVC compositions.1044 Direct examination of 1 I Identification Separation Granules and sheet I I XRF G LC Film I Determ'ination IR uv XRF GLC I I IR GLC Colorimetric uv. HPLC MS Ethanol solution Ethanol - KOH Ethanol - Ni02 Ethanol - Ni02 - KOH reagents Fig. 2. Analysis of polyolefines. Analysis of acrylic samples r I Dissolve in 1 I IR 'Pyrolysis - GC Opacifying I acetone XRF Dissolve Titration agents I Modifiers Plasticisers, UV absorbers Solution IR Filtrate 1 I I MS Polymer Plast icisers, Catalyst residues lubricants Fig.3. Analysis of acrylic samples. TABLE I EFFECT OF WINDOW MATERIAL: si Ka Parameter Window material. . Absorption coefficient Window thickness/ pm Counts per 1% 50, % of disc count-rate Form r Granules Granules Discs .. . . Polyester Polypropylene None .. * . 6 5 .. . . 44900 86 300 144 400 .. .. 31 60 100 A \ .. 573 345 .__ .. -October, 1981 RESIDUES IN PLASTICS MATERIALS 1045 TABLE I1 EFFECT OF GRANULE SIZE: Zn Ka, P Ka Element Phosphorus Parameter Zinc r A b Form .. .. .. .. Granules Sample cell window . . . . Polyester (6 pm) Coefficient of variation, % . . 1.7-2.3, depending Precision (20) .. .. . . f0.02yo at 0.4% Loss in sensitivity of disc Readings . . .. .. 3 X 3 X 3 on Zn concentration level count rate, yo . . .. 10 Granules Granules Polyester (6 pm) Polypropylene (5 pm) 3 X 3 X 3 3 X 3 X 3 7.8-15.9, depending 3.5-10.2, depending on P concentration on P concentration f 1 5 p.p.m. a t f7 p.p.m. at 70 p.p.m. level 70 p.p.m. level 54 33 Extraction Methods The main extraction procedures used are summarised in Table I11 and, although most of them can be carried out on material cut to a particle size of less than 2 mm diameter, it is often advantageous to produce material of a smaller size and with a larger surface area to mass ratio. This is conveniently done by grinding at the temperature of liquid nitrogen using an efficient and easily cleaned cutter mill.Full details of some of these methods are provided in reference 1. EXAMPLES OF EXTRACTION METHODS Polymer or Extraction method compound Single solvent (Soxhlet) . . PVC Single solvent (reflux) . . : . Polyethylene Solvent + reagent (reflux) . . Polypropylene Polyester Solutionlprecipitation . . . . Polyolefines Acrylics Steamlsolvent distillation . . Packaging films Vacuumlthermal extraction . . Nylon Fluorocarbon polymers Chemisorbed and Adsorbed Additives Solvent (s) Diethyl ether Chloroform, 1,Ll-tri- chloroethane Ethylene dichloride - trichloroacetic acid Methanol - Karl Fischer reagent Toluene - methanol, xylene - methanol Acetone - light petroleum Water - diethyl ether Additives or contaminant extracted Plasticisers Antioxidants Chemisorbed amidesl Water amines UV absorbers, anti- Plasticiser, lubricants oxidants, slip agents Odour- and taint- Water Process fumes forming additives The extraction of additives strongly adsorbed or chemisorbed on to the polymer - filler matrix must be carefully watched by the analyst, as a change in the method of manufacture of, for example, the filler or in the method of compounding the plastic formulation can also markedly alter the degree of adsorption bonding produced and hence invalidate an estab- lished quantitative extraction procedure.The use of “stronger” extraction reagents can cause complications at the measurement stage and hence each system must be carefully screened and frequently checked. Solution/Precipitation Methods This method of extraction involves dissolution of the organic phase of the plastic compo- sition in a suitable (sometimes hot) solvent, followed by precipitation of the polymeric constituents, often in a finely divided form in suspension with the inorganic fillers, by cooling or by means of another solvent in which the andyte compound is also soluble.This method is labour intensive but very effective, and if it does not completely release any chemisorbed1046 SQUIRRELL: ANALYSIS OF ADDITIVES AND PROCESS Analyst, Vd, 106 constituents from the polymer - filler matrix it will often leave them in a form very vulnerable to attack by the analytical reagent@) finally used in the determination. For example, the titration of aminelamide slip additives with perchloric acid in non-aqueous media can be carried out in the presence of suspended solids.Vacuum/Thermal/Displacement Extraction These procedures are used extensively for the direct isolation or release of volatile com- ponents from a polymeric matrix and may involve the combined use of vacuum and heat, as for example in the mass spectrometer direct insertion probe or during dry vacuum distillation. Alternatively, the volatiles may be swept from the heated sample by a flow of inert gas for concentration by freeze trapping and/or collection on to a solid adsorbent prior to thermal or solvent desorption for gas-chromatographic or mass spectrometric examination. A novel extension of the gas stripping technique was used when a roll of packaging film was the subject of an odour complaint. In this instance, the adsorption tube was made from a melting-point tube containing only 4-5 mg of Porapak Q, which was attached to a small low-flow pump.The reel of complaint film was slowly unrolled, moving the adsorption tube backwards and forwards like a vacuum cleaner across the two parting film surfaces (Fig. 4). The small sample of adsorbent was then inserted directly into the mass spectro- meter on the end of the insertion probe and heated. In addition to small amounts of the expected solvents, acetylacetone was detected. Previous methods of direct headspace analysis, not using this concentration technique, had failed to detect this solvent. Only the thermal decomposition products, acetaldehyde and diacetyl, had been detected. The source of this odorous contamination was subsequently shown to be a printing ink solvent containing titanium acetylacetonate as an adhesion promoter.This application is, of course, comparative rather than quantitative, but is none the less useful. Personal sampling Pump Fig. 4. Sampling volatiles from film roll. Secondary Methods of Analytical Separation When the above procedures for preliminary isolation of the analyte materials from the polymer matrix are complete, further separation is often required for identification and determination. Three fonns of chromatography are generally used. Gas Chromatography Gas chromatography in all its forms with appropriate detectors and, when necessary, temperature programming, heart cutting and back-flushing techniques, is used extensively for volatile components.The details are outside the scope of this paper but mention must be made of headspace methods. These are used extensively for the determination of residual monomers and other residues in polymer compositions after dissolution or dispersion in aOcto bey, 1981 RESIDUES IN PLASTICS MATERIALS 1047 suitable solvent and equilibration in a sealed vial at constant temperature prior to chromato- graphy of the headspace gas. For samples in the form of fine powders or thin films, the tech- nique can be applied directly to the solid samples. Some examples and experimental details are given in references 2 and 3 and are summarised in Figs. 5 and 6. Solids (direct examination) Coating and printing solvent residues Residual monomers I Packaging films Polyethylene . .. . . . . . . . . . Catalyst carrier Ethylene - vinyl acetate co-polymers . . Ethylene, vinyl acetate, Granular polymers . . . . . . . . . . . . Process residues acetic acid Polymer powders PVC and co-polymers . . . . . . . . Monomers, moisture { Polyethylene terephthalate . . . . . . Acetaldehyde Foodstuffs Biscuits, cakes and crisps , , . . , . i Migratory trace monomers Solids Polymer solutions PVC compositions, (indirect examination) (normally 10%) { ABS. . . . . . . . . . . . . . . . Residual monomers Fig. 5. Some examples of headspace gas-chromatographic procedures applied to solids. Cooking oils, fruit squashes, water, wine, spirits Solution extracts of adsorption tubes Migratory trace monomers . . . . Vinyl chloride in dichloromethane, acrylonitrile in benzyl alcohol, benzene in ethylene carbonate methacrylate monomers, dimers and co-dimers Foodstuffs { .. . . . . . . . . . . (atmospheric monitoring) Latices . . . . . . . . . . . . . . . . . . Butadiene, styrene, acrylate monomers, Liquids k \ \ Biological samples Whole blood . . . . . . . . . . . . . . . . Vinyl chloride from PVC transfusion Urine . . . . . . . . . . . . . . . . . . Trichloroethanol, trichloroacetic acid sachets, trichloroethylene, trichloroethanol Fig. 6. Some examples of gas-chromatographic procedures applied to liquids. Thin-layer Chromatography (TLC) TLC separations are now usually carried out only when some preliminary separation of non-volatile materials is required prior to mass spectrometric examination or to provide confirmatory or supplementary analytical evidence.For plasticisers and extenders from PVC, l-mm Kieselgel60G layers on glass plates are used with 95 + 5 toluene - ethyl acetate as elution solvent. Phthalate esters, aliphatic esters, phosphate esters, epoxy esters and chlorinated hydrocarbons separate readily when 0.05-0.1 g of extract is used. The separated bands are marked under ultraviolet light, removed from the plate and extracted with diethyl ether, followed by infrared andlor mass spectrometric examination. For additives extracted from polyolefins, usually with diethyl ether, the extract is refluxed with ethanol and the solution is decanted off the insoluble residual polymer. On cooling, additives such as dilauryl and distearyl thiodipropionate separate out and are identified by infrared examination.A 30-yl volume of the ethanolic solution is then spotted on to a thin Kieselgel60 TLC plate and eluted with a suitable solvent, usually 98.5 + 1.5 toluene - ethyl acetate. The eluted plate is dried and sprayed with colour-developing reagents and the spots are examined. Table IV illustrates R, values (based on Topanol OC = 1.00), colours and suitable solvent systems. If the spots are to be submitted to mass spectrometric exami- nation, methanolic iodine is used as the colour-developing reagent as this does not over- complicate the mass spectrometry. The normal spray procedure uses a 0.5% solution of 2,6-dibromo-~-benzoquinone-4- chlorimine in ethanol followed, after drying, by a 0.5y0 borax solution. After this spray the plate is dried at 120 "C for 5 min to develop the colours.1048 SQUIRRELL: ANALYSIS OF ADDITIVES AND PROCESS TABLE IV THIN-LAYER CHROMATOGRAPHY OF COMMERCIAL ANTIOXIDANTS Afialyst, VoZ.106 RF value* #- A % Additive Solvent 1 TopanolOC .. * . 1.00 DLTDP 1. . . 0.40 DSTDP .. . . 0.45 Distearyl disulphide . . 1.10 “DTB glycol ester” . . 0.00 Ionox 330 .. . . 1.05 Irganox 259 . . . . 0.30 Irganox 288 . . . . 0.25 Irganox 1010 . . . . 0.30 Irganox 1076 . . . . 0.85 Nonox WSP .. . . 0.85 Polygard Santanox R Tinuvin 326 Tinuvin 327 Tinuvin 328 Topanol CA UV531 .. Hoechst D55 .. . . 0.30, 0.15 .. . . 0.35 .. . . 1.05 * . . . 1.05 .. . * 1.00 .. . . 0.10 .. . . 0.75 .. . . 0.85, 0.75, 0.55, 0.35, 0.30, 0.20 Oleamide .. . . 0.00 Erucamide . . . . 0.00 Ethomeen TI2 . . . . 0.00 Stearic acid .. . . 0.00 Solvent 2 1.00 1.00 1 .oo 1.05 0.75 1.05 1.05 1.05 1.10 1.05 0.95 0.80, 0.70 0.80 1.05 1.05 1 .oo 0.60 1.00 0.90, 0.85 0.00 0.00 0.00 0.00 Solvent 3 1.00 0.05 0.10 1.20 0.00 0.80 0.00 0.00 0.00 0.60 0.45 0.10 0.10 0.95 1.05 I .oo 0.00 0.35 0.55, 0.45 0.25, 0.15, 0.10, 0.00 0.00 0.00 0.00 0.00 Solvent 4 Colour of spot 1 .oo 0.30 Yellow - brown 0.35 Yellow - brown 1.20 Bright yellow 0.00 Brown centre with mauve outer 0.80 Pinkish brown 0.65 Brown 0.60 Brown 0.60 Brown 0.95 Brown 0.35 Bright yellow centre Pale yellow centre with pink outer with pink outer 0.05 0.00 1.05 1.10 1.05 0.00 0.80 1.10, 0.45, 0.25. 0.15 Blue and red Purple Yellow Yellow Yellow Brown Blue Blue and red These additives give only a very faint brown * R g values quoted to the nearest 0.05.Solvent 1 = toluene - ethyl acetate (98.5 + 1.5); solvent 2 = toluene - isopropanol (88 + 12); solvent 3 = toluene - light petroleum (b.p. 60-80 “C) (1 + 1); and solvent 4 = cyclohexane - toluene - methanol (88 + 10 + 2). High-performance Liquid Chromatography (HPLC) HPLC has now become a very valuable tool for plastics analysis, particularly in the additive field. We use standard equipment with both reversed-phase and adsorption columns, isocratic and gradient elution and mainly a variable-wavelength ultraviolet detector. Two examples are appropriate. Fig. 7 shows a chromatogram of a mixture of phthalate esters separated on an SC6 column, with a methanol-water gradient mixture as solvent. The most important region is the C,-C, fraction, which is not completely separated but which does give characteristic patterns for commercial plasticiser mixtures and hence allows identification.Of course, quantitative analysis is carried out with suitable calibration. The second example (Fig. 8) shows a chromatogram of a mixture of antioxidants and ultraviolet absorbers. A reversed-phase system is again preferred because it can be washed clean with methanol, but this time an isocratic solvent system is used. With the variable- wavelength detector, the optimum wavelength can be set for the type of compounds being examined. Although gradient elution is not necessary for every analysis, we find it is of great help in setting up conditions for the particular additive mixtures. Isocratic solvent mixtures are preferred for quantitative analysis.The additives shown in Fig. 8 are almost all aromatic phenols and our remaining problem in this type of additive analysis is to detect and determine the aliphatic compounds. These are in two groups, the thioester synergistic stabilisers and the snrface-active modifiers such as oleamide and glycerol monostearate. We are now looking for post-column derivatisation reactions for these compounds to give species detectable by either fluorescence or ultraviolet methods.October, 1981 RESIDUES IN PLASTICS MATERIALS 1049 I M 4 H I F i i C / A d Fig. 7. Separation of phthalates by HPLC. Chromatographic conditions : column, Spherisorb SC6 (5 pm), 100 x S mm; solvent gradient, 60% V / V methanol in water to methanol in 20 min; flow-rate, 40 cma h-l; and detection, absorbance at 254 nm.A, Injection; B, methyl and methoxy- ethyl; C, ethyl; D, alkyl; E, isopropyl; F, phenyl; G, isobutyl and butylbenzyl; H, butyl; I, cyclohexyl; J, heptyl; K, impurity; L, isooctyl; M, sec-octyl and octyl; N, nonyl; 0, isodecyl; P, undecyl; and Q, tridecyl. Mass Spectrometry Like other forms of molecular spectroscopy, mass spectrometry may be used as a “finger- print” technique to identify the components of additive systems extracted from polymer compositions. The strengths of mass spectrometry are high sensitivity and the ability to distinguish between closely related compounds of differing relative molecular mass, e.g., the various alkyl thiodipropionates used as synergistic stabilisers in polyolefins and the ultra- violet-absorbing benzotriazole derivatives.Often it is not necessary to separate the com- ponents before examination as some separation may be achieved by careful variation of the sample probe temperature to produce, in effect, a fractional distillation of the components. The presence, however, of large amounts of low relative molecular mass polymers from polymers such as polyethylene and polypropylene can cause interference by producing a high hydrocarbon background extending to several hundred relative atomic mass units. In such instances a thin-layer chromatographic separation can be used as a clean-up procedure, the low relative molecular mass polymer normally moving with or close to the solvent front, leaving the additive components for separate examination. Complex mixtures can be resolved and the technique then used to isolate the relevant area of coating from the plate into a small melting-point tube (Fig.9). The additive is then eluted from the coating in the tube into the mass spectrometer sample tube or inlet probe with a suitable solvent, normally methanol. Compounds having the same relative molecular mass and different molecular formulae may be distinguished, provided that there is sufficient material for a high-resolution examina- tion, enabling the masses of the molecular and fragment ions to be measured to sufficient1050 Cotton- wool Plug SQUIRRELL : ANALYSIS OF ADDITIVES AND PROCESS Analyst, VOl. 106 t Fig. 8. Determination of addi- tives from polypropylene by HPLC. Chromatographic conditions : column, Spherisorb ODS (6 pm), 100 x 6mm; solvent, 90% V/V methanol; flow-rate 40 cm3 h-1; detection, absorbance a t 280nm and 0.2 a.u.f.s.A, Injection (10 pl); B, solvent front; C, Topanol CA, 10.12mg per 60cm3; D, Tinuvin 326,4.96 mg per 60 cms; E, UV 531, 9.82 mg per 60 cm3; and F, Irganox 1010, 10.10 mg per 50 cm3. Vacuum ?l Methanol Cotton-wool ’ Plug Probe tip Glass plate Fig. 9. Sampling of fractions from TLC plates for examination by mass spectrometry.October, 1981 RESIDUES IN PLASTICS MATERIALS 1051 accuracy for their chemical formulae to be deduced. An example of this, also demonstrating the advantages of a multi-technique approach, is that of an ultraviolet-absorbing additive extracted from a polyethylene terephthalate composition. The material had the same TLC R, value as UV531, but did not exhibit the right colour or ultraviolet characteristics for this component. Infrared, ultraviolet and nuclear magnetic resonance spectroscopy indicated that it was a benzophenone derivative.The low-resolution mass spectrum showed a benzophenone of relative molecular mass 468 and from the spectrum the following structure was postulated: OH C28H2007 Relative molecular mass = 468.1209 Synthetic considerations, however, suggested an alternative structure : OCH3 OH C2gH2406 Relative molecular mass = 468.1573 Further material was isolated and high-resolution mass spectrometry gave a relative mole- cular mass of 468.1544, showing C,,H,,O, to be the correct molecular formula, confirmation being obtained from the masses of the fragment ions: 241 -? 363-, OH I OCH3 ‘OCH, I bHI L- 227 L- 105 Formula Theoretical mass Measured mass C29H2406 468.157 3 468.1 544 C22H1905 363.123 2 363.1 244 c15H1303 241.0865 241 .OH5 C14H1103 227.070 8 227.0703 Gas Chromatography - Mass Spectrometry (GC - MS) Polymers contain trace amounts of residues of the catalyst used in their preparation and the identification of these is often necessary, particularly when a new catalyst is being evaluated.The use of gas chromatography in conjunction with mass spectrometry is required in order to separate the complex mixture of components that are extracted. For example, tetramethylsuccinodinitrile (TMSDN) has been detected in extracts of polymers prepared using azobisisobutyronitrile. This has been developed into a quantitative method using specific-ion GC- MS.In the development of this method it was discovered that substantial losses of TMSDN occur in the evaporation of methanolic solutions, which explained earlier difficulties in detecting residues of this catalyst. Even without concentration of the polymer extract it has been possible to achieve a lower limit of detection of 20 p.p.m. in the polymer.1052 SQUIRRELL: ANALYSIS OF ADDITIVES AND PROCESS Analyst, Vol. 106 A further family of catalysts often used are peroxides (e.g., benzoyl or lauroyl peroxide); these produce acids as residues which may be detected by mass spectrometry or by methyla- tion of the evaporated extract prior to GC - MS examination. Odour and Taint The analyst in the plastics industry may be required to trace the cause of odour and taint produced in foodstuffs packaged in plastic materials.This provides good examples of the use of high sensitivity GC - MS in the identification of such compounds. Two methods have proved useful for the concentration of these components. (a) Where the sample is a sealed plastic bag, the headspace gas from the bag (or many bags) is withdrawn through an adsorption tube normally packed with Porapak Q. The trapped organic species are then thermally desorbed for GC - MS examination. This has enabled residual printing and coating solvents in the bag headspace to be identified as causes of odour and taint. An interesting recent example was that of tainted sweets having a fatty taste. Examina- tion of the headspace gases from the polypropylene bags containing the tainted sweets detected aldehydes (pentanal and hexanal) in addition to the ff avourings (amyl acetate and limonene). Solvent washings from the internal surfaces of the bags were found to contain vegetable oil.Those from the most heavily tainted samples had a rancid odour and corre- sponded with the sample showing the highest concentration of aldehydes in the headspace. From these results it was deduced that the taint was not due to the film in contact with the sweets, but to the presence of rancid vegetable oil on the sweets themselves. The evolution of hexanal can apparently be used as a quantitative indicator of the rancidity of vegetable oils. (b) For containers and foods, the Likens and Nickerson combined solvent extraction - steam distillation procedure4 has proved useful.The advantage of this form of extraction is that large amounts of sample may be extracted with a small volume of organic solvent prior to further concentration by evaporation. In this procedure the evaporation of the organic solvent can concentrate trace amounts of solvent impurities to a significant proportion of the final extract and we therefore use redistilled solvents and distilled rather than de- ionised water, which can contain trace organics from the resin bed. A blank extraction is always carried out. This method was used effectively in the resolution of another tainting problem. About 300-400g of “good” and “bad” printed plastic lids were shredded and extracted. Direct gas-chromatographic analysis of the two extracts showed a large number of peaks and the two chromatograms differed in only one respect: the “bad” sample revealed one small extra peak that was not present in the “good” sample.GC - MS examination of this particular peak indicated an acetate ester of a polyol. Further extractions, followed by TLC, enabled sufficient of ,the component to be isolated to obtain infrared and nuclear magnetic resonance spectra, which, together with the mass spectrometric data, indicated that the contaminant was Butyl Carbitol acetate [2-(2-butoxyethoxy)ethyl acetate, CH,C02(CH2CH,0),C4H9], which is used in certain ink formulations. It was subsequently proved that as little as 2 p.p.m. of this acetate could cause a detectable taint in the foodstuff in question. In this example, the straightforward mass spectrum alone did not give a direct identifica- tion, as no molecular ion was produced.Shortly afterwards, however, this sample was used as a test of a spectrometer which could be employed in the chemical ionisation mode. Under these conditions a quasi-molecular ion of m/z 205 was formed, and had this technique been available at the time it is highly likely that the contaminant in question would have been identified without the considerable amount of extra work. We now make extensive use of chemical ionisation techniques and Fig. 10 shows electron impact and chemical ionisation mass spectra of Butyl Carbitol acetate. Processing Volatiles Recently there has been increased interest in the nature and amount of volatiles produced during the processing, i.e., drying, extruding, moulding, etc., of plastic compositions.In our laboratories work has been carried out to identify the volatiles produced in the drying and sintering of PTFE compositions. Mass spectrometry was used, the samples beingOctober, 1981 CH3 I C-CHZ- I c=o I 0 -CH3 RESIDUES IN PLASTICS MATERIALS - 1053 57 87 43 117 1144 50 100 150 200 2 rniz 22592 100.0 tb' 37 (b) 205 981 miz 10. Electron impact (a) and chemical ionisation (b) mass spectra of Butyl Carbitol acetate. heated in a stainless-steel tube and the volatiles passed directly to the mass spectrometer through a heated line. Programming of the heating from 40 to 400 "C enables process conditions to be simulated. Using a data system to scan repetitively the mass spectrum and store the data from all scans, the products may be identified and specific-ion displays enable the rate of evolution of these products to be followed.For example, the evolution of perfluoroalkenes from the breakdown of a fluorinated surfactant, and alkylphenols and ethoxylated phenols from polyethylene oxide condensates used in the emulsion polymerisa- tion have been detected. Pyrolysis - Gas Chromatography - Mass Spectrometry Pyrolysis - gas chromatography has been used for many years for the characterisation of plastics materials, particularly when they are of an intractable nature owing to cross-linking or are very heavily filled. The technique has now been extended to include mass spectro- metry and is of particular value when minor components need to be identified.An example has been described by Sharp and Paterson5 for the identification of small amounts (1-10%) of copolymerised unsaturated acids in acrylic polymers. The method can be summarised as follows : Jn R C-CHi c=o OH I 1 I - Propylation - . rn - n 1 J m R I CH3 I ~ y ~ ~ i ~ ~ i ~ &CH2 C=CH2 ------+I I + .alcohols, alkenes, COz, etc. F=O C==O O---CH3 OC3H7 I The copolymerised acid is propylated by treatment of the sample with Propyl 8 reagent, the resultant polymer is pyrolysed and the propyl ester of the acid, if present, is identified by GC - MS. By this procedure copolymerised acrylic or methacrylic acid has been identified1054 Analyst, VoZ. 106 in terpolymers with (a) butyl acrylate and styrene, (b) methyl methacrylate and ethyl acrylate and (c) ethylene and propylene. A methyl methacrylate - a-methylstyrene - maleic acid terpolymer, when examined by this propylation - pyrolysis procedure, yielded dipropyl fumarate and a smaller amount of dipropyl maleate.Attempts to detect copolymerised itaconic acid by this technique have failed, probably owing to the difficulty in esterifying the acid group. SQUIRRELL : ANALYSIS OF ADDITIVES AND PROCESS Ultraviolet and Infrared Spectroscopy The extensive use of ultraviolet and infrared spectroscopy in the analysis of antioxidants and ultraviolet-absorber systems used particularly in polyolefin products is well known196 and will not be dealt with in detail here. In many instances, of course, dual wavelength measurements are required to permit the resolution of mixtures and correction for overlap absorbances.A simple example of this in the ultraviolet region concerns the resolution of two of the oxidation products of Topanol 0 (butylated hydroxytoluene), which in some instances can cause troublesome yellowing of the polymer and which can be evident in process solvents : O*CH-CH - G O - 3,3‘,5,5‘-tetra (fert-buty llstilbenequinone (SO) X X 0 3,3:5,5’-tetra (tert-butyl )diphenoquinone (DPQ) X X X = tert-butyl The spectrum of the extracted oxidation products is measured in hexane over the wavelength range 550-325 nm. A base line is drawn as a tangent from 480 to 350 nm. The absorbance at 419 and 444 nm is measured and referred to separate calibrations prepared with SQ and DPQ. SQ shows two absorbance maxima, at 444 and 417 nm, and DPQ a single maximum, at 419 nm. Graphs are prepared at both 419 and 444 nm, permitting algebraic calculation of the concentration of each component.Infrared - Laser Raman Spectroscopy Although both of the above spectroscopic methods have a wide use in their own right, the example given below demonstrates well the complementary value of the two methods, taking advantage of the fact that elements of high atomic number, e.g., antimony and bromine, have relatively more intense Raman spectra but the lighter elements show up clearly in the infrared spectra. When examined by infrared spectroscopy the strongest bands (9.8 and 14.9pm) were due to a talc-type material and bands of medium intensity were assigned to polypropylene and possibly antimony trioxide (13.4 pm).Additional weak bands in the 7.3-7.7 pm region were possibly due to decabromodiphenyl ether. In the Raman spectrum, however, the strongest bands (250 and 185cm-1 shift) confirmed the presence of antimony trioxide and some bands of medium intensity confirmed the presence of decabromodiphenyl ether (doublet at 140, triplet at 220 cm-l shift) and polypropylene (800, 835, 1 150, 1325, 1450 and 2900 cm-l shift). The silicate bands that obscured regions of the infrared spectrum were not observed in the Raman spectrum. Polymer granules suspected to be PVC were submitted for identification.Octo bey, 1981 RESIDUES I N PLASTICS MATERIALS 1055 Other Spectroscopic Methods and Thermal Analysis Space does not permit a lengthy description of the other spectroscopic methods used in the analysis of plastics, but the main applications to additive and trace analysis are summarised in Table V.TABLE V OTHER TECHNIQUES APPLIED IN PLASTICS ANALYSIS Technique Principal use X-ray diffraction . . . . Identification of crystalline additives in situ (e.g., hydrates) Thermal analysis . . . . Detection of polymer phases with different melting-points NMR . . f . .. . . Identification of additives after extraction. Calibration of Identification of "bound" Identification of polymer phases infrared methods (structure). additives Wide-line NMR for liquid - solid phase blends ESCA.. .. .. . . Surface composition studies, e.g., presence of slip additive and antistatic agents a t polymer surfaces MOLE .. .. . . Particulate contamination a t or near film surfaces SIMS .. .. .. . . Ultra-surface analysis : catalyst studies Spectrofluorescence . . . . Degradation studies : surface oxidation ; study of scintillation additives Chemical Methods Chemical and colorimetric methods of analysis are used, of course, in both routine and special-case analyses. The differential non-aqueous titration of slip additives extracted from polyolefin compounds is fully described in references 1 and 7 and several variations of the titrant and solvent system are used in particular circumstances. The normal slip additives encountered are mixtures of a long-chain amide with a secondary amine such as Ethomeen T12 [R-N(CH,CH,OH),, where R is derived from tallow], and these aredifferenti- ated by a perchloric acid titration in an organic solvent containing glacial acetic acid and acetic anhydride. In some circumstances, particularly at the lower concentration levels and when alkali metal or calcium salts, which would interfere in the titration, are present, a colorimetric method is used for Ethomeen T12.This method relies on measurement of the chloroform- soluble complex formed by the reaction of the extracted additive with a mixed indicator solution containing bromophenol blue and bromocresol green at pH 3.5. Interesting variations on this method are being developed by Udris, particularly for the determination of polymeric amine-type additives that present problems of extraction and that form insoluble complexes with the indicator reagents. In such instances promising results have been obtained by measuring the decrease in colour in the aqueous phase rather than the increase in the organic phase.Instrumental methods for the determination of water in polymeric materials are numerous and many rely on the release of the water from the polymer matrix by heat. In some instances, however, particularly for polyethylene terephthalate, heating in the presence of parts per million amounts of water causes some hydrolysis of the polymer and consequent erroneous results. A simple Karl Fischer method is thus preferred and can be summarised by reference to Fig. 11. Dry methanol is placed in the sample flask, excess of dilute Karl Fischer reagent (1 ml = 1 mg of water) is added and the contents are boiled under reflux for 20 min to dry the apparatus. Slightly wet methanol is then added to the flask and the contents are immediately titrated to a visual end-point by means of the Karl Fischer reagent.The flask and contents are then refluxed for 6 h and any water picked up during this time is again titrated with the Karl Fischer reagent. This establishes that the equipment is giving a low blank value, which should be less than the equivalent of 2 ml of Karl Fischer reagent. For the analysis of the sample the apparatus is again pre-dried and the sample introduced via the 70-ml top cup. The cup is filled to a 50-g mark with polymer granules and then the latter are allowed to drop into the flask by releasing the clip on the wide-bore rubber tubing. In this way the exposure of the sample to the atmosphere is minimal. The 6-h reflux period is then repeated and the water content of the sample is calculated from the difference between the titre for sample and blank experiments.1056 SQUIRRELL Silicone tubing - (25 rnrn i.d., 2-rnrn wa I I) lsornantle Fig. 11. Apparatus for the determina- tion of water in polyethylene terephthalate. Conclusion In conclusion, I would stress that analyses for additives and trace organics in plastics materials can be made easier by a detailed understanding of the industry. Many of the modern analytical techniques are of great value in providing faster and more precise analyses, but much can also be done by the established chemical, chromatographic and spectroscopic techniques, particularly when these are supported by a degree of semi-automation for instru- ment control and data handling. The procedures described in this paper are the work of a team of analysts working in close co-operation. The assistance given by Messrs. L. H. Ruddle, J. Udris, A. R. Jeffs, P. L. Warren, P. E. Arnold and J. I. Sharp is particularly acknowledged. I am also indebted to the directors of ICI Petrochemicals and Plastics Division for permission to publish this paper. References 1. Haslam, J., Willis, H. A., and Squirrell, D. C. M., “Identification and Analysis of Plastics,” Second 2 . Thain, W., Editor, “The Determination of Vinyl Chloride-A Plant Manual,” Third Edition, Chemical 3. Jeffs, A. R., in Kolb, B., Editor, “Applied Headspace Chromatography,’’ Heyden, London, 1980, 4. Likens, S. T., and Nickerson, G. B., PYOC. Am. SOC. Brew. Chem., 1964, 5, 5. Sharp, J. L., and Paterson, G., A~alyst, 1980, 105, 517. 6. BS 2782: 1978, “Methods of Testing Plastics. Part 4. Chemical Properties,” British Standards 7. Squirrell, D. C. M., “Automatic Methods in Volumetric Analysis,” Hilger and Watts, London, 1964. Edition, Heyden, London, 1980. Industries Association, London, 1977. Chapter 12. Institution, London.

ISSN:0003-2654    

DOI:10.1039/AN9810601042

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Analyst, October, 1981, Vol. 106, pfi. 1057-1070 1057 Various Applications of Functional Group Analysis* G. D. B. van Houwelingen A kzo Research, Corporate Research Department Amhem, P.O. Box 60, Amhem, The Netherlands The importance of functional group analysis is demonstrated with a number of examples in various fields. Special attention is given to the determination of end groups in several kinds of high and low relative molecular mass materials. Methods for the determination of the following groups are described : hydroxyl groups (polymers, esters) , carboxyl groups (polymers, ester-inter- change products), unsaturated groups (vinyl ester end groups in polymers), anhydride groups (polymers) , amino groups (polymers, derivatised silica materials), epoxide groups (resins, derivatised silica materials) and quaternary ammonium compounds (surfactantq) .The methods include derivatisation procedures and several techniques for quantitative determination (spectrophotometry, X-ray fluorescence, coulo- metry and potentiometric and photometric titration). Keywords : Functional group analysis ; polymers ; derivatisation Functional groups are defined as particular reactive species present in an organic substance or mixture. They can be detected and sometimes measured directly by physical or chemical methods in a quantitative manner, but in many instances derivatisation is necessary and the substance formed or consumed is determined. Procedures for determining certain functional groups are usually not generally applicable ; owing to solubility properties, steric factors and the presence of interfering reactions, the use of various reagents and special solvents is necessary.There are about 100 possible functional groups and the determination of only a few of them will be discussed here. For a particular functional group, one or more applications for the quantitative determination of this group are given. The methods described deal mainly with procedures based on chemical derivatisation routes for the determination of hydroxyl, amino, carboxyl, anhydride, epoxide, unsaturated and quaternary ammonium groups in various products, including high relative molecular mass polymers, Especially in the end group analysis of high relative molecular mass materials, problems due to insolubility in common analytical solvents are encowtered, and finding suitable media in which derivatisation or titration procedures can be carried out is an important aspect of the development of analytical procedures.In another stage of the investigation the insolubility may be used advantageously for purification purposes. Finally, the con- centration of end groups varies inversely with the relative molecular mass, which makes high demands on the analytical techniques used in this field. Hydroxyl Groups Procedures developed for the determination of hydroxyl groups include titrimetric, spectro- photometric , nuclear magnetic resonance spectroscopic , gas-chromatographic, electrochemical and thermometric methods. In our laboratory the quantitative determination of hydroxyl groups in low and high relative molecular mass polymers is of importance.Examples of such polymers are 0 0 with n = 10-100 and x = 2 [polyethylene terephthalate (PETP)] or 4 Cpolybutylene terephthalate (PBTP)], and ester-interchange products of PBTB and polypropylene glycol * Plenary Lecture presented at the Joint NL - UK Symposium on Quantitative Organic Analysis, Noordwijkerhout, The Netherlands, April 22-24, 1981.1058 VAN HOUWELINGEN VARIOUS APPLICATIONS Analyst, Vol. 106 (PPG) (PBTP - PPG elastomers). These products have limited solubility but are soluble in phenols, hot nitrobenzene or aniline, dichloroacetic acid and hexafluoroisopropanol. The hydroxyl groups in these products are determined by acetylation with an excess of dichloroacetic anhydride (DCAA) in dichloroacetic acid (DCA) and measurement of the amount of acetylation by a chlorine determination.Owing to the low content of hydroxyl groups, especially in high relative molecular mass materials, the determination of the excess is inaccurate and the determination of the amount of reagent incorporated is much more attractive. This kind of procedure requires an effective elimination of the excess of reagent, which is normally carried out in dissolution and precipitation steps. As can be seen in Fig. 1, a reaction time of 1 h suffices. After the reaction the solution is poured into water and the precipitated polymer is washed out. To remove the last traces of solvent and acetylation agent, re-precipitation of the derivatised polymer from a hexafluoroiso- propanol (HFI) solution into water is carried out.For PETP and PBTP polymers with a low hydroxyl content (below 100 mmol k g l ) the re-precipitation is carried out from a solution in nitrobenzene into cold light petroleum, in order to obtain more reproducible results ; probably trace amounts of chlorine-containing impurities adhere to the polymer when using the HFI - water method. The content of hydroxyl groups is subsequently determined by measurement of the chlorine content of the purified derivative. This is done either by potentiometric titration with silver ions after combustion or by X-ray fluorescence spectroscopy after tabletting. The latter technique is particularly suitable for rapid analysis. Details of the method have been published e1sewhere.l By carrying out several consecutive precipitations it could be demonstrated that only one re-precipitation from a solution in HFI (low relative molecular mass materials) or nitro- benzene (high relative molecular mass materials) into water and light petroleum, respectively, is necessary (see Fig.2). We checked, as is necessary with this kind of procedure, that no loss of the lowest relative molecular mass fractions occurred during the precipitation step.1 The derivatisation is carried out in a 10% m/m solution of DCAA in DCA at 60 "C. W PBTP ' 4 0 t 6 1 2 3 4 Reaction time/h 680 f 200 8 160 ? \ \ \ *--. PBTP x 1 2 3 N Fig. 1. Esterification of various samples Fig. 2. OH content as a function of the number of re-precipitations ( N ) ; X = OH content of the sample after precipita- tion in water.with DCAA for different periods at 60 "C. The suitability of this method for a number of samples is demonstrated in Table I. The standard deviation of the method for high relative molecular mass PETP and PBTP materials is 0.7 mmol k g l , whereas for the low relative molecular mass materials it is about 10 mmol kg-l. For the determination of low contents of hydroxyl groups in low relative molecular mass esters we apply the frequently used method in which the sample is reacted with a solutionOctober, 1981 OF FUNCTIONAL GROUP ANALYSIS 1059 of vanadium 8-hydroxyquinolinate (VSHQ). According to Tanaka and Kojima,2 a coloured complex with the following structure is formed: 0 0 1 I 1 1 Q-V-Q + ROH - Q-V-Q + H20 OH bR After removal of the excess of reagent by extraction, the complex is acidified with dichloro- acetic acid and the blue colour formed is measured at 620nm.Calibration is carried out with an alcohol as internal standard. The procedure followed is that of Harrison3 with the following modifications: (a) to obtain a more stable reagent the solvent dimethylformamide - toluene is replaced with dimethylformamide - monochlorobenzene (see Table 11) ; and (b) removal of the excess of reagent by extraction with 4% V/V triethylamine solution instead of sodium hydroxide solution (favourable separation of the layers). TABLE I HYDROXYL END GROUP CONTENT OF SEVERAL PETP, PBTP AND PBTP - PPG SAMPLES Sample PETPA .. PETPB .. PBTPA .. PBTPB .. PBTP C PBTP - PPG'A PBTP - PPG B PBTP - PPG C PBTP - PPG D Viscosity Hydroxyl end groups/ ratio* mmol kg-l .* 1.80 30.9-29.8 30.6-30.4 . . 2.07 70.7-70.6 . . 2.08 50.6-49.3 . . 1.25 742-729 . . 1.34 329-345 . . 1.50 229-212 . . 1.82 14.4-14.8 . . 1.24 664-653 . . 2.38 5147 Remarks Total end groupst 76.6 mmol kg-l Total end groups 79.1 mmol kg-l Total end groups 95.2 mmol kg-I Total end groups 92.6 mmol kg-I MN (calc.)$ 2700; MN (measured) 24009 MN (calc.) 9000; MN (measured) 10100 MN (calc.) 35000; MN (measured) 44800 * Measured for a 1% m/m solution in m-cresol a t 25 "C (PETP and PBTP), or for a 1% na/V solution in t Total end groups is sum of OH + COOH + methyl ester end groups. 1 MN measured by gel-permeation chromatography in m-cresol as a solvent. 9 Calculated from OH + COOH content. o-chlorophenol a t 25 "C (PBTP - PPG).TABLE I1 STABILITY OF THE VANADIUM 8-HYDROXYQUINOLINATE REAGENT Decrease, yo * r A 3 Storage time/h In benzenet In toluenef In monochlorobenzene 24 50 75 3 48 15 72 33 * Determined with heptan-1-01 as the alcohol. t Values taken from reference 3. As presented in Table 111, the absorbance is only slightly dependent on the chain length for C,-C,, primary alcohols.1060 VAN HOUWELINGEN : VARIOUS APPLICATIONS Analyst, Vol. 106 TABLE I11 ABSORBANCE OF 10 pm01 OF ALCOHOL AT 620 nm (l-Cm PATH LENGTH) Alcohol .. . . Butan-1-01 Heptan-1-01 Dodecan-1-01 Octadecan-1-01 Absorbance. . . . . 0.49 0.54 0.56 0.58 The application of this method to some esters of new types of acids showed that a precise determination is possible (Table IV). The standard deviation of the V8HQ method is 1.5 mmol kg-1, which is about ten times more precise than that of the classical acetylation procedure.TABLE IV COMPARISON OF THE V8HQ METHOD WITH THE ACETYLATION METHOD OH content/mmol kg-1 r A > Sample V8HQ Acetylation I .. . . 47.9-47.6 50-70 I1 . . . . 35.5-34.9 35-50 IV .. . . 73.8-74.9 50-70 43.1-46.8 I11 . * . . 130.1-126.7 170-120 Carboxyl Groups The determination of these groups in high relative molecular mass polyesters and poly- amides gives, in combination with other data, important information about the degree of polperisation, chain branching, degradation and thermal stability. The choice of the method depends on the solubility of the polymer. Most methods are based on titration in non-aqueous media with visual, potentiometric or photometric indication of the end-point.A very convenient technique for the determination of carboxyl groups in PETP and polyamide 6 (PA6) is photometric titration. The sample is dissolved in o-cresol at 125 "C, cooled, diluted with chloroform and, after addition of bromophenol blue (for PETP) or bromo- cresol green (for PA6), the titration is carried out in a spectrophotometer with tetrabutyl- ammonium hydroxide (TBAH). The change in absorbance is recorded continuously and the end-point is determined graphically (see Fig. 3). A detailed description was given by Van Lingen.5 Colour change interval of bromophenol blue 1 Degree of neutralisation ---b Fig. 3 (a) Potentiometric and (b) photometric titration of PETP Table V surveys applications of the photometric titration method, together with a com- parison with a potentiometric titration method with TBAH in aniline (DMT and low relativeOctober, 1981 OF FUNCTIONAL GROUP ANALYSIS 1061 molecular mass PETP), o-cresol (high relative molecular mass PETP) or benzyl alcohol - methanol - water (PA6).Generally the agreement is very satisfactory. Photometric titration, however, is much faster than potentiometric titration and has been shown to be a very convenient technique in routine analysis. TABLE V APPLICATION OF THE PHOTOMETRIC TITRATION METHOD TO VARIOUS SAMPLES Content of carboxyl groups/mmol k g l Photometric Potentiometric titration titration -& Standard Standard Sample Z* deviation Z* deviation f A 3 DMT .. .. .. A 0.31 0.03 0.38 0.04 B 0.76 0.03 0.83 0.02 PETP (MN low) .. .. c 4.3 0.25 4.0t 0.10 D 12.0 0.34 12.0t 0.30 E 17.4 0.56 17.5t 0.49 PETP (MN high) . . .. F 36.8 0.32 35.2 0.9 G 68.3 0.29 68.2 1.2 H 113.4 0.52 112.8 1.8 PA6 .. .. .. I 73.3 0.31 72.4$ 1.0 53.2 0.22 52.5$ 0.37 JK 35.3 0.45 36.2$ 0.32 * Mean of four independent determinations. t Titration in aniline at 40 "C. $ Titration in benzyl alcohol - methanol - water. For aromatic polyamides, which are used for the spinning of high-strength and high- modulus fibres, the choice of solvents is very limited. Whereas in poly(m-phenylene iso- phthalamide) potentiometric titration in dimethylfonnamide is possible,6 direct titration of poly(9-phenylene terephthalamide) (PPD-T) is impossible as it is soluble only in strong acids such as concentrated sulphuric acid and methanesulphonic acid.An attractive method for determining carboxyl groups in PPD-T is based on the Schmidt reaction, in which the carboxyl group is subjected to reaction with sodium azide in sulphuric acid.? The acylazide formed rearranges to an arnine with the liberation of nitrogen and carbon dioxide : NaN3 m N H 2 + COj + NJ mCooH %,so,+ The carbon dioxide evolved is measured quantitatively by an automatic non-aqueous ti tra- tion. The apparatus for this determination is shown in Fig. 4. The suitability of this method has been investigated with the model compound NN'-bis- (9-carboxybenzoy1)-9-phenylenediamine (BCP) : Fig. 5 shows the conversion of BCP in 100-lOl~o sulphuric acid at 50 "C at two concentration levels. A reaction time of 2 h gives a quantitative conversion, which also suffices for polymer solutions (see Fig.6). As additions of known amounts of BCP to the viscous polymer solutions are recovered quantitatively,? the experimental conditions for complete removal of carbon dioxide are correct. The sulphuric acid concentration is very critical; as can be seen in Table VI, a concentration of 100~o is essential for quantitative conversion. The standard deviation for the method is 2.6 mmol k g l .1062 VAN HOUWELINGEN : VARIOUS APPLICATIONS Analyst, Vol. 106 Fig. 4. Apparatus for the determination of carboxyl end groups: A, nitrogen connection; B, buty rubber tubing; C . three-way tap; D, gas absorption bottle; E, reaction vessel; F, turnable spoon; G, nitrogen inlet tube; H, magnetic stirring bars; I, oil-bath with thermometer; J, magnetic stirrer with facilities for heating; K, adjustable support; L, glass column filled with MnO, on asbestos (length 50 em, diameter 2-4 cm); M, absorption and titration vessel; N, receiver; 0, ascarite tube; P, combined glass - calomel electrode; Q, inlet tube burette; R, drain; S, titration equipment; and T, gas absorption bottle.100 90 5 80 $ 70 2 60 50 0, 40 0 30 20 10 Y- 0 30 60 90 120 150 Reaction timelmin Fig. 5. Recovery of BCP after Schmidt reaction. Amount of BCP taken: A, 10.0pmol; and B, 100.0 pmol. 180 1 60 7 140 ET) Y 5 120 g 100 8 I 60 0 s 40 E 4- f 80 20 0 30 60 90 120150 180210240 Reaction time/min Fig. 6. Conversion of the carboxyl end groups of several samples for different periods at 50 "C. A, PPD-T (viscosity ratio = 2.09); B, yarn (viscosity ratio = 4.09); and C, PPD-T (viscosity ratio = 5.74).October, 1981 OF FUNCTIONAL GROUP ANALYSIS TABLE VI CONVERSION OF BCP AT VARIOUS SULPHURIC ACID CONCENTRATIONS 1063 Sulphuric acid concentration, yo .. . . 98.3 99.2 99.7 100.2 101.0 Conversion of BCP, yo . . .. . . 3.8 16.2 21.9 100.1 100.5 Unsaturation Chemical methods for the determination of unsaturation are often based on halogenation or hydrogenation. An elegant method for determining unsaturated compounds is the in s i t ~ generation of a halogenating agent (e.g., bromine) by constant-current coulometry. The amount of reagent consumed is proportional to the amount of electricity used to generate the reagent. The reaction between the double bond and bromine is catalysed by mercury(I1) chloride.The advantages of a coulometric method are that a standard reagent is not needed, minimum side-reactions (substitution) occur because the bromine concentration is kept low, and the method is precise, is suitable for low levels and can readily be automated. A survey of applications was given by Hirozawa.8 We have used coulometric bromination to determine vinyl ester end (VEE) groups in PETP formed by thermal chain scission and the source of subsequent liberation of acetaldehyde, 0 H Only a brief description of the method is given here; more details will be published el~ewhere.~ The constant-current generation of bromine is carried out in a medium of dichloroacetic acid (DCA) , water, potassium bromide and mercury(I1) chloride. To this medium an amount of the polymer, previously dissolved in HFI and diluted with anhydrous DCA, is added and bromine is generated.The suitability of this method was tested with the model compound methyl vinyl tereph- thalate (MVT) : The end of the reaction is detected biamperometrically. Additions of 14.2 and 1.0 pmol of MVT (corresponding to 30 and 2 mmol of VEE group per kilogram of polymer) were recovered quantitatively (recoveries of 99.8 ant1 98.5% , respec- tively). The coulometric analysis must be completed within 30 min, because after longer times the hydrolysis of the VEE group is no longer negligible. The VEE group is not the only reactive moiety in PETP that consumes bromine, as impuri- ties present also do so. To determine this background, in a second sample the VEE group is previously hydrolysed at 80 "C in a dichloroacetic acid - water medium.Fig. 7(a) shows the relationship between the bromine consumption and the hydrolysis time for MVT and Fig. 7 ( b ) represents the same relationship for two polymers containing a high and a negligible amount of VEE groups. It can be seen that the VEE group in MVT is hydrolysed completely after 6 6 h a t 80 "C. For polymer A a sharp decrease occurs, which can mainly be attributed to the hydrolysis of the VEE group in the polymer. In addition, a much smaller effect is observed due to the hydrolysis of the bromine-consuming impurities [see relationship for polymer B, Fig. ? ( b ) ] . This small effect is corrected by extrapolating this relationship from time 6h to time zero (Le., 0.7 mmol k g l ) .1064 VAN HOUWELINGEN : VARIOUS APPLICATIONS Analyst, Vol.106 (a) -, 20 E 2 16 C .- E 12 5 $ 4 8 - 00 1 J I 1 4 8 12 16 20 24 28 4 8 12 16 20 24 28 Hydrolysis time/h Fig. 7. (a) Hydrolysis of MVT. (b) Hydrolysis of VEE groups in PETP: A, PETP with a high VEE group level; B, PETP with a negligible VEE group level. The VEE group content is calculated by subtracting the background ( i e . , the value measured after 6-h hydrolysis + 0.7 mmol kg-1) from the content originally measured. The standard deviation of the method is 0.2 mmol kg-l. Application to some PETP polymers is illustrated in Table VII. It can be seen that samples with almost identical viscosity ratios (A-D) may have different levels of VEE groups, depending on the temperature at which the polycondensation is carried out.It is also apparent that, as expected, a higher level of VEE groups corresponds to a higher level of carboxyl groups. Polymerisation at higher temperatures also results in a higher background. TABLE VII VINYL ESTER END GROUPS IN PETP SAMPLES Viscosity VEE group/ Background/ Sample ratio* mmol kg-1 mmol kg-1 A .. . * 1.59 1.1,l.O 1.5,1.4 B .. . . 1.62 2.3 2.6 c * . . . 1.62 6. I 3.5 D .. . . 1.60 18.6,17.9 4.4,4.4 E .. .. 1.89 5.5,5.6 3.3,3.3 F .. .. 1.91 1.3,1.2 2.7,2.7 Carboxyl groups/ Polycondensation mmol kg-1 temperature/"C 20 33 56 80 50 22 280 2 84 294 302 280 2 80 * Measured for a 1% m/m solution in m-cresol at 25 "C. Amino Groups Chemical determination of amino groups is often carried out by titrimetric or spectro- photometric analysis.By proper selection of solvent and derivatisation techniques, differ- entiation between arnines (primary, secondary, tertiary and quaternary ammonium salts) is possible.lOpll Two examples of the determination of amino groups in insoluble or poorly soluble materials are described below. The first example relates to the determination of this group in derivatised silica materials used for affinity chromatography. This material is illustrated schematically in Fig. 8. The amino group can be determined directly in this material by heterogeneous titration with perchloric acid and acetic acid as titration medium. It is essential to apply an equilibrium titration. This is illustrated in Fig. 9. In Fig. 9(a), the first part of the titration was carried out continuously (0.2 ml min-l) with a reduced titration speed on the steep part of the curve.When the titration was stopped at A, a slow shift in potential to the more negative ( i e . , alkaline) direction occurred. When the titration was continued at B under carefully chosen equilibrium conditions (b., with a potential stability programmed addition) a second jump was found. The total consumptionOctober, 1981 OF FUNCTIONAL GROUP ANALYSIS 1065 A 8 C D T > E .- c.. C Q1 c.. a \"' B 0.01 N HClOJml ,-+ Fig. 8. Derivatised silica Fig. 9. Titration of silica material that contains amino material: A, silica; B, coat- groups with 0.01 N perhcloric acid under different conditions. ing; C, spacer group; and D, (a) Continuous and equilibrium titration and (b) equilibrium functional group.titration. Apparatus, Mettler DK 11 and DK 15. (C;) was equal to that of a direct equilibrium titration (Ci). Using this technique a co- efficient of variation of 274 was obtained. The accuracy of this determination was checked with results from a nitrogen analysis (Fig. 10). In the second example, the determination of amino groups in poly (p-phenylene tereph- thalamide), titration is impossible because this polymer is soluble only in strong acids. To determine the amino groups, use is made of a heterogeneous derivatisation with 1-fluoro-2,4- dinitrobenzene (FDNB)' : The polymer is treated for 4 h at 80 "C with FDNB in an ethanol-hydrogen carbonate medium. After washing out the excess of reagent, the dinitrophenyl group introduced is measured spectrophotometrically after dissolution of the derivatised polymer in methane- sulphonic acid.The measurement is carried out at 430nm against a solution of the same amount of underivatised polymer (Fig. 11). Fortunately, the molar absorptivity of un- reacted FDNB at this wavelength is very low, which means that trace amounts of non- washed FDNB will contribute to only a small extent. The spectrophotometric method was calibrated with the FDNB derivative of the model compound NN'-his@-aminophenyl) terephthalamide : NO2 NO2 and the correctness of the whole procedure was confirmed by the use of carbon-14-labelled FDNB and measurement of the incorporated activity (for details, see reference 7). The completeness of the derivatisation was checked carefully by comparing the hetero- geneous derivatisation with a homogeneous derivatisation for low relative molecular mass materials (see Fig.12) and by a re-derivatisation procedure for high relative molecular mass polymers, which showed that virtually no additional groups could be derivatised after the first derivatisation; details are given in reference 7.1066 VAN HOUWELINGEN : VARIOUS APPLICATIONS Analyst, Vol. 106 Z 40 80 120 160 NH2 content by titration/pmol g-' Fig. 10. Comparison of two methods for the determination of amino groups in silica. 2.0 1.4 P 5 1.0 .r - 8 0.8 - C ro LI $j 0.6 - ' 0.4 - 0.2 - 0 400 420 440 460 480 500 520 540 Wavelen g t h/n m Fig. 11. Absorption spectra of underi- vatised (A) and FDNB-derivatised (B) PPD-T. Concentration: 100 mg per 50 ml of methanesulphonic acid.Yarns are poorly accessible to FDNB and the derivatisation must be applied to precipitated polymer obtained from yarn dissolved in sulphuric acid (see Table VIII). TABLE VIII ACCESSIBILITY OF YARNS AND PPD-T TO FDNB NH,/mmol kg-' I A -l Sample Without precipitation After precipitation Yarn 1 . . .. 7.4 24.0 Yarn2 .. .. 25.6 48.4 Yarn 3 .. .. 11.5 59.0 PPD-T 1 .. .. 19.9 23.0 PPD-T2 .. .. 30.7 29.4 Quaternary Ammonium Salts For the manufacture of galvanic metal - polytetrafluoroethane (PTFE) layers on metal, the PTFE is dispersed in galvanic baths by using two types of surfactants, vix., one cationic, [C,Fl,SO&HC,H,N+(CH3),],S0,Z- (CAT), and the other non-ionic, C,F,,SO,NC,H,- (CH,CH,O) ,H (NI). Ion-selective electrodes are often used for the determination of surfactants, either via direct measurement of the potential or as an indicator electrode in a titration.l2,l3 The electrodes usually have a specially prepared membrane containing the active phase.Recently Seligl4 described the use of a calcium electrode for this purpose. A well known reagent for quaternary ammonium salts is tetraphenyl borate (TPhB), which forms insoluble complexes with cationic surfactants.15 Applications of this reagent were described by Vytras.l* For the determination of CAT in baths we apply a potentiometric titration with TPhB in which a home-made electrode is used as the indicator electrode. The construction of the electrode is shown in Fig. 13.October, 1981 OF FUNCTIONAL GROUP ANALYSIS 1067 40 80 120 160 NH2 content by heterogeneous derivatisation/mmol kg- ' Fig.12. Derivatisation of PPD-T under heterogeneous and homogeneous conditions. Fig. 13. Electrode for cationic sur- factants : A, membrane consisting of PBTP (20%) and polytetrahydrofuran (PTHF) (80%); B, internal solution M CAT+ - TPhB- complex in cyclohexanone) ; C, internal reference electrode (Ag - AgC1) ; and D, electrode housing. Fig. 14 shows the relationship between the potential and the negative logarithm of the It can be seen that the electrode follows the Nernst equation in the At The drift of the elec- M An example of a titration of an aqueous solution of CAT is presented in Fig. 15. Titrations CAT concentration. region p(CAT) 4-6. concentrations exceeding low3 M, micelle formation probably occurs.trode is about 3 mV h-l, whereas the response for a concentration change from lo--* to (achievement of 95% of the maximum signal) is 8 s. were carried out with an equilibrium titrator (Mettler DK 15). The slope (63 mV) is somewhat higher than the theoretical value. - 100 > , -200 1 I I , 3 4 5 6 7 p(CAT) Fig. 14. Relationship between p(CAT) and potential. T > E . - m C 0 a .- c c 0.01 M TPhB/ml + Fig. 15. Titration of about 2 mg of CAT with 0.01 M TPhB. The non-ionic surfactant can also be determined by TPhB titration. The ethylene oxide According to Levens and Ikedal' group is charged positively by the addition of barium ions. the following structure is formed:1068 VAN HOUWELINGEN : VARIOUS APPLICATIONS Analyst, Yo,?. 106 VW(CH20CH20),H + Ba2+ ___+ This positively charged complex also reacts with the TPhB ions to form an insoluble complex.In this way separate titrations of CAT and NI are obtained, i.e., first the CAT is titrated and after the equivalence point has been passed barium chloride is added, followed by NI titration (see Fig. 16). Ethoxylated non-ionic surfactants often contain free polyethylene oxide ; this impurity is detected by the occurrence of two vague jumps in the titration (Fig. 17). As they do not correspond to the free and bound polyethylene oxide, quantitative measurement is impossible. The surfactants are collected by extraction from the galvanic baths after removal of the PTFE. The limits of detection of CAT and NI are 2 and 5 mg l-l, respectively. I TPhBiml + Fig. 16.Titration of a mixture of CAT and NI with TPhB; * indicates addition of BaC1,. L 0.01 M TPhB/ml --+ Fig. 17. Titration of NI con- taining free polyethylene oxide with 0.01 M TPhB. Anhydride Groups The example discussed here is the determination of anhydride and carboxylic groups in a The resin is derived from octadecene-1 and maleic commercial polyanhydride resin. anhydride : Lr c-c-c--c- Y 2 ; n C16H33 A h 0 0 0 The acid groups are formed by hydrolysis of the anhydride group. The common method involving reaction with an excess of aniline and subsequent back- titration of the excess1* is unsuitable, as the reactivity of the anhydride group is low. Even after hydrolysis with aqueous pyridine (containing 40% V/V of water) in a Parr bomb at 150 "C for 4 h, anhydride groups are still seen in the infrared spectrum.A suitable method for determining the anhydride group is titration with aqueous potassium hydroxide in pyridine after previous esterification of the carboxyl group with diazomethane. This esterification is carried out in diethyl ether-methanol (9 + 1); after methylation, which takes about 10 min for 0.5 g of sample, the solvents are removed by evaporation and a portion of the derivatised polymer is dissolved in pyridine and titrated.October, 1981 OF FUNCTIONAL GROUP ANALYSIS 1069 Fig. 18(a) and (b) show the infrared spectra of the resin before and after methylation, respectively. It can be seen that the absorption band of the acid group at 1710 cm-l disappears and a carbonyl band of the ester at 1740 cm-l is formed.The acid content of the sample is found from the difference in titres of an unmethylated and a methylated product. Wavenumberlcm- ' Fig. 18. Infrared spectra of polyanhydride resin: (a) before and (b) after methylation. The suitability of this method was checked by an independent method, vix., proton nuclear magnetic resonance spectroscopy. To a solution of the methylated resin in deuterochloro- form a known amount of dimethyl terephthalate was added as an internal standard and the amount of methyl ester groups (corresponding to the acid content) was determined quantita- tively. The content of the alkyl part was determined by measuring the terminal methyl group content of the alkyl chain. As the resin is prepared with identical amounts of the two mono- mers, this amount corresponds to the sum of acid and anhydride.As can be seen in Table IX, the agreement between the methods is satisfactory. TABLE IX ANALYSIS OF POLYANHYDRIDE RESIN Z = Mean value; n = number of determinations; and s = standard deviation. Method (C=O),O, % COOH, y-, n = 5; s = 0.29 KOH titration . . . . Z = 16.0 Z = 5.1 NMR .. 1 . . . 15.7 5.4 n = 5 ; s = 0.21 Epoxy Groups Titrimetric methods for epoxides are mainly based on titration with a halogen acid.l@ Epoxy groups in derivatised silica materials (for composition see Fig. 8) can be titrated with perchloric acid in acetic acid containing cetyltrimethylammonium bromide. As in the determination of amino groups, in this instance also an equilibrium titration is necessary (see Fig. 19). The coefficient of variation of the titration method is 2.5% at a level of about 20-200 pmol g-1.Low contents of epoxy groups are determined in our laboratory by reaction with 2,4- dinitrobenzenesulphonic The reaction product is made alkaline and the resulting orange colour measured at 498 nm. Amounts down to 0.01 pmol of epoxide can easily be1070 VAN HOUWELINGEN measured by this method. Applications to epoxide-containing polymers were described by Cramer and van Houwelingen.% 0.01 N HC104/ml --+ Fig. 19. Titration of equal amounts of silica material that contains epoxide groups under different conditions: A, equi- librium titration; and B, con- tinuous titration. Our thanks are due to Mrs. J. M. B. Pelgrim, Mrs. M. Cramer, Mr. A. H. M. Schotman and Mr. J. G. M. Aalbers for their considerable contributions to the many practical aspects of the work described. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. van Houwelingen, G. D. B., Peters, M. W. M. C.. and Huysmans, W. G. B., Fresenius Z . Anal. Tanaka, M., and Kojima, I., Anal. Chim. Acta, 1968, 41, 75. Harrison, S., Hinchcliffe, H., and Wdroffe, G. L., Analyst, 1974,99,491. Dutch Normalisation Institute, “Vegetable and Animal Oils and Fats,” Methods of Test, No. NEN Van Lingen, R. L. M., Fresenius 2. A n d . Clkpm., 1969, 169, 410. Kreshkov, A. P., Svetsora, L. N., and Emelin, E. A., Sov. Plast., 1968, 10, 53. van Houwelingen, G. D. B., Aalbers, J. G. M., and De Hoog, A. J., Fresenius 2. Anal. Ckem., 1980, Hirozawa, S. T., in Kolthoff, I. M., and Elving, P. J., Editors, “Treatise on Analytical Chemistry,” Aalbers, J- G. M., and van Houwelingen, G. D. B., to be published. Gyenes, I., “Titration in Non-aqueous Media,” Ilifle Books, London, 1967, p. 316. Kaka, B., and Vejd&lek, 2. J., “Handbook of Photometric Analysis of Organic Compounds,” Part Gavach, C., and Bertrand, C., A d . Chim. Ada, 1971, 55, 385. Fogg, A. G., Pathan, A. S.. and Bums, D. T., Anal. Chim. Actu, 1974, 69, 238. Selig, W., Fresenius 2. A d . Chem., 1980, 300, 183. Cross, J. T., A d y s t , 1965, 90, 315. Vytras, B. K., Am. Lab., 1979, February, 93. Levens, R. J., and Ikeda, R. M., Anal. Chem., 1965, 37, 671. Gyenes, I., “Titration in Non-aqueous Media,” Iliffe Books, London, 1967, p. 261. Gyenes, I., “Titration in Non-aqueous Media,” Iliffe Books, London, 1967, p. 400 Urbanski, J-, Plaste Kautsch., 1968, 15, 260. Cramer, M., and van Houwelingen, G. D. B., Analyst, to be submitted. Chem., 1978, 293, 396. 1046, 1961, p. 62. 300, 112. Part 11, Volume 14, John Wiley, New York, 1971, p. 23. I, Verlag Chemie, Weinheim, 1974, p- 408 (in German).

ISSN:0003-2654    

DOI:10.1039/AN9810601057

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Analyst, October, 1981, Vol. 106, pp.1071-1075 Aspects of the Analysis of Drugs and Drug Metabolites by Hig h-performance Liquid Chromatography* 1071 G. G. Skellern Drug Metabolism Research Unit, Department of Pkrmuceutical Chemistry, University of Strathclyde, Gtasgow, G1 1XW The biotransformation of a drug may result in a diversity of metabolites with widely differing physico-chemical properties. With the aid of suitable examples the applicability of adsorption, partition and paired ion high- performance liquid chromatography to the measurement of drugs and their metabolites is discussed in relation to their lipophilicity and polarity. Keywords : Drug analysis ; drug metabolite analysis ; high-perfwmance tiquid chvomatography Most drugs are by design lipphilic weak electrolytes whose relative molecular masses range from low to high.Essentially the metabolic biotransfonnations of a drug have been divided into two distinct types of reaction: functionalisation (phase 1) reactions involving enzymically controlled oxidations, reductions and hydrolyses, and conjugation reactions (phase 2).1 The functionalisation reactions often produce metabolites which have physico- chemical properties similar to those of the parent drug, although there is usually an attendant increase in polarity and a decrease in lipophilicity. The metabolism of the 1,4-benzo- diazepine chlordiazepoxide provides an example of a compund that is initially primarily metabolised to phase 1 metabolites (Fig. 1). However, conjugation reactions involve enzyme-catalysed reactions of the drug or its metabolite with endogenous compounds, producing polar metabolites whose physico-chemical properties are usually markedly different from those of the parent drug.As with phase 1 reactions, this often results in increased polarity and decreased lipophilicity; but the converse can be observed. fi-Hydroxy- acetanilide (paracetamol) is metabolised to a variety of polar conjugates; pKB, values range from 9.5 for the parent drug to less than 2 for its sulphate conjugate (Fig. 2). Chlordiazepoxide Oxidative dearnination I d T o 4 reduction N-Oxide ci -N CI I Ph Fig. 1. Chlorodiazepoxide metabolism. * Keynote Lecture presented at the Joint NL - UK Symposium on Quantitative Organic Analysis, Noordwijkerhout, The Netherlands, April 22-24, 1981.1072 SKELLERN : ASPECTS OF THE ANALYSIS OF DRUGS Analyst, Vol.106 NHCOCH, NHCOCH3 Fig. 2. Paracetamol metabolism. Because of the aforementioned physico-chemical properties of drugs and their metabolites, high-performance liquid chromatography (HPLC) is a satisfactory method for their measure- ment at low concentrations in biological fluids, as it combines the advantages of other bio- analytical methods in providing rapid separation and quantitation, but at room temperature. Adsorption, partition, ion-exchange and ion-pair chromatography are all applicable in drug metabolism and pharmacokinetic studies. Adsorption chromatography is limited to the quantitation of metabolites whose physico- chemical properties are similar to the parent drug, whereas partition chromatography (reversed-phase) using either alkyl- or phenyl-bonded supports is suitable for the measurement of compounds with wide-ranging pK, values and lipophilicities. A variety of methods (Table I) have been used to study in humans the elimination of the antithyroid drug methimazole (1-methylmercaptoimidazole) (Fig.3), which is also a meta- bolite of the anti-thyroid drug carbimazole, its N-carboethoxy derivative. Those methods which do not have a chromatographic separation are non-specific, as they could include sulphur-containing metabolite concentrations in the observed drug concentrations, which would account for the longer biological half-lives of methimazole. TABLE I LITERATURE VALUES FOR THE BIOLOGICAL HALF-LIFE OF METHIMAZOLE AND CARBIMAZOLE Dose administered/ Drug mg [35S]Methimazole .. 10 [35S]Methimazole . . 10 Methimazole . . .. 10 Methimazole . . * . 10 Methimazole . . .. 60 Carbimazole . . .. 60 Methimazole Method of assay half-life/h f S.E.M.* Reference Total plasma radioactivity 13.0 (n = 1) 2 Total plasma radioactivity 7.9 4: 0.4 3 Spectrophotometric 6.9 f 0.6 (n = 4) 4 Gas - liquid chromatography 3.7t (m = 1) 5 HPLC 3.2 f 0.29 (n = 5) 6 HPLC 2.96 j, 0.34 (n = 5) 6 * S.E.M. = standard error of the mean; n = number of patients in each study. 7 Calculated from information given in the paper.October, 1981 AND DRUG METABOLTTES BY HPLC 1073 Methimazole 3-Methyl-2-thiohydantoin I-Methyl-2-thiohydantoin Fig. 3. Methimazole and 1- and 3-methyl-2-thiohydantoin. In the course of our development of an HPLC method’ to measure quantitatively methi- mazole in biological fluids obtained from patients receiving either methimazole or carbi- mazole,g a sulphur-containing metabolite was identified and characterised with the aid of HPLC and spectral techniques as 3-methyl-2-thiohydantoin (Fig.3).8 Both alumina (10 pm) and silica (10 pm) columns (10 x 0.46 cm i.d.) were examined for their suitability for the quantitative measurement of methimazole and 3-methyl-2-thiohydantoin in plasma using an internal standard at the extraction stage. The possibility that 1-methyl-2-thio- hydantoin (Fig. 3) might be a metabolite was also considered. Although methimazole and the 1- and 3-methyl isomers of 2-thiohydantoin were satisfactorily resolved on a silica column (Table 11) with a chloroform-based mobile phase, exogenous purines present in biological fluids, which were co-extracted, interfered with the HPLC separation.With basic alumina, 1-methyl-2-thiohydantoin has a high capacity factor, whereas methimazole, 3-methyl-2- thiohydantoin and the internal standard fi-toluamide are satisfactorily separated from each other (Table 11) in a reasonable time (10 min) with no interference from co-extracted endo- genous compounds. This different behaviour of the methyl isomers of 2-thiohydantoin on silica and alumina could partly be explained by the fact that their pK, values differ by 2 units. TABLE I1 CAPACITY FACTORS FOR METHIMAZOLE, 3-METHYL-2-THIOHYDANTOIN AND 1-METHYL-2-THIOHYDANTOIN ON SILICA AND ALUMINA HPLC COLUMNS Capacity factor (k’) f 1 Compound PKa Silica* Aluminat$ Methimazole .. .. .. . . 11.5 2.9 3.2 3-Methyl-2-thiohydantoin . . . . 10.6 1.4 1.8 * 10-pm silica with 0.75% methanol in chloroform as mobile phase. t 10-pm alumina with 1.0% methanol in chloroform as mobile phase. f k’ for toluamide (internal standard) = 7.4. 1-Methyl-2-thiohydantoin . . . . 8.6 1.9 20 For the reversed-phase systems, the capacity factor, a property of the column, is related to the partition coefficient (log P), a measure of the lipophilicity of the compound to be measured. The quantitation of chlordiazepoxide and its metabolites (Fig. 1) using an octadecylsilane (ODS) support illustrates the applicability of reversed-phase HPLC systems to the separation of a drug and its metabolite^.^ The separation, by virtue of their different lipophilicities, of chlordiazepoxide and metabolites at a particular pH, and their quantitation, without prior derivatisation or hydrolysis to their corresponding benzophenones, illustrate the advantage of reversed-phase HPLC in drug metabolism studies.Operation at room temperature is beneficial for analysing thermally labile compounds such as oxazepam,lo another 1,4-benzodiazepine. It is in the measurement of polar metabolites, and drugs and metabolites of high relative molecular mass that HPLC has its widest application. Paracetamol is an example of a drug that is eliminated from the body via conjugation reactions (Fig. Z), predominantly as its highly polar glucuronide and sulphate conjugates. Using silanised ODs-silica and octyl- amine in the mobile phase to act as a basic counter ion to form ion pairs with the weakly acidic paracetamol and strongly acidic metabolites, satisfactory separation of paracetamol from its glucuronide, its sulphate and its cysteine and mercapturic acid metabolites has been1074 SKELLERN: ASPECTS OF THE ANALYSIS OF DRUGS Analyst, VOl.106 obtained.ll With this method it is possible to measure paracetamol and its metabolites quantitatively in urine, by injecting urine samples directly on to the HPLC column.1~ Not only can HPLC be used for the quantitation of drugs and metabolites from in vivo studies, but we have also been examining its applicability in enzyme kinetic studies, in particular the in Vitro glucuronidation of paracetamol (Fig. 4) and p-nitrophenol by liver UDP glucuronyltransferase.The formation of the glucuronides of these substrates in incubations containing enzyme, substrate md uridine diphosphoglucuronic acid (UDPGA) was monitored by injecting samples of incubation mixture directly on to a column packed with ODSsilica (5 pm), eluted with a mobile phase containing an ion-pairing reagent, tetra- butylammonium phosphate in aqueous methanol.lS 11 - 5 10 (a' I' 0 ;I L 0 5 10 Time/min Fig. 4. Chromatograms of samples from incuba- tion medium for studying the in vitro rate of glucu- ronidation of paracetamol. Sample taken 2 min (a) and 30min (b) from the start of incubation. Con- ditions: column (10 x 0.46cm) packed with 6-pm Spherisorb C6; mobile phase, 6 lll~t tetra- butylammonium phos- phate in methanol - water mlmin-1; and detection, 264nm.Peaks: P, para- cetamol glucuronide. (12 + 88); flow-rate, 1.2 ~etam01; a d PG, para- The diastereoisomeric glucuronide metabolites of the 1 ,&benzodiazepine oxazepam have been quantitatively measured by HPLC1* after their separation by preparative ion-exchange chromatography. The rates of enzyme hydrolysis of the individual diastereoisomers were studied by HPLC, subsequent to their purification. The anthracycline cytotoxic antibiotic adriamycin, which has a high relative molecular mass (5434, its major metabolite adriamycinol and its aglycone metabolites have been separated and quantitatively measured with the aid of an internal standard, using an ODS- silica column and a fluorescence detector.l5October, 1981 AND DRUG METABOLITES BY HPLC 1076 In conclusion, HPLC has been clearly shown to be a versatile technique effective in either in vivo or in in vitro drug metabolism studies. In particular it is in the measurement of polar metabolites, thermally labile drugs and metabolites and drugs and metabolites of high relative molecular mass that HPLC has its widest application.1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References Testa, B., and Jenner, P., “Drug Metabolism: Chemical and Biochemical Aspects,” Marcel Dekker, Alexander, W. D., Evans, V., MacAulay, A., Gallagher, T. F., and Londono, J., BY. Med. J., 1969, Crooks, J.. Hedley, A. J.. MacNee, C., and Stevenson, I. H., Br. J. Pharmacol., 1973, 49, 156P. Vesell, E. S., Shapiro, J. R., Passananti, G. T., Jorgensen, H., and Shively, C. A., Clin. Pharmautl. Bending, M. R., and Stevenson, D., J . Chromatogr., 1978, 154, 267. Skellern, G. G., Knight, B. I., Low, L. C. K., Alexander, W. D., McLarty, D. G., and Kalk, W. J., Skellern, G. G., Knight, B. I., and Stenlake, J. B., J . Chromatogr., 1976, 124, 405. Skellern, G. G., Knight, B. I., Luman, F. M., Stenlake, J. B., McLarty, D. G., and Hooper, M. J., Skellern, G. G., Meier, J., Knight, B. I., and Whiting, B., BY. J . Clin. Pharmautl., 1978, 5, 483. Kabra, P. M., Stevens, G. L., and Morton, L. J., J . Chromatogr., 1978, 150, 355. Knox, J. H., and Jurand, J., J. Chromatogr.. 1978, 149, 297. Howie, D., Adriaensens, P. I., and Prescott, L. F., J . Pharm. Pharmacol., 1977, 29, 235. Knight, B. I., and Skellern, G. G., J . Chromatogr., 1980, 192, 244. Rueluis, H. W., Tio, C. O., Knowles, J. A., McHugh, S. L., Schillings, R. T., and Sisenwine, S. F., Pierce, R. N., and Jatlow, P. I., J. Chromatogr., 1979, 164, 471. New York, 1976, Chapter 1, p. 3. 2, 290. Ther., 1975, 17, 48. BY. J. Clin. Pharmawl., 1980, 9, 137. Xenobiotica, 1977, 7, 247. Drug Metab. Dispos., 1979, 7 , 40.

ISSN:0003-2654    

DOI:10.1039/AN9810601071

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

1076 Analyst, October, 1981, Vo1. 106, pp. 1076-1081 Determination of Mercury Vapour in Air Using a Passive Gold Wire Sampler J. E. Scott and J. M. Ottaway Analytical Laboratory, Philips Hamilton, Wellhall Road, Hamilton, Scotland, ML3 9BZ Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow, G1 I X L The determination of atomic mercury vapour in air in the range 10-120 pg m-3 is described. Mercury vapour was collected by exposing a 1-cm length of gold wire to the air for 5min, then thermally desorbed from the wire for measurement by atomic-fluorescence spectrometry. The results obtained showed good agreement with those obtained from an acid permanganate wet sampling system. Keywords : Mercury determination ; air analysis ; gold wire ; atomic-fluorescence spectrometry There are several methods available for the sampling and determination of mercury vapour in air in the region of the Threshold Limit Value (TLV) (currently1 50 pg m-3) and below.The mercury may be collected by drawing a measured volume of air through an absorbing solution such as iodide - iodine2 or potassium permanganate - sulphuric acid solution,3 or through solid adsorbents such as iodised charcoal,* gold-wire clippings5 or silver foil.6 These techniques require cumbersome apparatus incorporating a flow meter, vacuum pump and valves. Several continuous monitors based mainly on ultraviolet atomic-absorption spectrometry are available commercially, but these instruments, although sensitive, are bulky and expen- sive. Passive devices, in which a gold foil’ or vacuum-deposited gold layer on a glass substrate8 was exposed to the air for a fixed length of time, have been described recently. These devices are passive in that they depend upon diffusion of mercury from the air to the adsorbing surface.It is assumed that the mass of mercury deposited on to the gold surface in a given time is proportional to the concentration of mercury vapour in the air. No attempt is made to remove all of the mercury from a measured volume of air. In the work described in this paper a 1-cm length of gold wire was exposed to the draught- free atmospheres to be measured for 5min, then removed to the laboratory where the adsorbed mercury was thermally desorbed and determined by atomic-fluorescence spectro- metry (AFS).g The air was simultaneously sampled with a potassium permanganate - sulphuric acid absorption system followed by analysis by AFS, and the results of the two techniques were compared.Experimental Reagents water. All materials were of analytical-reagent grade, and all solutions were prepared using distilled Potassium permanganate absorption solution, 0.01 M in 5% V/V sulphuric acid. Hydroxylammonium chloride solution, 20% VlV. Tin(I1) chloride solution, 10% V/V in hydrochloric acid. in 80 ml of “concentrated” 35% hydrochloric acid. metallic tin and warm the solution until it clears. Dissolve 20 g of tin(I1) chloride If a precipitate appears add a trace of Dilute to 200 ml with distilled water. SuZphurzc acid, 8% V/V. Reduction solution. Mix 20 ml of the tin(I1) chloride solution with 60 ml of 8% sulphuric A continuous stream of argon should be bubbled through the reduction solution to acid.prevent atmospheric oxidation.SCOTT AND OTTAWAY 1077 Apparatus Fluorescence signals were measured using a Baird Atomic, Model A 3000, spectrometer with all lenses removed. Signals were recorded on a Philips, Model 8000, flat-bed chart recorder. Gold Wire Sampler A 2-cm length of gold wire, diameter 0.125 mm, purity 99.99% (Goodfellow Metals Ltd.), was securely fixed vertically into a slotted rubber base in such a way that 1 crn of the wire remained exposed. A glass flask into which the base fitted protected the gold wire in transit to and from the sampling area. The exposure commenced when the wire was removed from the flask, and ended when replaced into the flask.After exposure to the draught-free air for 5min the sampler was taken to the laboratory, where the gold wire was removed from the rubber base and thoroughly rinsed in two successive baths of chloroform. The cleaned gold wire was placed into a quartz tube (150mm x 6mm i.d.) through which flowed a regulated stream of argon at 1 1 min-l. The quartz tube was heated using a fan-tail burner, and the desorbed mercury vapour swept into the cold vapour AFS systems (Fig. 1). The chloroform cleaning stage was found to be necessary to remove trace amounts of oil that had deposited from the air on to the gold wire during sampling. (This contaminant was identified as vacuum pump oil (from a nearby mercury vacuum distillation apparatus) by injecting the effluent from the heated quartz tube into a gas chromatograph.Further con- firmation was obtained by injecting small amounts of the oil into the quartz tube and observing the massive fluorescence signals obtained when the quartz tube was heated.) Stainless-steel forceps were used to handle the gold wire to prevent contamination from skin oils, etc., which were found to cause positive errors during the subsequent AFS measurement, The AFS system used was that described by Thompson and Godden.9 t II Argon Fig. 1. The desorption and AFS system: A, lamp housing; B, spectrometer; C, quartz tube; D, fan-tail burner; E, mercury generation cell ; F, fluorescence measurement head ; G, flow control rotameters; H, argon shield gas; and J, lamp cooling argon.Potassium Permanganate - Sulphuric Acid Sampler3910 Five litres of air were drawn at 1 1 min-l through two sintered-glass absorbers11 connected in series, each containing 20 ml of permanganate absorption solution. The second absorber served as a guard to indicate that the first absorber had not been overloaded. After sampling, the inlet tube of each absorber was rinsed with 2 ml of water, 3 rnl of hydroxylammonium chloride solution were added, the liquid level in the absorber was made up to a 25-ml calibration line if necessary and the absorber contents were mixed until a clear solution was obtained.1078 SCOTT AND OTTAWAY: DETERMINATION OF MERCURY VAPOUR Analyst, rd. I06 A 1.0-ml volume of the analyte solution was mixed with 1 ml of the reduction solution in the mercury generation cell and the mercury vapour liberated was swept into the cold vapour AFS system.Calibration was achieved by pipetting 1 .O-ml aliquots of appropriate standard mercury solutions (prepared in permanganate absorption solution and reduced with hydroxyl- ammonium chloride solution) into the mercury generation cell and reducing with tin( 11) chloride reduction solution as above. Results and Discussion Comparison of Gold Wire Sampler and Permanganate Absorption Solution In order to determine the precision of the permanganate system, 20 samples were collected using in each instance five independent identical permanganate samplers operating simul- taneously. During these tests the concentration of mercury in air ranged from 10 to 15Opgm-3 and the results showed relative standard deviations ranging from 18% at 10 pg m-3 to 9% at 150 pg m-3 (Fig.2). .3 I , , , I CI 50 100 150 c[l K - @ O Mean concentration from 5 absorberdwg m-3 Fig. 2. Variation of precision of the permanganate absorption system with the concentration of mercury in the air. Comparison of the results obtained from both sampling methods applied simultaneously over a series of 100 measurements made in draught-free room atmospheres showed a recti- linear relationship between the signals obtained from the gold wire system and the mass of mercury collected in the permanganate solution (and thus the concentration of mercury in the air) over the range 04.4 pg (0-80 pg m-3), with a useful curve up to 0.6 pg (120 pg m-3). Linear regression analysis of the results from 0 to 80 pg mW3 gave a correlation coefficient of 0.965 at the 99.9% level, and slope of 0.693 chart divisions per pg m-3, with an intercept on the y-axis (gold wire signal) of 2.17 chart divisions.This intercept, equivalent to 3.1 pg m-3, was considered insignificant. The same gold wire sampler was used for all of the tests, and showed no evidence of deterioration after 100 cycles. Delays of up to 24 h between sampling and thermal desorption produced no discernible difference in the signals obtained. These results clearly indicated that the gold wire sampler could provide a simple means of determining the concentration of mercury in air, provided that a suitable calibration system could be devised. Determination of Mass of Mercury Deposited on the Gold Wire It was clear from the shapes of the signal traces that the rate of thermal desorption of mercury from the gold wire was different from the rate of gas partition desorption of mercury from the solution. Thus, although there was clear proportionality between the signals from gold wire desorption and the signals from solution desorption for given concentrations ofOctober, 1981 IN AIR USING A PASSIVE GOLD WIRE SAMPLER 1079 mercury in air, it was not possible to determine the relationship between the concentration of mercury in the air to which the gold wire had been exposed and the mass of mercury deposited on the gold wire from this comparison.The solution sampling system was readily calibrated using standard mercury solutions, but calibration of the gold wire system required the additional steps of dissolving the adsorbed mercury from the gold wire in 50% nitric acid solution, followed by solution analysis with calibration against standard mercury solutions. A series of tests was carried out in which the gold wire was exposed to draught-free atmospheres of various known mercury concentrations (as determined by the permanganate absorption method), then immersed for 15 min in 1.5 ml of 5070 nitric acid solution contained in a narrow-form centrifuge tube.A 1.0-ml volume of the nitric acid solution was subse- quently taken for analysis by the AFS system, using appropriate calibration standards prepared in 50% nitric acid. This method of desorption of mercury from the gold wire was found to be effective provided that it was carried out within Wmin of exposure.If the gold wire was allowed to stand for a longer period before nitric acid desorption, low recoveries of mercury were obtained, presumably owing to diffusion of mercury into the gold wire. After each nitric acid desorption, the gold wire was rinsed with distilled water and subjected to the thermal desorption process in the quartz tube. When nitric acid immersion had been delayed for longer than 30 min, residual mercury was thermally desorbed from the wire. No signals were obtained from the thermal desorption process when the nitric acid immersion had been carried out within 30 min. A rectilinear relationship between mass of mercury in the nitric acid solution and con- centration of mercury in the air during exposure was observed in the range WOpgm-3, with a useful curve up to 120 ,ug m-3 (Fig.3). No deterioration of the gold wire sampler was observed during these tests, provided that the thermal desorption stage was carried out after each nitric acid stage. + ++++ ~ ++ + + + I 50 100 150 Concentration of mercury in air/pg md3 Fig. 3. Relationship between the mass of mercury deposited on the gold wire and the concentration of mercury in air. Preparation of Standard Atmospheres Aliquots of saturated mercury vapour in air at known temperatures were injected through a rubber septum into a sealed 5-1 glass flask. The flask was shaken vigorously for 5 min to ensure homogeneity, and the gold wire sampler was introduced into the flask via a ground- glass connection.After exposure for 5 min the wire was removed from the flask and the mercury desorbed by the thermal technique. The exposure was repeated at the same concentration (using a freshly prepared standard atmosphere) and the mercury desorbed by the nitric acid method. The relationship between mass of mercury in the nitric acid solution and signal obtained from the thermal desorption technique was rectilinear up to 15 ng of mercury (corresponding to a concentration of mercury in air of 120 pg m-7, although the relationship between thermal desorption signal and concentration of mercury in air was rectilinear only from 0 to 80 pg m--3 and curved from 80 to 120 pg m-3, as before.1080 SCOTT AND OTTAWAY: DETERMINATION OF MERCURY VAPOUR Analyst, VoZ. 106 Comparison of signals from both thermal and nitric acid desorption techniques, obtained from analysis of standard atmospheres of calculated concentrations, with those obtained from “real” atmospheres standardised by the permanganate absorption technique, showed good agreement in the range 0-120 pg m-3 of mercury.In order to determine the precision of the gold wire system, ten standard atmospheres with mercury concentrations in the range 0-120 pg m-3 were prepared and the gold wire was exposed to each ten times, then thermally desorbed as previously described The gold wire was thus exposed to each standard atmosphere (freshly prepared) ten times. The results showed a relative standard deviation of 19% at 10 pg m-3 of mercury, which decreased to about 6% between 80 and 1 2 0 ~ g m - ~ (Fig.4). The detection limit, corre- sponding to twice the standard deviation at the 10 pg mW3 level, was 3.6 pg m-3. It should be noted, however, that the results “spread” at the 120 pg m-3 level showed an artificial bias as the mass of mercury deposited approached saturation. 50 100 Concentration of mercury in air/pg m-3 (standard atmospheres) Fig. 4. Variation of the precision of the gold wire sampling system with the concentration of mercury in air. Time of Exposure The gold wire sampler was exposed to standard atmospheres, prepared in the 5-1 flask, in the range 0-150 pg m-3 for varying lengths of time, then subjected to the thermal desorption procedure. In each instance the relationship between mass of mercury adsorbed by the gold wire and time of exposure was rectilinear in the range 0-12 ng of mercury (Fig.5). Conclusion It has been shown that a l-cm length of 0.125 mm diameter gold wire, exposed for 5 min to mercury vapour in air in the range 0--12Opgn1-~, will adsorb a mass of mercury pro- portional to the concentration of mercury in the air. This principle forms the basis of a simple technique for the determination of mercury vapour in air. The method is rapid and sensitive, although not particularly precise. It has further been shown that the l-cm length of gold wire, when exposed to a fixed concentration of mercury in air, will adsorb mercury at a constant rate until a mass of 12 ng has been adsorbed, irrespective of the time of exposure up to at least 35 min. Further work is proposed in which a continuously moving length of gold wire will be progressively exposed to the air over an 8-h period.Subsequent progressive desorption of the mercury from this wire should thus provide a record of the variations of the mercury in air concentration over the 8-h period. It is hoped that this system will provide the basis for a personal mercury exposure monitor. In this work mercury concentrations were determined by AFS because of the simplicity of this technique and its freedom from the use of optical cells, the windows of which could “mist up” during the thermal desorption process. It is expected, however, that atomic- absorption spectrometry would be equally effective with few modifications.October, 1981 entering the photomultiplier. that described by Hutton and Preston,12 would be suitable for this purpose.IN AIR USING A PASSIVE GOLD WIRE SAMPLER 1081 The principal function of the spectrometer used in this work was to minimise stray light It is expected that a simpler, non-dispersive system, such as 0 5 10 15 20 25 30 35 Time of exposure/min Fig. 5. Variation of the mass of mercury adsorbed by the gold wire with time of exposure for various standard mercury atmospheres. Concentration of mercury in standard atmos- pheres: A, 160; B, 80; C, 40; D, 20; and E, 10 pg m-3. J.S. thanks his employers, Philips Hamilton, for use of facilities, and for permission to publish this work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References American Conference of Governmental Industrial Hygienists, “Threshold Limit Values for Chemical Substances and Physical Agents in the Workroom Environment,” American Conference of Governmental Industrial Hygienists, Cincinnati, 1978. Barnes, H., J . Ind. Hyg. Toxicol., 1946, 28, 256. Drew, R. G., and King, E., Analyst, 1957, 82, 461. Sergeant, G. A., Dixon, B. E., and Lidzey, R. G., Analyst, 1957, 82, 27. Head, P. C., and Nicholson, R. A., Analyst, 1973, 98, 53. Kalb, G. W., At. Absorpt. Newsl., 1970, 9, 84. McCammon, C. S., and Woodfin, J. W., Am. Ind. Hyg. Assoc. J . , 1977, 38, 378. Main, C., and Lenihan, J. M. A., J . Phys. E, Sci. Instrum., 1978, 11, 1123. Thompson, K. C., and Godden, R. G., Analyst, 1975, 100, 544. Hanson, N. W., Reilly, D. A., and S;,agg, H. E., Editors, “The Determination of Toxic Substances Analytical Chemists Committee, ICI, Heffers Printers Ltd., Gage, J. C., J . Sci. Instrum., 1952, 29, 409. Hutton, R. C., and Preston, B., Analyst, 1980, 105, 981. in Air, A Manual of ICI Practice, Cambridge, 1965, p. 162. Received February 23rd, 1981 Accepted April 30th, 1981

ISSN:0003-2654    

DOI:10.1039/AN9810601076

出版商:RSC    

年代:1981

数据来源: RSC