摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYEDITORIAL ADVISORY BOARD"Chairman: J. M. Ottaway (Glasgow)R . B e I c h e r (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exefer)L. R. P. Butler (South Africa)E. A. M. F. Dahmen (The Netherlands)A. C. Docherty (Billingham)D. Dyrssen (Sweden)W. T. Elwell (Birmingham)J. Hoste (Belgium)H. M. N. H. Irving (Leeds)M. T. Kelley (U.S.A.)W. Kemula (Poland)"J. H. Knox (Edinburgh)G. W. C. Milner (HarweN)G. H. Morrison (U.S.A.)"H. J. Cluley (Wembley)H. W. Nurnberg (West Germany)E. Pungor (Hungary)D. I . Rees (London)*R. Sawyer (London)P. H. Scholes (Sheffield)"W. H. C. Shaw (Greenford)S. Siggia (U.S.A.)"D. Simpson (Thorpe-le-Soken)A. A. Smales, O.B.E.(Harwell)*A. Townshend (Birmingham)A. Walsh (Australia)T. S. West (Aberdeen)*J. Whitehead (Stockton-on- Tees)A. L. Wilson (Medmenham)P. Zuman (U.S.A.)"G. E. Penketh (Billingham)*Members of the Board serving on The Analyst Publications CommitteeREGIONAL ADVISORY EDIT0 RSDr. J. Aggett, Department of Chemistry, University of ,4uckland, Private Bag, Auckland, NEW ZEALAN D.Professor G. Ghersini, Laboratori CISE, Casella Postale 3986,201 00 Milano, ITALY.Professor L. Gierst, Universit6 Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50,Professor R. Herrmann, Abteilung fur Med. Physik., 63 Giessen, Schlangenzahl 29, W. GERMANY.Professor W. A. E. McBryde, Faculty of Science, University of Waterloo, Waterloo, Ontario, CANADA.Dr.W. Wayne Meinke, KMS Fusion Inc., 3941 Research Park Drive, P.O. Box 1567, Ann Arbor,Dr. 1. Rubegka, Geological Survey of Czechoslovakia, Kostelni 26, Praha 7, CZECHOSLOVAKIA.Dr. J. R6Fi6ka. Chemistry Department A, Technical llniversity of Denmark, 2800 Lyngby, DENMARK.Professor K. Saito, Department of Chemistry, Tohoku University, Sendai, JAPAN.Dr. A. Strasheim. National Physical Research Laboratory, P.O. Box 395, Pretoria, SOUTH AFRICA.Bruxelles, BELGIUM.Mich. 481 06, U.S.A.Published by The Chemical SocietyEditorial: The Director of Publications, The Chemical Society, Burlington House,London, W1 V OBN. Telephone 01 -734 9864. Telex No. 268001Advertisements: Advertisement Department, The C,hernical Society, Burlington House, Piccadilly,London, W1 V OBN. Telephone 01 -734 9864Subscriptions (non-members) : The Chemical Society, Distribution Centre, Blackhorse Road,Letchworth, Hert!;., SG6 1 HNVolume 103 No 1231 October 1978@ The Chemical Society 197

ISSN:0003-2654    

DOI:10.1039/AN97803FX037

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

ANALAO 103 (1 231 ) 1 009-1 088 (1 978)ISSN 0003-2654October 1978THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTS10091021Methods f o r the Quantitative Determination of Asbestos and Quartz in BulkSamples Using X-ray Diffraction-M. TaylorApplication o f the Faraday Effect t o the Trace Determination of Cadmium byAtomic Spectroscopy with an Electrothermal Atomiser-K. Kitagawa, T.Shigeyasu and T. TakeuchiFlotation of Sub-microgram Amounts of Arsenic Coprecipitated with Iron(ll1)Hydroxide from Natural Waters and Determination o f Arsenic by Atomic-absorption Spectrophotometry Following Hydride Generation-SusumuN a kas h ima1037 Extraction - Spectropkotometric Determination of Tin in Lead and Lead-based Alloys with 5.7-Dichloroquinolin-8-ol-A. Sanz-Medel and A.M.Gutidrrez Carreras1046 Method f o r the Determination of Methanol in Binary Methanol - WaterMixtures by Use of lon-selective Electrodes-G. J. Kakabadse, H. AbdulahedMaleila, M. N. Khayat, G. Tassopoulos and A. VahdatiPotentiality o f Seaweed as a Resource: Analysis o f the Pyrolysis Products o fFucus serratus-Phillip J. Morgan and Keith SmithDetermination o f Theophylline in Plasma : Comparison o f High-performanceLiquid Chromatography and an Enzyme Multiplied lmmunoassay Tech-nique-M. L. Eppel, J. S. Oliver, Hamilton Smith, A. Mackay and L. E. RamsayOptical Emission Spectrometry w i t h an Inductively Coupled RadiofrequencyArgon Plasma Source and Sample Introduction with a Graphite RodElectrothermal Vaporisation Device.Part 1. Instrumental Assembly andPerformance Characteristics-A. M. Gunn, D. L. Millard and G. F. Kirkbright1031105310611066SHORT PAPERS1074 Determination of Antimony by Stibine Generation and Atomic-absorptionSpectrophotometry Using a Flame-heated Silica Furnace-D. L. Collett,D. E. Fleming and G. A. TaylorQuantitative Determination of Steroids in Semi-solid Pharmaceutical Prepara-tions by Using High-performance Liquid Chromatography-Monir Aminand Peter W. SchneiderDetermination by Gas Chromatography - Single-ion Monitoring Mass Spectro-metry o f Phthalate Contaminants in Intravenous Solutions Stored in PVCBags-G. A. Ulsaker and R. M. Hoem1083 Thin-layer Chromatographic Behaviour o f Barbiturates Under Various Con-ditions-R. Abu-Eittah, A.Osman and A. El-Behare1088 Book Reviews10761080Summaries o f Papers in this Issue-Pages iv, v, viii, i xPrinted by Heffers Printers Ltd Cambridge EnglandEntered as Second Class at New York, USA, Post OfficAnnual Reports on AnalyticalAtomic SpectroscopyVOLUME 6,1976This comprehensive and critical report of developments in analytical atomicspectroscopy has been compiled from over 1650 reports received fromworld-wide correspondents who are internationally recognised authorities inthe field and who constitute the Editorial Board. In addition to surveyingdevelopments throughout the world published in national or internationaljournals, a particular aim has been to include less widely accessible reportsfrom local, national and international symposia and conferences concernedwith atomic spectroscopy.Paperbound 282pp 8;"x 6" f 18 (CS Members f 13.50)(Still available: Vols.2-5 covering 1972 to 1975)Obtainable from : The Chemical Society, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HNNOTICE TO SUBSCRIBERS(other than Members of the Society)Subscriptions for The Analyst, Analytical Abstracts and Proceedings shouldbe sent to:The Chemical Society, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HNRates for 1978The Analyst, Analytical Abstracts and Proceedings (including indexes):. . . . f99.00. . . . f105.00(a) The Analyst, Analytical Abstracts and Proceedings . . . .Proceedings . . . . . . . . . . . . . .(b) The Analyst, Analytical Abstracts printed on one side of the paper, andThe Analyst and Analytical Abstracts without Proceedings (including indexes):(c) The Analyst, and Analytical Abstracts . . .. .. . . . . . . f87.00(d) The Analyst, and AnalyticalAbstracts printed on one side of the paper . . f93.00(Subscriptions are NOT accepted for The Analyst and/or for Proceedings alone)Analytical Abstracts only (two volumes per year, including indexes) :(e) Analytical Abstracts . . * . * . . . .. .. .. . . f67.00. . . . €73.00 (f) Analytical Abstracts printed on one side of the paper . . .

ISSN:0003-2654    

DOI:10.1039/AN97803BX039

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

... Octobcr, 1978 THE ANALYST 111TECHNIQUES OF CHEMISTRYVol. 14: Thin-Layer Chromatography 2nd Ed.Series Editors: A. Weissberger and E.S. Perry;Volume Author: J.G. Kirschner.l An up-dated, comprehensive treatise on thin-layer chromatography, covering techniques and , applications to many types of organic and inorganic compounds. Thin-layer chromatography is valuable1 in separating solids or liquids when only small amounts are available. Contains over 6,000 citationscovering all fields of TLC.~ 0471 93264 7 approx. 900 pages In Press approx. $72.60/f38.80~ TREATISE ON ANALYTICAL CHEMISTRY 2nd Ed.Part 1 : Theory and Practice Vol. 1l~1by I . M. Kolthoff, University of Minnesota,and P.J. Elving, University of Michigan.A complete and definitive information source for ail analytical chemists, designed t o stimulatefundamental research i n pure and applied analytical chemistry.Coverage includes aspects of classicaland modern analytical chemistry, and the scientificand instrumental fundamentals of analytical methods.0471 03438 X approx. 91 2 pages In Press approx. $76.50/f40.85DIOXIN: Toxicological and Chemical Aspectsedited by F. Cattabeni, University of Milan;A. Cavallaro, Laboratory of Hygiene and Profilaxis, Milan,and G. Galli, University of Milan.On July 10, 1976, a mixture of chemicals containing 2,3,7,8-Techlorodibenzo-p-dioxin (TCDD) wasreleased from an industrial chemical plant over a wide and densely populated area at Seveso, near Milan,Italy. The impact of the subsequent chemical epidemic and scientific insights derived from it w i l l have along lasting effect on future developments in toxicology and biomedicine.The event has posed manydramatic questions on health and environmental risks to the world population for government authoritiesand the scientific community i n general. This book is an important and extremely necessarycontributionto current knowledge of the chemistry, analysis, toxicology and decontamination of TCDD.(Monographs of the Giovanni L orenzini Foundation Series)047026361 X 236 pages September 1978 $26.50/f 14.50Published by Spectrum Publications lnc., and distributed by John Wiley & Sons Ltd.CHEMICAL ANALYSIS BY MICROWAVE ROTATIONALS P E CTR OSCOPYby R . Varrna. Argonne National Laboratory, Argonne, Illinois,and L.W.Hrubesh, Lawrence Livermore Laboratory, Livermore, California.Represents the current state of the art and provides a detailed, comprehensive tteatment of basic theoryand a complete coverage of techniques. The characteristics for qcalitative and quantitative chemicalanalysis are presented with examples of actual applications; numerous other applications, such asisotope radiation measurements and engine exhaust analysis, are discussed with projections for futureuse. (Chemical Analysis Series Vol. 52)0471 03916 0 approx. 192 pages In Press approx. $24.75/f 13.75Available from all good booksellers or from Wiley. If you wish to use American Express, Diners Club,Barclaycard or Access, please quote your card and number1v SUMMARIES OF PAPERS I N THIS ISSUE October, 1978Summaries of Papers in this IssueMethods for the Quantitative Determination of Asbestos andQuartz in Bulk Samples Using X-ray DiffractionProcedures are described for the quantitative determination of the asbestosand cc-quartz contents of bulk samples by use of X-ray powder diffracto-metry.The method gives satisfactory results for several different types ofasbestos and for mixtures of two or more different types. Problems withsample grinding and preferred-orientation effects have been largely overcome.An effective procedure has been developed for grinding samples to a suitableparticle size for accurate quantitative work. This procedure works equallywell for all the types of asbestos studied and the sample is intimately mixedwith the internal standard, nickel(I1) oxide, at the same time.A samplepress has been designed that enables the same pressure to be applied to eachsample when sample holders are filled for the diffractometer, giving the samedegree of preferred orientation each time. Calibration lines have beencalculated for chrysotile, amosite, crocidolite and anthophyllite, and resultsare given for mixtures containing two or more types of asbestos as well asother commonly occurring minerals.Similar techniques are used to grind samples containing quartz and to mixthem with internal standard. Work on both synthetic and real samples isdescribed and results are compared with those obtained by use of an infraredspectroscopic method.Keywords: Asbestos determination ; quartz determinution ; X-ray diflractionM.TAYLORHealth & Safety Executive, The Occupational Medicine and Hygiene Laboratories,403-405 Edgware Road, London, NW2 6LN.Analyst, 1978, 103, 1009-1020.Application of the Faraday Effect to the Trace Determination ofCadmium by Atomic Spectroscopy with an Electrothermal AtomiserAs a technique for the trace determination of elements, the Faraday effecthas been applied to atomic spectroscopy with the use of electrothermalatomisers. The atomiser was located between two crossed plane-poIarisers(called the polariser and the analysei-). An electromagnet magnetised theatomised sample and, as aresult of the Faraday effect, the source radiationcould, in the presence of the atomic vapour, pass through the optical system.A simple theoretical treatment was developed to explain the dependenceof the transmitted intensity on the magnetic field strength and the sourceintensity and the ability to eliminate the effects of background scattering.Measurements on cadmium were carried out a t a wavelength of 228.8 nm,dispensing into the atomiser 5-pl samples, which, after drying, were atomiseda t 1000-1 800 "C.The calibration graphs demonstrated a square-lawdependence. At high concentrations the calibration graphs were correctedfor atomic absorption. The technique gave a detection limit 0fL5 x 10-13 gfor cadmium.Keywords : Magneto-ofitical rotatio+z ; atomic spectroscopy ; electrothevmalatomisation ; Faraday effect ; cadmium determinationK.KITAGAWA, T. SHIGEYASU and T. TAKEUCHIDepartment of Synthetic Chemistry, Faculty of Engineering, Nagoya University,Furo-cho, Chikusa-ku, Nagoya, Japan.Analyst, 1978, 103, 1021-1030October, 1978 SUMMARIES OF PAPERS IN THIS ISSUEFlotation of Sub-microgram Amounts of Arsenic Coprecipitatedwith Iron(II1) Hydroxide from Natural Waters and Determinationof Arsenic by Atomic-absorption Spectrophotometry FollowingHydride GenerationA method is described for the flotation and determination of sub-microgramlevels of arsenic in water. A sub-microgram amount of arsenic(II1, V) in a500-ml sample of water is coprecipitated with iron(II1) hydroxide a t pH 8-9.The precipitate is floated with the aid of sodium oleate and small air bubbles,then separated and dissolved in 5 M hydrochloric acid, and finally the arseniccontent is determined by generation of arsine using sodium tetrahydro-borate(II1) as a reductant and atomic-absorption spectrophotometrywith a long absorption cell (60 x 1.2 cm i.d.).This separation techniquehas been successfully applied to the determination of sub-microgram amountsof arsenic(II1, V) in natural waters.Keywords : Arsenic determination ; flotation ; natural waters ; atomic-absorption spectrophotometry ; hydride generationSUSUMU NAKASHIMAInstitute for Agricultural and Biological Sciences, Okayama University, Kurashiki-shi$10, Japan.Analyst, 1978, 103, 1031-1036.Extraction - Spectrophotometric Determination of Tin in Lead andLead- based Alloys with 5,7-Dichloroquinolin-8- 01A method is described for the direct spectrophotometric determination ofmicro-amounts of tin, by extraction into a chloroform solution of 5,7-dichloro-quinolin-8-01 from a solution containing sulphuric acid.The influence ofthe different experimental parameters on the formation and extraction ofthe complex were studied and optimum conditions for the determination oftin were established. The precision of the extraction - spectrophotometricprocedure, expressed in terms of relative standard deviation, was 1.4%.It is shown that two different complexes (with A,,,. 403 or 390nm) canbe extracted into the chloroform, depending on the presence or absence ofchloride and on the pH of the solution.The method has been tested on six standard lead-based samples with tincontents ranging from 0.05 to 1%.The average relative error (mean error)of the results lies within the range &l.4y0, which shows that the accuracyis good and that systematic errors are absent.Keywords ; Tin determination ; 5,7-dichloroquinolin-8-01 extraction ; spectro-photometry ; lead-based alloysA. SANZ-MEDEL and A. M. GUTIRRREZ CARRERASDepartamento de Quimica Analitica, Facultad de Ciencias Quimicas, UniversidadComplutense, Ciudad liniversitaria, Madrid-3, Spain.Analyst, 1978, 103, 1037-1045.VMethod for the Determination of Methanol in BinaryMethanol - Water Mixtures by Use of Ion-selective ElectrodesFor a given concentration of indicator ion X (X = F-, C1-, Br-, I-, OH-,S“, Ag+ or H+), the systematic change of cell potential, E, with variation inthe concentration of methanol provides a graphical method for the rapiddetermination of methanol in methanol - water mixtures.Readings obtainedby direct potentiometry show good reproducibility and stability. In 99.0-99.9% m/m methanol, containing M hydrochloric acid, trace amountsof water can be determined accurately owing to a potential “anomaly.”Keywords : Methanol determination ; methanol - water mixtures ; ioiz-G. J. KAKABADSE, H. ABDULAHED MALEILA, M. N. KHAYAT, G.TASSOPOULOS and A. VAHDATIDepartment of Chemistry, University of Manchester Institute of Science andTechnology, P.O. Box 88, Manchester, M60 lQD.Analyst, 1978, 103, 1046-1052.selective electrodes ; trace water determinatioVi THE ANALYST October, 1978 RS solvents forUV and IRAcetone UVand IRAcetonitrile UV and IRBenzene UV and IRCarbonium sulfideUV and IRCarbonium tetrachlorideUV and IRChloroform UV and IRCyclohexane UV and IRN-N-DimethylformamideUV and IRDichloroethane IRDimethylsulfoxide UVDioxane UVEthyl acetate IREthyl alcohol UV950 and abs.Ethyl ether UVn-Heptane UVn-Hexane UVlsoctane UV and IRlsoptopyl alcohol UVMethylene chlorideUV and IRMethyl Alcohol UVn-Pentane UVPotassium bromide IRTetrachloroethylene IRTetra hyd rof uranUV and IRToluene IRTrichloroetilene IRCHEMICALS DIVISIONP.O. Box 3996120159 MilanolVla lmbonatl 24 [Italy)Telex Erba Mi 36314lTel. 6995a MONTEDISON S p A REG TRADEMAR

ISSN:0003-2654    

DOI:10.1039/AN97803FP093

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

viii SUMMARIES OF PAPE:RS IN THIS ISSUEPotentiality of Seaweed as a Resource: Analysis of the PyrolysisProducts of Fucus serratusOctober, 1978As a prelude to the investigation of the potentiality of seaweeds as a futuresource of organic chemicals or fuel, the products of pyrolysis under nitrogenof Fucus serratus (serrated wrack) have been analysed. The pyrolysis pro-duces large amounts of charcoal, water and carbon dioxide with smalleramounts of an oil, pitch, hydrocarbon gases, carbon monoxide, ammonia andcarboxylic acids. The oil proved to be a complex mixture of heterocyclicbases, phenols, aromatic hydrocarbons, nitrogen and oxygen heterocycliccompounds and small amounts of aliphatic compounds. A variety of analyticaltechniques have been employed to analyse the pyrolysis products, with gaschromatography and mass spectrometry being the most widely applicable.Keywords : Seaweed pyrolysis ; Fucus serratus ; gas chronzatography ; massspectrometryPHILLIP J.MORGAN and KEITH SMITHDepartment of Chemistry, University College of Swansea, Singleton Park, Swansea,SA2 8PP.Analyst, 1978, 103, 1053-1060.Determination of Theophylline in Plasma: Comparison ofHigh-performance Liquid Chromatography and an EnzymeMultiplied Immunoassy TechniquePlasma theophylline concentrations have been determined by both high-performance liquid chromatography and the enzyme multiplied immunoassaytechnique (EMIT). Comparison of the results showed a good correlationbetween the techniques. The faster processing time and smaller sample sizemakes EMIT the preferred technique when routine batch assays for theo-phylline are required.Keywords : Theophylline determination ; plasma ; high-performance liquidchromatography ; enzyme multiplied immunoassy techniqueM.L. EPPEL, J. S. OLIVER and HAMILTON SMITHDepartment of Forensic Medicine, Glasgow University, Glasgow, G12 8QQ.A. MACKAYMRC Blood Pressure Unit, Western Infirmary, Glasgow, G11 6NT.and L. E. RAMSAYDepartment of Medicine, Western Infirmary, Glasgow, G11 6NT.Analyst, 1978, 103, 1061-1065.Optical Emission Spectrometry with an Inductively CoupledRadiofrequency Argon Plasma Source and Sample Introductionwith a Graphite Rod Electrothermal Vaporisation DevicePart I. Instrumental Assembly and Performance CharacteristicsA system is described in which a graphite rod electrothermal vaporisationdevice is employed for the introduction of microlitre liquid samples, afterdesolvation, into an inductively coupled argon plasma source for atomisationand excitation for optical emission spectrometry.The analytical per-formance of the system has been studied and detection limits for 16 elementsat the sub-nanogram level are presented.Keywords : Optical emission spectromt:try ; inductively coupled radiofrequencyargon plasma ; graphite rod electrothermal atomisationA. M. GUNN, D. L. MILLARD and G. F. KIRKBRIGHTChemistry Department, Imperial College, London, SW7 2AY.Analyst, 1978, 103, 1066-1073October, 197'8 SUMMARIES OF PAPERS IN THIS ISSUEDetermination of Antimony by Stibine Generation andAtomic-absorption Spectrophotometry Using a Flame-heatedSilica FurnaceShort PaperKeywords : Antimony determination ; atomic-absorption spectrophotonzetry ;hydride generation ; $awe-heated silica furnaceD. L.COLLETT, D. E. FLEMING and G. A. TAYLORGovernment Chemical Laboratories, Food and Industrial Hygiene Division, 30 PlainStreet, Perth, Western Australia 6000.Analyst, 1978, 103, 1074-1075.Quantitative Determination of Steroids in Semi- solidPharmaceutical Preparations by Using High-performanceLiquid chromatographyShort PaperKeywords: High-performance liquid chromatography; quantitative steroiddetermination ; semi-solid pharmaceutical preparationsMONIR AMIN and PETER W. SCHNEIDERSchering A.G., Department Galenik/Department Allgemeine Physikochemie,Mullerstrasse 170-178, 1000 Berlin 65, Germany.Analyst, 1978, 103, 1076-1079.Determination by Gas Chromatography - Single-ion MonitoringMass Spectrometry of Phthalate Contaminants in IntravenousSolutions Stored in PVC BagsShort PaPevKeywords : Phthalate contaminants ; intravenous solutions ; poly (vinyl chloride)bags ; gas chromatography - single-ion monitoring mass spectrometryG.A. ULSAKER and R. M. HOEMNational Centre for Medicinal Products Control, Sven Oftedals vei 8, Oslo 9, Norway.Analyst, 1978, 103, 1080-1083.Thin-layer Chromatographic Behaviour of BarbituratesUnder Various ConditionsShort PaperKeywords : Barbiturate determination ; thin-layer clzronzatographyR.ABU-EITTAH, A. OSMAN and A. EL-BEHAREDepartment of Chemistry, Faculty of Science, Cairo University, Giza, Egypt.Analyst, 1978, 103, 1083-1087.iOctober, 19’18 x THE ANALYSTAnalytical IChemistWe are seeking an analyst, with ahigh level of practical ability, toprovide an analytical service on acommercial basis to a range ofclients in the paper, plastics andeffluent fields. Chemical, micro-scopical and physical analyticalmethods, including infra-red,atomic absorption and gas chroma-tography are involved.For the first year the successfulapplicant will understudy our chiefanalyst who retires in a year’s time.He/she will then assume responsi-bility for a small team and will beexpected to develop the service on acommercial basis.This is a challeng-ing appointment requiring at leastfive years‘ analytical experience inthe above fields, and the minimumqualification considered will beHNC.Please apply in confidence, givingdetails of experience and technicalqualifications, together with em-ployment history and present salary,to:E. Moss,Personnel Manager,Reed Engineering & Develop-ment Services Limited,E 8 D Centre,Aylesford.Maidstone,Kent.A Reed International company -IFOR SALE. The Analyst, volume 78 (1933) to volume 98 (1973) in-clusive, unbound. (Index for 1968 missing.)Also Analytical Abstracts with index, volume 1 (1964) to volume 13(1966) inclusive, unbound.Offers to Box number 1418.C S Publicationslllodern Classicsn AnalyticalChemistry Vol. 2:ompiled by Alvin L. Beilby4 selection from the best featureirticles that appeared in issues of3nalyfical Chemistry from 1970 to1975. Ideal assupplementaryreadingor theadvancedstudent of analytical:hemistry. Includes spectroscopy,?lectrochemistry, chromatography,1 utomation and instrumentation,neasurement techniques, analyticalnethods, and art conservation.Paperbound 314pp I l ” x 8 $ ”0 841 2 0332 6 f 6.505th EditionACS Specifications for 320 r agentchemicals; includes 50 pages ofdefinitions, tests, and reagentsolutions. Features flame and flame-less atomic absorption methods;new polarographic and chromato-graphic procedures; and newcolorimetric test for arsenic.Clothbound 685pp 9 % ” ~ 6%”0 841 2 021 0 9 f 30.00both available from:The Chemical Society,Distribution Centre,Blackhorse Road,Let c h worth,Herts. SG6 1 HNEnglan

ISSN:0003-2654    

DOI:10.1039/AN97803BP097

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

OCTOBER 1978 Vol. 103 No. 1231 The Analyst Methods Asbestos for the Quantitative Determination of and Quartz in Bulk Samples Using X-ray Diffraction M. Taylor Health E Safety Executive, The Occueational Medicine and Hygiene Laboratories, 403-405 Edgware Road, London, NW2 6LN Procedures are described for the quantitative determination of the asbestos and a-quartz contents of bulk samples by use of X-ray powder diffracto- metry. The method gives satisfactory results for several different types of asbestos and for mixtures of two or more different types. Problems with sample grinding and preferred-orientation effects have been largely overcome. An effective procedure has been developed for grinding samples to a suitable particle size for accurate quantitative work. This procedure works equally well for all the types of asbestos studied and the sample is intimately mixed with the internal standard, nickel(I1) oxide, at the same time.A sample press has been designed that enables the same pressure to be applied to each sample when sample holders are filled for the diffractometer, giving the same degree of preferred orientation each time. Calibration lines have been calculated for chrysotile, amosite, crocidolite and anthophyllite, and results are given for mixtures containing two or more types of asbestos as well as other commonly occurring minerals. Similar techniques are used to grind samples containing quartz and to mix them with internal standard. Work on both synthetic and real samples is described and results are compared with those obtained by use of an infrared spectroscopic method.Keywords : A sbestos determination ; quartz determination ; X-ray diflraction Inhalation of asbestos dust can cause fatal lung diseases such as cancer and asbestosis and also mesothelioma of the pleura and abdominal cavity linings. X-ray diffraction (XRD) is an established procedure for the qualitative determination of asbestos, by which it is possible to identify the different crystalline types (chrysotile, amosite, crocidolite and anthophyllite), and to distinguish crocidolite from amosite even though the two have very similar diffraction patterns. The technique is less satisfactory when accurate quantitative results for asbestos are required; one of the main problems is that the preferred-orientation effects of fibrous materials cause large variations in diffraction-peak intensities1 Also, it is difficult to obtain a truly representative portion from the bulk of the sample submitted for analysis because sample mixtures are often inhomogeneous and the fibrous nature of asbestos inhibits mechanical mixing and stirring.Asbestos generally, and the chrysotile form in particular is difficult to grind to the optimum size for accurate quantitative work (Crable and Knott used a particle size of less than 3.5 pm in their work on amosite and crocidolite2). Prepara- tion of an intimate mixture of the ground asbestos with an internal standard, e.g., calcium fluoride or nickel(I1) oxide, is also difficult. The method described here substantially over- comes these problems.Crable and Knott used a powder diffractometer in their work on amosite and crocidolite, and also on chrysotile.3 Goodhead and Martindale used a powder camera technique.4 The proposed method is substantially different and is applicable to all types of asbestos and also to mixtures containing several types in conjunction with other minerals. Nickel(I1) oxide as purchased and without pre-sizing is used as the internal standard. Mixing and grinding are carried out reproducibly in grinding mills. The grinding procedures adopted reduce the length of the asbestos fibres to below 5.0pm, irrespective of asbestos type or the complexity of the mixture being ground, which helps to reduce preferred- orientation problems. Crown Copyright. 10091010 TAYLOR : QUANTITATIVE DETERMINATION OF ASBESTOS Analyst, VOl.103 It was found that the intensities of the asbestos diffraction peaks were dependent upon the pressure used in preparing the samples for scanning in the diffractometer. A sample press has been constructed to enable the same pressure and thus the same degree of preferred orientation to be applied to each sample. The method has been used to analyse a large variety of asbestos-containing samples, including complex mixtures. Analytical results obtained by analysts experienced at this type of work gave good agreement with those obtained by others new to the method. A similar analytical procedure has been used to prepare quartz samples for XRD analysis. Inhalation of quartz particles in the respirable range of below 5.0 pm can, over many years, cause the lung disease silicosis, so dust samples are analysed in order to help predict if the use of a particular material in a factory could cause a quartz hazard.The analysis of asbestos-containing samples will be described first and then the related procedures for the analysis of quartz will be discussed. Determination of Asbestos Apparatus Philips PW 1011/00 X-ray generator. A vertical powder diffractometer fitted with a graphite-crystal focusing monochromator and proportional counter. Copper Kcc X-radiation is produced with generator settings of 44 kV, 36 mA, 1" divergence slit and 2" anti-scatter used as receiving slit. Chart recorder. The chart recorder used for recording the XRD traces has a zero suppression facility, which has proved invaluable when traces at very low attenuations are recorded in order to measure very low percentages by mass of asbestos. Glen Creston microhammer mill.Fitted with a 3.0 mm mesh diameter screen. The microhammer mill has been modified by replacing with a screw the retaining clip that normally holds the hammers in place on the central shaft, thus facilitating removal of the hammers for cleaning. Anahtical mixer mill. Obtained from Glen Creston Ltd. With a 5-ml hardened stainless- steel grinding cylinder and steel ball pestle, and a 5-ml agate grinding cylinder and matching ball pestle. Sample press. Specially constructed (see below). Reagents Union Internationale Contre le Cancer ( UICC) standard reference asbestos samples. Chrysotile A, chrysotile B, amosite, crocidolite and anthophyllite.NickeZ(1.I) oxide (NiO) . Johnson Matthey Specpure. y-Aluminium(III) oxide (y-Al,O,) . Aluminium oxide 90 active, neutral for column chromatography (E. Merck). Acetic acid, 1 N. Cyclohexane. Analytical-reagent grade. Celldose powder. Whatman CF 11. Analytical-reagent grade diluted with de-mineralised water. Procedure First, a microhammer mill fitted with a 3 mm mesh screen is used to grind all, or a substantial portion (several grams), of the sample material to an average fibre length of 1-2mm. Asbestos fibres in bulk samples are usually much longer than this and must be reduced in length before the mixer mill can be used to grind them to a size suitable for analysis. After the first grinding a representative portion can be taken from the bulk sample and a qualita- tive XRD pattern obtained. Comparison with reference patterns shows the types of asbestos present.If chrysotile is detected a rough visual estimate is made of whether the percentage by mass is likely to exceed 60%. Chrysotile is the most difficult type of asbestos to grind and a larger mass of y-aluminium(II1) oxide must be used in the second grinding to give the required fibre length if the percentage by mass is over 60%. Any carbonates present, e.g., calcium carbonate, should be removed by treatment with 1 N acetic acid. Samples as received are often very inhomogeneous and are ground in two stages.October, 1978 AND QUARTZ I N BULK SAMPLES USING X-RAY DIFFRACTION 101 1 For the second grinding the steel cylinder of the mixer mill is used.An amount of sample weighing 0.1000 i. 0.0005 g, 0.1000 rfi: 0.0005 g of y-aluminium(II1) oxide [0.1500 -& 0.0005 g of y-aluminium(II1) oxide should be used if the percentage by mass of chrysotile is estimated to be above SO%], 0.0333 & 0.0001 g of nickel(I1) oxide and 0.5 ml of cyclohexane (added by pipette) are placed in the cylinder with the steel ball pestle, the ends of the cylinder are taped with poly(viny1 chloride) tape in order to prevent leakage and the mixture is ground for 25 min (see Note). The mixer mill is thus used both to grind the sample and to mix i t intimately with the nickel(I1) oxide, the internal standard, in one operation. If the specified amount of sample cannot easily be weighed out because of handling problems, a slightly smaller amount is weighed and the combined mass of sample plus y-aluminium(II1) oxide is made up to 0.2000 & 0.001 0 g (or 0.2500 & 0.001 0 g for samples containing more than 60% of chrysotile) by adding more y-aluminium(II1) oxide.The necessary correction is applied to the result. NOTE- The y-aluminium(II1) oxide acts as a grinding agent and is effective in reducing the asbestos fibres to the required length, particularly with cyclohexane present. It also helps to ensure that the grinding is uniform; without y-aluminium(II1) oxide present the asbestos is not uniformly ground and a sizeable proportion of it merely becomes packed into the end caps of the grinding cylinder. y-Aluminium(II1) oxide gives a very weak diffraction peak, which does not interfere with that of asbestos.Close control of the mass loading of the grinding cylinder and of the length of the grinding period helps to ensure that samples are reproducibly ground so that most of the fibres are of a length below 5.0 pm. The material is removed from the cylinder and the cyclohexane is allowed to evaporate. The dry residue is ground in the mixer mill in an agate grinding cylinder with an agate ball pestle for a further 5 min in order to break up the lumps left when the cyclohexane has evaporated, and to improve homogeneity. Samples are pressed in standard Philips sample holders with a specially constructed sample press (see below). A 0.620 rfi 0.001-g amount of cellulose powder is used for each pressing. The sample is then scanned in the diffractometer over the relevant asbestos peaks and also over two nickel(I1) oxide peaks. The primary peaks for chrysotile, amosite, crocidolite and anthrophyllite are 7.36, 8.33, 8.42 and 3.24 A, respectively.The secondary peaks are 3.66, 3.07, 3.09 and 3.05 A, respectively. These peaks are used in conjunction with the nickel(I1) oxide 2.09- and 2.41-A peaks (1 A = 10-lo m). Scanning speed is normally Q" min-l but should be slower if the percentage by mass of asbestos is low. Three separate portions of each sample mixture are scanned. The asbestos to nickel(I1) oxide peak-height ratios are calculated and the percentage by mass of asbestos in the sample is determined from calibra- tion graphs. Four values are generally obtained for each type of asbestos present in the sample, from the permutations of (asbestos primary or secondary peak height)/[nickel( 11) oxide 2.09- or 2.41-A peak height].These four values are averaged to give the concentration of asbestos. That the grinding is reproducible has been verified by optical microscopy. Sample Press Standard Philips sample holders for the powder diffractometer were used and sample mixtures were pressed in place with cellulose powder. In the early stages of the work two analysts prepared calibration mixtures, pressing sample slides in the usual way and independently constructing calibration graphs. It was found, however, that calibration graphs produced by one analyst could not be used by the other; each analyst had to produce calibration graphs for his own use only. It was realised that the asbestos to nickel( 11) oxide peak-height ratios obtained were dependent upon the pressure used to press the samples.To ensure that the same pressure was applied to each sample, a press (Fig. 1) based on a small portable electric drill stand was designed and constructed. The press is made of steel and is plated to prevent corrosion. Two locating pins, B, fit into holes in the base plate, A, and the locating plate, C, for the metal sample holder, H, and the block, D, fits on to these. A fixed mass of cellulose powder is pressed to a fixed volume each time, the volume being controlled by two adjustable stops, E, which screw into the block and which arrest the downwards movement of the plunger. It was found by trial and error that a press with the dimensions shown required 0.620 6 0.001 g of cellulose powder to obtain good pressings, and this mass of powder was used each1012 TAYLOR : QUANTITATIVE DETERMINATION OF ASBESTOS AnaZyst, VoZ.103 3.5 cm '2.6 cd 9 cm Side view Front view Diagrams showing the construction of the sample press (not to scale). Fig. 1. (a), Side view. ( b ) , Exploded view. A, Base plate, bolted to the base of the drill stand; B, locating pins; C, locating plate for sample holder; D, block; E, adjustable stops; F, plunger; G, strengthened arm for the drill stand; H, sample holder. (c), Locating plate with sample holder in situ (plate thickness and sample-holder thickness both approximately 1 mm). (d), Cross-sections through the block and plunger. Operation of the press is as follows.The locating plate is placed over the locating pins and the sample holder is fitted into place. A thin layer of sample mixture is spread evenly over the hole in the sample holder. The block is fitted into place, the cellulose powder is poured over the sample through the hole in the block and its surface is roughly levelled withOctober, 1978 AND QUARTZ I N BULK SAMPLES USING X-RAY DIFFRACTION 1013 a microspatula. Lowering the handle of the drill stand moves the plunger down until its progress is arrested by the stops. The bevelled edges prevent the sample from sticking within the block. The plunger is slowly withdrawn and the block removed from the locating pins: the pressed sample can then be removed for scanning in the powder diffractometer. The cellulose becomes compacted by the action of the press and it does not stick to the plunger when the latter is withdrawn.Calibration Graphs The UICC standard reference samples supplied by the Pneumoconiosis unit of the Medical Research Council5 were used to prepare mixtures for the construction of calibration graphs. These samples were received already ground to an average fibre length of approximately 1-2 mm, which makes them directly comparable with real samples that have been ground by the microhammer mill fitted with the 3.0 mm mesh diameter screen. I t was therefore only necessary to grind them in the mixer mill by means of the procedure given above. Calibration graphs were constructed for all of the reference materials; at least nine mixtures were prepared for each material.For any mixture the total mass of asbestos plus y-aluminium(II1) oxide as primary diluent was 0.1000 & 0.0005g. [For example, a 60% amosite mixture would contain 0.0600 & 0.0003g of amosite diluted with 0.0400 -J= 0.0002g of y-aluminium(II1) oxide.] To this would be added 0.0333 & 0.0001 g of nickel(I1) oxide and a further 0.1000 -j= 0.0005 g of y-aluminium(II1) oxide as grinding agent; chrysotile is a special case as it is particularly difficult to grind and for mixtures containing more than 0.0600 -+ 0.0003g of chrysotile (i.e., more than 60% of chrysotile), the amount of y-aluminium(II1) oxide as grinding agent had to be increased to 0.1500 & 0.0005 g. The absorption coefficient of the asbestos being analysed differs from that of the matrix.If an internal standard is added in a constant proportion, the concentration of the asbestos component is a linear function of the asbestos to internal standard peak-height ratio.6 Three separate portions of each mixture were scanned and the peak-height ratios calculated. Average values for the ratios were used to construct the graphs. A calibration graph is shown for the 7.36-A chrysotile A diffraction peak in Fig. 2. All the calibration graphs were straight lines and the coefficients m and c for the straight-line equationy = mx + c, calcu- lated by linear regression, are given in Table I for each line. 0 Concentration of chrysotile A, % Fig. 2. Calibration graphs for chrysotile A using the 7.36-A chrysotile A peak. Wavelength of nickel(I1) oxide peak: 1, 2.41 A ; 2, 2.09 A.Amount of y-aluminium(II1) oxide added as grinding agent: 0 and 0, 0.1OOg; and a, 0.150g. For each mixture the combined mass of chryostile A and y-aluminium( 111) oxide added as primary diluent is 0.1000 f 0.0005 g, with 0.0333 & 0.0001 g of nickel(I1) oxide added as internal standard.1014 TAYLOR : QUANTITATIVE DETERMINATION OF ASBESTOS Analyst, VoZ. 103 Calibration lines do not pass through the origin as the peak heights are measured from a “noisy” base line; a significant peak cannot be detected until a few per cent. of asbestos are present. Fig. 3 shows how the base lines were drawn for the relevant peaks in order to measure the peak heights. Asbestos does not give particularly intense diffraction peaks. Chrysotile Crocidolite 3.66 a 2 --b 2.09 8, Nickel ( 1 1 ) oxide Anthophyl lite AmositeOctober, 1978 AND QUARTZ I N BULK SAMPLES USING X-RAY DIFFRACTION 1015 Goodhead and Martindale* state that samples of a particular type of asbestos mined from different locations can have slightly different chemical compositions and that this can alter the diffraction pattern, so that calibration lines may be somewhat dependent upon the origin of the sample.The calibration lines obtained for UICC chrysotile A mined in Rhodesia and UICC chrysotile B (which originates in Canada) demonstrate this effect. Ideally, therefore, specimens of pure asbestos should be taken at each sampling site and should be used to determine whether or not the existing calibration graphs can be used without correction. In practice, however, it is rarely possible to obtain these specimens of “pure” asbestos when sampling and the unmodified graphs must be used.TABLE I VALUES OF THE COEFFICIENTS m AND c IN THE ASBESTOS CALIBRATION L I N E S Y = mX + C y = asbestos to nickel(I1) oxide peak-height ratio and x = percentage of asbestos. Type of asbestos Chrysotile A Chrysotile B Amosite Crocidolite Anthopyilite .. . . .. .. .. .. .. .. .. .. Asbestos peak/a 7.36 7.36 3.66 3.66 7.36 7.36 3.66 3.66 8.33 8.33 3.07 3.07 8.42 8.42 3.09 3.09 3.24 3.24 3.05 3.05 Nickel ( 11) oxide peakla 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 2.41 2.09 m 0.00309 0.001 97 0.002 58 0.001 65 0.002 77 0.001 73 0.00255 0.001 58 0.008 88 0.00535 0.00706 0.004 24 0.009 64 0.005 66 0.006 10 0.003 58 0.001 85 0.001 15 0.002 64 0.001 63 c -0.0128 -0.009 48 -0.00491 -0.004 27 -0.0148 - 0.008 74 - 0.009 05 -0.004 88 - 0.020 6 - 0.013 4 -0.003 72 - 0.002 56 - 0.030 6 -0.0173 -0.0146 -0.007 92 -0.00264 - 0.002 02 + 0.014 9 + 0.009 46 Factors afecting the determination of chrysotile These results arise from the permutations of the primary and secondary asbestos peaks with the two nickel(I1) oxide peaks.The cellulose used to hold the sample in place can give a peak that partially overlaps the 3.66-A chrysotile peak. If cellulose shows on the surface of the sample pressing it is safest to use only ratios based on the 7.36-A peak to calculate the result, because the height of the 3.66-A chrysotile peak may have been affected. The height of the 3.66-A peak is usually slightly less than that of the 7.36-A primary peak. The mineral kaolinite gives a diffraction pattern which overlaps that of chrysotile, and no way has yet been found of overcoming its interference with the chrysotile trace.Generally four results are calculated for each type of asbestos in the sample. Mixtures containing crocidolite and amosite If crocidolite and amosite occur together in a sample, results must be based on the 3.09-A secondary peak for crocidolite and the 3.07-A secondary peak for amosite. The primary peaks for crocidolite and amosite at 8.42 and 8.33 A, respectively, cannot readily be resolved. Removal of carbonates A large proportion of asbestos-containing samples received by this laboratory are taken during de-lagging operations and often contain a high proportion of calcium carbonate or other carbonates. Calcium carbonate gives an intense diffraction peak at 3.035 A, which1016 TAYLOR : QUANTITATIVE DETERMINATION OF ASBESTOS Analyst, Vd.103 overlaps the 3.09- and 3.07-A secondary peaks of crocidolite and amosite, respectively. If, therefore, it is not removed from samples containing amosite or crocidolite, results must be based on the primary asbestos lines only. Treatment with 1 N acetic acid removes the carbonates, so that the sample mixture is concentrated and simplified and secondary asbestos lines can be used in the determination. It was not known, however, whether the different types of asbestos were completely immune to attack by this reagent. Portions of UICC standard reference asbestos samples were subjected to the treatment with 1 N acetic acid and were then analysed for asbestos content.It was found that the percentage remaining of amosite, chrysotile A, crocidolite and anthophyllite was 85.5, 92, 94 and 97, respectively. These results suggest that amosite in particular is not completely free from attack by this acid. Analysis of syntlzctic mixtwes Test mixtures containing two or three different types of asbestos, and some containing asbestos plus quartz or calcium carbonate, were analysed and the results are shown in Table 11. For mixtures containing both amosite and crocidolite their concentrations have been calculated using the 3.07- and 3.09-A secondary peaks and no allowance has been made for the overlap between these two peaks. I t can be seen that the largest deviations between actual and calculated values occur for chrysotile and in some instances the actual value is not within &lo% of the value determined.TABLE I1 RESULTS OF THE ANALYSIS OF TEST MIXTURES CONTAINING ASBESTOS Mixture Composition, % nz/m 1 Amosite, 18 Chrysotile A, 28.5 Calcium carbonate, 46.6 y-Aluminium(II1) oxide, 6.9 Chrysotile B, 74 y-Aluminium(II1) oxide, 17.7 Amosite, 14 Calcium carbonate, 73 2 Crocidolite, 8.3 3 Quartz, 13 4 Quartz, 14.6 Chrysotile A, 32 y-Aluminium(II1) oxide, 53.4 Chrysotile A, 36 Amosite, 48 5 Crocidolite, 16 6 Crocidolite, 38 Chrysotile B, 24.5 Amosite, 15 y-Aluminium(II1) oxide, 22.5 Chrysotile B, 25 Amosite, 28 y-Aluminium(II1) oxide, 5 7 Crocidolite, 42 Amount of asbestos determined, % mlnz Amosite, 17 Chrysotile, 22.5 Crocidolite, 11 Chrysotile, 73 Amosite, 13.5 Chrysotile, 25.5 Crocidolite, 23.5* Chrysotile, 37 Amosite, 53 Crocidolite, 42* Chrysotile, 29.5 Amosite, 17 Crocidolite, 44.5* Chrysotile, 23 Amosite, 27 * There was considerable overlap between the 3.07-A amosite and 3.09-A crocidolite It was invariably found with the test mixtures that the 7.36-A chrysotile peak had the non-symmetrical profile shown in Fig.3 (iii), making it impossible to measure the peak height accurately. The peak shapes shown in Fig. 3 (i) and (ii) were found with the chrysotile calibration mixtures. It was thought that the non-symmetrical peak shape was due to some peaks used.October, 1978 AND QUARTZ IN BULK SAMPLES USING X-RAY DIFFRACTION 1017 overlap between the chrysotile peak and the primary amosite and crocidolite diffraction peaks, but it occurred even with mixture 4 in Table 11, which contained no amosite or crocidolite.Limits of detection The lowest percentages by mass of asbestos used to plot the calibration graphs were approximately 10%. These percentages were readily measurable for all the different types of asbestos. The average primary peak heights obtained for the mixtures of lowest asbestos content are shown in Table 111. TABLE I11 PEAK HEIGHTS OBTAINED WITH THE CALIBRATION MIXTURES OF LOWEST ASBESTOS CONTENT Peak height/ Peak position/ Asbestos type and percentage counts s-1 A 11% Chrysotile A . . .. .. 36 7.36 13.5% Chrysotile B . . .. 38 7.36 10.5% Amosite . ... .. 80 8.33 12.8% Crocidolite . . .. .. 138 8.42 10% Anthophyllite . . .. 23 3.24 10% Anthophyllite . . .. 62 3.05 The limit of detection is less good for chrysotile than for the other forms of asbestos, i.e., about 6% by mass of chrysotile should be just detectable whereas 2 or 3% of amosite or crocidolite is clearly detectable. Six per cent, of chrysotile should give a rather diffuse peak about 20 counts s-l in height and a very slow scanning speed and relatively long associ- ated time constant are needed to measure it. Two or three per cent. of amosite or crocidolite would give a more clearly defined peak. Anthophyllite is rarely found in the absence of interfering minerals but its limit of detection should again be about 3%. If the mixture can be cleaned with acetic acid detection limits will correspondingly improve. The latest diffraction tubes are designed for use at up to 2700 W.Use of a more powerful generator that could take advantage of this higher tube rating would considerably increase the signal and improve the detection limits. The X-ray generator used had a maximum power rating of 1600 W. Determination of Quartz The main use of XRD at the Health & Safety Executive laboratories in London is in the analysis of samples for quartz. The samples are largely of airborne dust collected on filters, but bulk samples are also examined and these are prepared in much the same way as asbestos samples. Nickel(I1) oxide is again used as internal standard' and y-aluminium(II1) oxide is also used, acting as a buffer to even out the grinding so that the quartz and nickel(I1) oxide particles will be ground to the same extent for each sample, independently of sample composition.The sample press is not used as there is no problem of preferred orientation with quartz particles. Gordon and Harris* have demonstrated that the determination of quartz by use of XRD is dependent on particle size. Nenadic and Crableg have suggested that for quantitative work the quartz mean particle diameter should be of the order of 5pm. Hand grinding and mixing with a mortar and pestle is subjective and a grinding procedure based on the analytical mixer mill was found to give more reliable results and to make it possible to grind consistently to the recommended particle size. Procedure The microhammer mill can be used to grind the bulk sample so that a representative portion can be taken.A qualitative XRD scan on the sample shows the presence of quartz and other minerals. Chemical clean-up procedures are used to concentrate the quartz in the sample and to remove minerals that interfere with the quartz 3.343-a diffraction peak.1018 TAYLOR : QUANTITATIVE DETERMINATION OF ASBESTOS Analyst , VoZ. 103 Samples from foundries commonly contain iron; this is removed by treatment with a warm solution of hydrochloric acid - water (1 + 2). The mixture is filtered through a pulp pad, which is then ignited. Silicates such as mullite (3A1,0,.2Si02) , which give a strong diffrac- tion peak at 3.39 hi, close to the 3.343-,k a-quartz pnmary peak, are removed by fusing the sample with analytical-reagent grade potassium hydrogen sulphate and dissolving the melt in hot water to which a few millilitres of hydrochloric acid - water (1 + 2) have been added.The solution is filtered through a pulp pad, which is then ignited, and again any quartz is left as a residue. The cleaned sample is ground initially with a mortar and pestle in order to remove extreme roughness. A 0.200 & 0.001-g amount of sample, 0.200 & 0.001 g of nickel(I1) oxide and 0.1000 0.0005 g of y-aluminium(II1) oxide are placed in the steel cylinder of the analytical mixer mill with 0.5 ml of cyclohexane and ground for 30 min with the steel ball pestle. The ends of the cylinder are taped to prevent leakage. At the end of the grinding period the material is scraped from the cylinder, the cyclohexane is allowed t o evaporate and the lumps in the dried mixture are gently broken up with a pestle and mortar but without further grinding.Optical microscopy has shown that this grinding procedure reduces the mean diameter of the quartz and nickel(I1) oxide particles to below 5.0 pm. As with asbestos, cellulose powder is used as the backing material to hold the sample in place in the diffracto- meter sample holder. Results are not pressure dependent and so samples can be pressed by hand. The same diffractometer scanning conditions are used for quartz as are described for the quantitative work on asbestos. A scanning speed of 4'min-l is usual. The 2.09- and 2.41-,k nickel(I1) oxide peaks are again used, in conjunction with the 3.343-A quartz peak; quartz to nickel(I1) oxide peak-height ratios are calculated and the percentage of quartz in the sample is determined from a straight-line calibration graph.Quartz sample mixtures are scanned in duplicate as the precision of peak-height measurements is better for quartz than for asbestos. The sample mixture is then ready for scanning on the diffractometer. Calibration Graph The calibration graph used for quartz was constructed with mixtures of quartz, Y- aluminium(II1) oxide and nickel(I1) oxide. Two types of pure quartz were used in 21 standard mixtures; in each instance the total mass of quartz plus y-aluminium(II1) oxide as diluent was 0.2 g [for example, a 10% quartz mixture would be 0.02 g of quartz diluted with 0.18 g of y-aluminium(II1) oxide].To this would be added 0.2 g of nickel(I1) oxide and a further 0.1 g of y-aluminium(II1) oxide as grinding agent. Duplicate portions of each mixture were scanned and the average values for the relevant peak-height ratios were used to construct the calibration graphs. The calculated values of m and c in the straight-line equation y = mx + c were m = 0.0199, c = -0.0196 for the a-quartz 3.343-&nickel(II) oxide 2.41-A line and m = 0.012 3, c =- 0.0073 for the or-quartz 3.343-hi/nickel(II) oxide 2.09-A line. Peak-height ratio was again plotted along the y-axis and percentage of quartz along the x-axis. The spread of the points about these graphs was less than in Fig. 5 , though the gradient was virtually the same. Both graphs have correlation coefficients of 0.995 and in 95% of cases the true quartz concentration should lie within -+6% of the value determined. The mean quartz and nickel(I1) oxide particle size for these mixtures was again below 5.0 pm; optical microscopy showed no noticeable difference between mean particle size in mixtures used to prepare the calibration graph and those used to prepare Fig.5. The degree of grinding would appear to be the same for both graphs; the wider spread of the points in Fig. 5 is attributed to the difficulty of obtaining a uniform mixture when more components are present. This demonstrates the advisability of carrying out clean-up procedures to reduce the number of components to be mixed. No iron was present. Results and Discussion Analysis of synthetic mixtwes In order to determine the effects of y-aluminium(II1) oxide on the grinding process, synthetic mixtures were prepared from two different samples of relatively pure quartz.The first, known as Snowit X6403, comes from a Belgian deposit and is used by the SafetyOctober, 1978 1019 in Mines Research Establishment (SMRE) as a reference quartz. The other is a commercially available Redhill sand, reference T3/300. Both samples showed a wide range of particle sizes. Figs. 4 and 5 demonstrate the effect of the y-aluminium(II1) oxide on the grinding process for samples containing quartz. Fig. 4 shows results obtained with synthetic mixtures of quartz and iron powder, but with no y-aluminium(II1) oxide present. The combined mass of quartz and iron was 0.2 g for each mixture and the mass of nickel(I1) oxide was 0.2 g in each instance.The graph is not a straight line for the entire concentration range up to 100% of quartz, which suggests that the quartz particles are ground to a different extent at the upper and lower regions of the graph. For higher concentrations of quartz there is less iron present to cushion the grinding. Optical microscopy showed that the mean size of the quartz particles in both the upper and lower region was below 5.0pm, but in the upper region of the graph the quartz was more vigorously ground and the spread of particle size was very narrow: all the particles of quartz observed in this region were below 5.0 pm. In the lower region of the graph there was a wider spread of quartz particle size. For Fig. 5 the total mass of quartz and iron used for each mixture was again 0.2 g, with 0.2 g of nickel(I1) oxide, but 0.1 g of y-aluminium(II1) oxide was added to each mixture before the grinding and mixing.The presence of y-aluminium(II1) oxide evens out the grinding so that the resulting graph is linear, but the spread of the points about the lines is still fairly wide. The mean particle size for quartz and nickel(I1) oxide in these mixtures was again below 5.0 pm. AND QUARTZ IN BULK SAMPLES USING X-RAY DIFFRACTION - (I) 5 1.6 - 1.4 - 1.2 .- +.’ - .$ 1.0 - Y 0.8 - 2 0.6 - 0) .- , > I 20 40 60 80 100 Y Concentration of quartz, % Effect of grinding quartz with no y-aluminium(II1) oxide present. For each mixture the combined mass of quartz and iron is 0.2 g, with 0.2 g of nickel(I1) oxide added as internal standard.Wavelength of quartz peak, 3.343 A. Wavelength of nickel(I1) oxide peak: 1, 2.41 A; 2, 2.09 A. Quartz sample: and 0, Snowit X6403; 0 and 0, T3/300 ground silica. Fig. 4. Concentration of quartz, % Fig. 5. Effect of grinding quartz with y- aluminium(II1) oxide. Details as for Fig. 4, except that 0.1 g of y-aluminium(II1) oxide is added to each mixture. Analysis of real samples Fourteen real samples were analysed for quartz content by the above method, and also independently at SMRE, Sheffield, by an infrared spectroscopic technique. The samples included dusts from foundries, brickworks, potteries and roofing-felt manufacture. Particle size is even more critical for infrared methods than it is for XRD. The presence of kaolinite in some of the pottery samples and talc in the roofing-felt samples caused the analyst who was carrying out the infrared spectroscopy some difficulty.SMRE claim an accuracy of 10% for their infrared deterrninations on samples with quartz content greater than 30%.1020 TAYLOR Table IV compares the results obtained with the two techniques. An optical examination using optical microscopy of several of the samples prepared for XRD showed that the quartz and nickel(I1) oxide particle size was uniform and generally below 5 ,urn. TABLE IV COMPARISON OF RESULTS FOR QUARTZ OBTAINED ON REAL SAMPLES BY XRD AND INFRARED SPECTROSCOPY Amount of quartz, % r A \ Factory type XRD Infrared spectroscopy Foundry Brickworks Pottery Roofing felt (Sand) .. . . .. .. .. .. .. .. .. .. 76 57 69 66 79 18 39 8 15 42 7 26 94 White powder (source unknown) . . 85 84 59 78 69 70 38 8 12 49 2 27 93 12-16 93 The author gratefully acknowledges the help of various members of the Health & Safety Executive staff, particularly Messrs. Black, Evans and Revell. This paper is contributed by permission of the Director of the Research and Laboratory Services Division, Health & Safety Executive. 1. 2. 3. 4. 5. 6. 7. 8. 9. References Klug, H. P., and Alexander, L. E., “X-ray Diffraction Procedures for Polycrystalline and Amorphous Crable, J. V., and Knott, M. T., Am. Ind. Hyg. Ass. J . , 1966, 27, 449. Crable, J. V., and Knott, M. T., Am. Ind. Hyg. Ass. J . , 1966, 27, 383. Goodhead, K., and Martindale, R. W., Analyst, 1969, 94, 985. Timbrell, V., and Rendall, R. E. G., Powder Technol., 1971/72, 5, 279. Klug, II. P., and Alexander, L. E., “X-ray Diffraction Procedures for Polycrystalline and Amorphous Klug, H. P., Alexander, L., and Kummer, E., J . Ind. Hyg. Toxicol., 1948, 30, 166. Gordon, R. L., and Harris, G. W., Natuve, Lond., 1955, 175, 1135. Nenadic, C. M., and Crable, J. V., in Simmons, I. L., and Ewing, G. W., Editors, “Progress in Received February 23rd, 1978 Accepted April 25th, 1978 Materials,” Second Edition, Wiley-Interscience, New York, 1974, p. 368. Materials,” Second Edition, Wiley-Interscience, New York, 1974, p. 536. Analytical Chemistry,” Volume 6, Plenum, New York, 1972, p. 81.

ISSN:0003-2654    

DOI:10.1039/AN9780301009

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, October, 1978, Vol. 103, pp. 1021-1030 1021 Application of the Faraday Effect to the Trace Determ i nation of Cadmium by Atomic Spectroscopy with an Electrothermal Atomiser K. Kitagawa, T. Shigeyasu and T. Takeuchi Department of Synthetic Chemistry, Faculty of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan As a technique for the trace determination of elements, the Faraday effect has been applied t o atomic spectroscopy with the use of electrothermal atomisers. The atomiser was located between two crossed plane-polarisers (called the polariser and the analyser) . An electromagnet magnetised the atomised sample and, as a result of the Faraday effect, the source radiation could, in the presence of the atomic vapour, pass through the optical system.A simple theoretical treatment was developed to explain the dependence of the transmitted intensity on the magnetic field strength and the source intensity and the ability to eliminate the effects of background scattering. Measurements on cadmium were carried out at a wavelength of 228.8 nm, dispensing into the atomiser 5-p1 samples, which, after drying, were atomised at 1000-1 800 "C.' The calibration graphs demonstrated a square-law dependence. At high concentrations the calibration graphs were corrected for atomic absorption. The technique gave a detection limit of 5 x 10-13 g for cadmium. Keywords : Magneto-optical rotation ; atomic spectroscopy ; electrothermal atomisation ; Faraday effect ; cadmium determination Atomic-fluorescence spectroscopyl-3 has been developed as a useful technique for the determination of trace amounts of elements. It requires no thermal excitation of atoms and consequently can reduce the known inter-element effects due to variations in the excitation temperature or the degree of ionisation of atoms, that may be encountered in atomic-emission ~pectroscopy.~ Compared with atomic-absorption spectroscopy, the noise in the light source does not contribute to the noise on the base line.Further, the intensity of the re-emitted radiation increases with increasing source intensity and the development of intense light sources may therefore enable the analytical sensitivity to be improved. As the re-emitted light has a very narrow band width, the construction of a non-dispersive detection system and multi-element detectors is feasible.However, two problems in the practical application of atomic-fluorescence spectroscopy must be pointed out : the difficulty of collecting the re-emitted radiation necessitates the use of a sensitive detector, such as a photoelectron counter, and the background signal produced by scattering of the source radiation by non-atomic species produced in the atomiser, for example water droplets in a total consumption burner or smoke generated by pyrolysis of organic substances in the sample matrix in a furnace atomiser. In this work, a spectroscopic method utilising the Faraday effect is proposed for over- coming these problems. A magnetic field is applied to an electrothermal atomiser in the direction parallel to the propagation of the source radiation, as shown in Fig.1. The electrothermal atomiser is located between two plane-polarising prisms in a crossed con- figuration. When the atomic vapour is absent or not magnetised, the source radiation is blocked by the analyser. Owing to the Faraday effect, the magnetised atomic vapour rotates the plane of polarisation of the incident light, which then passes through the analyser. By this means we can detect the presence of analyte atoms. Afacaluso and Corbino,S in 1898, stated that a rotation of the plane of polarisation was observed when a magnetic field was applied to sodium vapour. They employed solar light or an arc discharge as the source. A stainless-steel tube containing sodium metal was placed between crossed Nicol prisms and a magnetic field was applied by means of a solenoid coil.They heated the tube with a burner to vaporise the sodium metal and measured the1022 KITAGAWA et al. : APPLICATION OF THE FARADAY EFFECT TO THE Analyst, VoZ. 103 Analyser Atorniser Po I ariser - . ,. ::. : .. . . q \m &Iqq . . +.;. .. . .:. Fig. 1. Schematic diagram of the Faraday configuration. Source * angle of rotation by revolving the analyser. The found that magnetic rotation occurred in the atomic vapour of sodium as well as in the molecular vapour of Na,. Later, Ladenburg6 discussed magnetic rotation by hydrogen atoms. More recently, Corney et aZ.7 applied dispersion theory to this phenomenon and to the Voigt effect for atomic vapour. They called the phenomenon “forward scattering of resonance radiation.” Making use of the fact that the forward-scattered light is coherent, Hackett and Seriess studied the line crossings (see under Principle) of the isotope-shifted lines in Hg I 253.65nm. Church and Hadeishi9 studied the line crossings and pointed out the possibility that the phenomenon was applicable to the trace determination of mercury vapour in air.The Faraday effect applied t o molecular spectroscopy is called magneto-optical rotation (MOR), and is expected to be useful for the determination of molecules that are optically inactive and difficult to examine by optical rotatory dispersion. Buckingham and StephenslO described the theoretical aspects of MOR in detail. Although this principle has been used for the modulation of a laser beam, its development for analytical purposes has, to date, been limited.The Faraday effect applied to atomic spectroscopy can be called “atomic magneto-optical rotation” (AMOR). Principle The phenomenon that the plane of polarisation rotates owing to asymmetric carbon atoms in certain organic substances when located between crossed Nicol prisms is called optical rotatory dispersion (ORD) and has been developed for molecular analysis.ll In certain instances, and only if external magnetic forces are applied to the substance, the source radiation can pass through the optical system. The associated effects are named after the original workers involved. In the Faraday effect, the plane of polarisation rotates through the substance to which the magnetic force is applied in the direction parallel to the propaga- tion of the source radiation.Another example in which the magnetic force is applied transversely is the Cotton - Mouton effect or the Voigt e€fect.12 In this instance, the source radiation can pass through the optical system because the propagating light wave is ellipti- cally polarised through the magnetised substance. An application of the Voigt effect to the trace determination of elements has been reported.13 As illustrated in Fig. 2, when the atomic vapour is magnetised the absorption line splits into 0-, T and of components due to the Zeeman effect, which correspond to the transitions AM = -1, 0 and +1, respectively, where AM is the difference in magnetic quantum number between the upper and lower energy levels. When the optical system is viewed in the direction of the magnetic field, the a- and (T+ components are circularly polarised with clockwise and counter-clockwise motion, respectively.The TT components are not observed from this position as they vibrate along the lines of magnetic force. Because plane-polarised light can be considered to be the resultant of opposite circular motions of the same frequency, the plane-polarised source radiation can be resolved into clockwise and counter-clockwise rotations. Fig. 3 shows the correlation of the refractive indices of the cr- and a+ components of the source radiation with the absorption coefficient, a, and the angle of rotation, 4. As the refractive index is inversely proportional to the propagating velocity of the motion, the clockwise motion travels slower outside and faster inside the region between the cr- and (T+ components.For the counter-clockwise motion, the reverse applies. Owing to the coherence, the two circular motions are synthesised to yield a plane-polarised wave. Because of the difference in velocity, the plane of the resulting light is at an angle to that of the incident light. Hence the rotatory power arises from the difference in refractive index or, in other words, it is caused by the double refraction of two circularly polarised waves through the Zeeman com- ponents a+ and cr-. If there is no difference in frequency between the CT+ and a- components, The Faraday effect is related to the longitudinal Zeeman effect.October, 1978 TRACE DETERMINATION OF CADNIUM BY ATOMIC SPECTROSCOPY 1023 the refraction curves are superimposed on each other, and there is no difference between them, resulting in no rotation.This is called line crossing. The line crossings among hyperfine components and isotope-shifted components in the mercury line a t 253.65 nm have been studied in .f, Transmitted intensity U- U+ PE 9 Fig. 3. Correlation of the angle of rotation, 0, with the absorption co- efficient, CL, and the refractive index, n. k = 2nv, where v is the frequency of the radiation. ,/-- J (k,H) 6 I w) --+k Fig. 2. Relation between the Faraday effect and the Zeeman effect. Theoretical According to Malus's law, the fraction of the light energy passing through the analyser is expressed as sin2#, where is the angle of rotation compared with the plane of the incident polarised light.As factors determining the transmitted intensity we must take into account the profile of the source radiation, p ( k ) , and the damping by atomic absorption. Thus, if the frequency is expressed in terms of angular velocity k = 2 m , the transmitted intensity at k, J ( k ) , can be expressed as follows: where ~ ( k ) is the absorption coefficient and L the length of the atomic vapour in the atomiser. In practice, we detect the integrated intensity, I : I = Jsp(k) I sin#@) j2exp[---a(k)L] dk . . . . - - (2) where s is the band pass of the monochromator. According to the analysis based on dispersion theory by Corney et aZ.7 and Buckingham and Stephens,lo the rotatory power is proportional to the number of atoms in the light beam, NL, where N is the number of atoms per cubic centimetre, and the oscillator strength, f .As stated above, the rotatory power arises from the difference in refractive index, so it includes the difference between the two dispersion functions, which is called the Faraday function, F(K,H). In addition, applying the Zeeman splitting to the dispersion functions, # is expressed as follows : .. .. * * (3) #W) = NLfF(k,H) . . where H is the magnetic field strength.1024 KITAGAWA et d. : APPLICATION OF THE FARADAY EFFECT TO THE Analyst, vd. 103 If the number of atoms in the light beam, NL, is small enough, we can make approxi- Consequently, the transmitted intensity mations such as sin + m y5 and exp[--cc(k)L] w 1. becomes As p(k) can be normalised as p'(k) = j(k)/Io, where lo is the source intensity integrated over the band pass of the monochromator, J,p(k)dk, the transmitted intensity is expressed as follows : ..* * (5) I cc (NL)2fV"M(H) . . .. .. where M ( H ) = Jsp'(k) [ F(k,H) I dk. Thus, the transmitted intensity is theoretically expected to be proportional to the square of the number of atoms in the light beam and to the source intensity, and to change as a function of the applied magnetic field strength and the line profile of the source radiation. In experimental calibration of the instrument, it is convenient to use an intensity normalised by the source intensity: T = I/I, or T = K(NL)2f2M(H)/Jsp(k)dk . . .. where K is a constant defined by the instrumental factors. If NL is so large that the atomic absorption is not negligible, the calibration should be made after a correction for the absorp- tion has been applied.The corrected transmitted intensity, Tc, is given by substituting the reduced source intensity for the source intensity lo: In order to apply this principle in practical analysis, we combined the Faraday configura- tion with an electrothermal atomiser, and studied several fundamental characteristics of the system for cadmium, namely the dependence of the transmitted intensity on the magnetic field strength and its proportionality to the square of the number of atoms in the light beam and to the source intensity. Use of the Faraday effect is expected to have the following advantages: 1. Those similar to atomic-fluorescence spectroscopy. As found in the above equations, the transmitted intensity increases proportionally to the source radiance, and the base-line noise is independent of the source noise.2. Unless depolarisation occurs, the light scattered by non-atomic species is blocked by the analyser. Therefore, it has the ability to remove the positive background signal, which is a significant problem in non-dispersive atomic-fluorescence spectroscopy. 3. As the optical system is aligned, it is easy to measure the source radiation. Experimental It consists of a light source, polariser, electromagnet , atomiser, analyser, monochromator, photomultiplier tube (RCA 1P28), power supply for the atomiser, d.c. amplifier (Sanei Sokki Co.), power supply for the coils of the electromagnet, power supply for the light source, flow controller for argon gas and strip-chart recorder.A cadmium hollow-cathode lamp (Hitachi) and a cadmium electrodeless discharge lamp (EMI) were used as light sources. The former was operated in the d.c. mode and the latter mounted in a tuned cavity14 through which microwave power of 2450 MHz was supplied. The electromagnet generated a maximum field strength of 17 kG between the pole pieces 10 mm apart with a coil current of 10 A. The pole pieces have a diameter of 20 mm and a central aperture of 3 mm diameter. The strength of magnetic field was calibrated against the coil current by using a gauss meter (Yokokawa Co.). Glan-type prisms were employed as polariser and analyser (Karl Lambrecht Co.). A prism monochromator was chosen for the dispersing element in order Fig.4 illustrates the apparatus constructed for this work.October, 1978 TRACE DETERMINATION OF CADMIUM BY ATOMIC SPECTROSCOPY 1025 to prevent the polarisation that can arise with some grating monochromators. The slit width of the mckochromator was 2 mm, giving a band $ass <f 2.5 nm. due to black-body radiation from the atomiser was negligibly small. The photocurrent A H Fig. 4. Schematic diagram of the optical system: A, microwave power supply for the electrodeless discharge lamp ; B, microwave cavity ; C, electrodeless discharge lamp ; D, lens ; E, polariser ; F, insulator ; G, transformer; H, variable transformer; I, pole piece; J, graphite tube atomiser; K, graphite cone; L, solenoid coil; M, analyser; N, gas inlet; 0, power supply for the solenoid coil; P, chamber; Q, prism mono- chromator; R, stabiliser for high-voltage supply for the photomultiplier tube; s, d.c.amplifier; T, strip- chart recorder; and U, photomultiplier tube. Two types of electrothermal atomiser were located in turn between the pole pieces. One was an M-shaped strip atomiser of tungsten, 2 mm in width,15 and the other was a graphite tube atomiser of 3 mm i.d., 5 mm 0.d. and 18 mm long. The strip atomiser was attached to brass rod electrodes 4 mm in diameter, through which the atomiser was supplied with electric current. An aliquot of the sample solution was placed on the bottom of the M-shape. The incident beam was focused on the atomic vapour. The graphite tube atomiser was fitted to the pole pieces of the electromagnet through graphite cones, which served as thermal insulators, as shown in Fig.5. The atomiser was supplied with electric current through the pole pieces. The sample solution was dispersed on to the bottom of the graphite tube with a syringe through a sampling port of diameter 1 mm in the middle of the lateral face of the tube. Argon gas was passed into the chamber and the graphite tube for purging at flow-rates of 1 1 min-l and 50 ml min-l, respectively. The latter plays an important role in reducing the contamination on the inside surface of the graphite tube by the atomic vapour. The flow of argon gas carries the atomic vapour generated by electrical heating out of the tube through the sampling port. As the temperature of the tube atomiser is highest at the port, the atomic vapour does not condense there.Thus, contamination at the edges of the tube can be prevented. Both atomisers were enclosed in a chamber of 70 mm i.d.1026 KITAGAWA et d. : APPLICATION OF THE FARADAY EFFECT TO THE Analyst, VOJ. 103 Graphite tube Graphite cone Fig. 6 . Schematic diagram of the graphite tube All dimensions atomiser fitted with graphite cones. are given in millimetres. A stock solution with a cadmium concentration of 1000 pg ml-1 was prepared by dissolving cadmium metal in 6 N hydrochloric acid purified by distillation. Standard solutions were prepared by dilution and the acidity was adjusted to 0.1 N. Sample solution (5 pl) was removed with a microsyringe (SGE), dried at 120 "C for 1 min and atomised at an equilibrium temperature of 1000 "C for the strip atomiser and 1800 "C for the carbon tube atomiser. Because of the large capacity of the latter, its equilibrium temperature was chosen to be higher than that of the strip atomiser.A cylindrical air - acetylene flame was also examined for comparison purposes. A burner was constructed by fitting a stainless- steel tube of 6 mm i.d. to a mounting block of 30 mm 0.d. The burner was connected to a conventional nebuliser for atomic-absorption spectroscopy and operated at flow-rates of 7 and 2 1 min-l for air and acetylene, respectively. The Cd I line at 228.8 nm was chosen as the analytical line. Results and Discussion Dependence of Transmitted Intensity on Magnetic Field Strength Fig. 6 shows the dependence of the transmitted intensity ( T ) on magnetic field strength.The curve A presents the results obtained with the hollow-cathode lamp (discharge current 10 mA) and curve B those for the electrodeless discharge lamps (100 W; reflected power <5 W). These curves have a maximum value at a field strength of about 3-5 kG, which we interpret as follows. When no magnetic field is applied to the atomic vapour, the frequency of the (T+ and (T- Zeeman components are the same, so line crossings occur. For this reason, the source radiation is not transmitted by the optical system. As the field strength is increased, the rotatory power is enhanced owing to the change in the relative velocity of the of and (T- components, corresponding to the change in the term $(K,H) in equation (2). Thus, an increasing amount of source radiation is allowed to pass through the optical system.How- ever, too strong a field causes such a large displacement of the cr components that the source line cannot superimpose upon them. In this event, there is no region of frequency in which the source radiation is affected by the influence of the (T components, which leads to a decrease in the transmitted intensity. This corresponds to a decrease in the product p(R) [~in#@,If)]~ in equation (3). Thus, the curve presents a maximum value, depending on the profile of the source radiation. On comparing curves A and B, it is evident that the maximum of the latter lies at a higher field strength and that its profile is broader than that of the former. Recalling the superimposition effect, the results suggest that the line profile of the radiation from the electrodeless discharge lamp is broader than that from the hollow-cathode lamp under the given conditions.This was also examined by calculating curves C and D according to equation (2). The parameters for curve C are as follows: the width of the source line was 0.2 pm, that of the absorption line 1.8 pm, the Doppler width of the absorption line 2.3 x lo0 rad s-l and the number of atoms 5 x lo7 cm-2. The parameters for curve D were the same as those for curve C, except that the width of the source radiation was larger (0.5 pm).October, 1978 TRACE DETERMINATION OF CADMIUM BY ATOMIC SPECTROSCOPY 1027 2 4 6 8 Magnetic field strength/kG ? Fig. 6. Dependence of the transmitted intensity ( T ) on the magnetic field strength: A, observed curve with the hollow-cathode lamp ; B, with the electrodeless discharge lamp; C and D, the calculated curves.These calculated curves also indicate that a broader source line gives a broader curve and that the curve maximum shifts to a stronger field. Another factor that determines the curve profile is line crossing. The effect of hyperfine structure on line crossing was taken into account and the Zeeman splitting pattern estimated for the analytical line at 228.8 nm. The natural abundance of cadmium is 0.875y0 of 108Cd, 12.39% of IloCd, 12.75% of ll1Cd, 24.07% of l12Cd, 12.2670 of 113Cd, 28.86% of l14Cd and 9.58% of 116Cd. The splitting pattern can be readily calculated for the isotopes of even relative atomic mass as they possess no nuclear spin.However, the situation is complicated for those of odd relative atomic mass because their spin momenta contribute to the Zeeman splitting. In weak fields, the nuclear spin momentum, I, is coupled with the electronic momentum, J, to yield a resultant momen- tum, F, leading to 2F + 1 Zeeman levels in the magnetic field, where F is the total momentum quantum number. On the other hand, in a strong field, they cannot be coupled together owing to the Back - Goudsmit effect,l6 but independently yield their magnetic momenta. In practice, an intermediate case exists between the two extremes. Fig. 7 shows the calculated splitting pattern for the case based on the perturbation method.17 It is expected from Fig. 7 that a line crossing may occur at a field strength near 0.2 kG.However, contrary to expectation, no minima can be found on the curves in Fig. 6. Presumably because of Lorentz broadening of the absorption line at atmospheric pressure, the zero-field line crossings cannot be separated from the line crossings in question. In addition, because the natural abundance of the isotopes of odd relative atomic mass is only a quarter of the total, they have no significant effect on the shape of the curve. Another problem concerned with line crossings is created by isotope shifts.8 If any isotope shift exists with a displacement comparable to the broadening of the absorption line, the fall might be found on the observed curve. The results suggest that such a large shift does not occur in the cadmium line at 228.8 nm. Calibration Graphs and Atomic Fluorescence amount of cadmium in the atomiser, i.e., the calibration graphs.Fig. 8 shows the relationship between the square root of the transmitted intensity and the The lower graph (un-1028 KITAGAWA et at. : APPLICATION OF THE FARADAY EFFECT TO THE A%a&.d, VOl. 103 0 5 10 Magnetic field strength/kG Fig. 7. Zeeman splitting pattern calculated for cadmium isotopes with the nuclear spin quantum numbers, I = 4 (A) and I = 0 (B). corrected) bends at high cadmium concentrations. The 1/T, graph is the corrected graph defined by equation (7), which includes the correction for atomic absorption. The atomic absorption was measured by setting the polariser and the analyser in a parallel configuration. The results indicate that the transmitted intensity is proportional to the square of the number of atoms.The relationship is important in measuring the phenomena observed in this work. From the above theoretical discussion, only the rotatory dispersion induced under the magnetic field is taken into account. Atomic fluorescence of the Zeeman components might contribute to the phenomena observed. If the fluorescent radiation yields plane-polarised light synthesised in a fashion similar to that in the Faraday effect, the expression for the detected light should be Two cases are possible. where y is the quantum yield. The first term corresponds to the Faraday effect and the second to rotation via atomic fluorescence. If the number of atoms is small enough, the second term on the right-hand side can be reduced to the expression I , = constant x (NL)3f3y10 (under a constant field).Thus, if the fluorescence has an important role, the transmitted intensity should change as a function C,(NL), + C,(NL)3 under a constant field, where C, and C, are constants. The results indicate that the transmitted intensity is almost proportional to (NL),, in contrast to the above expression. The other case, which may be the more likely, is partial depolarisation by inelastic collision between photons and atoms. For this reason, the plane-polarised radiation may become elliptically polarised and could pass through the analyser. However, this process could also occur under zero magnetic field. The results (Fig. 6) indicate that no radiation is detected under such conditions and it therefore appears that either the fluorescent energy is negligible compared with that due to the Faraday effect, or the plane of polarisation of the fluorescent radiation is the same as that of the exciting radiation and hence is blocked by the analyser.October, 1978 TRACE DETERMINATION OF CADMIUM BY ATOMIC SPECTROSCOPY 1029 This discussion is valid for small N , but under conditions when the number of atoms in the light beam is large the fluorescence could predominate.In the future, a detailed examina- tion of this question will be made by checking a decay curve, using a temporal disperser of picosecond response. Proportionality of the Transmitted Intensity to the Source Radiance Fig. 9 shows the dependence of the transmitted intensity on the intensity of the source radiation.Contrary to the prediction by equation (5), the resulting graph is not rectilinear but is concave a t high source radiance, probably because the increase in lamp current causes an increase in line width. As the line width of the source radiation increases, the convolution term p(k) I sin$ I2 in equation (2) becomes larger, leading to an increase in the transmitted intensity. Consequently, it could be suggested that a source radiation of higher band width is more suitable for this technique. However, theoretical calculations using equation (2) predict that too broad a line causes an increase in the rotatory power of molecular species and thus results in an unfavourable background. This was verified by introducing cigarette smoke into the graphite tube atomiser and using a D, lamp as the light source.Macaluso and Corbino5 also demonstrated that magneto-optical rotation of a continuum source occurred through magnetised diatomic sodium (Na,) as well as through the atomic vapour of sodium. Mass of cadrnium/pg t- Id 2 4 1 0 2 4 6 Source intensity/arbitrary units Fig. 9. Dependence of the transmitted intensity on the source intensity. The figures on the curve indicate the discharge v Fig. 8. Calibration graphs : 1/T, uncorrected for the lamp current in milliamperes' atomic absorption and d T c , corrected. Elimination of Background Cigarette smoke was introduced into the graphite tube atomiser with the flow of argon in order to examine the ability of the system to overcome background effects. The source radiation was decreased to half of the incident power, but no background magnetic rotation signal was detected.In this technique, the depolarisation of the incident radiation through the scattering process is significant for the background. However, in practice, it was found that the transmitted radiation was negligible. When 5 pl of 25 mg ml-1 sodium chloride solution was introduced in order to generate background, the result was the same. Thus, this technique proved to be insensitive to those background components of scattered light from which conventional atomic-absorption or atomic-fluorescence spectrometers may suffer. In this technique, there is a problem associated with a loss of incident power and transmitted energy attributable to background components, which leads to a negative interference.Nevertheless, the correction method, using Tc, enables us to overcome the interference. The operational circuits for a correction system will be described elsewhere.1030 KITAGAWA, SHIGEYASU AND TAKEUCHI Detection Limit and Reproducibility On the basis of a signal to noise ratio of 2, the cylindrical flame atomiser gave a detection limit for cadmium of 1 pgml-l, the tungsten strip atomiser with the hollow-cathode lamp 5 x 10-12 g and the graphite tube atomiser with the electrodeless discharge lamp 5 x 10-13 g, the discharge current for the hollow-cathode lamp being 15 mA and the microwave power for the electrodeless discharge lamp 120 W. In the flame atomiser, the number of atoms in the light beam is much smaller than that in the electrothermal atomisers and therefore, owing to the dependence of the transmitted intensity on the number of atoms, i.e., I = constant x (NL)2, the latter atomiser is the more suitable for the present technique, as it can produce a pulse of condensed atomic cloud in the light beam.Although the tungsten strip atomiser had the advantage of less contamination, it gave less reproducible signals (the relative standard deviation was 10%) than the graphite tube atomiser (4%). Finally, it should be emphasised that the present technique is applicable to trace analyses for elements other than cadmium and applications to copper and lead are being developed. This work was supported by a grant from the Ministry of Education of Japan. We thank Dr. J. B. Dawson and Professor T. S. West for helpful discussions and advice. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 1032 Winefordner, J. D., and Vickers, T. J., Analyt. Chem., 1964, 36, 161. Dagnall, R. M., Thompson, K. C., and West, T. S., Analytica Chim. Acta, 1966, 36, 269. Marshall, G. B., and West, T. S., Analytica Chim. Acta, 1970, 51, 179. Kitagawa, K., and Takeuchi, T., Analytica Chim. Acta, 1972, 60, 309. Macaluso, D., and Corbino, 0. M., C. R. Hebd. Skuanc. Acad. Sci., Paris, 1898, 127, 548. Ladenburg, R., Artnln Phys., 1912, 38, 249. Corney, A., Kibble, B. P., and Series, G. W., PIOG. R. Soc., 1966, A248, 701. Hackett, R. Q., and Series, G. W., Opt. Cornmun., 1970, 2, 93. Church, D. A., and Hadeishi, T., Phys. Rev., 1973, AS, 1864. Buckingham, A. D., and Stephens, P. J., A. Rev. Phys. Chem., 1971, 22, 259, Wong, K. P., J . Chem. Educ., 1975, 52, A83. Wood, R. W., “Physical Optics,’’ Third Edition, Dover, New York, 1961. Ito, M., Murayama, S., Kayama, K., and Yamamoto, M., Spectrochim. Ada, 1977, 32B, 347. Broida, H. P., and Chapman, M. W., Analyt. Chem., 1958, 30, 2049. NAKASHIMA : FLOTATION AND AAS DETERMINATION OF AnaZyst, VoZ. 103 A J

ISSN:0003-2654    

DOI:10.1039/AN9780301021

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, October, 1978, Vol. 103, pp. 1031-1036 1031 Flotation of Su b-microgram Amounts of Arsenic Coprecipitated with Iron(ll1) Hydroxide from Natural Waters and Determination of Arsenic by Atomic- absorption Spectrophotometry Following Hydride Generation Susumu Nakashima Institute for Agricultural and Biological Sciences, Okayama University, Kurashiki-shi 7 10, Japan A method is described for the flotation and determination of sub-microgram levels of arsenic in water. A sub-microgram amount of arsenic(II1, V) in a 500-ml sample of water is coprecipitated with iron(II1) hydroxide a t pH 8-9. The precipitate is floated with the aid of sodium oleate and small air bubbles, then separated and dissolved in 5 M hydrochloric acid, and finally the arsenic content is determined by generation of arsine using sodium tetrahydro- borate(II1) as a reductant and atomic-absorption spectrophotometry with a long absorption cell (60 x 1.2 cm id.).This separation technique has been successfully applied to the determination of sub-microgram amounts of arsenic(II1, V) in natural waters. Keywords : Arsenic determination ; fEotation ; natural waters ; atomic- absorption spectrophotometry ; hydride generation Arsenic probably exists at the 1 pgl-1 level in natural water samples. At such concentra- tions a precise direct determination is impracticable even by atomic-absorption spectro- photometry of arsine, which has a high sensitivity.1-4 Accordingly, the arsenic must be concentrated from the water prior to determination. Coprecipitation with iron(II1) hydroxide is commonly used as a pre-concentration technique for the determination of arsenic in water.5 This bulky amorphous precipitate, however, is difficult to filter, and centrifugation is cumbersome for larger volumes.Therefore, a flotation technique6.7 in which the precipitate of iron(II1) hydroxide is floated with the aid of sodium oleate and small air bubbles (0.1-0.5 mm diameter) has been used for the pre-concentration of arsenic. The precipitate is readily separated from the mother liquor and then dissolved in hydrochloric acid for the atomic-absorption spectro- photometry of arsine using sodium tetrahydroborate(II1) as a reductant. Various para- meters, such as the pH of the solution, the amounts of iron(II1) and surfactant added, stirring time, sample volume and foreign ions, have been investigated in order to obtain optimum conditions for the flotation and determination of the arsenic.For the atomic- absorption determination of arsenic a technique involving a long absorption cell and the argon - hydrogen flame system has been used. The proposed method is simple and rapid, and applicable to the determination of arsenic a t the low parts per billion (109) level in water. Experiment a1 Apparatus A Nippon Jarrell-Ash, Model AA-1, Mark I1 atomic-absorption spectrophotometer equipped with a Westinghouse arsenic hollow-cathode lamp and a custom-made silica absorption cell (60 x 1.2 cm i.d.) was used with a Beckman burner supplied with argon and hydrogen. The apparatus used for hydride generation was a modified Nippon Jarrell-Ash, Model ASD-1 A, arsenic measurement unit coupled to a custom-made hydride generating cell approximately 40 ml in volume.A schematic diagram of the analytical system is illustrated in Fig. 1.1032 NAKASHIMA : FLOTATION AND AAS DETERMINATION OF APzaZyst, VoZ. 103 Arsenic hollow- cathode lamp / Flame C K- Fig. 1. Schematic diagram of analytical system. A, Light from hollow- cathode lamp; B, Beckman burner; C,, silica absorption tube (60 x 1.2 cm i d . ) ; D, outlet from tube; E, argon gas flow containing arsine; F, by-pass; G, four-way stopcock, alternative gas passages shown by broken lines; H, gas flow with valve in sweep position; I, argon supply; J, flow meter; K, sodium tetrahydroborate(II1) solution; L, hydride generating cell ; M, sample solution; and N, syringe. All pH readings were made with a Hitachi-Horiba, Model M-5, pH meter together with a The flotation and separation apparatus was similar to that described by Mizuike and The flotation cell, shown in Fig.2, was a glass cylinder, 40 x 6.5 cm i.d., The air that served as the inert gas was supplied by a Nippon Jarrell-Ash, Model combined glass electrode. co-workers.6:' which was fitted with a sintered-glass filter (No. 4) to generate small bubbles. AMD-B1, air pump unit. Reagents All reagents were of analytical-reagent grade except for sodium oleate. Aqueous reagents were prepared in de-ionised, distilled water. Arsenic standard solutions were freshly prepared by diluting stock solutions before use. Dissolve 1.321 g of diarsenic trioxide in a mini- mum amount of 20% m/V sodium hydroxide solution, acidify with 5 M hydrochloric acid and dilute to 1000 ml with water.Dissolve 4.165 g of sodium arsenate (Na,HAsO,.- 7H,O) in 1000 ml of water. Arsenic(II1) stock solution, 1 mg ml-1. Arse.nic(V) stock solution, 1 mg ml-1.October, 1978 SUB-MICROGRAM AMOUNTS OF ARSENIC IN NATURAL WATERS 1033 Iron(I11) solution, 5 mg ml-l. Dissolve 43.17 g of ammonium iron(II1) sulphate [Fe2(SO,),(NH~),S0,.24H,O] in water and dilute to 1000 ml in 1 M hydrochloric acid with hydrochloric acid and water. Sodium oleate solution, 1 mg ml-l. Dissolve sodium oleate (powder, extra-pure reagent, Wako Pure Chemicals Co.) in 99.5% V/V ethanol with magnetic stirring. Sodium tetra~ydroborate(III) solution, 5% m/V. Dissolve sodium tetrahydroborate(II1) in water.1 Foam layer containing iron ( 1 I I 1 hydroxide and arsenic Sintered-glass disc, 'porosity 4 Fig, 2. Flotation cell for pre-concentration of arsenic. 0 0.02 0.04 0.06 0.08 0.10 Arsenic content/pg Fig. 3. Typical calibration graph for determination of arsenic in 1 ml of solution. Recommended Procedure for the Determination of Arsenic Transfer 1 ml of freshly prepared 5% m/V sodium tetrahydroborate(II1) solution into an arsine-generating cell and attach the cell to the apparatus. Insert the needle of a plastic syringe containing 1 ml of sample solution that contains less than 0.10 pg of arsenic through the side-arm seal of the cell. Turn the four-way stopcock of the apparatus to the sweep position in order to introduce argon into the system and inject the sample into the cell. Sweep the arsine thus generated into the long absorption cell with the argon so that it is atomised in the argon - hydrogen flame and record the absorption signal on a recorder.Return the stopcock to the by-pass position. Rinse the cell carefully with distilled water and re-charge with sodium tetrahydroborate(II1) solution ready for the next sample. If the concentration of arsenic exceeds 0.10 pg rnl-l, dilute the solution further, adjusting the hydrochloric acid and iron(II1) concentrations accordingly. Construct a calibration graph using 5 M hydrochloric acid solutions containing 1 mg ml-1 of iron(II1) and 04.10 pg ml-l of arsenic(II1). A typical calibration graph for arsenic is illustrated in Fig. 3, which is linear up to 0.10 pg of arsenic.The same result was obtained by using an arsenic(V) solution containing 1 mg ml-l of iron(II1) as a standard. The coefficient of variation based on 10 replicate runs of a solution containing 0.05pgml-1 of arsenic was within 1.5%. The atomic-absorption equipment was operated under the following conditions : wave- length, 193.7 nm; lamp current, 16 mA; gas flow-rates, argon 1.5, hydrogen 1.5 and auxiliary argon 6 1 min-l; slit (spectral band width), 1 nm. The 60-cm tube system described in this paper cannot be used with most commercial atomic-absorption spectrophotometer units. However, the arsenic in a sample solution can also be determined by atomic-absorption spectrophotometry, using an arsine generation - electrically or flame-heated silica tube.Moreover, nitrogen can be substituted for the more expensive argon.1034 NAKASHIMA : FLOTATION AND AAS DETERMINATION OF Analyst, VOt?. 103 Procedure Recommended for the Flotation Step Place 500 ml of the water sample in a 500-ml beaker and add 2 ml of iron(II1) solution and 1 ml of sodium oleate solution. Adjust the pH to 8-9 with aqueous ammonia solution (5 and 0.1 M) in order to precipitate iron(II1) hydroxide, while stirring magnetically, and stir the solution for 15 min. Transfer the contents of the beaker (excluding the stirring bar) to a flotation cell and wash the residue in the beaker into the cell by using two small portions of water. Pass air at a flow-rate of 50mlmin-l from the lower end of the cell for about 30s, in order to obtain complete mixing and flotation of the precipitate.Suck off the mother liquor through the sintered-glass disc and wash the precipitate with 30 ml of water. Add 5 ml of 5 M hydrochloric acid to the cell t o dissolve the precipitate, collect the filtrate by suction in a 10-ml calibrated flask, wash the sintered-glass disc with hydrochloric acid, add the washings to the flask and dilute to the mark with 5 M hydrochloric acid. Results and Discussion Determination of the Optimum pH for Collection of Arsenic The effect of the pH of the 500ml of sollution containing 0.5pg of arsenic(II1, V), 10 mg of iron(II1) and 1.0 mg of sodium oleate on the coprecipitation of arsenic was studied. Hydrochloric acid and aqueous ammonia solution were used to adjust the pH to values within the range 4-10.As a result satisfactory recoveries of both trivalent and pentavalent states of arsenic were obtained over this range. The most stable layer of surface foam supporting the precipitate of iron(II1) hydroxide was formed within the pH range 7-9.5; the pH range of 8-9 was used throughout the work. At a pH below about 6.5, a stable surface-foam layer was obtained by using sodium lauryl sulphate as a surfactant. Determination of Optimum Amounts of Iron( 111) and Surfactant added to the solution. iron(II1). gated. oleate and 1.0 mg in 500 ml of solution was adopted in further work. Table I shows the percentage of arsenic recovered as a function of the amount of iron(II1) Quantitative recoveries of arsenic were obtained above 2.5mg of The amount of sodium oleate required for complete flotation of the precipitate was investi- Quantitative recoveries of arsenic were obtained between 0.2 and 4.0 mg of sodium In this work 10 mg of iron(II1) were added to 500 ml of the solution.TABLE I RELATIONSHIP BETWEEN AMOUNT OF IRON(XII) ADDED AND RECOVERY OF ARSENIC Solution contained 0.6 pg of As(II1) and 1 mg of sodium oleate; pH, 8-9; volume, 600 ml. Fe(II1) added/mg . . 2.6 6.0 7.5 10.0 16.0 20.0 As recovered, yo . . 98 100 101 100 98 99 Stirring Time are shown in Table 11. stirring for 15 min was found to be best. The relation between stirring time and recovery of arsenic was investigated. The results Coprecipitation was quantitative over the range 540 min and TABLE I1 RELATIONSHIP BETWEEN STIRRING TIME AND RECOVERY OF ARSENIC Solution contained 0.5 p g of As(II1) and 10 mg of Fe(II1) ; pH, 8-9; volume, 500 ml.1 B 20 25 30 40 101 100 98 99 97 Stirring timelmin . . .. 5 As recovered, yo . . .. 99 101October, 1978 SUB-MICROGRAM AMOUNTS OF ARSENIC IN NATURAL WATERS 1035 Solution Volume The effect of variation of the volume of solution containing 0.5 pg of arsenic(III), 10 mg of iron(II1) and 1.0 mg of surfactant was studied. Arsenic was recovered quantitatively from volumes of up to at least 1000 ml. Taking the arsenic content (1 pg 1-1 level) in natural waters and the sensitivity of analytical equipment into account, 500ml of water sample was considered to be a suitable volume. Effect of Foreign Ions By following the recommended procedure, the effect of various ions on the separation and determination of arsenic was investigated.Table I11 shows permissible amounts of foreign ions for the determination of 0.5pg of arsenic(II1) in 500ml of solution with 10mg of iron(II1) added. As can be seen, the determination of arsenic is scarcely affected by the amounts of foreign ions normally present in natural waters. Of these ions, hydride-forming elements such as selenium(IV), antimony(II1) and tin(1V) are coprecipitated with iron(II1) hydroxide in the same way as arsenic and would have a relatively strong effect on the arsine generation. However, these ions occur at extremely low levels in natural waters in corn- parison with arsenic. Therefore, this method can be employed for the determination of arsenic in natural waters without any interference from co-existing ions.TABLE I11 PERMISSIBLE AMOUNTS OF FOREIGN IONS FOR DETERMINATION OF ARSENIC Solution contained 0.5 pg of As(II1) and 10 mg of Fe(II1) ; volume, 600 ml. Ion Na+ K+ Caa+ c1- NOs- SiOi- Sr2+ Ba2+ Mg2+ sot- Limit 20 000 20 000 20 000 20 000 20 000 20 000 20 000 20 000 2 000 2 000 [Ion1 / [As1 Ion Cd2+ ha+ Mna+ Cr3+ Cr6+ Mos+ Pb2+ Hga+ See+ ~ 1 3 + Limit 2 000 2 000 2 000 2 000 2 000 2 000 2 000 2 000 2 000 2 000 [Ion1 / [As1 Ion cog+ Poi- V=+ Bi3+ Te4+ Ni2+ cu2+ Sb3+ Sn4+ Se4+ Limit 2 000 2 000 1000 400 400 400 200 00 60 10 [Ion31 [As1 Recovery of Arsenic Solutions (500 ml) at pH 8-9 containing 10 mg of iron(III), 1.0 rng of sodium oleate and 0.2, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, 7.0 and 10.0 pg of arsenic(II1, V) were analysed by the recom- mended procedure.Recoveries of the arsenic that had been added were greater than 95% in all instances. Recommended conditions, therefore, appear to be optimum for 500-ml volumes of solutions containing up to at least 10 pg of arsenic. The relative standard deviation of 10 analyses of solutions containing 0.5 pg of arsenic(II1) per 500 ml was less than 3%. Determination of Trace Amounts of Arsenic in Natural Water The analyses were carried out on 500-ml aliquots of clear, uncontaminated water (river, well and tap waters), filtered through 0.45-pm Millipore filters after addition of 10 ml of hydrochloric acid per 1000 ml of sample immediately after collection. Table IV presents analytical results for natural water samples by the procedure recom- mended. A known amount of arsenic(II1) was added to water samples and the recovery was measured.In these instances the recoveries were from 94 to 103%. The arsenic concentrations in the waters analysed were 1.59, 0.63 and 0.71 pg 1-1 for Takahashi River water (Okayama Prefecture, Japan), well water and laboratory tap water, respectively.1036 NAKASHLIMA TABLE IV DETERMINATION OF ARSENIC IN NATURAL WATER SAMPLES Volume of sample, 500 ml. Amount of arsenic/ pg -- Sample Added Found Recovered River water . . . . None 0.795 0.200 1.001 0.206 0.400 1.200 0.405 Well water . . .. None 0.315 0.200 0.502 0.187 0.400 0.700 0.385 Tap water . . . . None 0.357 0.200 0.560 0.203 0.400 0.734 0.377 Arsenic in 1.59 Recovery, % sample/pg 1-1 103 101 0.63 94 96 0.71 102 94 Conclusions The flotation of sub-microgram amounts of arsenic coprecipitated with iron(II1) hydroxide is useful as a pre-concentration technique for extraction of arsenic from a large volume of water, and subsequent atomic-absorption spectrophotometry in a long absorption cell of the arsine generated from the arsenic is an accurate method for the determination of arsenic. The method offers a rapid and precise procedure for the routine determination of arsenic in natural fresh waters.This method is also applicable to the flotation of arsenic in artificial sea water. With suitable modifications, the procedure can be applied to deter- mining sub-microgram levels of metals, such as tin(IV), antimony(II1, V) and selenium(IV), which readily form volatile hydrides in water. The author thanks Professor Atsushi Mizuike and Dr. Masataka Hiraide of Nagoya Uni- versity for their helpful advice on the flotation technique, and Assistant Professor Fuji Morii of Okayama University for useful discussions. References 1. 2. 3. 4. 5. 6. 7. Yamamoto, Y., Kumamaru, T., Hayashi, Y., and Kamada, T., Bunseki Kagaku, 1973, 22, 876. Yamamoto, Y., Kumamaru, T., Edo, T., and Takemoto, J., Bunseki Kagaku, 1976, 25, 770. Thompson, K. C., and Thomerson, D. R., Analyst, 1974, 99, 595. Goulden, P. D., and Brooksbank, P., A7zalyt. Chenz., 1974, 46, 1431. Sugawara, K., Tanaka, M., and Kanamori, S., Bull. Cltem. SOC. Japan, 1956, 29, 670. Hiraide, M., and Mizuike, A., Bunseki Kagaku, :1977, 26, 47. Hiraide, M., Yoshida, Y., and Mizuike, A., Anaiytica Chim. Acta, 1976, 81, 185. Received February 9th, 1978 Accepted April 12th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9780301031

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, October, 1978, Vol. 103,pp. 1037-1045 1037 Extraction - Spectrophotometric Determination of Tin in Lead and Lead-based A l l ~ y s with A. Sanz-Medel and A. M. Gutierrez Carreras Departamento de Quimica Analitica, Facultad de Ciencias Quimicas, Universidad Coqblutense, Ciudad Universitaria, Madrid-3, Spain A method is described for the direct spectrophotometric determination of micro-amounts of tin, by extraction into a chloroform solution of 5,7-dichloro- quinolin-8-01 from a solution containing sulphuric acid. The influence of the different experimental parameters on the formation and extraction of the complex were studied and optimum conditions for the determination of tin were established. The precision of the extraction - spectrophotometric procedure, expressed in terms of relative standard deviation, was 1.4%.It is shown that two different complexes (with Am=. 403 or 390nm) can be extracted into the chloroform, depending on the presence or absence of chloride and on the pH of the solution. The method has been tested on six standard lead-based samples with tin contents ranging from 0.05 to 1%. The average relative error (mean error) of the results lies within the range f1.4y0, which shows that the accuracy is good and that systematic errors are absent. Keywords : Tin determination ; 5,7-dichloroquinolin-8-01 extraction ; spectro- photometry ; Eead-based alloys Most chromogenic reagents for the spectrophotometric determination of tin (dithio1,l phenylfiuorone,2 catechol violet ,3 haematoxylin, quercetin, gallein, e t ~ .~ ) lack selectivity. As a result, the spectrophotometric determination of tin in metallic samples usually requires a preliminary separation of the tin. For example, distillation of the bromide or chloride is frequently used,lp5 although this procedure is troublesome and not reproducible when micro- gram amounts of tin have to be separated. Solvent-extraction methods have been extensively used for this purpose, especially the extraction of tin(1V) into toluene as tin(1V) iodide, which gives very good recoveries and has been applied successfully as a preliminary step in the spectrophotometric determination of tin in steel.6 Other techniques used to achieve a preliminary separation are based on coprecipitation [e.g., with manganese( IV) oxide] and ion exchange.However, the advantage of liquid-liquid extraction is that it permits not only the separation of the desired constituent of the sample, or an increase in the sensitivity of the determination by a simple process of concentration, but also spectrophotometric determina- tion directly in the non-aqueous phase by selection of an organic reagent that forms coloured neutral chelate complexes with the cation, e.g., the extraction - spectrophotometric methods using quinolin-8-01 and its derivatives. Gentry and Sherrington7 reported that tin (IV) is extracted quantitatively with quinolin-8-01 in chloroform over the pH range 2.5-5.5. Owing to the poor selectivity of this procedure, Eberle and Lernel-8 recommended the extraction - spectrophotometric determination of tin(1V) with quinolin-8-01 at pH 0.85 in the presence of chloride ions.The use of this acidic medium made the extraction procedure much more selective. It is well known that substitution by negative radicals of the hydrogen atoms in the quinolinol nucleus can result in increased selectivity of the reagent, as was shown for the dihalogen derivatives.9 Several of these derivatives have been used for the gravimetric or spectrophotometric determination of cations.lO In particular, the 5,7-dibromo derivative has been investigated for the analytical separation a i d spectrophotometric determination of tin .11P Previous work by the present authors13 on the behaviour of five different dihalogen deriva- tives of quinolinol as reagents for niobium(V) showed that tin(1V) can also be extracted with an excess of the 5,7-dichloro derivative in chloroform from 2 M hydrochloric acid. This1038 SANZ-MEDEL AND CARRERAS : SPECTROPHOTOMETRIC DETERMINATION Analyst, Vol.103 acidity differs significantly from the optimum value of pH 1 given by Ruffl and Matsuo and Funayama12 using 5,7-dibrornoquinolin-8-01 and so could produce increased selectivity. In a more strongly acidic medium the tendency for the weaker metal complexes to be formed is minimised because the ionisation of the reagent is virtually suppressed. This fact, along with the better solubility of the dichloro derivative in chloroform,13 encouraged the investi- gation of the 5,7-dichloroquinolin-8-01 as an extraction - spectrophotometric reagent for tin(1V). A method for the selective spectrophotometric determination of tin(1V) by extracting it from 2 M sulphuric acid with 5,7-dichloroquinoliin-8-ol in chloroform and direct measure- ment of the absorbance of the organic extract at.403 nm has been developed. The method has been applied successfully to the determination of tin in lead and lead-based alloys. Experimental All of the reagents were of analytical-reagent grade. Tin(IV) stock solution, 1 000 pg ml-l. Reagents Dissolve 1 .OOO g of pure tin in 50 ml of concentrated When it has dissolved, add a further 150 ml of the acid and Cool this solution and dilute it Dilute standard tin(1V) solutions are prepared by diluting this stock solution with the 5,7-Dichloroquinolin-8-ol solution, 1 yo m/V in chloroform. Ammonium chloride solution, 2 M.Ammonia solution, 10 M. Nitric acid, 50% VlV. Tartaric acid solution, 50% mlV. Nitric acid - tartaric acid mixture. sulphuric acid with heating. carefully pour the solution into 500ml of distilled water. to 1 1 in a calibrated flask. necessaxy volume of 3.25 M sulphuric acid. Mix 100 in1 of 50% VlV nitric acid with the same volume of 50% m/V tartaric acid solution. Equipment Spectro9hotometers. A Beckman, Model DU, single beam, for absorbance measurements, and a Beckman, Model DK-2A, for automatic recording of the spectra, each with 10-mm glass cells. Metrohm, Model E-516, with glass electrode and saturated calomel reference electrode. p H meter. Separating funnels. Capacity 100 ml. Procedure Dissolution of lead samples Dissolve an appropriate amount (0.2-2g for samples with tin contents ranging from 3 to 0.3%) of lead by gently warming it with 20 ml of the nitric acid - tartaric acid mixture.Heat to fuming and boil until fumes are no longer evolved. Cool the solution and dilute to 100 ml with distilled water in a calibrated fla~Ik.1~ Determine the tin content as described under Gelzeral procedure. General procedure Pipette a portion of the sample solution containing 40-15Opg of tin into a separating funnel and add sufficient sulphuric acid to ensure a final concentration of it in the aqueous phase of 1.5-2 M (better selectivity is attained by using a concentration of 2 M, which can be produced by adding 6 - n ml of 3.25 M sulplhuric acid to n ml of the solution) and 1 ml of the ammonium chloride solution. Adjust the final volume of the aqueous phase to approxi- mately 10 ml with distilled water.Add 10 rnl of 1% 5,7-dichloroquinolin-8-01 solution in chloroform with gentle manual shaking to achieve a rapid distribution of the complexing reagent between the two phases, allow to stand for 5-10 min and finally extract the tin by shaking the funnel for 3 4 m i n and allowing tlhe phases to separate. Filter the chloroform solution through a dry Whatman No. 1 filter-paper and measure the absorbance at 403 nm against a blank prepared by extracting all of the reagents in the absence of sample.October, 1978 OF TIN IN LEAD AND ALLOYS WITH ~,7-DICHLOROQUINOLIN-8-OL 1039 Prepare a calibration graph by taking portions of dilute standard tin solutions containing 40, 50, 60, 70, 80, 100, 120 and 140 pg of this metal, and extracting each of these by the procedure described above.8 0.6 < 0.5 0.4 0.3 0.2 0.1 c a 0 - Results and Discussion Spectral Characteristics of Complex As can be seen from Fig. 1, the absorption spectrum of the complex tin(1V) - 5,7-dichloro- quinolin-8-01 extracted by following the General procedure exhibits a maximum at 403 nm when measured against a similar blank. However, a hypsochromic effect was observed when decreasing the hydrogen-ion concentration of the aqueous phase before the extraction step. Effect of pH on Extraction of Tin and Colour of Complex A set of experiments was carried out in order to study the influence of pH on the extrac- tion of tin. The results are given in Fig. 2 and demonstrate that when the pH was increased above pH 1 the maximum of the absorption spectrum started to shift towards shorter wavelengths, and hence from pH > 3 the maximum appeared at 390 nm.The magnitude of such a variation for different pH values is given in Table I. - - - - - - :::I 0.7 I I I Wavelengthhm Fig. 1. Absorption spectra of: A, tin(1V) - 6,7-dichloroquinolin-8-01 com- plex, extracted into chloroform, measured against chloroform as reference; B, reagent blank (extracted under the same conditions) against chloroform ; and C, tin( IV) - 5,7-dichloroquinolin-8-ol com- plex, extracted into chloroform, measured against reagent blank. 350 400 450 500 Wavelengthhm Fig. 2. with pH. Variation of the absorption spectra All of the tests were carried out by using 1OOpg of tin(IV), a 0.6-g total amount of ammonium chloride and 10 ml of the 1 yo 5,7-dichloroquinolin-8-01 solution in chloroform.The pH was adjusted with ammonia solution as the final step before shaking the separating funnel (this order of addition produces a higher reaction rate).13 As shown in Table I, the amount of tin(1V) extracted remained virtually constant (for the compound absorbing at 403 nm) down to pH 1 and the effect of higher acid concentrations wasinvestigated, but sulphuric acid was used because the use of hydrochloric acid for this purpose gave too high absorbances for the blanks. The results are plotted in Fig. 3, which shows that over the range 0.5-2 M in sulphuric acid the amount of tin(1V) extracted reaches a virtually constant value. A sulphuric acid concentration of 1 M was initially used to optimise the experimental conditions.1040 SANZ-MEDEL AND CARRERAS : SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 103 TABLE I EFFECT OF pH ON WAVELENGTH OF' MAXIMUM ABSORBANCE AND ABSORBANCE AT THAT WAVELENGTH PH 4.00 3.10 2.35 2.10 1.10 0.70 0.60 0.40 0.20 0.10 hmax.l nm 390 390 397 400 403 403 403 403 403 403 7- Sample:* 1.400 1.0001 0.7201 0.620 0.536; 0.520 0.517' 0.536 0.550 0.626 Absorbance Blank* 0.280 0.250 0.160 0.100 0.055 0.057 0.072 0.075 0.085 0.145 -- 7 Samplet 1.120 0.790 0.547 0.530 0.442 0.435 0.445 0.469 0.465 0.462 * Absorbance of sample and of blank measured against chloroform t Absorbance measured experimentally against blank as reference.as reference. Tin(1V) Extraction in the Absence of Chloride The effect of eliminating the ammonium chloride addition step in the General procedure was examined.The spectra of the organic e:xtracts were similar to that obtained at pH 2 3 with chloride ions present (Amax. 390 nm, as shown in Fig. 2). Tin(1V) was found to be extracted with 1% 5,7-dichloroquinolin-8-01 in chloroform in the total absence of chloride ions over the pH range 1-9, as shown in Fig. 4.. The molar absorptivity of the extracted compound (Amax. 390 nm) is approximately double that of the compound extracted in the presence of halide ions (Amax. 403 nm). All of the literature available claims that the presence of halide is necessary in order to achieve quantitative extraction of micro-amounts of tin(1V) with quinolin-8-ols or with dibromoquinolin-8-ol.12 Although extraction in the absence of halide shows better spectrophotometric sensitivity, the higher pH needed for the quantitative extraction (Fig.4) leads to a less selective extrac- tion procedure than that with chloride present. This finding was confirmed experimentally by interference tests. The method given in the General procedure was chosen for the analysis of the actual lead samples. Effect of Chloride Concentration on the Extraction of Tin(IV) The extraction of tin(1V) in the presence of increasing halide concentrations was investi- gated by following the General procedwe except that the sulphuric acid concentration was \ 4 5 0 1 2 3 Concentration of su I phu ric acid/M Fig. 3. Influence of the siilphuric acid concentration on the extraction of the tin(1V) complex.October, 1978 OF TIN I N LEAD AND ALLOYS WITH 5,7-DICHLOROQUINOLIN-8-OL 1041 0.5 $ 0.4 42 2 t m 0 0.3 0.2 0.1 0 PH Fig.4. Effect of pH on the extraction of the tin(1V) complex in the absence of chloride. 1 M (Table 11). Tin(1V) is extracted in the absence of chloride ions as a complex with Am=. at 390nm. As the chloride content of the aqueous phase increases, A,,,. moves towards longer wavelengths to reach the value of 403nm at a 1 0 - 2 ~ chloride concentration. A further increase in the halide concentration has no effect on the wavelength, although the absorbance increases slightly. A concentration of 2 x 10-1 M in chloride ions was selected as giving a sufficient excess to ensure the extraction of the maximum amount of tin as the complex with Am= 403 nm.TABLE I1 EFFECT OF CHLORIDE CONCENTRATION ON THE EXTRACTION O F TIN FROM 1 M SULPHURIC ACID Concentration Absorbance of ammonium Amax./ r A -l chlorideIM nm Sample* Blank* Samplef 0 390 0.740 0.160 0.630 10-3 395 0.600 0.055 0.550 10-2 403 0.610 0.020 0.495 10-1 403 0.555 0.030 0.525 1 403 0.620 0.100 0.510 * Absorbance of sample and of blank measured against chloroform t Absorbance measured experimentally against blank as reference. as reference. Rate of Extraction and Stability of Colour The plot of absorbance versus time of shaking the funnels reaches a constant, reproducible, maximum value after a relatively short time of about 3 min. The colour produced in the chloroform layer remained constant for at least 24 h against a similar blank when protected from exposure to direct sunlight. Effect of Reagent Concentration The influence of the concentration of 5,7-dichIoroquinolin-8-01 in the chloroform was studied by extracting 100 pg of tin from 1 M sulphuric acid in a single extraction step.The absorbances were measured at 403 nm and plotted against the concentration of 5,7-dichloro- quinolin-8-01 in chloroform (Fig. 5). The horizontal portion of the curve represents virtually 100yo extraction. Although the use of 0.75% of reagent is apparently sufficient for the proposed procedure, a 1% concentration was selected in order to ensure an adequate excess.1042 SANZ-MEDEL AND CARRERAS : SPECTROPHOTOMETRIC DETERMINATION Analyst, Yd. 103 Concentration of 5,7 - dichloroquinolin - 8 - 01, % Fig. 5. Effect of 5,7-dichlor~~quinolin-8-ol concentra- tion on the extraction of the tin(1V) complex in the presence of chloride.Nature of the Complex Published results8 on the composition of the complex formed in the extraction of tin(1V) with quinolin-8-01 (oxH) in chloroform in the presence of chloride ions showed that the probable composition of the complex is SnCl,(ox),. This formula is in agreement with the results reported by Hamaguchi et al. ,15 who determined large amounts of tin gravimetrically using quinolin-8-01 as the precipitating reagent. On the other hand, recently reported experi- mentsf6 using different anionic compounds, e g . , trichloroacetic acid instead of chloride ions, suggested the extraction of a mixed complex with a similar formula, Sn(ox),(CC1,C00),.An analogous relationship between tin anion and reagent was found by Matsuo and Funayama12 when extracting tin(1V) into carbon tetrachloride with 5,7-dibromoquinolin-8-01. As described above we have observed the presence of two different complexes when extracting tin( IV) with 5,7-dichloroquinolin-8-ol (C1,oxH) into chloroform, depending on the presence or absence of the halide anion. The analysis of the compound which had been obtained by following the procedure proposed in the present work (chloride present) suggests the formula SnC1, (C1,ox) ,. Precipitation experiments to establish the composition of the complex formed in the absence of halide, in which the SnCl,(Cl,ox), complex was isolated, resulted in a mixture of SnO,.utH,O and excess of reagent absorbed on the oxide.Continuous-variation and molar- ratio spectrophotometric methods using relatively concentrated tin(1V) solutions (2 x 10-3 M) and a pH of 2 for extraction in the absence of halide, led eventually to a ratio of tin to reagent of 1:3, which suggests the formula Sn(Cl,ox), for this complex, It therefore appears that at higher pH values the two monodentate halide ions are replaced by a third bidentate reagent molecule (see Fig. 2). Calibration Graph The solution of the complex obeys Beer’s law over a range of concentrations corresponding to 1-12 pg ml-l of tin in the organic phase. The sensitivity of the determination, expressed in terms of molar absorptivity at 403 nm, was 4.7 x lo3 1 mol-l cm-l. The reproducibility of the determination under the optimum conditions was established by carrying out 11 determinations on 100 pg of tin in a freshly prepared test solution.The set of absorbance values obtained was used to evaluate the precision of the method, which, expressed in terms of relative standard deviation, was found to be &1.4%. Interferences The effect of those potentially interfering elements that are present with tin in the lead samples was initially investigated by extracting 50 pg of tin from 1 M sulphuric acid in the presence of 500 pg of aluminium(III), copper(II), antimony(III), iron(III), lead(I1) or bismuth(II1). Aluminium, bismuth and lead did not interfere, even when present in greater excess, but iron, antimony and copper did interfere.October, 1978 OF TIN IN LEAD AND ALLOYS WITH ~,7-DICHLOROQUINOLIN-8-OL 1043 It was thought that there could be a gain in selectivity13 by increasing the acid concentra- tion as it was possible to extract tin from a more acidic medium without losing sensitivity (Fig.3) and all further tests were carried out using a 2 M sulphuric acid medium. Table I11 lists 15 foreign elements investigated under these conditions and their effect on the determination of 1OOpg of tin. As can be seen, the procedure is highly selective, iron(III), copper(I1) and antimony(II1) being the only interfering elements that formed a coloured extractable reaction product.! The use of masking agents to eliminate these interferences was investigated to see if it would be possible to overcome them without recourse to a preliminary separation technique.The masking agents investigated were fluoride, EDTA and tartaric acid. As shown in Table 111, fluoride and EDTA, when added to the aqueous phase, inhibited the tin(1V) extraction with the reagent and thus only tartaric acid could be used as masking agent. The concentration of tartaric acid in the acid solution could be as high as 1 M without affecting the determination of 100 pg of tin. In the presence of 1 M tartaric acid iron(II1) and antimony(II1) could be masked when present in small amounts (up to 200 pg), but the masking action was insufficient if the amount of interferent was greater than five times the tin(1V) concentration. TABLE I11 EFFECT OF SOME COMMON IONS ON THE DETERMINATION OF TIN BY THE GENERAL PROCEDURE Amount of 50 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 tin/ Clg Element or anion added A1 Pb Bi Ag Mn Zn Ni co Si Ti Sb Sb Fe c u Nitrate Tartrate Sulphate Chloride EDTA Fluoride Mg Compound A1(N03) ,.9H20 BiONO,.H,O MnSO, ZnSO, Pb (NO,) 2 AgNO, WNO3)2 Mg(NO3) ,*6H20 Co(NO3) a SiO, TiOSO, SbC1, CuS0,.5H20 53403 3 Pb(N03) 2 (CHOH.COOH) a H2S04 NH,C1 Na, .EDTA.2H20 NaF Amount tolerated/mg* 5.0 100 60 0.5 1.5 0.5 0.5 1.5 0.5 0.05 0.25 0.50 Interfered Interfered Interfered 0.05 M 1.00 M 2.00 M ' 1.00 M Interfered Interfered * The tolerances given correspond to the maximum amount of the foreign element or anion investigated in each instance. In order to eliminate interferences in such instances, a selective pre-washing step was devised, taking advantage of the fact that tin(1V) is not extracted from a 2 M sulphuric acid medium in the absence of chloride ions (Fig.4). By pre-extracting the sample solution in 2 M sulphuric acid, in the total absence of chloride ions, with a 1% 5,7-dichloroquinolin-8-01 solution in chloroform, a clear colourless extract is eventually obtained and large amounts of iron(II1) or copper(I1) can be removed. A single washing step eliminated the interference of up to 100 p g of copper(I1) and 100 pg of iron(II1) in the determination of 100 pg of tin(1V). The ammonium chloride was added after the washing step and tin(1V) was finally extracted with a fresh chloroform solution of the reagent by following the General procedure. Antimony(II1) showed similar extraction behaviour to tin(1V) and was extracted only in the presence of halide; it therefore could not be removed by the pre-washing technique.It was demonstrated that antimony(V) was not extracted under the General procedure con- ditions and thus the interference of antimony could be avoided by oxidising it to the non- interfering state before carrying out the extraction of tin(1V). On account of the practical1044 SANZ-MEDEL AND CARRERAS : SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 103 importance of antimony in commercial tin samples, a preliminary oxidation based on the use of sodium nitrite was tested for increasing antimony to tin ratios, the procedure being as follows: 0.5 g of sodium nitrite was added to the acidic sample solution containing the mixture of antimony(II1) and tin(1V).After boiling foir 10-15 min, 0.2 g of urea was added to destroy nitrogen oxides and excess of nitrite. After the solution had cooled, tin was deter- mined by following the general extraction - splectrophotometric procedure. The results obtained are given in Table IV and show that amounts of antimony of up to twenty times the tin content can be tolerated by this procedure. 0.326 0.490 ' 0.491 0.606 0.0941 ' 0.097 8 0.100 8 0.970 ' 0.984 1.00 0.078 ' TABLE IV r > ELIMINATION OF INTERFERENCE FROM: ANTIMONY BY A PRELIMINARY OXIDATION WITH S0:DIUM NITRITE Sb addedlpg 0 100 200 600 1000 2 000 6 000 10 000 Solution contained 1.00 pg of tin. Sb: Sn ratio 0 1: 1 2: 1 5: 1 10: 1 20: 1 60: 1 100: 1 Absorbance 0.412 0.412 0.415 0.425 0.420 0.430 0.455 0.600 Amount of Sn foundlpg 100.0 100.0 100.7 103.1 101.9 104.3 110.4 145.6 Determination of Tin in Lead-based Samples the determination of tin in six lead samples.different samples. The recommended extraction - spectrophotometric procedure was tested by applying it to The results are given in Table V along with the certified values of the tin content in the It can be seen that the accuracy of the method, considered on the basis TABLE V TIN CONTENTS OF CERTI:FIED LEAD SAMPLES Tin content, % I - Sample number I I1 I11 IV V VI Mass takenlg 1.678 4 1.767 8 1.636 4 1.020 3 0.945 6 1.214 0 2.429 6 2.030 8 2.788 0 0.487 6 0.602 6 0.647 1 1.250 4 1.136 3 1.243 7 2.001 0 1.802 6 1.601 1 Determined average 0.319 0.496 0.097 6 0.986 0.078 0.050 Relative Certified error, % 0.315* + 1.3 0.600* -1.0 0.loot - 2.0 1.oot - 1.5 0.079t -1.3 0.0621 -3.8 * Commercial lead alloy for pipes. (Composition : lead containing Bi, Ag, Cu, Sb, Fe, Zn, Cd in amounts ranging from 0.006% for Ag to 0.1% for Sb, in addition to the Sn contents specified above.) Certified by Instituto del Hierro y del Acero (CENIM), Spain. 7 Lead melted with the appropriate amount of pure tin, obtained from tin contents certified by Instituto del Hierro y del Acero (CENIM), Spain. $ Commercial lead from Peiiarroya (Spain).October, 1978 OF TIN I N LEAD AND ALLOYS WITH ~,7-DICHLOROQUINOLIN-8-OL 1045 of the calculated average relative error, was satisfactory for tin contents above 0.05y0. There was a variation within -+1.4%, corresponding to the mean relative error, for the five lead samples analysed that had tin contents above 0.05y0. For samples with tin contents below 0.05% the method gave low results, probably caused by adsorption of these small amounts of tin on the surface of the precipitated lead sulphate.As the precision of the extraction - spectrophotometric procedure itself was 1.4y0, as shown above, it can be seen that there are no systematic errors in the tin contents. When dissolving samples with high tin contents it is recommended that the excess of nitrogen oxides formed is destroyed by addition of urea. Unless this is done the deep orange colour produced by the oxidation of the reagent causes an error in the results. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. References Farnsworth, M., and Pekola, J., Analyt. Chem., 1964, 26, 736. Luke, C. L., Analyt. Chem., 1966, 28, 1276. Ross, W. J . , and White, J. C., Analyt. Chem., 1961, 33, 421. Marczenko, Z., “Spectrophotometric Determination of Elements,” John Wiley, New York, 1976, Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Second Edition, Interscience, Ashton, A., Fogg, A. G., and Thorburn Burns, D., Analyst, 1973, 98, 202. Gentry, C. H. R., and Sherrington, L. G., Analyst, 1950, 75, 17. Eberle, A. R., and Lerner, M. W., Analyt. Chem., 1962, 34, 627. Irving, H., Butler, E. J., and Ring, M. F., J . Chem. SOC., 1949, 1489. Hollingshead, R. G. W., “Oxine and its Derivatives,” Volume 4, Butterworths, London, 1966. Ruf, E., 2. Analyt. Chem., 1958, 162, 9. Matsuo, T., and Funayama, K., J . Chem. SOC. Jaflan, Pure Chem. Sect., 1966, 87, 433. Sanz-Medel, A., Revta Acad. Cienc. Zaragoza, 1973, Serie 2, 28, No. 2, 208. Jimenez, J. L., Gomez, A., and Dorado, M . T., IZev. Metal. Madrid, 1969, 5, 603. Hamaguchi, H., Ikeda, N., and Osawa, K., Bull. Chem. SOC. Japan, 1959, 32, 656. Rakovskii, E. E., and Krylova, T. D., Zh. Analit. Khim., 1974, 29, 910. p. 546. New York, 1959, 862. Received May 27th, 1977 Amended November 28th, 1977 Accepted March 17th, 1978

ISSN:0003-2654    

DOI:10.1039/AN9780301037

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

1046 Analyst, October, 1978, Vol. 103, pp. 1046-1052 Method for the Determination of Methanol in Binary Methanol = Water Mixtures by Use of Ion-selective Electrodes G. J. Kakabadse, H. Abdulahed Maleiia, IVI. N. Khayat, G. Tassopoulos and A. Vahdati Department of Chemistry, University of Manchester Institute of Science and Technology, P.O. Box 88, Munchester, M60 1QD For a given concentration of indicator ion X (X .= F-, C1-, Br-, I-, OH-, S2-, Ag+ or Hf), the systematic change of cell pote:ntial, E, with variation in the concentration of methanol provides a graphical method for the rapid determination of methanol in methanol - water mixtures. Readings obtained by direct potentiometry show good reproducibility and stability. In 99.0- 99.9% m/m methanol, containing M hydrochloric acid, trace amounts of water can be determined accurately owing to a potential “anomaly.” Keywords : Methanol determination; methanol - water mixtures ; ion- selective electrodes ; trace water determination The effect of organic solvents on the potentials of ion-selective electrodes is well known.1-6 The use of this effect in the determination of solvents, first reported in 1975,’ has been tested on a number of binary water - organic solvent mixtures and the method is explained for the methanol - water system.In the system ISE I X, 0-n% methanol] RE (ISE is the ion-selective electrode = AgC1, AgBr, AgI, Ag,S, LaF, or pH glass electrode X = F-, C1-, B r , I-, OH-, S2-, Ag+ or H+; n <loo; RE is the reference electrode), for a given concentration of indicator ion X and a given reference electrode the electrode potential decreases systematically with an increase in the concentration of methanol when X is an anion and the reverse is observed when X is a cation.With the pH glass electrode an initial decrease in potential is reversed at higher concentrations of methanol. Experimental and Results Apparatus and Experimental Conditions Measurements of potential were made with solutions stirred magnetically (at a constant rate throughout) at a temperature of 25 0.5 “C by using an Orion, Model 801, digital pH/ millivolt meter with a potential range of &lo00 mV and a discrimination of hO.1 mV. Potentials were recorded on a Servoscribe potentiometric recorder, Type RE 511. A Hewlett-Packard 9862A calculator/plotter was used for drawing the graphs.Orion solid-state lanthanum fluoride, silver halide and silver sulphide electrodes, and EIL and Beckman pH glass electrodes were used. The concentration of X was about l o - 4 ~ and that of the background electrolyte was 10-1 M. Analytical-reagent grade chemicals were used throughout. Preparation of Calibration Graphs Direct method A series of solutions having the same concentration of X (and of background electrolyte, when used) and different concentrations of methanol (O-n% m/m) were prepared by accurate weighing. Measurements of potential were made in an enclosed system in order to prevent the evaporation of solvent. Each solution was placed in a 250-ml three-necked round- bottomed flask fitted with indicator and reference electrodes and measurements were recorded over a period of several minutes so as to allow for equilibration (see Table I and Figs.1-3).KAKABADSE, MALEILA, KHAYAT, TASSOPOULOS AND VAHDATI TABLE I POTENTIAL RESPONSE OF METHANOL TO DIFFERENT ION-SELECTIVE ELECTRODES 1047 Measured range Corresponding of methanol change in concentration, potential, ISE Xion yo mlm AE*/mV LaF, . . .. .. F- 0-95 - 129 AgCl . . .. . . c1- 0-75 -81 0-94 97 AgBr .. .. . . Br- 0-75 - 65 0-75 32 Ag,S . . .. .. &.E+ 0-64 - 32 0-87 54 0-80 - 33 80-98 4 63-99 94 (AEJ 99-99.9 20 ( A E J Ag+ AgI . . .. 1 . I- 0-75 - 33 Ag+ EIL pH glass electrode OH- EIL pH glass electrode H+ 0-63 -11 (AEJ * AE for 0% mlm methanol is taken arbitrarily as zero, Range of methanol con- centration showing linear response to E , % mlm 0-84 < 50 < 50 < 50 (50 0-54 0-64 0-65 0-60 0-60 99-99.9 Corresponding change in potential, AE*/mV -119 21 - 32 34 -28.5 - 10 20 Indirect method The reversibilities of the systems were checked by constructing Nernstian-type graphs for solutions of the same concentration of methanol and different concentrations of X (in the approximate range 10-6-10-3 M X).Known increments of a stock solution of X were added to the methanol - water mixture (plus background electrolyte) in a 250-ml three-necked round-bottomed flask. After each addition potentials were recorded for a period of 3-5 min so as to allow for equilibration. The electrode slopes are listed in Table 11. From a series of 7-10 Nernstian-type graphs (corresponding to a series of methanol - water mixtures) potentials 100 F 3 80 ae .60 40 - 0 5 f 20 0 0 ' 10 20 30 40 50 AEImV Fig. 1. Change in potential, AE, in solutions M in silver nitrate, 0.1 M in sodium nitrate and of various methanol concentrations, measured by using silver sulphide and double-junction electrodes. t \ t t+ \+ 300 320 340 360 380 400 280 300 320 340 360 380 PotentiaVmV Fig. 2. Potential response in solutions of various methanol concentrations and M in hydrochloric acid measured by using: (a), an EIL pH glass electrode and a double-junction reference electrode; and (b), a Beckman pH glass electrode and a double-junction reference electrode.1048 KAKABADSE et al. : DETERMINATION OF METHANOL IN BINARY Analyst, Vol. 103 400 0 70 20 30 40 50 60 Timdmin Fig. 3. Graphs of time veyszcs potential for methanol- water mixtures in M hydroclhloric acid, Potential mea- sured by using an EIL pH glass electrode and a double-junc- tion reference electrode.Concentration of methanol: A, 0 ; €3, 85.7; C, 92.6; D, 96.3; and E, 99.9%. for the same concentration of X were abstracted and then plotted against the percentage by mass of methanol. From a series of such graphs the one with the best linearity provided the optimum value for the concentration of X for use in the direct method. Fig. 4 shows a typical graph. PotentiaVmV Fig. 4. Potential response (obtained by the indirect method) in solutions l o - 4 ~ in sodium fluoride, 0.1 M in potassium chloride and of various methanol concentrations, measured by using lanthanum fluoride and silver - silver chloride electrodes.Reproducibility Measurements Measurements of potential on solutions used in the direct method were repeated 7-10 times, allowing 5 min for each measurement. The results, expressed as standard deviations, are shown in Table 11. Stability of Potentials potential for a period of 60 min and the rnillivollt readings were recorded on a chart. shows typical results. All of the solutions used in the direct m,ethod were submitted to measurements of Fig. 3 Discussion Several experiment a1 parameters must be considered when assessing the suitability of the proposed method for the determination of meth,anol in a binary mixture with water. As the systems investigated were restricted to solutions used for the preparation of calibration graphs no systematic bias was observed.October, 1978 METHANOL - WATER MIXTURES WITH ION-SELECTIVE ELECTRODES 1049 Electrochemical Reversibility As non-aqueous solvents have a pronounced effect on ionic activitiesapg and, hence, on electrode potentials, the electrochemical reversibility was checked by establishing the electrode slope for each system.The results in Table I1 show that the mean values of the slopes for fluoride ion, silver ion and hydrogen ion are close to the theoretical Nernstian values. TABLE I1 ELECTRODE SLOPES FOR SOLUTIONS HAVING A CONSTANT CONCENTRATION OF METHANOL AND ELECTROLYTE AND A VARYING CONCENTRATION OF X (INDIRECT METHOD) AND STANDARD DEVIATIONS OF POTENTIAL READINGS FOR SOLUTIONS METHANOL (DIRECT METHOD) HAVING A CONSTANT CONCENTRATION OF x AND A VARYING CONCENTRATION O F Concentration of methanol, 0 10 20 30 40 50 60 70 80 87 99.9 % m/m Mean values X ion I A \ F- (LaF,) Ag+ ( A m H+ (Glass electrode) r-1 Electrode Standard Electrode Standard Electrode Standard slope/ - mV, deviation, slope/mV deviation, slope/mV deviation, per decade u/mV per decade o/mV per decade o/mV 58.5 58.0 58.5 57.5 57.5 58.0 57.0 57.0 58.5 - - 57.8 0.69 0.41 0.69 0.69 0.84 0.75 0.75 0.81 0.69 - - 0.70 59.5 58.5 59.0 58.5 58.5 58.5 58.0 58.0 58.0 58.0 58.5 - 0.43 0.02 0.31 0.59 0.14 0.21 0.84 0.30 0.16 0.45 0.35 - 60.0 60.0 59.5 61.5 60.25 61.5 60.0 60.0 - - - 60.3 0.13 0.36 0.22 0.28 0.34 0.45 0.45 0.48 0.52 0.36 - - Response Time and Stability of Potentials In general both response time and stability were satisfactory (Fig.3). Steady potentials were obtained after 1-2 min but the response time increased to 2-5 min at high concentrations of methanol. Long-term stabilities over periods of more than 1 h were also good (Fig. 3). Reproducibility and the Magnitude of A E The standard deviations given in Table I1 clearly indicate the reproducibilities of the results, which were always reasonable when compared with the performance of the same ion-selective electrodes in aqueous solutions. For a given concentration of methanol and a given standard deviation, the magnitude of AE is important: the greater the magnitude the more accurate the results. Coefficient of Variation, Sensitivity and Relative Uncertainty These three parameters, listed in Table 111, are a useful guide when comparing the per- formance of different ion-selective electrodes.While the coefficient of variation indicates the relative error in the determination of methanol (lOOa/AE~o, varying from &0.6y0 for the fluoride system to h3.4y0 for the hydrogen ion system in the O-6Uy0 methanol range), it does not take into account the sensitivity of a given cell to the change in concentration of methanol. A more realistic estimate of the electrode performance is therefore given by the relative uncertainty parameter, incorporating both CT and AE/(concentration of methanol). For linear regions of the graph the best performance is shown by the hydrogen ion system in the 99.0-99.9% m/m methanol range, followed in decreasing order of accuracy by the fluoride ion and silver ion systems, all three having relative uncertainty values of less than 1.1050 KAKABADSE et al.DETERMINATION OF METHANOL I N BINARY Analyst, Vd. 103 TABLE I11 COMPARISON OF COEFFICIENT OF VARIATION, SENSITIVITY AND RELATIVE UNCERTAINTY FOR METHANOL - WATER MIXTURES IN THE PRIESENCE OF A CONSTANT CONCENTRATION OF FLUORIDE, SILVER AND HYDROGEN IONS USING FLUORIDE, SILVER SULPHIDE AND GLASS ELECTRODES, RESPECTIVELY Range of Sensitivity, methanol Coefficient S[AE/(con- Relative Indicator concentration, of variation centration of uncertainty Electrode ion % m/m (&IOOo/AE), % methanol)J/mV (4s) Fluoride . . . . F- 0-84 0.69 1.42 0.49 Silver sulphide , . Ag+ 0-65 1.03 0.62 0.67 65-87* 1.75 0.91 0.38, Glass . .. H+ 0-60 3.43 63-99* 0.38 99-99.9 1.80 0.17, 2.06 2.61 0.14 22.22 0.02 * Non-linear region of the graph.The pH glass electrode can also be used for the determination of trace amounts of water in methanol, which is of considerable practical interest. The problem of overlapping ranges of potential, AEl and AE,, in the hydrogen ion system (see Table I and Fig. 2) can be resolved as the addition of water to a methanol- water mixture would increase the potential in AE, but decrease it in AE,. “Hypersensitivity” of the pH Glass Electrode Measurements obtained with an EIL pH glass electrode were repeated in 1 0 - 2 ~ hydro- chloric acid, using a Beckman pH 0-14 glass electrode. While the over-all response was similar [Fig. 2(b)], differences occurred in the magnitude of AE measured in millivolts for the individual ranges of methanol concentration (EIL data in parentheses) : 0-63% methanol, AE, = -19 (-11); 63-99y0 methanol, AE, = 83.5 (94); 99-99.9Yo methanol, AE, = 30.5 (20).The observed large increase in AE in acid solution at high concentrations of methanol is in agreement with results obtained by other workers.10-12 Several factors may contribute to the hypersensitivity phenomenon of the pH electrode. There is an increase in conductivity in the range 80-100 mol yo methanol13 in a methanol - water mixture. A comparison of the “acidity potentials” of acid - base conjugate pairs in methanol and water has shown CH30H,+ to lie more acidic than H30+.l4 Further, an increase in the concentration of methanol may lead to gradual dehydration of the gel layer on the outer glass surface,17 causing in turn a large shift in the equilibrium between the gel layers of the outer and inner glass surfaces. The effect of methanol on the potential of the hydrogen cell, EH, is well documented.15-17 Using a combined glass - hydrogen cell over the range O - l O O ~ o of methanol, the observed changes in potential, AEH (where AEH = EH H20 - EEH,,,), were between 40 and 100mV, depending on the glass composition of the pH e1ectr0de.l~ Effect of Reference Electrode The presence of a liquid junction (between thle solution inside the reference electrode and the test solution) of variable potential (liquid-junction potential, El) gives rise to some uncertainty in measurements of p~tential.l*,~~ ’The change in the liquid-junction potential, AEl , with the concentration of methanol is probably a substantial f a ~ t o r .~ , ~ ~ Further, there appears to be a considerable variation in El betvveen reference electrodes of a different type, as illustrated in Table IV. In the fluoride system, replacing a silver - silver chloride electrode (Beckman 39403 Futura) with a conventiona.1 mercury - mercury( I) sulphate electrode caused a reduction of -AE by nearly 31 mV (from -119.0 to -88.2 mV). On the otherOctober, 1978 METHANOL - WATER MIXTURES WITH ION-SELECTIVE ELECTRODES 1051 hand, the variation between different reference electrodes of the same type has, in the authors’ experience with several Orion double-junction reference electrodes, no significant effect on AE. TABLE IV EFFECT OF REFERENCE ELECTRODE ON POTENTIAL CHANGE, hE, FOR (0-80)% mlm METHANOL IN THE SYSTEM 2 x 10-4 M SODIUM FLUORIDE + 10-1 M SODIUM CHLORIDE Ag - AgCl Ag - AgCl Ag - AgCl S.C.E Hg - Hg,SO* Reference (Beckman 39403 (Orion double- (Orion single- electrode Futura) junction) junction) AEIrnV ..119.0 114.1 113.4 106.6 88.2 Effect of X A given ion X can be used over a fairly wide concentration range, the upper limit of which is set by the solubility of the ion in methanol and the lower limit by the limit of the Nernstian response of the ion-selective membrane in question. If the test (methanol) sample already contains a particular X ion as contaminant, one can determine the concentration of this ion and either prepare standard solutions with an identical concentration of X ion or increase the concentration of X ion in the test solution to match that present in the already existing standard solutions.Alternatively, a known amount of a different X ion (absent in the test sample) can be added to methanol and an appropriate ion-selective electrode used. Different Types of Graph The potential v e y s m percentage by mass of methanol graphs show linear and non-linear regions (Figs. 1 and 2 and Tables I and 111). While the observed linearity is useful analytically, it must be accepted as empirical in nature. The extent of the linear range is probably a function of the properties of ions in solution and of liquid-junction potentials. Conclusions When limited to binary methanol - water mixtures, the proposed method is fast, reasonably accurate and simple to operate. Judging by the literature data, this method compares favourably with several other methods for the determination of methanol in methanol - water mixtures, e.g., colorimetric,20 dichromate,21 mass-spectrometric,22 infrared23 and ~pecific-gravity~~ methods, but is inferior to gas chr~matography.~~ An added advantage of the proposed method is its adaptability to continuous monitoring. The authors gratefully acknowledge discussions with the following : Professor K.Burger (L. Eotvos University of Budapest); Professor J. de 0. Cabral (University of Oporto), Mr. J. Dwyer (UMIST), Professor M. C. R. Symons (University of Leicester) and Dr. J. D. R. Thomas (UWIST). References 1. 2. 3. 4. 5 . 6. 7. 8. 9. Lingane, J. J., Analyt. Chem., 1968, 40, 935. Wells, C. F., J . Chem. Soc., Faraday Trans.I , 1974, 70, 694. Kazarjan, N. A., and Pungor, E., Analytica Chim. Acta, 1970, 51, 213. Covington, A. I<., and Thain, J. M., J . Chem. Soc., Faraday Trans. I , 1975, 71, 78. Ficklin, W. H., and Gottshall, W. C., Analyt. Lett., 1973, 6 , 217. Bennetto, H. P., and Spitzer, J. J., J . Chem. SOC., Faraday Trans. I , 1973, 69, 1491. Elbakai, A. M., Kakabadse, G. J., Khayat, M. N., and Tyas, D., Proc. Analyt. Div. Chern. SOC., 1976, Bates, R. G., and Robinson, R. A., in Conway, B. E., and Barradas, R. G., Editors, “Chemical Parsons, R., “Handbook of Electrochemical Constants,” Buttenvorths, London, 1959. 12, 83. Physics of Ionic Solutions,” John Wiley, New York, 1966, p. 211.1052 KAKABADSE, MALEILA, KHAYAT, TASSOPOULOS AND VAHDATI 10. 11. 12. 13. 14. 15.16. 17. 18. 19. 20. 21. 22. 23. 24. 25. DeLigny, C. L., and Rehbach, M., Recl Trav. Chim.. Pays-Bas Belg., 1960, 79, 727. Johansson, G., Karlberg, B., and Wikby, A., Talawta, 1975, 22, 953. Nikolski, B. P., Schul’z, M. M., and Beljustin, A. A, Wiss. 2. Tech. Hochsch. Chem. Leuna-Merseb., 1976, 18, 573. Bockris, J . O’M., and Conway, B. E., Editors, “Modern Aspects of Electrochemistry,” No. 3, Butterworths, London, 1964, p. 86. Bates, R. G., in Lagowski, J. J., Editor, “The Chemistry of Non-Aqueous Solvents,” Volume 1, Academic Press, London, 1966, p. 103. Oiwa, I. T., J . Phys. Chem., 1956, 60, 754. Paabo, M., Bates, R. G., and Robinson, R. A., Analyt. Chem., 1965, 37, 462. Shul’z, M. M., and Ivanovskaya, I. S., Sou. Electrochem., 1967, 3, 506. Covington, A. K., in Durst, R. A., Editor, “Ion-Selective Electrodes,” National Bureau of Standards, Bailey, P. L., “Analysis with Ion-Selective Electrcides,” Heyden, London, 1976, p. 23. Boos, R. N., Analyt. Chem., 1948, 20, 964. Welcher, F. J., Editor, “Standard Methods of Chemical Analysis,” Sixth Edition, Volume 11, Part B, Gifford, A. P., Rock, S. M., and Comaford, D. J., Analyt. Chem., 1949, 21, 1026. Kaye, W., Spectrochim. A d a , 1954, 6, 257. Snell, F. D., and Ettre, L. S., Editors, “Encyclopedia of Industrial Chemical Analysis,” Volume 16, Interscience, New York, 1972, p. 107. Blustein, C., and Posmanter, H. N., Analyt. Chem., 1966, 38, 1866. Washington, D.C., 1969, p. 127. D. Van Nostrand, New York, 1963, p. 2143. Received July 27th, 1977 Amended April llth, 1978 Accepted May 17th 1978

ISSN:0003-2654    

DOI:10.1039/AN9780301046

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, October, 1978, Vol. 103, pp. 1053-1060 1053 Potentiality of Seaweed as a Resource: Analysis of the Pyrolysis Products of Fucus serratus Phillip J. Morgan and Keith Smith Department of Chemistry, University College of Swansea, Singleton Park, Swansea, SA2 8PP As a prelude to the investigation of the potentiality of seaweeds as a future source of organic chemicals or fuel, the products of pyrolysis under nitrogen of Fucus serratus (serrated wrack) have been analysed. The pyrolysis pro- duces large amounts of charcoal, water and carbon dioxide with smaller amounts of an oil, pitch, hydrocarbon gases, carbon monoxide, ammonia and carboxylic acids. The oil proved to be a complex mixture of heterocyclic bases, phenols, aromatic hydrocarbons, nitrogen and oxygen heterocyclic compounds and small amounts of aliphatic compounds.A variety of analytical techniques have been employed to analyse the pyrolysis products, with gas chromatography and mass spectrometry being the most widely applicable. Keywords : Seaweed pyrolysis ; Fucus serratus ; gas chromatography ; mass spectrometry There has recently been concern over the rate of depletion of world reserves of fossil fuels, especially crude oil. In addition to making the major contribution to world energy pro- duction, fossil fuels provide the basic feedstock for most of the organic chemicals industry. In the future it will be necessary to utilise alternative organic feedstocks and renewable biomass would appear to be an attractive prospect. In view of the large areas of ocean, the rapid growth rates of some species of kelp (sometimes more than 1 f t per day) and the success being achieved in the artificial cultivation of seaweeds,l the latter may be considered as a promising source of biomass for large-scale utilisation. We have therefore considered means of converting seaweed into more readily usable products.While others are investi- gating the microbiological production of methane,2 we have mainly considered the possi- bilities for chemical conversion using combinations of elevated temperatures and/or pressures, with or without additional reactants or catalysts, and with or without preliminary treatment with other reagents. It has been known for over a century that dry distillation of seaweed produces oily products,3 but it was in the period prior to about 1930 that attempts were made to identify the nature of the distillation products,4-7 the incentive at that time being a possible increase in profits in the potash production industry.Although some of the compound classes were recognised,4-7 the detailed composition of the distillates remained obscure and, when sea- weed was no longer used as a source of potash, research along these lines was discontinued, We have therefore carried out a much more detailed analysis, using modern techniques, of the pyrolysis products from Fucus serratus in order to establish a basis for comparison when seaweed is treated by different methods. Experimental Samples Fucus serratus (serrated wrack) was collected from Limeslade Bay, Swansea. The whole plants were air dried, crushed manually and then ground in an electric grinder.Thenno- gravimetric analysis and oven drying showed that the air-dried seaweed contained 15% of residual moisture. Gas-chromatographic supports and stationary phases were standard commercial materials. Apparatus A cylindrical (about 30 x 5cm) electrically heated furnace (Severn Science) with a Eurotherm 021 temperature controller (adjustable to 1000 "C) was equipped with a B3410% MORGAN AND SMITH: POTENTIALITY OF SEAWEED AS A RESOURCE: Analyst, VOl. 103 jointed quartz test-tube. The test-tube was connected by standard Pyrex glassware to a series of cooled traps. Gas-chromatographic analyses were carried lout using a Pye 104 gas chromatograph equipped with a flame-ionisation detector and facilities for link-up to an AEI MS9 mass spectrometer.The thermal analyses were carried out on a Linseis thermogravimetric analysis apparatus and the X-ray powder diffraction patterns were obtained on a para- focusing Guinier de Wolff No. I1 camera. Pyrolysis Procedure The quartz test-tube was charged with about 60g of dried, crushed seaweed and then connected to a series of three standard traps (ea.ch with a capacity of about 150ml), which were cooled successively in solid carbon dioxide, liquid nitrogen and liquid nitrogen. The outlet of the third trap was connected to a rotary pump-nitrogen cylinder-mercury bubbler arrangement leading to a soda-lime tower, and from there to a large (10-1) bell-jar for collection of the gases over water. The whole apparatus was evacuated and filled with nitrogen prior to the start of pyrolysis.The desired maximum temperature was pre-set and the distillation was allowed to continue until no further volatile products were formed. In the high-temperature (800 "C) distilla- tions, this point was reached at a temperature of about 600 "C, but the temperature was allowed to increase to the pre-set maximum. When pyrolysis was complete, the apparatus was allowed to equilibrate to ambient temperature. A solid residue remained in the quartz test-tube and the gases were collected over water, carbon dioxide being removed in the soda-lime tower. The liquid distillates were directly redistilled from the traps under nitrogen to a maximum temperature of 300 "C. This procedure gave a light oil and an aqueous solution, which formed two layers in the receiver, and a pitch which remained in the traps.The distillates from all of the traps were combined. Separation and Analysis of the Pyrolysis Products details are given below. An over-all scheme of the methods of separation of the products is given in Fig. 1, and Solid residue The solid residue was extracted in a continuous extractor (Soxhlet) using water as the solvent. On cooling, the aqueous solution precipitated a solid, which was shown by X-ray powder diffraction to be calcium carbonate, while the mother liquor contained primarily potassium chloride and sodium chloride (identified by X-ray powder diffraction after evapora- tion). The insoluble residue was termed charcoal, although it undoubtedly also contained a small amount of ash.Liquid distillates Redistillation of the liquid distillates at 300 "C gave three fractions: a residual pitch and a liquid consisting of an aqueous fraction and an oil. The aqueous- oil mixture was separated by extraction with diethyl ether into the aqueous fraction and an ether solution (A) that contained the oil. Aqueous fraction (B) The aqueous fraction (B) was found to be alkaline (pH about 8.5) and was made more alkaline by addition of excess of sodium hydroxide. Extraction of this strongly alkaline solution with diethyl ether gave a fraction containing organic bases, which was combined with a similar fraction obtained from the ether layer (A) (see below). Distillation of the remaining aqueous layer under reduced pressure gave an aqueous ammonia distillate, which was analysed by titration and converted into ammonium chloride, this being identified by X-ray powder diffraction.The residual salts after the last distillation were acidified with excess of concentrated orthophosphoric acid a:nd then a further distillation under reduced pressure was carried out to give a concentrated aqueous distillate containing carboxylic acids,October, 1978 ANALYSIS OF THE PYROLYSIS PRODUCTS OF FUCUS serratus Solid residue Liquid distillates Gases 1055 Carbon dioxide Ether layer (A) Aqueous layer (B) (pH approx. 8.5) Water extraction 4l L Addition of excess of NaO H , ,I,, , , cai:;yipitation NaCl Extraction with 10% NaOH Distillation up Strongly alkaline Heterocyclic solution (pH 14) bases to 300 "C Ether layer (Cl NaOH layer i? Pitch Reduced Conc.HCI pressure distillation + Aqueous liquid + oil Extraction with 2M HCI 1 Extraction with diethyl ether Sodium salts Ammonia so I u t i on Ether layer (D) Oil The ether layer (A) was extracted with 10% sodium hydroxide solution to give an alkaline layer from which phenols were extracted on acidification. These phenols were identified by + Cone. H3P04 reduced pressure H C ~ distil lation fraction Conc. Aqueous carboxylic Ammonium acids chloride1056 MORGAN AND SMITH: POTENTIALITY OF SEAWEED AS A RESOURCE : AnaZyst, Yd. 103 their mass spectra and determined by gas chromaiography, using cyclohexanol as an internal standard. Extraction of the remaining ether layer (C) with dilute hydrochloric acid gave an acid layer from which heterocyclic bases were liberated on basification.These bases were combined with those obtained from the aqueous fraction (B) and were analysed by gas chromatography - mass spectrometry using a gas-chromatographic column packed with Carbowax 20M - potassium hydr~xide.~ The residual ether layer (D) contained a large number of neutral components, including aromatic hydrocarbons, heteroaromatic compounds and some aliphatic compounds. The major components were identified by gas chromatography - mass spectrometry using standard mass spectra documentation tables and co-inject ion of samples when possible. Pitch The pitch was briefly examined using an Iatroscan quantitative thin-layer chromatography system and by gas chromatography, but it showed few discrete compounds, consisting mainly of polar polymeric material.Gases Carbon dioxide was removed from the gases by the use of a soda-lime tower and was determined by direct measurement of the increase in mass. The other gases were found to consist of a mixture of carbon monoxide, hydrocarbon gases, nitrogen and small amounts of hydrogen sulphide. Carbon monoxide was determined by titration of the iodine liberated when aliquots of the gases were passed over diickdine pentoxide heated at 150 "C in a stream of nitrogen,lO the iodine formed being trapped in potassium iodide solution and determined by titration against standard thiosulphate solution. Hydrocarbon gases present in the mixture were determined by gas chromatography on a column of alumina deactivated with liquid paraffin-ll Mass spectrometry indicated the virtual absence of hydrogen.Gas-chromatographic Identification and Determination are given in Table I. Details of the columns used for the gas-chromatographic identifications and determinations TABLE 1 GAS-CHROMATOGRAPHIC CO:LUMNS AND CONDITIONS Stationary phase OV-17 (5%) . . .. .. Apiezon N (8%) . . .. Carbowax 20M - TPA (5%) Carbowax 20M (8%) - KOH (2%) .. .. .. Liquid paraffin (476) . . Support Chromosorb G, 60-80 mesh, AW, DMCS Chromosorb G, 100-120 mesh, AW, DhJCS Chromosorb G, 60-80 mesh, AW, DMCtS Chromosorb G, 60-80 mesh, AW, DMC'S Alumina, 100-120 mesh Conditions A f 7 Tempera- Column ture/OC length/m Uses and comments 100 1.5 Phenols. Cyclohexanol is a convenient standard 65-200 3.5 Neutral compounds 110 3.5 Carboxylic acids.Cyclohexanol is a convenient standard Results and Discussion 100 1.5 Organic bases 50 1.5 Light hydrocarbons Pyrolyses were carried out at three maximum temperatures (300, 450 and SOOOC) and Table I1 records the breakdown of products by class at the three temperatures.October, 1978 ANALYSIS OF THE PYROLYSIS PRODUCTS OF Fucus sewatus TABLE I1 PYROLYSIS PRODUCTS OBTAINED AT DIFFERENT TEMPERATURES 1057 Figures are expressed as percentage of air-dried seaweed. as percentages of the organic material in seaweed. Figures in parentheses are expressed Air-dried seaweed contains 15% of moisture and 15% of inorganic salts. Product Hydrocarbon gases . . .. Carbon monoxide . . .. Carbon dioxide . . .. Nitrogen and oxygen? . . Total gases . . Phenolics .. ,. Neutral compounds: Oil (F) Bases . . .. .. .. .. .. .. .. { Pitch Ammonia . . Carboxylic acids . . Water? . . .. .. Total liquid distillates Inorganic salts . . .. Charcoal7 .. .. .. Total residue . . .. .. .. .. .. .. .. .. ,. .. .. .. .. .. .. .. goo* 1.7 (2.4) 1.5 (2.1) 12.0 (17) 10.8 (15.6) Temperaturel'C 450 A 0.74 (1.0) 1.3 (1.9) 11.0 (16) 5.0 (7.1) 1 300 0.50 (0.70) 7.0 (10) 5.0 (7.1) <0.01 (0) 26 (37) 18 (26) 12.5 (17.8) 0.16 (0.23) 0.16 (0.23) 5.0 (7.1) 0.18 (0.26) 0.75 (1.1) 32 (24.3) 3.7 (5.3) 0.16 (0.23) 0.13 (0.19) 3.7 (6.3) 4.09 (5.7) 0.19 (0.27) 0.04 (0.91) 36 (28.6) 0.0 (0.0) 0.01( (0.0) 0.1 (0.14) 2.5 (3.6) 0.10 (0.14) 0.67 (0.96) 29.1 (20.1) 42 (38.5) 44 (41.2) 15 17 (24.6) 15 23 (32.8)** 32.5 (25.0) 15 40 (57.2)** 32 (24.5) 38 (32.8) 55 (57.2) * Distillation complete by about 600 "C.t By difference. $ By difference, fraction contaminated by polymer formation and precipitation, making direct 9 A further 1% of pitch remained with the residue. 11 Does not distil but about 1% of pitch was extracted from the residue. 7 In high-temperature pyrolyses this is mainly carbon, but includes undecomposed organic ** Includes about 1% of pitch. determination difficult. material in low-temperature pyrolyses. The products of the distillation at 800 "C were subjected to a more detailed study. Table I11 gives the relative amounts of components within each fraction. The neutral and organic base fractions are complex mixtures, but most of the major components have been identified by gas chromatography - mass spectrometry (comparison with literature standard mass spectra) and when possible confirmed by co-injection with authentic materials.Similar detailed studies of the components of the various fractions from pyrolyses at 300 and 450 "C showed qualitatively similar results, although quantitative differences were noted. Phenols were found to be almost absent from the products of pyrolysis a t 300 "C, whereas there was a large increase in the amount of acetylfuran, which became the major component in the neutral fraction. Small amounts of cyclic dienes such as cyclohexadiene and methylcyclopentadiene were indicated by the mass spectral data. These compounds would be aromatised at the higher temperatures. Hydrocarbon gases were essentially absent. These features indicate a lower degree of decomposition of the seaweed and its initial decomposition products at the lower temperature. Microanalysis of the raw air-dried Fucus sewatus, after correction for moisture and inorganic salts, gave the following results: C 45.5, H 5.7, N 3.3 and 0 45.5%.These figures are similar to those for cellulose, which is reasonable for a material that contains large amounts of carbohydrates. The bulk of the organic compounds formed in the pyrolyses have a higher C : H ratio, with a large part of the hydrogen (about 50%) present in the sea- weed being used in the formation of large amounts of water. The major organic products are either aromatic or contain large amounts of oxygen (e.g., formic acid). This is not Chromatograms are shown in Figs. 2 and 3.1058 MORGAN AND SMITH : POTENTIALITY OF SEAWEED AS A RESOURCE: Analyst, 'Vd.103 TABLE I11 PRODUCTS OF THE DISTILLATION AT 800 "C Product Fraction Oil . . .. . . Phenolics Neutral* Heterocyclic basest - Pitch . . .. Aqueous layer (B) Carboxylic acids Ammonia Water - Charcoal . . .. Salts . . . . Soluble Insoluble Hydrocarbons Gases . . . . co Nitrogen and oxygen - co* .. .. See Fig. 2. t See Fig. 3. Components Phenol Cresols Xylenols - - Acetic acid Formic acid Propionic acid Higher acids - - E l } CaCO, Methane Ethane Ethylene Propane Propene - Proportion of fraction, % 40 40 20 - I Combined - 80 12 6 1 1 Product as a Proportion of proportion of product, yo total, yo 1.6 4 1.6 0.8 92 4 6 2 33 0.6 97.6 97 3 10.6 11.6 - - 17 16 14 77.8 - 12 surprising as the oxygen content of seaweed is high, and this is also responsible for the considerable production of carbon dioxide and water, which together constitute over 40% of the decomposition products from the organic material.60 50 40 30 20 10 0 Time/min Chromatogram of neutral fraction on 8% Apiezon N on 100-120-mesh Chromosorb G. Tem- perature programmed from 60 to 200 "C at 4 "C min-1; carrier gas (nitrogen) flow-rate, 10 ml min-l. Peaks: 1, butanone; 2, benzene; 3,, pyrrole; 4, toluene; 6, acetylfuran; 6, xylene; 7, xylene + styrene; 8, tri- methylbenzenes; 9, indene; 10, unknown; 1 1, naphtha- lene; 12, indole; 13, 2-methylnaphthalene; 14, 1- methylnaphthalene; and 15, dimethylnaphthalenes. Fig. 2.October, 1978 ANALYSIS OF THE PYROLYSIS PRODUCTS OF FZLCUS serratus 1059 25 20 15 10 5 0 Time/min Fig.3. Chromatogram of organic base fraction on 8% Carbowax 20M + 2% KOH, on 60-80- mesh Chromosorb G a t 100°C; carrier gas (nitrogen) flow-rate, 40 ml min-1. Peaks : 1, pyridine; 2, a-picoline; 3, 2,6-lutidine; 4,p- + y-picolines ; 5, a dimethylpyridine; 6, 2,4,6- collidine + a dimethylpyridine; 7, a trimethyl- pyridine* + a trimethyldiazine*; 8, a tri- methyldiazine* ; 9, a dimethylpyrazine* ; and 10, te tramethylpyrazine. Compounds marked with asterisks were identified by their mass spectrum only. The results obtained in these pyrolyses broadly match those of earlier reports but provide greater detail. However, we were unable to detect the significant amounts of hydrogen recorded in earlier work. The results of pyrolyses at 800 and 450°C were very similar, indicating that a pyrolysis temperature of about 500 "C would suffice to maximise product fonnation.Pyrolysis temperatures of about 300 "C appear to be of little value. Conclusion An extensive analysis of the components of seaweed pyrolysates should serve as an important basis for comparison in future studies of the possibilities of the conversion of seaweed into useful chemical components. Direct pyrolysis produces useful organic pro- ducts, such as hydrocarbon gases, phenols, carboxylic acids, organic bases and neutral mole- cules, and also some valuable inorganic by-products, such as ammonia, carbon monoxide and potassium chloride. However, the larger organic fractions (e.g., neutral distillates) are complex mixtures and would probably be of little direct value, and in any event the total amount of organic materials is small.It therefore seems unlikely that simple pyrolysis would be an economic prospect in the foreseeable future. A large proportion of the carbon content is converted into hydrogen-deficient species such as charcoal, carbon dioxide and carbon monoxide, which is not unexpected in view of the high oxygen and low hydrogen content of the raw seaweed. It is likely that the products would have greater economic value if reducing agents, such as hydrogen or an alcoholic solvent, were present in the pyrolysis mixture. Pre-treatment of the seaweed is another possibility for improving the economics of the process. We are currently investigating such possibilities. We thank the Science Research Council and BP Ltd. for a CASE studentship (to P. J . M.), and Mr. E. V. Whitehead, Dr. D. Brooks and Dr. A. Banvise (BP Ltd.) for valuable assistance and discussions. We are grateful to Dr. J. A. Ballantine of University College, Swansea, for guidance and assistance with all gas chromatographic - mass spectrometric work reported.1060 MORGAN AN11 SMITH 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Jackson, G. A., and North, W. J., Final Report, Contract No. N60530-73-MV176, US Naval Wilcox, H. A., and Leese, T. M., Hydrocarb. Process., 1976, 86. Stanford, E. C. C., Chem. News, Lond., 1862, 5, 167. Hoagland, D. R., J . Ind. Engng Ch.em., 1915, 7, 673. Turrentine, J . W., and Shoaff, P. S., J. Ind. Enpg Chem., 1919, 11, 864. Spencer, G. C., J . Ind. Engng Chem., 1920, 12, 786. Tupholme, C. H. S., Chem. MetaZE. Engng, 1926, 33, 81. Baker, R. A., J . Gas Chromat., 1966, 4, 418. Smith, E. D., and Radford, R. D., Analyt. Chem., 1961, 33, 1160. Scott, W. W., and Furman, N. H., “Standard Methods of Chemical Analysis,” Fifth Edition, Technical Press, London, 1939, p. 2406. Jeffery, P. C., and Kipping, P. J., “Gas Analysis by Gas Chromatography,” Second Edition, Pergamon Press, Oxford, 1972, p. 114. Received Februar-y 20th, 1978 Accepted Apvil 14th, 1978 Weapons Center, China Lake, Calif., October 1973.

ISSN:0003-2654    

DOI:10.1039/AN9780301053

出版商:RSC    

年代:1978

数据来源: RSC