摘要:

iv SUMMARIES OF PAPERS I X THIS ISSUE January, 1978Derivative Formation in the Quantitative Gas-chromatographicAnalysis of Pharmaceuticals. Part IA ReviewSummary of ContentsPavt IIntroductionXcylationBase-catalysed alkylationDiazoalkane alkylationExtractive alkylation1;lash alkylationXrylationAcid-catatysecl esterification of acidsBoron trihalide - alcohol esterihcation of acidsHoronate ester formationPavt 11SilylationHydrolysisInteraction o f amino and carbonyl groupsIJse of ,VA’-dimethylformamide dialkyl acetalsOxime formationCarbamate formationPyrolysisReductionOxidationIsothiocyanate and isocyanate formationCyclisation of biguanides and guanidinesPhosphorus-containing derivativesAmidc formationI’henylhydrazone formationKeywovds ; Review; plzavmnncez~tical a%alj,sis ; duivatii’e fowzation; gasclwomatogifaplz?J.D. NICHOLSONhIedicines Testing Laboratory, Pharmaceutical Society of Great Britain, 36 YorliPlace, Edinburgh, EH1 3HU.A n n l y t , 1978, 103, 1-28.The Resurgence of Analytical ChemistryThe historical development of analytical chemistry during the 19th and 20thCenturies is outlined. Particular attention is given t o the history of Cniver-sity Chairs in Analytical Chemistry in Europe, and to the dearth of similaracademic positions in the United Kingdom.The numerous early British contributions t o analytical chemistry arc proofof the importance of the subject as a field of scientific endeavour t h a t lastedwell into the present century.The apathetic view of analytical chemistrywhich developed during the years between the t n o V’orld Wars led t o itsdecline, perhaps most noticeable in academic institutions. The new awarenessof the role t h a t analytical chemistry has t o play in industry and technology isdiscussed and the need for clieiiiists trained over a wide field of analyticalmethods is stressed.Keywovds ; Analytical chemistvy ; histiivyR. RELCHERDepartment of Chemistry, University of 13irminghan1, P.0. 13ox 363, Birmingham,I315 2TT.A ~ a l y s t , 1978, 103, 29-36January, 197s THE AlriALYST VTABLES OF STANDARD ELECTRODE POTENTIALSedited by G. Milazzo, Instituto di Clzimica della Facoltd di Ingegneria dell’Unirersitcidi Roma; S. Caroli, Laboratorio di Tossicologie, Instituto Superiore di Sanita, Roma,and V.K. Sharma, Department of Chemistry of the University of Jaipur, IndiaA set of reference tables and electrode potentials. Each entry is given a referencenumber to facilitate location of a system for any given potential. Notes and referencesaccompany many of the entries.0471 99534 7 approx. 440 pages In Press approx. El 1.00/$20.00PRACTICE OF THIN LAYER CHROMATOGRAPHYby J. C. Touchstone and M. F. Dobbins, both of the University of PennsylvaniaSchool of MedicineThis is a fundamental text which explains the basics of performing thin layerchromatography. It minimizes nonapplicable theory and applications to emphasizethe basics needed to perform the separation and assorted tasks related to separation,such as applying the sample, selecting the mobile phase, and quantitation.0471 88042 6 approx. 375 pages In Press approx.&12.70/$21.55ATLAS OF METAL-LIGAND EQUILIBRIA IN AQUEOUSSOLUTIONby J. Kragten, Natuurkundig Laboratorium, University of AmsterdamThis Atlas provides directly accessible information about the behaviour of 45 mostcommonly used metals in the presence of 29 common ligands. The graphs forthe separate metal-ligand combinations are superposable, which means that the graphfor a system containing more than one ligand can be composed from the variousgraphs of the separate metal-ligand combinations. This Atlas is the first to fill thegap between daily laboratory practice and literature data, and its use saves thelaboratory worker long hours of extensive calculations and tedious graph plotting.(Ellis Horwood Series in Analytical Chemistry: Editors, Dr.R. A. Chalmers andDr. M. Masson, University of Aberdeen).085312 084 6 approx. 500 pages In Press approx. f25.00/$47.50Published by Ellis Horwood Ltd., Chichester, and distributed by John Wiley & Sons Ltd.TRACE ANALYSIS OF ATMOSPHERIC SAMPLESby K. Oikawa, Japan Environmental Sanitation Center, Japananalysis of the metal content of atmospheric particulate matter. The text dealsmainly with the most recent methods for pretreatment and analysis. Analyticalprocedures are also included.0470 99013 9 166 pages September 1977 &16.00/$28.60Published by Kodansha Ltd., and distributed by John Wiley & Sons Ltd.This book describes methods for sample collection and pretreatment leading toJohn Wiley & Sons Ltd.,Baffins Lane, Chichester, Sussex PO19 lUD, Englanv i J a w a r y , 1978 SUIVIM~~RIES OF PAPERS I N THIS ISSUEHighlights from Contemporary Analytical Liquid ChromatographyRecent work on the mechanism of solute - phase interactions is discussedand used to explain the mechanism of distribution between the mobile andstationary phases.Examples of solute - solvent - stationary phase inter-actions are considered for silica gel, non-polar bonded phases and ion-pairchromatography. The use of silica gel for exclusion chromatography is alsodescribed in conjunction with the preparation of very high efficiency liquid-chromatographic columns and the trend towards compatible low dead volumedetectors. A practical liquid chromatograph - mass spectrometer system isdescribed in detail and results from commercially available liquid chromato-graph - mass spectrometer equipment are given.The automation of liquid-chromatographic analysis involving automatic injection devices and computerdata acquisition is discussed and the precision that can be obtained from suchequipment given. Finally, an example of the use of high precision as analternative to resolution in a liquid-chromatographic separation is demon-strated.Keywords ; Liquid chvowaatogvaphy ; solute - solvent - stationary phase intev-action ; high-ejj5ciency co1uim.s ; liquid chvontatograph - naass spectvometevR. P. W. SCOTTChemical Research Department, Hoffman-La Roche Inc., Nutley, S. J., USA.Analyst, 1978, 103, 37-55.Hybrid Methods of AnalysisMethods of separation and determination are usually combined and theircombinations are diverse.The separation and determination stages are notoften connected closely, but in a number of instances the properties of aseparation product (concentrate) have a significant effect on the determina-tion. Such a combination is not a simple sum of separation and determinationtechniques; sometimes virtually new methods arise. An example is thecombination of solvent extraction and flame atomic-absorption spectrophoto-metry (if the extract is introduced directly into the flame). There is atendency to carry out the separation and determination in the same device(gas chromatography, high-performance liquid chromatography).In gaschromatography a group of analytical methods was developed that werehybrids of separation and determination techniques.Kew methods of such a type have been developed in this and in otherlaboratories. These methods include a combination of solvent extractionpre-concentration and spark-source mass spectrometry. Trace elementsfrom a sample are collected on high-purity alumina after separation of thematrix by solvent extraction; the alumina is then analysed by mass spectro-metry. Copper oxide has been used as a collector for the determination oftrace amounts of noble metals. Other hybrid methods have been suggestedthat are based on combinations of solvent ext action with polarography andspectrochemical analysis. Organic chelate-forming reagents, which are freeradicals, have been used in a combination of separation and electron-spinresonance spectroscopy; this allows the latter technique to be made auniversal analytical technique and the sensitivity to be increased. High-performance liquid chromatography can be used for inorganic analysis incombination with solvent extraction. Some new reagents have been pro-posed for the group pre-concentration of trace elements.Keywovds ; Hybrid methods of analysis ; five-concentration ; physical methodsof analysisYU. A. ZOLOTOVVernadsky Institute of Geochemistry and Analytical Chemistry, Academy of Sciences,Moscow 117334, USSR.Analyst, 1978, 103, 56-07

ISSN:0003-2654    

DOI:10.1039/AN97803FP001

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

ANALAO 103 (1222) 1-112 (1978)ISSN 0003-2654January 1978THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYCONTENTS12937566872798493101104106REVIEW. Derivative Formation in the Quantitative Gas-chromatographicAnalysis of Pharmaceuticals. Part I-J. D. NicholsonPLENARY LECTURE. The Resurgenc.e o f Analytical Chemistry-R. BelcherPLENARY LECTURE. Highlights from Contemporary Analytical Liquid Chrom-atography-R. P. W. ScottPLENARY LECTURE. Hybrid Methotls o f Analysis-Yu. A. ZolotovCalcium lon-selective Electrodes Based on Calcium Bis[di(p-I ,I ,3,3-tetramethyl-butylphenyl)phosphate] Sensor and Trialkyl Phosphate Mediators-G. J.Moody, N. S. Nassory and J. D. R. ThomasAutomatic Determination o f the Specific Gravity o f Alcohol and Water Mixtures.Part 1.Development and Evaluation o f a Simple Sensing System-W.Bunting and P. B. StockwellRapid Spectrophotometric Determination o f Copper in Steel and Non-ferrousAlloys After Synergistic Extraction-P. S. Patil and V. M. ShindePreparation of Trace Element Reference Materials by a Co-precipitated GelTechnique-Alan R. DateREPORT BY THE ANALYTICAL METHODS COMMITTEEPrecise Standardisation of Disodium Dihydrogen EthylenediaminetetraacetateDihydrate by Spectrophotometric Titration Against Pure Zinc MetalSHORT PAPERImprovement in the Determination of Total Arsenic by Arsine Generation andAtomic-absorption Spectrophotometry Using a Flame-heated SilicaFurnace-D. E. Fleming and G. A. laylorCO M M U N ICATIO NPreparation of a Stable Borohydride Solution for Use in Atomic-absorptionStudies-J.R. Knechtel and J. L. FraserBook ReviewsSummaries of Papers in this issue-Pages iv, w i t viii, xPrinted by Heffers Printers Ltd, Cambridge, EnglandEntered as Second Class at New York, USA, Post OfficNewEU RO -STAN DAR Dnow availableE. S .877-lFurnace Dust (LD Converter)Certified for t h e following elements :Fe, Si, Ca, Mg, Al, Ti, Mn, P, S, Na,K, F, V, Cr, Ni, C, Zn, Pb, Cu and AsFull details obtainable from :Bureau of Analysed SamplesLtd.Newham Hall, Newby,Nliddlesbrough, Cleveland TS8 9EATelephone: Middlesbrough 31 721 6ADVERTISERSPLEASE NOTEAll advertisements forTHE ANALYSTshould from now on be addressed toour ownAdvertisement Department,The Chemical Society,Burlington House,Piccadilly, London W1V OBNTel: 01-734 9864Please send all space orders, copy,enquiries etc.to this address.NOTICE TO SUBSCRIBERS(other than Members of the Society)S u bsc r i pt i o n s for The Analyst, Analytical Abstracts a n d Proceedings s h o u 1 dbe sent to:The Chemical Society, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HNRates for 1978The Analyst, Analytical Abstracts and Proceedings (including indexes):(a) The Analyst, Analytical Abstracts and Proceedings . . . . . . . . €99.00(b) The Analyst, Analytical Abstracts printed on one side of the paper, andProceedings . . . . . . . . . . . . . . . . . . f105.00The Analyst and Analytical Abstracts without Proceedings (including indexes) :The Analyst, and Analytical Abstracts printed on one side of the paper(c) The Analyst, and Analytical Abstracts . . .. . . . . . . . . f87.00(d) . . 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(f) Analytical Abstracts printed on one side of the paper . . . . . . . . f73.0

ISSN:0003-2654    

DOI:10.1039/AN97803BX003

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

January, 1978 THE ANALYST viiCLASSIFIED ADVERTISEMENTSThe Rate for Classified Advertisements i s L2.30 per sin&column centimetre (manimum L4.60)Box Numbers are charged an extra 50p.Deadline for classified copy is 20th of the month precedingmonth of issue.All space orders, copy instructions and enquiries should beaddressed to The Advertisement Department,The Chemical Society, Burlington House, Piccadilly,London WrV oBNTelephone 01-734 9864 Telex 268001WORKSHOP COURSEELEMENTS OF ANALYTICALINSTRUMENT ELECTRONICS3-7 April 1978A Workshop Course for novicesInfo.: Wolfson Bioanalytical Centre,University of Surrey, Guildford GU2 5XH CAI 990Selected Annual Reviewsof the Analytical SciencesVolume 4'Advances in Voltammetric Techniques,' byB.Fleet and R. D. Jee'High -frequency Electrodeless PlasmaSpectrometry,' by B. L. SharpCONTENTSPp. vi + 73 f 9.50ISBN 0 85990 204 8CS Members' price €3.00Orders should be sent direct, with remittance, orthrough your usual bookseller to-THE CHEMICAL SOCIETYDistribution Centre,Blac khorse Road, Letc hworth,Herts. SG6 IHN, EnglandCS Members must write direct to the above addressenclosing the appropriate remittance.SENIOR ANALYSTand QUALITY CONTROLLERrequired by a small but expanding North London based Company involved in theresearch and manufacture of a range of pharmaceutical products.Applicants should be pharmacy or chemistry graduates with at least several yearsexperience in pharmaceutical analysis and quality control and will be responsiblefor the testing of raw materials finished products, analytical methods and newformulations.Excellent salary offered and Pension/Life Assurance Scheme in operation.Pleaseapply to the Company Secletary, Biorex Laboratories Limited, Biorex House,Canonbury Villas, London N1 2HB.PUBLICATIONNEW DIRECTORY OF SURFACE ACTIVE AGENTS (1st Edition)Having widespread application in Chemical Processing Surface active agentsSURFACTANTS UK available in the UK including Property & Usage Summary, Trade Names,Gordon L. Hollis Chemical Identity, Active Strength, etc.Now available: f7.50 UK, f8.00 overseas, including p 8 p. Remittance with order please to G. L. Hollis,Tergo-Data, 34 Edinburgh Drive, Darlington, Co. Durham DL3 8AT. Phone (0325) 6899...Vlll SUMMARIES OF PAPERS I N THIS ISSUECalcium Ion-selective Electrodes 13ased on Calcium Bis[di(p-1,1,3,3-tetramethylbutylphenyl)phosphate] Sensor and Trialkyl PhosphateMediatorsCalcium ion-selective electrodes based on PVC matrices with membranescomposed of calcium bis[di(p-l,1,3,3-tetramethylbutylphenyl)phosphate]sensor and trialkyl phosphate solvent mediator are described and comparedwith electrodes of the same sensor with di-n-octyl phenylphosphonate mediator.Trialkyl phosphates of sufficiently high viscosity, for example, tri-n-butylphosphate, tri-n-amyl phosphate, tri( 1,1,3,3-tetramethylbutyl) phosphateand tri-n-octyl phosphate make good solvent mediators and maintain thecalcium ion selectivity of the sensor for the range of alkali metal and bivalentions studied, namely, Na+, K+, Mgz+, !ir2+, Ra2+, %In2+, Cu2+, Ni2+ and Zn2+.The trialkyl phosphate mediators provide commercially available alternativesto di-n-octyl phenylphosphonate, the best in the range examined being tri-n-amyl phosphate.January, I978Keywords ; Ion-selective electrodes ; calcium ion-selective electrodesG.J. MOODY, N. S. NASSORY and J. D. R. THOMASChemistry Department, University of TVa.les Institute of Science and Technology,Cardiff, CF1 3NU.Analyst, 1978, 103, 68-71.Automatic Determination of the Specific Gravity of Alcohol andWater Mixtures. Part 1. Development and Evaluation of aSimple Sensing SystemThe design of a simple sensing system to determine the specific gravity ofliquids is described. The evaluation of a prototype system to measure thespecific gravity of alcohol - water mixtures is discussed.The system is fullyautomatic and data processing is achieved in an off-line manner. Furtherpossible areas of application are discussed.Keywords ; Specific gvaiiity ; automatic detevmination ; data processingW. BUNTING and P. B. STOCKWELLDepartment of Industry, Laboratory of the Government Chemist, Cornwall House,Stamford Street, London, SE1 9NQ.Analyst, 1978, 103, 72-78.Rapid Spectrophotometric Determination of Copper in Steel andNon-ferrous Alloys After Synergistic ExtractionA simple and rapid procedure is described for the separation and deter-mination of copper in steel and alloys comprising extraction of the copperinto chloroform with isonitrosoacetylacetone (H-INAA) in combinationwith pyridine from an aqueous solution of pH 6.2.The extracted speciesabsorbs a t 450 nm; the molar absorptivity is 2.44 x lo3 1 mol-l cm-l andSandell’s sensitivity is 2.6 x p g cm-2. Beer’s law is obeyed over therange 20-125 pg of copper per 10 ml of organic phase. The complex is stablefor 48 h and the extracted species is probably Cu(INAA),.2C6H,N. Copperis separated from associated elements. Results of the analysis of syntheticmixtures and standard samples are reported.Keywords : Copper deteymination ; steel analysis ; alloy analysis ; spectro-photometry ; synergistic coppev extvactionP. S. PATIL and V. M. SHINDEDepartment of Chemistry, Shivaji University, Kolhapur 41 6004, India.Analyst, 1978, 103, 79-83Japiuavy, 1978 THE ANALYST ixRS solvents forUV and IRspectrophotometryAcetone UV and IRAcetonitrile UV and IRBenzene UV and IRCarbonium sulfideUV and IRCarbonium tetrachlorideUV and IRChloroform UV and IRCyclohexane UV and IRN-N-DimethylformamideUV and IRDic hloroet hane I RDi met h ylsu If oxide U VDioxane UVEthyl acetate IREthyl alcohol UV9 5 O and abs.Ethyl ether UVn-Heptane UVn-Hexane UVlsoctane UV and IRlsopropyl alcohol UVMethylene chlorideUV and IRMethyl Alcohol UVn-Pentane UVPotassium bromide IRTetrachloroethylene IRTetra hydrofuranUV and IRToluene IRTtichloroetilene IRCHEMICALS DIVISIONP.O.Box 3996/20159 MilanolVia lmbonali 24 (Italy)Telex Erba Mi 36314lTel.6995MON 1 LDISON S 1) A REG. TRADEMARX SUMMARIES OF PAPERS I N THIS ISSUEPreparation of Trace Element Reference Materials by aCo -precipitateti Gel TechniqueA method is described for the preparation of trace element reference materialsby a co-precipitated gel technique. 'The technique is applied to the manu-facture of multi-element calibration standards for the analysis of geologicalmaterials by direct-reading emission spectrometry. The quality of gelreference material is considered in terms of homogeneity, major element andtrace element accuracy and purity. Results are presented for the analysisof the gels by a variety of techniques, and for the analysis of internationalstandard geological materials by direct-reading emission spectrometry withgel reference material calibration.precipitated gel techniqueJanuary, 19 7 8Keywords : Refevence material preparation ; geochemical trace elements ; co-ALAN R.DATEInstitute of Geological Sciences, Analytical and Ceramics Unit, 64/78 Gray's InnRoad, London, WClX 8NG.Annlyst, 1978, 103, 84-92.Precise Standardisation of Disodium Dihydrogen Ethylenedi-aminetetraacetate Dihydrate by Spectrophotometric Titration AgainstPure Zinc MetalReport Prepared by the Compleximetric Standards Panel of the AnalyticalStandards Sub-Commi ttee.Keywords : ED T A standavdisation ; zinc ; dithizone indicator ; spectvophoto-metric titrationANALYTICAL METHODS COMMITTEEThe Chemical Society, Burlington House, London, W1V OBN.Analyst, 1978, 103, 93-100.Improvement in the Determin:ation of Total Arsenic by ArsineGeneration and Atomic-absorption Spectrophotornetry Using aFlame- heated Silica FurnaceShort PaperKeywords : Arsenic determination ., atomic-absovption spectroplzotometvy ;Jaydride generatioit ; silica fuvnaceD. E. FLEMING and G. A. TAYLORDepartment of Mines, Government Cheniical Laboratories, 30 Plain Street, Perth,Western Australia, 6000..4nalyst, 1978, 103, 101-104.Preparation of a Stable Borohydride Solution for Use inAtomic-absorption StudiesCommunicationKeywords : Stable borohydride solution ; sodium borohydride reduction ; atomic-absorption spectrophotometryJ. R. KNECHTEL and J. L. FRASERWastewater Technology Centre, Canada Centre for Inland Waters, P.O. Box 5050,867 Lakeshore Road, Burlington, Ontario, L7R 4A8, Canada.Analyst, 1978, 103, 104-105

ISSN:0003-2654    

DOI:10.1039/AN97803BP005

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, January, 1978, Vol. 103, $9. 29-36 29 The Resurgence of Analytical Chemistry Plenary Lecture* R. Belcher Department of Chemistry, University of Birmingham, P.O. Box 363, Birmingham, B15 2TT The historical development of analytical chemistry during the 19th and 20th Centuries is outlined. Particular attention is given to the history of Univer- sity Chairs in Analytical Chemistry in Europe, and to the dearth of similar academic positions in the United Kingdom. The numerous early British contributions to analytical chemistry are proof of the importance of the subject as a field of scientific endeavour that lasted well into the present century. The apathetic view of analytical chemistry which developed during the years between the two World Wars led to its decline, perhaps most noticeable in academic institutions.The new awareness of the role that analytical chemistry has to play in industry and technology is discussed and the need for chemists trained over a wide field of analytical methods is stressed. Keywords : Analytical chemistry ; history When planning this Plenary Lecture, I sought a general title into which almost anything could be fitted, for I wanted to be as far-ranging as possible over the whole field of analytical chemistry. Eventually, I decided on the present title, for if there has been a resurgence, it must have been preceded by a period when analytical chemistry declined, and before that when it flourished. This title, therefore, provided me with as much scope as I needed. Special reference is made to analytical chemistry in the United Kingdom, for this is where the ebb and flow can be seen most clearly.Reference is made to the contributions made by British chemists from the 17th Century to fairly recent times, and to the contributions made by Birmingham and Midlands chemists long before our own times. Chairs of Analytical Chemistry are also considered, because this subject is associated with research and teaching in analytical chemistry and also with its further progress and future direction. The Rise of Analytical Chemistry The earliest chemistry that was practised, as far back as written records go, was industrial chemistry.1 The ancients knew seven metals and some of their oxides, acetic acid and certain alkalis ; the processes of crystallisation, evaporation, distillation and filtration were widely used.This was closely followed by a subject which we would hesitate to call analytical chemistry, but which was undoubtedly chemical analysis. This was a necessary operation for purposes of quality control. There are many examples of analytical methods being used by the ancients and these are too well known to require repetition here. These are the two oldest branches of chemistry and, as I have indicated elsewhere,2 other branches are newcomers by comparison, for they took shape only in the last century. These facts have been clearly recognised on the Continent of Europe and in the USA, because the earliest named chemistry Chairs were of Technical Chemistry and Analytical Chemistry. This was reflected in Europe by the pattern of the naming of the constituent institutes of University Chemistry Departments, viz., Industrial, Analytical, Organic, Inorganic and Physical Chemistry.When the International Union of Pure and Applied Chemistry formed its Divisions in 1949, this was the pattern it followed, with the addition of Biochemistry to provide six Divisions. Several years later Biochemistry was to secede to develop its separate Union. At least four chemists have been awarded the Nobel Prize for the development of analytical techniques, but it is significant that in each instance the particular technique chosen accelera- ted the progress of biochemistry. * Presented a.t the Fourth SAC Conference, Birmingham, July 17th to 22nd, 1977.30 BELCHER: THE RESURGENCE Analyst, VoL 103 In earlier times, inorganic chemistry was closely allied with analytical chemistry and so it was possible to find in the last century and later in this century combined Chairs of Analytical and Inorganic Chemistry.Curiously, there was only ever one combined Chair of Organic and Analytical Chemistry (in the USSR), even though analytical chemistry is of the utmost importance in the progress of this associated branch of chemistry. The earliest titled Chair devoted exclusively to analytical chemistry appears to be that of C. F. Chandler,3 created in 1859 at Union College, Schenectady. However, before that time, K. F. Plattner in 1842 held the title of Professor of Mineralogy and Blowpipe Analysis at the Mining Academy of Freiberg. I have attempted to establish when the first Chairs of Analytical Chemistry (alone or combined) were founded in Europe.This was a considerably difficult task, partly because some of the information kindly provided by many European Universities is contradictory and some of the dates do not fit properly. I remarked many years ago how difficult it was to establish events that had occurred only 30 years earlieI-4; it is even more difficult when one tries to go back nearly 200 years. Table I presents a summary of some of the informa- tion; some of the material has been omitted because of its obvious contradictions. Combined Chairs are included, but not Assistant Professorships or even famous analytical chemists who held a non-titular Chair of Chemistry. For example, Torbern Bergman, who is said to be the founder of modern chemical analysis, was made Professor of Chemistry at Uppsala in 1767.This appears to be the first Chemistry Chair established in Sweden, but it did not carry any title. TABLIE 1 CHAIRS OF ANALYTICAL CHEMISTRY Country No. of Chairs* Date of first Chair Sweden .. .. .. 10 (1975) 1960 Finland . . .. . . - 1952 Bulgaria . . .. .. - 1907 Italy .. .. .. .. 24t 1937 Belgium (Liege): . . .. - 1876 Romania .. .. .. 3 1962 Netherlands . . .. .. 7 1905 France.. .. .. .. - 1895 USSR .. .. .. .. - 1899s Switzerland . . .. .. - 1855 Spain . . .. . . . . 22 1924 Austria .. . . . . - 1875 * A dash indicates that reliable information is not available. t Six of these are Chairs of Instrumental Analysis. Only information for Liege is available. This was the first Chair of Analytical Chemistry.Although some combined Chairs existed before that time, the exact dates are not known. The first Chair in Britain, which was a combined Chair of Inorganic and Analytical Chemistry5 and was created in 1920, was held by R. M. Caven at what is now the University of Strathclyde and 200 years ago was called the Andersonian University. The Chair dis- appeared when the incumbent died in 1934. A Chair of Analytical Chemistry was not established until 1958 (C. L. Wilson at Queen’s University, Belfast). This Chair at a latter date was combined with that of Inorganic Chemistry, but was finally divided into two separate Chairs. This Chair at Queen’s University is the only remaining established Chair in the UK and is currently held by Professor D.Thorburn Burns. Although the number of Chairs of Analytical Chemistry in Britain reached a total of four in 1974, only two others now remain beside the Belfast Chair. These a.re Personal Chairs and the fate of one of these is still in the balance.* Professor Lloyd-Smythe has already pointed out, in another * In 1940, the TJniversity of St. Andrews offered the late Fritz Feigl a Chair after he had lost his position at the University of Vienna. It is not known if this would have been a named Chair, but with Feigl as the incumbent there is no doubt what type of research would have been undertaken. I t would surely have had a profound effect on the progress of analytical chemistry in Britain. Feigl had to decline as he had already accepted another offer.January, I9 78 OF ANALYTICAL CHEMISTRY 31 context,g that Sweden, a country with a population of 8.5 million, has ten Chairs of Analytical Chemistry.It seems likely that if we were able to collect all of the figures for the leading countries and relate the number of Chairs to the population, the UK would easily come bottom of the league. Despite the reluctant acceptance of academic analytical chemistry by British universities, it is still regarded (apart from by an enlightened few) as a “trendy” or even “off-beat” subject. Its origin and its venerable place in the history of chemistry are not realised, nor is its status outside the UK. Even in the University of Birmingham, which to its credit and in keeping with its earliest traditions created the very first Chair devoted exclusively to analytical chemistry in Great Britain, there are Professors in the Faculty of Science who think of a Chair of Analytical Chemistry as a new, trendy specialised subject comparable perhaps with some of the peculiar and probably useless Chairs that are established in the pseudo-sciences ; whereas the truth is that British university chemistry has been functioning minus a limb, or with only part of one, for over a century.Nevertheless, as I have pointed out el~ewhere,~ despite these formidable obstacles, the UK has made outstanding contri- butions to analytical chemistry. The earliest outstanding British analytical chemist was Robert Boyle, most of whose work was concerned with analytical chemistry. He was the first to suggest indicators, he used sulphide for separations and he was the first to work out the sensitivity of a chemical reacti0n.l Although he is often considered as a physical chemist, because of Boyle’s law, even this is the basis of classical constant-volume gas analysis apparatus.Francis Home (1756) and, somewhat later, William Lewis must be ranked amongst the founders of titri- metric analysis. I have ref erred to Lewis’ pioneer contributions elsewhere,’ but should record once more that he was the first to use an indicator in titration, the first to use a primary standard and the first to use a weight burette. The foremost teacher in Britain in the 18th Century was Joseph Black and many students from abroad were trained in his laboratories. He was probably the first chemist to use a “blank” determination.He noted that litmus indicator contained a slightly alkaline impurity, for which he made allowance. Although the first recorded use of a “blank” determination is by Winterl, almost a quarter of a century earlier, this was part of Winterl’s method for the determination of phlogiston and the “blank” was to allow for the trace “phlogiston” content of the water used in the method. Although Winterl had the right idea in allowing for “ blanks,” Black’s correction was concerned with the determination of substances that actually existed .1 In the first half of the 19th Century, Griffin devised the first hydrogen sulphide generator. The better known Kipp generator was a modification of the Griffin generator. Griffin also proposed the use of “Centigrade solutions” and Ure the use of normal solutions.At about this time, the titrimetric method for the determination of the hardness of water was developed by Thomas Clark, Professor of Chemistry at the University of Aberdeen. It was he who differentiated between temporary and permanent hardness and who developed the method for evaluating hardness. Clark had first studied under Liebig at Giessen and it was there he learned the technique of titrimetric analysis. In this same period, William Penny developed his famous titration method using dichromate. Penny, like Ure before him and Dittmar after him, taught chemistry at the Andersonian University, Glasgow. This is, of course, now the University of Strathclyde, and it is perhaps fitting that the first Chair in Britain connected with analytical chemistry was founded there.These investi- gators continued the British contributions to the development of titrimetric analysis made by Home, Lewis and Black. In 1836 Marsh described his test for arsenic, which became universally used and the basis of further generation methods that are still practised (Fig. 1). Marsh was a chemist at the Arsenal, but had also been Michael Faraday’s assistant. At Queen’s University, Belfast, Thomas Andrews in 1852 developed certain microchemical tests and was probably the first to develop a quantitative method on the micro-scale. In the 19th Century, one of the most difficult determinations, despite its importance, was that of sodium, for before 1890 only one sparingly soluble compound was known, namely sodium antimonate. At that time, this was the only basis on which a method could be developed. British chemists addressed themselves to this problem and three new reagents were discovered over the next decade.These were dihydroxytartaric acid (one of the first32 BELCHER : THE RESURGENCE Analyst, Vol. 103 organic reagents) developed by Fenton, fluoroaluminic acid by Wilks, both of Cambridge University, and potassium caesium bismuthate by Ball, who was a Lecturer in Chemistry a t Guy’s Hospital and later worked in the War Office Laboratories. He also used the reagent Fig. 1. Marsh’s apparatus for arsenic testing, 1836. in reverse to determine caesium. Unfortunately, this reagent was unstable and had to be kept under an inert atmosphere. This problem might easily have been overcome, but a few years later the well known zinc uranyl acetate reagent was developed and led to the standard method for the determination of sodium.Of course, this method has been generally super- seded by spectroscopic methods, but it is still the preferred method when a classical method has to be applied. However, it is interesting to note that only a few years ago a reagent corresponding to Ball’s reagent with thiocyanate substituted for nitrite is claimed to be the most sensitive reagent so far developed for sodj.um. Notable discoveries were made in the field of spectroscopy by 19th Century British scientists and their massive contributions have been dexribed recently.* Table I1 lists their names and periods; the order given is according to the dates of their particular contributions.British scientists have also played a prominent part in the development of electroanalytical and radiochemical methods. TABLE I1 NINETEENTH-CENTURY BEITISH SPECTROSCOPISTS Name T. Young .. a . J. F. W. Herschel , . W. H. F. Talbot a . W. H. Miller . . .. D. Brewster . . .. C. Wheatstone . . .. W. A. Miller . . .. G. G. Stokes . . .. W. Swann .. .. J. N. Lockyer . . .. W. N. Hartley . . .. J. W. Draper . . .. Dates 1773-1829 1792-1871 1800-1877 1801-1892 1781-1868 1802-1875 1811-1882 18 17-1870 1819-1903 1 8 1 8- 1894 1836-1920 1843-1913 There were many other contributions to analytical chemistry in the later part of the century, and some of those mentioned and some others are recorded in Table 111. In more recent times the work of Adams and Holmes, who produced the first ion-exchange resins that were of practical value,l and the discovery of partition chromatography and of gas chromatography by Martin and Synge and Martin and James,l may be noted.Before commenting on the decline of analytical chemistry, it is appropriate to include contributions made by Birmingham and Midlarids chemists ; this heading includes not only those who were born in the area, but also those who lived and practised there.January, 19 7’8 OF ANALYTICAL CHEMISTRY 33 Probably the first analytical chemist of any note was William Withering, who lived at Edgbaston Hall, less than a mile from the University, and who was a medico, mineralogist, botanist and chemist. It was he who first discovered the use of digitalis for the control of heart disease.The mineral witherite is named after him. It was Withering who first introduced barium salts as tests for sulphate and it has been stated that he was the first to develop the gravimetric method for the determination of sulphate as barium sulphate. He was also a member of the Lunar Society. TABLE 111 SELECTED BRITISH CONTRIBUTIONS OF THE NINETEENTH CENTURY Date Event 1813 1835 1836 1841 1850 1852 1871 1878 1893 1894 1908 Ure proposed “normal” solutions D. B. Reid-Micro-scale teaching Marsh-Arsenic test Clark-Hardness of water Penny-Dichromate titrations T. Andrews-Microchemical methods Griess-Nitrite test Bayley-Spot tests on paper Thomson-Glycerol- boric titrations Fenton-Dihydroxytartaric acid test for sodium Sand-Controlled-potential internal electrolysis Because of the great industrial developments, for example, in the gas, glass-making, pottery, non-ferrous and ferrous metallurgy and the many other industries of Birmingham and the towns around, there must have been a good deal of chemical analysis practised, especially when there were such men around as Murdoch, Kerr, Watt, Wedgewood and Boulton.Unfortunately, there is little on record; and yet analytical methods must have been used extensively. It is probable that analytical methods were developed and not published immediately but passed almost from “father to son.” The situation may have been very similar to that which existed in the United States during the same period. Twentieth century writers ascribe the method for the determination of phosphorus in steel to Blair. Blair wrote a famous text book on the analysis of iron and steel in the last century in which the method is described, but he states that he “got the method from a Mr.Wood, who in turn got it from a Mr. J. H. Nicholls of the Homestead Works.” Probably it never occurred to any of these chemists to publish their methods, which became known outside their particular Works only when they were included in books by writers such as Blair. Griess developed his well known test for nitrite in 1871 at the brewery town of Burton, a few miles to the north of Birmingham. He tried various coupling reagents and eventually recommended l-naphthylamine to couple with diazotised sulphanilic acid. This is one of the most sensitive tests a~ailable.~ Percy Frankland, the Second Mason Professor at the University of Birmingham, amongst other activities, set up a laboratory for the analysis of water.His successor, Sir Gilbert Morgan, although not directly concerned with analytical chemistry, was instrumental in preparing and examining the compounds he termed “chelate compounds” and which were later to play such an important part in analytical chemistry. Many of the compounds that Morgan made and studied later became of value in analytical Chemistry. In the time of Sir Norman Haworth, one notable contribution to analytical chemistry was the setting up of the first organic microanalytical laboratory in Britain, and it was in the Frankland Building of Birmingham University that the compound ascorbic acid was first synthesised.Although it has many other well known uses, one of its important applications is as an analytical reagent. At about this time, Main-Smith in the Chemistry Department developed the crucible named after him (Fig. 2). I t is not generally appreciated that Aston, who invented the mass spectrometer, was a Birmingham man and worked in the Department of Physics at Birmingham University before he left for Cambridge. He was born in Harborne, an adjoining suburb to Edgbaston. In the 19th Century, one of the greatest achievements of chemical analysis was in exposing the illegal adulteration of foods. How- This application was initiated by Frederick Accum.34 BELCHER : THE RESURGENCE Analyst, Vol. 103 ever, he made so many enemies as a result of his revelations that they trumped up false charges against him and he had to flee the country.Despite this temporary victory, how- ever, eventually laws were passed to control the adulteration of food. These early applica- tions led to the strict control of the purity of other daily necessities of life. Undoubtedly, analysis played a major part in quality control in various industries, but until titrimetric analysis had been sufficiently developed many of the methods must have been comparatively crude. Some well known books were published in the latter half of the century concerned with metallurgical analysis, for by then metallurgical processes were completely controlled by the analyst. Some interest was shown in the pollution of rivers, for we know Michael Faraday was consulted on this problem.In the early years of the 20th Century, chemical analysis continued to expand and was undoubtedly applied extensively in industry and in Ministry laboratories during the First World War. It is probably at this point that the peak was reached, and subsequently there was a decline in Britain, even though the first Chair was founded shortly after the War. It seemed to have no influence on the develop- ment of analytical chemistry, probably because it was a combined Chair and the incumbent was more interested in inorganic chemistry. Fig. 2. Main-Smith crucible with interior- fitting serrated lid. The Decline A. Chaston Chapman, who was the President of the Society of Public Analysts, 1914-1915, and in 1920 was President of the Institute of Chemistry and also Chairman of the Committee of Chemical Education set up jointly by the Board of Education and the Institute of Chemis- try, continually appealed for the creation of Chairs in Analytical Chemistry10 and his prophetic views on the future developments of analytical chemistry are well worth reading after the passage of 60 years.Some impetus was given to the progress of analytical chemistry in Europe by the spectacu- lar results achieved by the Pregl methods of organic microanalysis, but more than a decade was to pass before the methods were practised :in Britain. In fact, they were first practised in Britain in the University of Birmingham in 1926, as stated earlier. At the time of the outbreak of the Second World War, there were very few organic microanalytical laboratories in Britain and most of them still used the original Pregl methods, which were then almost 30 years old.Another example of the lack of awareness in Britain was illustrated when Briscoe and Matthews gave lectures a t the Royal Institute of Chemistry describing the technique of inorganic microanalysis. These lectures were published in an Institute Mono- graph in 193411 and immediately aroused widespread and enthusiastic interest, not only in Britain, but throughout the British Empire. Yet these methods in many instances had been known for more than a quarter of a century on the Continent of Europe and their description was readily available in the literature. Looking back a t the post-First World War period, it would seem that this was the time when the universities should have expanded and included analytical chemistry as a major part of their curriculum.That they did not is probably related to the political and economic decline of Britain at that time. Undoubtedly, the demands of the Second World War showed the glaring shortcomings of a university system that did not train research analytical chemists when so many new analytical problems arose during the War effort. It is not intended to provide here an account of some o.F the applications of analytical chemistry in the War effort, but perhaps the most spectacular achievement was in the examination of the trans-uranium elements in the USA, when the methods of submicroanalysis were used to study the chemistry of these metals, which were then only available in very limited amounts.January, 1978 OF ANALYTICAL CHEMISTRY The Resurgence 35 It can be said that the resurgence slowly and imperceptibly began in the post-War years; however, it took a long time before this became recognisable. It was more noticeable in other countries than Britain and so what I say refers mainly to Britain.Chalmers remarked recently1* that “. . . while it is true that there has been a resurgence in analytical chemical practice, if there has been a renaissance in the teaching of the subject, it exists only in the minds of analytical chemists, and has passed undetected by other kinds of chemists.” This means that there has been a resurgence in every way except in the teaching of analytical chemistry.It is clear that pace has not been kept with the general resurgence. When one sees the facilities for teaching analytical chemistry abroad or, for that matter, even in some of the polytechnics, compared with what our own universities can provide, it is surprising that graduates ever desire to follow analytical chemistry as a profession. There are also still many universities in Britain that do not teach analytical chemistry. I should not need to list all of the industries and newer industries that are so dependent on chemical analysis. It is necessary for monitoring the various health hazards, it is essential in clinical chemistry, in oceanography, in space research and so on. Above all, there need to be chemists available who are capable of devising new methods to meet the ever increasing problems that arise.There has been a resurgence in analytical chemistry because its ancient and venerable partner, industrial chemistry, has also expanded. Dr. Chalmers’ observation then raises a further query-has there really been a resurgence apart from teaching in the universities! There has in other countries because of the great expansion in the number of Chairs of Analytical Chemistry in the last quarter of a century. I have stated elsewhere that most of the great developments in analytical chemistry have originated in universities and, despite the many handicaps, the British universities have made significant contributions which reach back into the last century. The question can be answered affirmatively with hardly any reservation; the resurgence has even had an impact on the attitude of British universities.In my Plenary Lecture5 delivered at the Centenary Celebrations of the Society for Analytical Chemistry in London in 1974, I referred to the impressive amount of analytical research work then proceeding in British universities and I indicated the contributions made a t the annual Research and Development Topics meeting in that year. I pointed out that no less than 17 universities had contributed papers to that meeting. In the pre-War period and in the immediate post-War period, not more than three universities could have contributed. At the end of the Second World War there were probably only two Lecturers in Analytical Chemistry in the whole of Britain. Now there axe probably a dozen Readers in Analytical Chemistry and at least twice as many Senior Lecturers.Although the effect of the resurgence on the universities may not have been as far-reaching as some of us would have liked, it cannot be denied that considerable strides have been made. The resurgence in British universities arose mainly because the dangers of lack of training in this subject became apparent during the Second World War. When certain leading scientists who had been engaged in the War effort obtained influential positions in universities, they exerted that influence to extend the teaching of analytical chemistry. This expansion did occur, of course, in other countries, but there was already a sound basis on which to build. Another significant reason for the resurgence of analytical chemistry was the great development in new techniques, most of which were in instrumentation.As this area expanded, so did specialisation. Moreover, very often specialists in certain branches of instrumentation were chemists from other disciplines, brought on to a particular instrument because of its great value in providing analytical data for that particular discipline. As a result, the boundaries of analytical chemistry (or, as some may prefer to call it, analytical sciences) have expanded a t a great pace. If we try to establish the causes of the resurgence or renaissance of analytical chemistry, there are many reasons, some of which may not yet be completely apparent, but we can say that they are due to (a) the expansion of industry and technology in general and the associated problems that arose from that expansion, such as pollution; (b) the Second World War showed the necessity for research analytical chemists, both in Britain, where the requirement was the most urgent, and abroad; and (G) the great expansion in new analytical techniques.It has been said that chemical industry could abandon its old partner completely, for by36 BELCHER exact computerised and other automated controls at the preparative stage it could produce materials that fell within the required specifications with machine-like precision. This situation might well be reached in the not tmoo distant future, but computers and other automated equipment, as all of us know, can also go wrong, and I cannot visualise any responsible industrialist selling his products unless there has been careful analytical checking of its specification.Even if he were to take that risk, one would hope that the law would compel him to subject his product to chemical analysis. As industry and technology expand, the need for analytical chemistry will continue. However, one must utter a word of warning. There are many analytical chemists of wide interest, experience and knowledge still available. They are capable of assessing which particular technique is the best for a particular purpose and within a particular set of conditions. But they are from another generation, and they will not automatically be replaced. The high degree of specialisation is increasing and it would be disastrous for analytical chemistry if only specialists eventually remained. They would be unable to judge objectively whether or not their own technique was the best for a particular analysis. This danger can be avoided only as long as the universities continue to produce postgraduates who have had a sound general training in analytical chemistry. I thank Dr. W. I. Stephen for a number of fruitful discussions, and also Mr. Solomon Jackson for undertaking much of the literature and historical research. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References SzabadvAry, F., “The History of Analytical Chemistry,’’ Pergamon Press, Oxford, 1966. Belcher, R., Chemy. Brit., 1976, 12, 12. Belcher, R., Proc. Analyt. Div. Chem. SOC., 1977‘. 14, 161. Belcher, R., “Proceedings of the International Symposium on Microchemistry, 1958,” Pergamon Belcher, R., Analyst. 1974, 99, 802. Smythe, L. E., ANalyt. Lett., 1976, 9, vii. Belcher, R., Analytica Chim. Ada, 1976, 86, 1. West, T. S., Proc. Analyt. Div. Chem. Soc., 1977, 14, 177. Stephen, W. I., Proc. Analyt. Div. Chem. Soc., 1977, 14, 183. Chirnside, R. C., and Hamence, J , H., “The ‘Practising Chemists’,’’ Society for Analytical Chemistry, Briscoe, H. V. A., and Matthews, J. W., “Microchemical Methods,” Irzstitute of Chemistry Mono- Chalmers, R. A., Plenary Lecture delivered a t the International Symposium on Microchemistry, Press, Oxford, 1958, p. 658. London, 1974, p. 97. graph, Institute of Chemistry, London, 1934. Davos, May 22-27, 1977.

ISSN:0003-2654    

DOI:10.1039/AN9780300029

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, January, 1978, Vol. 103, pp. 37-55 37 Highlights from Chromatography Contemporary Analytical Liquid Plenary Lecture* R. P. W. Scott Chemical Research Department, Hoffmann-La Roche Inc., Nutley, N.J., USA Recent work on the mechanism of solute - phase interactions is discussed and used to explain the mechanism of distribution between the mobile and stationary phases, Examples of solute - solvent - stationary phase inter- actions are considered for silica gel, non-polar bonded phases and ion-pair chromatography. The use of silica gel for exclusion chromatography is also described in conjunction with the preparation of very high efficiency liquid- chromatographic columns and the trend towards compatible low dead volume detectors. A practical liquid chromatograph - mass spectrometer system is described in detail and results from commercially available liquid chromato- graph - mass spectrometer equipment are given.The automation of liquid- chromatographic analysis involving automatic injection devices and computer data acquisition is discussed and the precision that can be obtained from such equipment given. Finally, an example of the use of high precision as an alternative to resolution in a liquid-chromatographic separation is demon- strated. Keywords : Liquid chromatography ; solute - solvent - stationary phase inter- action ; high-eBciency columns ; liquid chromatograph - mass spectrometer Liquid chromatography was discovered over 75 years ago but it is only in the last decade that the technique has developed to an efficient, effective separation method and analytical procedure.The renaissance of liquid chromatography was catalysed by the introduction of sensitive, linear, in-line detection systems. Such detectors permitted the accurate measurement of the concentration profiles of an eluted peak and from the results obtained, paved the way to the development of modern high-efficiency columns. It is ironical to note, however, that today column technology has surpassed detector development and the modern liquid-chromatographic detectors are again beginning to impede the progress of column technology, Solutes eluted from the high-efficiency columns available today have band widths of only a few microlitres. This volume is commensurate with that of the cell and connecting tubes of contemporary detectors.Thus , in many instances, the efficiencies realised from a given column are significantly below those which the column is capable of producing, owing to a loss in efficiency caused by dispersion in the detector system itself. The high efficiencies obtained from modern chromatographic columns have resulted from the careful development of the stationary phases employed. Most stationary phases used in liquid chromatography are based on silica gel and it would be difficult to imagine liquid chromatography, in its present form, without the novel characteristics of silica gel. Silica gel is a polymer of silicic acid and is usually produced by the acidification of sodium silicate solution. Hydrolysis of silyl esters can also be used but the most common method of preparation is from sodium silicate.On treatment of a sodium silicate solution with acid, orthosilicic acid is released and immediately starts to condense with itself and produce primary particles of silica ge1.l These primary particles can be up to 10 pm in diameter. On achieving a certain size these primary particles begin to condense with one another, forming the hydrogel. The size of these primary particles is very important because they control the surface area and the porosity of the final product. The smaller the primary particles, the higher is the surface area and the smaller the pore diameter. The size of these primary particles is controlled during the production of the hydrogel by adjusting the pH with suitable buffers or by changing the temperature.Thus, the method of preparing the hydrogel can be extremely important from the chromatographic point of view. After the * Presented at the Fourth SAC Conference, Birmingham, July 17th t o 22nd, 1977.38 SCOTT HIGHLIGHTS FROM CONTEMPORARY Analyst, Vol. 103 hydrogel has been formed it is converted into the xerogel by heating it a t 120 "C for a number of hours. During this heating period further condensation of the primary silica particles occurs and the result is the rigid silica gel that is familiar to most chromatographers. The xerogel itself can be further modified by treatment with water vapour at high pressure or by treatment with an electrolyte, which can increase the size of the primary particles and thus reduce the surface area of the resulting silica gel and increase its pore diameter.2 It follows that silica gels used in liquid chromatography can be tailored to provide specific chromatographic properties for specific applications.Many manufacturers provide silica gels with a range of surface areas, which have been shown3 to provide a range of loading capacities, different selectivities and a range of scope. Silica gel is now frequently used as an exclusion medium in exclusion chromatography and this development has been possible because the pore diameters of the silica gel can be controlled and silicas having different ranges of pore diameters can be selected for specific separations.* Just over a decade ago the efficiencies obtainable from 25 or 50 cm long columns were of the order of 200-800 theoretical plates and the,y were packed with silica gel having particle diameters ranging between 30 and 60 pm. It was well known that smaller particle diameters would provide higher efficiencies but at that time they could not be packed efficiently.Majors5 and Kirkland6 developed the slurry packing technique for microparticulate silica and the column efficiencies available increased1 by an order of magnitude. Today, 25-cm columns having efficiencies varying between 8 000 and 12 000 theoretical plates are readily available. The diameter of the particles of the packing is now 5 or 10 pm. The introduction of stable bonded phases Wa:; the next important development in column packings. These were obtained by reacting appropriate silanyl reagents with the hydroxyl groups on the silica gel to provide a permanently bonded organic moiety.If the organic moiety is a hydrocarbon chain then a reversed-phase packing is obtained. However, the bonded phase moiety can be polar or contain ion-exchange groups and thus production of ion-exchange packings based on a silica gel is possible. Today a wide choice of stationary phases is available, giving characteristics that range from highly polar, as with silica gel, to dispersive, with the hydrocarbon bonded phase. The introduction of the bonded phase has opened up the technique of liquid chromatography to even wider fields of application. Interactive Mechanisms Silica Gel Silica gel can display a wide range of activities depending on the nature of the solvent employed as the mobile phase.It has been shown7~* that in the presence of a hydrogen- bonding solvent such as ethyl acetate, tetrahjTdrofuran or propan-2-01, the polar solvent hydrogen bonds to the silica. In fact, the interacting surface is no longer that formed by the silanol groups of the silica gel but rather the surface occupied by the hydrogen-bonded solvent. In effect, the absorbed layer of polar solvent on the stationary phase is acting as a hydrogen-bonded phase. In Fig. 1 the adsorption isotherm of ethyl acetate on silica gel is shown as the concentration of ethyl acetate in grams per gram adsorbed on the silica 0.1 5 1 0 5 10 15 20 30 40 Concentratiom of ethyl acetate in the mobile phase, % m/V D Fig. 1. Adsorption isotherm for ethyl acetate on silica gel. Solvent, heptane.January, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 39 against the concentration of ethyl acetate in heptane.It is seen that, initially, the silica gel removes the ethyl acetate from the mobile phase forming the coating. A t about a level of 0.3% of ethyl acetate in heptane the curve starts to flatten and at the 3% level it is almost horizontal. It has been showns that solutes eluted from a column a t k' values up to 10 do not displace the ethyl acetate but merely associate with it. It is uncertain whether the solute is interacting directly with the ethyl acetate or is interacting with the silanol groups owing to long-range forces; however, the latter appears to be unlikely. For substances incapable of hydrogen bonding to the silica gel, such as chloroform and dichloromethane, a layer is still built up but becomes stable only at significantly higher concentrations.The hydrogen bonding between the polar solvent and the silica gel has been shown7 to cause difficulties when used with gradient elution. If a chloroform - heptane solution is initially used as the mobile phase to elute a sample mixture and then propan-2-01 - heptane is introduced, the large change in polarity resulting from the hydrogen bonding of the propan-2-01 on to the silica gel displaces the solutes and results in them being eluted as a bunch close together almost in the form of displacement development. It follows that for effective gradient elution a range of solvents needs to be employed with each solvent having a small polarity step between itself and the next.A series of 12 solvents has been suggested and proved to be satisfactory for gradient elution between the extremes of heptane and water.' The 12 solvents, together with their k' values with respect to heptane, are shown in Table I and two examples of the use of these solvents for the separation of blood lipids and some polypeptides are shown in Figs. 2 and 3. The separations were carried out employing a 25 cm x 4.6 mm i.d. column packed with 10-pm Partisil and a flow-rate of 2 ml min-l. Each of the solvents in Table I was passed through the column for 5 min to give a 1-h chromatogram. Very effective separations can be obtained for substances that have a wide polarity range and few or no displacement steps can be seen in the chromatogram. The first group of peaks in the blood-lipid chromato- gram consists of glycerides, the second group consists largely of cholesterol and cholesterol esters and the final group is made up largely of sphingolipids.In the series of solvents used there are several that are completely opaque to ultraviolet radiation. Thus, the ultraviolet detector could not be employed and these solvents had to be used with the moving-wire transport detector. This detector, although bulky, expensive and relatively insensitive (approximately g ml-l), can tolerate any solvent in the gradient elution system provided that it is reasonably volatile. This layer is approximately one molecule thick. TABLE I BASIC SOLVENTS USED FOR INCREMENTAL GRADIENT ELUTION AND k' VALUES WITH RESPECT TO HEPTANE Solvent Carbon 1,2-Di- Metha- tetra- Chloro- chloro- 2-Nitro- Nitro- Propyl Methyl Solute Heptane chloride form ethane propane methane acetate acetate Acetone Ethanol no1 Water Heptane ..0 Carbon tetrachloride: , . 0.144 Chloroform . . .. 0.651 1,2-Dichloroeth'ane . . . . 1.800 2-Nitropropane . . . . - Nitromethane .. .. - Propyl acetate _. .. - Methyl acetate .. . . - Acetone .. _ . .. - Ethanol .. .. .. - Methanol .. _ . .. - Water . . - Average value of k'&)jk'(n$;) 2.34 - - - - - - - - - 0 0.286 0 0.750 0.233 0 - - 0.450 0.300 0 - 2.15 0.612 0.148 - - 2.812 1.003 - - - 1.162 - - - - - - 0 0.485 0 0.565 0.232 1.082 0.638 - 1.483 - - - - - 2.06 1.72 - - 0 0.370 0.863 1.161 1.56 - - - - 0 0.512 0.812 1.463 1.96 - - - - 0 0.366 2.48 - - 0 0.402 0.882 2.48 - - - - 3.22 3.34 2.44 2.06 Once the polar solvent contained in the mobile phase is in excess of 3%, the surface of the silica gel is completely covered and its properties remain constant at all concentrations of polar solvent in excess of this value.It follows that the interactions of any solute with the stationary phase will remain constant and the effect on solute retention, by increasing the concentration of polar solvent, will be due entirely to increased interactions in the mobile phase. The nature of the interactions will remain the same but, as the concentration of polar solvent increases, the probability of interaction increases. The distribution coefficient,40 SCOTT : HIGHLIGHTS FROM CONTEMPORARY Analyst, Vol. 103 and thus the retention of a solute, is directly proportional to the interactions in the stationary phase and inversely proportional to the interactions in the mobile phase.The corrected retention volume reflects the value of the distribution coefficient and thus, as the probability of interaction increases in the mobile phase with the concentration of polar solvent, the L Fig. 2. Chromatogram of blood lipids. Fig. 3. Chromatogram of protected di-, tri-, penta- and octapeptides. corrected retention volume will decrease. It has been shown1* that the reciprocal of the corrected retention volume under these circumstances is directly related to the concentration of polar solvent in the mobile phase. Fig. 4 shows graphs relating the reciprocal of the corrected retention volume of a series of solutes with the concentration of polar solvent. This reciprocal relationship is closely held and permits the retention of any substance to be predicted for any concentration of polar solvent, provided that the retention at two known concentrations is available.This relationship also holds for thin-layer chromatography and in Fig. 5 the reciprocal of the capacity ratio of a series of solutes calculated from their R, values is shown plotted against solvent concentration. For retained peaks the same linear I I 0.4 0.3 0.2 0.1 0 2 4 6 8 Amount of ethyl acetate in heptane, % rnl V Fig. 4. Graphs of the reciprocal of the corrected retention volume against the concentration of ethyl acetate in the mobile phase. A, 2-Ethylanthra- qumone ; B, 2-methylanthraquinone : C, anthraquinone ; D, methylphenyl- methanol; E, benzyl alcohol; and F, 3-phenylpropan- 1-01.1 .o 0.8 Q" 0.6 2 0.4 0.2 - oc" u 1 II -.. c 0 5 10 15 20 25 30 35 Mobile phase composition, amount of ethyl acetate in heptane, % m/V Fig. 5. Graph of 1 /k' against concentra- tion of polar solvent in the mobile phase from thin-layer separation of dyes. A, 4-Dimethylaminoazobenzene ; B, di- hydroxyazobenzene ; and C , benzo- quinone-4-h ydrox yphen ylimine.January, 19 78 ANALYTICAL LIQUID CHROMATOGRAPHY 41 relationship is shown but for the solute moving close to the solvent front the relationship breaks down at low concentrations of polar solvent. This effect is caused by the polar solvent being extracted by the silica gel and thus the character of the mobile phase is continually changing.Solutes that travel close to the front of the solvent in thin-layer development are, in effect, being developed under gradient elution conditions. Provided that the interactions in the stationary phase are constant then the probability of interaction in the mobile phase will be conditioned by the concentration of the interacting moieties. Thus, in ion-exchange chromatography, at a constant pH, the reciprocal of the corrected retention volume of any solute will be linearly related to the buffer concentration and this relationship is demonstrated in Fig. 6. 1 .o 0.8 0.6 5 --. P 0.4 0.2 I 1 I 0‘ 0.1 5 0.20 Concentration of potassium dihydrogen orthophosphate,% mlV Fig. 6. Graph of 1/V’ against buffer concentration for different solutes. Column, Partisil 10 SAX; pH, 4.80.A, Adenosine 3’,5’-cyclic monophosphate ; B, uridine 5’-mono- phosphate ; C, cystidine 5’-monophosphate ; D, adenosine 5’-monophosphate ; E, uridine 2’,3’-diphosphate ; F, adenosine 2’,3’-diphosphate ; and G, guanosine 5’-monophosphate. The introduction of bonded phases to the field of liquid chromatography has expanded its capability greatly and many mixtures that were not amenable to separation by silica gel can now be separated on bonded phases. One of the most useful bonded phases is the reversed phase, which is a hydrocarbon chain linked directly to a silicon atom, which is itself linked via an oxygen atom to a silicon atom on the surface of the silica gel. The chromatographic properties of a hydrocarbon bonded phase vary extensively with the chain length of the hydrocarbon moiety.All of the hydroxyl groups on the surface of silica gel cannot be reacted, owing to steric hindrance, and it appears that a maximum of about 70% of the hydroxyl groups available can be linked to the hydrocarbon moiety. In Fig. 7 chromato- grams are shown from three different bonded phases all derivatised to the extent of about 70%. Chromatogram (a) is from a C, chain bonded phase, chromatogram (b) from a C, chain bonded phase and (c) from a C,, chain bonded phase. The solutes are a mixture of phenol, benzene, ethylanthraquinone, methylanthraquinone and anthraquinone. For bonded phases having maximum derivatisation the selectivity and retention increase with the carbon chain length. Fig. 8 shows separations of the same mixture but with bonded phases that have different degrees of derivatisation and different carbon contents but all having a C,, carbon chain.Chromatogram (a) is from a bonded phase having only 5.0% of carbon and thus it is very lightly derivatised. Selectivity and retention are relatively poor. Chromatogram ( b ) , from a bonded phase that has a carbon content of 16.6y0, shows a greater retention than chromatogram (c), which was from a bonded phase that had 19.8% of carbon. Chromatograms (b) and (c) demonstrate the effect of different types of derivatisation. It is probable that the bonded phase with 16.6% of carbon was derivatised using octadecyl-42 SCOTT HIGHLIGHTS FROM CONTEMPORARY Analyst, Vol. 103 L 0 10 20 Retention ti me/m i n Fig. 7. Chromatograms demonstrating the effect of hydrocarbon chain length on solute retention.Mobile phase : water in acetonitrile (35% V / V ) . Solutes : phenol, benzene, anthraquinone, methylanthraquinone and ethylanthraquinone. (a), Carbon chain length 2, carbon content 5y0, OH groups reacted 68%; ( b ) , carbon chain length 8, carbon con- tent 12.2%, OH groups reacted 75% ; and (c), carbon chain length 18, carbon content 19.8%, OH groups reacted 68%. LAAL 15 0 Retention time/min Fig. 8. Chromatograms demonstrating the effect of the carbon content of C,, bonded phases on solute retention. Mobile phase : water in acetonitrile (35% V / V ) . Solutes : phenol, benzene, anthraquinone, methylanthraquinone and ethylanthraquinone. (a), Bonded phase ODs, carbon content 5.0%; ( b ) , bonded phase ODS 2, carbon content l6.9Y0 ; and (c), bonded phase RP 18, carbon content 19.8%.0 15 Retention timehin 3 Fig. 9. Forward- and reversed-phase chromatograms from an incompletely deriva- tised octadecyl bonded phase. (a), Mobile phase, water in acetonitrile (35% V / V ) ; solutes, benzene, anthraquinone, methyl- anthraquinone; and ( b ) , mobile phase, ethyl acetate in heptane (7% m/ V ) ; solutes, benz- ene, acetophenone, 2-phenylethanol and benzyl alcohol.January, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 43 trichlorosilane in the presence of adsorbed moisture on the surface of the silica. Under these conditions a polymeric or bulk materialll is formed and this phase obviously exhibits greater selectivity and greater retention. The bonded phase with 19.8% of carbon was of the brush-type material12 and was probably made from octadecyldimethylmonochlorosilane with the aid of hexylmethyldisilazane as catalyst.Chromatogram (a), from material con- taining only 5% of carbon, shows that it contains a considerable amount of unreacted silanol groups and thus could act to some extent as a normal gel column. Fig. 9 shows the same bonded phase used both in the reversed-phase and forward-phase manner. Effective separ- ations are obtained in both modes of operation although, when used as a silica gel column, some peak tailing is observed. This result introduces the interesting concept of a packing that could be used for both forward- and reversed-phase separations. The bonded phase, partially derivatised and optimally reacted to provide forward- and reversed-phase separa- tions, could be extremely useful for the analyst who wished to determine the best mode of separation.It is far easier to change solvents than to change columns and therefore optimum conditions could be determined rapidly. TABLE I1 WETTING CHARACTERISTICS OF FIVE REVERSED PHASES Chain length Carbon content, of bonded Reversed phase % m/m material ODS . . . . 5.0 18 ODSZ.. . . 16.9 18 R P 2 . . . . 5.0 2 R P 8 .. . . 12.2 8 KP 18 . . .. 19.8 18 Maximum water content of solvents to permit complete wetting, yo V / V f A \ Methanol Acetonitrile Propan-2-01 100 100 100 32 64 74 50 76 84 55 68 82 50 68 84 Solute interactions with the surface of a hydrocarbon phase depend upon whether the mobile phase wets the surface or whether the water content is high enough to produce a hydrophobic system.In Table 11, the wetting concentrations of a series of commonly used reversed phases for the solvents methanol, acetonitrile and propan-2-01 are given. Under al - .- 5 Amount of acetophenone in mobile phase, % m/V Amount of acetophenone in mobile phase, % m/V Fig. 10. Adsorption isotherm for acetophenone between R P 18 reversed phase and water in aceto- nitrile (40.4% m/V) at 25 OC. Line A shows calculated concentration level if all of the acetonitrile was dis- placed by the solute into the mobile phase.44 SCOTT : HIGHLIGHTS FROM CONTEMPORARY Analyst, VoZ. 103 conditions of wetting it appears that a layer of the solvent one molecule thick coats the surface of the bonded phase.13 Further, those solutes eluted at a k’ value up to 10 do not displace the solvent layer.In Fig. 10 the concentration of acetophenone on the RP 18 reversed phase is plotted against the concentration of acetophenone in the mobile phase. The mobile phase composition was such that it wetted the stationary phase. It is seen that as the concentration of solute on the stationary phase increases up to a level of about 25 mg g-1, the concentration of acetonitrile in the mobile phase remains constant. This means that no acetonitrile has been displaced by the solute into the mobile phase. Under circumstances where displacement did occur the concentration of acetonitrile in the mobile phase would have increased. It has also been shown13 that if an ion-pair reagent is introduced into a mobile phase that wets the reversed phase it resides solely in the mobile phase, Under such circumstances ionic interactions are introduced in the mobile phase, which elutes the solute ion more rapidly.Under non-wetting conditions, however, ion-pair reagents are adsorbed on to the stationary phase and act as an adsorbed ion exchanger. Under these conditions an appropriate ionic solute will interact with the stationary phase and thus its retention is increased in the presence of the ion-pair reagent. 100 al - 5 8 0 - 9 60- a Lc a 40 m rrJ c. C g 2 0 - 0, a 0 - Exclusion Chromatography Silica gels can be obtained with a wide range of pore diameters and pore volumes and therefore can be used very effectively for exclusion chromatography. They can be manu- factured to provide pore diameter ranges from 1 to 10 nm and from 10 to 500 nm.In Fig. 11 graphs relating the pore distribution as a percentage of the total pore volume for a number of different commercially available silica gels are shown. There appear to be three groups of adsorbents: Partisil 10, LiChrosorb 10, Biosil A, Biosil HA and Sil-LC, all having pore diameters ranging from about 1.5 to about 15 nm; Silarex 2 and Porasil A together with CPG 10 have pore diameters ranging between 2 and 50 nm; and Porasil C and Spherosil 10 have pore diameters that range from about 5 to about 400 nm. - - lo-’ 100 10’ 1 o3 Fore diarneter/nm Fig. 11. Graphs showing pore size distribution as percentage of total pore volume for different silica gels, The permeation volume of a silica gel column is only approximately half the dead volume; thus, for a 25 cm x 4.6 mm i d .column, having a dead volume of 3 ml, the total chromato- graphic volume is only 1.5 ml. I t follows that only a limited number of substances can be separated within the total retention volume of 1.5 ml. Thus, for silica gel columns to be effective for exclusion chromatography, high efficiencies must be obtained, so that as many peaks as possible can be fitted into this relatively small retention volume. Fortunately, silica gel columns can provide very high efficiencies so they can be used successfully for exclusion chromatography. Although other types of packing can provide much higher permeation volumes the efficiencies from such packings are very much less than silica gel and the columns have over-all lower resolution.One drawback to silica gel as an exclusion medium is its instability in the presence of electrolytes and at extremes of pH. As statedJanuary, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 45 above, this instability causes a modification of the silica gel, the packing shrinks and the efficiency is lost. This effect is particularly unfortunate as many of the substances of interest that could be separated by exclusion chromatography, for instance biological materials, polypeptides and proteins, can be maintained in solution only if electrolytes are present. Polystyrene gels have also been used for exclusion chromatography but do not produce columns of very high efficiency. The controlled-pore glass packings are now available in a range of pore sizes and can, in microparticulate form, produce high-efficiency columns but have pore volumes even lower than that of silica gel.Recently, ion-exchange resins have been manufactured in microparticulate form and can provide efficiencies similar to that obtained from silica gel. Although not an exclusion medium they have found extensive use in the separation of materials of biological origin. High - efficiency Columns Most columns today are supplied 25 cm in length and 4.6 mm i.d. and from such columns efficiencies of between 6 000 and 12 000 plates can be expected, depending on whether they are packed with 5 or 10pm diameter particles. If the columns are packed well, they can be connected in series and Kirkland14 described columns providing an efficiency of 25 000 theoretical plates, which were connected in this way.Kirkland used these columns for exclusion chromatography where, as already stated, high efficiencies are essential. Connec- ting columns in series, however, does not always produce a linear increase in efficiency with column length and, even if it does, this process is usually limited to about three or four 25-cm columns in series. The reason that the combined efficiency does not increase linearly with column length appears to be due to variations in the permeability across the diameter of the column. If the permeability across the diameter of the column is not homogeneous, more mobile phase will flow through certain portions of the cross-section than others and this difference produces a serious multi-path effect.It can, in 1-in diameter columns that are badly packed, produce two distinct peaks from a single solute. In preparative columns this effect is usually caused by part of the solute band travelling near the wall of the column and part travelling down the centre of the column; as the permeabilities at the edge are different from those at the centre, two discrete peaks are often eluted. To provide very high efficiency columns, therefore, it is necessary either to pack very homogeneous columns or to reduce the diameter of the columns to that level where the effects of inhomogeneity are minimal. Scott and Kucera15 packed 1 m x 1 mm i.d. columns with Partisil 20 (20 pm particle diameter). Each column was tested and then connected in series to produce one 10-m column.This column was found to provide 250 000 theoretical plates a t a flow-rate of 25 p1 min-l, which was equivalent to an HETP of 40 pm, i.e., the theoretical limit. A flow- rate of 25 p1 min-l through a 1 mm bore column is equivalent to a linear velocity of about 0.6 mm s-l, which is the same velocity one would obtain by applying a flow-rate of 0.5 ml min-l to a column 25 cm long x 4.6 mm i d . It should be pointed out that a column with a dead volume of about 7 ml having an efficiency of 250 000 plates will provide a peak with a standard deviation of only 28 pl at the dead volume. I t follows that detectors having volumes of 8 or 10 pl cannot be used with such columns. The 1 mm bore columns were employed with an LDC ultraviolet monitor, the cell of which had been modified and reduced to 2.5 p1 in volume and the connecting tubes to 1 p1 in volume.For columns having higher efficiencies than this, smaller detector volumes would be necessary. For the normal 25 cm x 4.6 mm i.d. column packed with 5 pm diameter particles and giving about 12 000 theoretical plates with a dead volume of 3 ml the standard deviation of the dead volume peak will be about 55 pl. It follows that even for those columns which are commercially available, detectors having volumes of 8 or 10 pl or more can seriously impair the column efficiency and the true performance of the column will not be realised. I t is hoped that detector design will be improved in the near future so that they can accommodate the advanced column technologies that are being developed.In Fig. 12 a chromatogram is shown of the separation of a series of alkylbenzenes on the 10 m x 1 mm i.d. column. This separation is achieved entirely by exclusion and the solutes have relative molecular mass differences that are equivalent to only two carbon numbers. It is seen that these peaks are well resolved and it would be possible to differentiate between two substances that have single carbon number differences. The relative molecular mass46 SCOTT : HIGHLIGHTS FROM CONTEMPORARY Analyst, Vd. 103 Fig. 12. Chrornatogram of a series of alkylbenzenes from a column of 250 000 theoretical plates. Column, 10 m x 1 mm; packing, Partisil 20; mobile phase, tetrahydrofuran ; flsw-rate, approxi- mately 30 p1 miri-l. Solutes: benzene, ethylbenzene, butyl- benzene, hexylbenzene, octyl- benzene and decylbenzene.I I l . I I k-4 A L A i i 'B' Fig. 13. Chromatogram of cinnamon bark oil from a column of 160 000 theoretical plates. Column, 10 m x 1 mm; packing, Partisil 20; mobile phase, ethyl acetate in heptane (3% m / V ) ; sample volume, 0.5 pl; and flow-rate, 38 p1 min-'.January, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 47 of decylbenzene is 218 and thus one methylene group would represent a 6.4% discrimination in relative molecular mass. High-efficiency columns can be used in the normal elution mode and in Fig. 13 a chromatogram demonstrating the separation of cinnamon bark oil is shown. The chromatogram is reminiscent of those obtained by capillary-column gas chromato- graphy but obviously can only depict those substances that absorb in the ultraviolet region.The insert shows the chromatogram expanded between 120 and 140min. It can be seen that extremely useful separations can be obtained and that the resolving power of the column is maintained over short retention distances. The chromatogram shown in Fig. 13 was obtained at a flow-rate of 50plmin-l, which gave an efficiency of about 160000 plates. The second insert shows the enlarged minor peak eluted late in the chromatogram. It demonstrates that the symmetry of the peak is maintained up to a k’ valve of nearly 20. Scott and Kucera’s work, demonstrating columns of 250 000 plates, indicates that columns with lo6 plates are possible. It remains to be seen whether lo7 plates or more are equally feasible.Much of the work today in liquid chromatography is carried out adequately with columns having efficiencies of 6 000-12 000 plates. However, for multi-component mixtures high efficiencies are still required. Today, such efficiencies are novel and are required only in a minority of circumstances. As the technique develops and expands into the field of biological materials, very high efficiency columns will no longer be a luxury but a necessity. It is well known that liquid chromatography has three attributes : speed, load and resolution. The development time for the high-efficiency column may be 20-30 h and thus the high efficiency is achieved by sacrificing load and speed. However, to improve column technology, it is necessary first to obtain the efficiencies and then attempt to reduce the time required for them.In chromatographic terms the high-efficiency column produces about 23 plates per second but this is largely the result of using the relatively large particle diameter of 20 pm. Liquid Chromatograph - Mass Spectrometer Systems The introduction of liquid chromatography - spectroscopic systems was a natural develop- ment following the prior complementary development of gas chromatograph - infrared and gas chromatograph - mass spectrometer systems. However, the majority of solvents used in liquid chromatography absorb strongly in the infrared region, the use of liquid chromatography - infrared systems has limited possibilities for structural elucidation of eluted solutes. The combination of the liquid chromatograph with the ultraviolet spectrometer was the most straightforward and easiest to accomplish.Commercial models providing on-line ultraviolet spectral monitoring of h u i d chromatograDh eluates are available, but unfortu- nately the ultraviolet spectrum has very limited use for structural analysis. The most useful spectroscopic technique that can be coupled with the liquid chromatograph is the mass spectrometer. The mass spectrometer has adequate sensitivity, can provide spectral data rapidly and is very useful in structural elucidation. However, there is the problem of taking the solute contained in a solvent at atmospheric pressure and introducing it into the ion source of the mass spectrometer at a pressure of about mmHg. Two successful approaches have been developed, one by McLafferty and co-workersl6--ls and the other by Scott et aZ.19920 McLafferty and co-workers took a portion of a column eluate and fed it directly into the mass spectrometer, the solute being accompanied by the mobile phase.The solute and mobile phase were vaporised and the solvent used as the chemical ionisation agent to provide chemical ionisation spectra of the solute. The system worked satisfactorily providing the solute had sufficient vapour pressure and the solvent was suitable for the production of chemical ionisation spectra. The disadvantages of the system were that the solute had to be reasonably volatile, which excluded a large proportion of the solutes normally separated in liquid chromatography. Further, the system restricted the range of solvents that could be used and excluded gradient elution development.The alternative method developed by Scott et al. utilised a wire trans- port system to bring the sample into the mass spectrometer. The eluate was allowed to flow over the wire, leaving a coating of mobile phase and solute adhering to it. The mobile phase was evaporated, leaving the solute coated on the wire, which was then passed through the ion source of a quadrupole mass spectrometer and the solute thermally vaporised. The quadrupole mass spectrometer had to be employed as the electrical potential of the source was close to earth potential and the earth wire would not interfere with the ion optics of the source. Owing to the high voltages involved in the conventional mass spectrometer ion48 SCOTT HIGHLIGHTS FROM CONTEMPORARY Analyst, "01. 103 source, an earthed wire would produce electrical breakdown.Success of the wire transport system depended on a suitable interface that would permit a wire to pass from atmospheric pressure, through the ion source and out again, in a continuous manner while still permitting pressures of lov6 mmHg to be maintained in the ion source. The over-all system is shown in Fig. 14 (a and b) and one of the two interfaces used is shown in Fig. 14 (c). The interface consists of two chambers separated and terminated by ruby jewels, each of which has a 0.01-in hole through the centre. Stainless-steel wire was used as the carrier, and the jewels were necessary to eliminate abrasion due to contact at the surface of the apertures.The first chamber was connected to a rotary pump, which reduced the pressure in it to about 1 mmHg. The second chamber was connected to an oil diffusion pump backed by a rotary pump, which reduced the pressure to about 5 pmHg. In this way the pressure difference between the final aperture and the ion source was only 5pmHg and permitted the ion source to be maintained at mmHg. The advantage of this system was that any solvent could be used in the chromatographic development providing it was volatile and, further- more, electron-impact spectra could be obtained, which are far more useful for structure elucidation than chemical ionisation spectra. An example of the use of the liquid chromato- , MS body Flange I I MS source It MS f l a n g e d /" ring seal Interface body / Vacuum p o r t 1 dl \ --+ To ion source N \ O_ ring seal ire Internal Cacuum p o r t Fig.14. (a), Layout of wire train system. (b), Diagram of ( G ) , The liquid interfaces located in mass spectrometer source. chromatograph - mass spectrometer interface.January, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 49 graph - mass spectrometer system is shown in Fig. 15. The chromatogram (a) was obtained by using incremental gradient elution employing the moving-wire detector. The fractions numbered were collected, evaporated to small bulk and spotted on to the moving wire of the liquid chromatograph - mass spectrometer system. The diagram (c) shows the results of spotting these samples on the wire using a total ion current monitoring procedure and in (e) peak 7 is picked out by monitoring on mass 327 a characteristic of the solute eluted.The gradient elution was then repeated using liquid chromatography with the mass spectro- meter system as the detector and chromatogram (b) was obtained from total ion current monitoring. The separation is not as good owing to the lower sensitivity of the mass spectro- meter system requiring a larger charge. The chromatogram obtained by monitoring on mass 327 is shown in Fig. 15 (d), which again picks out the characteristics of solute 7 in the chromatogram (a). The system was shown to function well and in addition to behaving as an effective liquid chromatograph - mass spectrometer system also worked well as a rapid automatic probe sampling device. The disadvantage of the system was that the sensitivity was limited and was only about 4 x g ml-1.(a 1 Fraction 1 2 2 7 Fig. 15. (a), Chromatogram of a fermentation extract by incremental gradient elution (IGE). (b), Total ion current chromatogram by liquid chromatography - mass spectrometry of fermentation extract, (c), Total ion current trace of fractions sampled directly on to wire. ( d ) , Ion current trace for mass 327 from liquid chromatography - mass spectrometry chromatogram. (e), Ion current trace for mass 327 from fractions sampled directly on to wire. The system was further developed by McFadden et aLZ1 and their modifications are now proved and the device is commercially available. The interface system used is shown in Fig. 16. A continuous ribbon loop is used as a transport system as opposed to the wire- spool device.The advantage of the ribbon system is that it can take much more of the column eluate and therefore more solute into the mass spectrometer. In the wire-transport device the column eluate is sampled at a rate of 10 pl min-l whereas the ribbon device will sample a t a rate of 850 pl min-l. This increase means that the entire eluate can be taken on to the transport system if a microbore column is used. The same type of vacuum lock interface is used except that the jewelled orifices are now designed to take a ribbon as opposed to a wire. The solvent is removed in the vacuum locks and the solute passes on to the ion source. In the ion source the ribbon passes over a flash vaporiser in the form of a heated coil and the vapour diffuses directly into the electron beam.The ribbon then passes by another heater coil, which burns off any residue and cleans the ribbon prior to passing through the column eluent again. A photograph of the interface system devised by McFadden et al.50 SCOTT HIGHLIGHTS FROM CONTEMPORARY Analyst, Vol. 103 LC effluent vaporiser source Clean-up heater Fig. 16. Diagram of liquid chromatograph - mass spectrometer interface system. and now marketed by Finnigan Inc. is shown in Fig. 17. This system retains the advantages of the wire transport system in that it permits the use of any solvent or gradient elution and provides electron-impact spectra. At the same time it has increased the sensitivity of the over-all system to 10-7-10-8gml-1. In Fig. 18 are shown two chromatograms, (a) obtained from the ultraviolet monitor and (b) obtained from the total ion current measured by the mass spectrometer.The solutes eluted numbered 1 , 2 , 3 and 4 are propoxur, carbaryl, malathion and aldrin, respectively. In Fig. 18 (c) are shown the spectra obtained from each Malathion Background - d . ' k - 1, 149 Fig. 18. (a), Chromatograms of propoxur (l), carbaryl (2), malathion (3) and aldrin (4) obtained by use of an ultraviolet monitor. (b). The same four peaks as in (a), obtained from total ion current measured by mass spectrometer. ( c ) , The spectra of each of these four peaks as plotted from the mass spectrometer. The graph for malathion was recorded a t one tenth of the sensitivity used for the other four traces.Fig.17. Liquid chromatograph - mass spectrometer interface system.January, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 51 peak as presented by the plotter from the data handling system of the mass spectrometer. It can be seen that clear and unambiguous mass spectra are obtained that easily permit identification. The spectrum at the bottom of Fig. 18 (c) gives an indication of the back- ground and this can usually be eliminated, provided that suitable software is available, by subtracting the background from the spectrum of each peak. The liquid chromatograph - mass spectrometer system manufactured by Finnigan has only recently become available and its use in the field has yet to be established. However, the advantages of such a system are obvious and it is likely to find considerable use in the field of liquid chromatography as gas chromatograph - mass spectrometer systems have found use in the field of gas chromato- graphy.The liquid chromatograph - mass spectrometer transport system acts as a very effective rapid-probe sampling device and the total samples spotted on to the wire that are shown in Fig. 15 were carried out in just over 25 s per sample. This rapid sampling rate linked with the speed of modern data handling systems of mass spectrometry means that a probe injection sample can be made to provide a complete mass spectrum for identification in 2-3 min. One particular advantage of the system should again be emphasised. Automation of Liquid-chromatographic Systems and Analytical Precision In the past year, automatic liquid-chromatographic injection systems have been made available which permit 20-30 liquid-chromatographic analyses to proceed automatically and continuously, overnight if necessary.Basically, the system consists of a device similar to that used in gas chromatography where the samples are contained in a number of phials. The phials are regularly sampled on a time basis and the sample either drawn up into a sample valve or forced into it via positive gas pressure on the phial. When the sample valve is filled it is automatically actuated and the sample injected on to the column. The automatic injection devices are best used in conjunction with computer data acquisition systems, which will then provide the analytical report automatically printed out. Precise results, however, can be obtained only if the chromatographic system is designed to control the necessary variables within the required limits and, secondly, if the data acquisi- tion system is suitably organised.Retention data can be obtained with good precision and repeatability only if both the temperature and the solvent concentration are made satisfactorily constant and if a good quality pump is employed that provides a consistent flow-rate. By determining the change of retention times with solvent concentration the tolerance of the solvent mixture composition can be assessed to give a specified precision. In Table I11 tolerances in solvent composition are given for a 44% m/V solution of 1-chlorobutane in heptane to maintain retention time precisions of 1 and 0.1%.The solutes examined have retention times ranging from 2.85 to 17.3 min, which are equivalent to values of k' for the three solutes of 0, 1.8 and 6.06, respectively. It can be seen from the table that for a 1% precision the solvent concentration has to be kept constant to &O.lyo. For precisions of 0.1% the tolerance in the solvent composition is &O.Olyo. In a similar manner the effect of temperature on retention time precision can be determined and in Table IV the temperature control to obtain 1 and 0.1% precision is given. For 1% precision the temperature must be maintained constant to h0.35 "C and for a 0.1% precision the temperature must be maintained constant to k0.04 "C. Thus, very stringent temperature control is necessary and for high precision, the column and detector should be thermostatically controlled.In the author's experience an air thermostat is inadequate and the column has to be immersed in a thermostat bath. The precision of the pump is directly related to the precision of the retention times TABLE I11 SOLVENT CONCENTRATION TOLERANCES FOR RETENTION TIME PRECISION Retention time Concentration Concentration at 44% V/V of tolerance for tolerance for 1-chlorobutane in lo/ recision, 0.1 recision, Solute heptane/min "$ m / v GPm,V 4-Chlorophenetole . . .. .. 2.85 f 0.14 &0.014 2-Methylnaphthalene .. .. 5.01 kO.10 ,to.010 1,2-Dinitrobenzene . . .. .. 17.27 fO.10 fO.01052 SCOTT : HIGHLIGHTS FROM CONTEMPORARY Analyst, Vol. 103 TABLE I V TEMPERATURE TOLERANCES FOR RETENTION TIME PRECISION Corrected retention volume Temperature control Temperature control Solute at 23.8 "C/ml k' for 1 yo precision/"C for 0.1 % precision/"C 4-Chlorophenetole .. 3.072 0.946 f0.35 f 0.04 2-Methylnaphthalene . . 4.925 1.519 f 0.35 f 0.04 1,2-Dinitrobenzene . . 17.185 5.301 f 0.33 k0.03 obtained. In the author's experience the Waters 6000A pump can, if operated with care, provide flow-rates that are constant to within &0.04% over a period of many hours. This means that the pump would deliver a flow-rate of 1 ml min-l with a standard deviation of 0.4 p1 min-l. The precision of the analytical results obtained from a liquid-chromatographic analysis with computer data acquisition is also dependent on the sampling rate of the computer and the noise level of the system.High-efficiency columns available today will provide peaks close to the dead volume having a peak width of only a few seconds. It follows, therefore, that the sampling rate of the data acquisition system controls both the precision of the efficiencies measured and also the peak areas. In Fig. 19 the discrimination in efficiency possible at different computer sampling rates is shown for a standard 25 cm x 4.6 mm i d . column, having a dead volume of 3 ml and an efficiency of 5 000 theoretical plates, and operated with a flow-rate of 1 ml min-l. It can be seen that if an efficiency discrimination of 1-2% is required the computer sampling rate must not be less than 20 samples per second. Most minicomputers available today can sample detector output at rates in excess of 1 000 per second. To provide a wide linear dynamic range the computer has to sample via a multi-ranging amplifier and the multi-ranging amplifiers available today limit the frequency of sampling to about 250 samples per second.This sampling rate has to be shared between whatever number of chromatographs are associated with the computer. Thus, if the computer is time-sharing between 10 liquid chromatographs the maximum sampling rate that can be taken from any chromatograph is 25 data points per second. There is a further limit on the sampling rate imposed by the frequency of the electrical supply. If data are sampled at or in excess of the electrical supply's frequency considerable noise occurs and, as will be seen later, this noise can also affect the precision of measurement.Thus for a 50-H~ supply, irrespective of the number of instruments being time-shared with the computer, the maximum sampling rate is 50 samples per second. However, it is not the computer that limits the sampling rate that is possible. I I I 1 .- 5 10 15 20 n Computer sampling ratehamples s-' 4.4 s Time - Fig. 20. Peak crests re- Fig. 19. Graph of minimum effici- constructed by the computer ency discrimination against computer (99.9-100% peak height). data sampling rate. Column, 25 cm x The marker bar A shows the 4.6 mm i.d. ; dead ;volume, 3 ml; flow- actual difference between the rate, 1 ml min-1; k' of solute, 1.0; and retention times as 2.1 s and column efficiency, 5 000 theoretical B shows the time as recorded plates. by the computer as 4.4 s.The next important factor that can affect precision when using computer data acquisition Experiments to determine the precision of computer measurement of is signal noise.January, 1978 ANALYTICAL LIQUID CHROMATOGRAPHY 53 retention time included 12 replicate analyses. The peak tops between 99.9 and 100% of the peak height of the two most diverse values are shown in Fig. 20. The centres of the peaks are only 2.1 s apart, but the computer measured between the two maximum values and, as there was a noise spike on the front of peak 1 and a noise spike on the rear of peak 2, the difference between the two retention values became 4.4 s. I t follows that it is extremely important to eliminate noise either by the use of active filters at the analogue output of the detector or exponential software smoothing by the computer.In Table V the precisions of retention time and peak-width measurements are shown for three peaks eluted at k’ values TABLE V PRECISION OF RETENTION TIME AND PEAK WIDTH MEASUREMENTS Parameter k‘ value .. .. .. .. Retention time : Mean/min . . .. .. .. Standard deviationis . . .. Mean/min . . .. .. .. Standard deviationls . . .. Standard deviation, yo of the mean Peak width : Standard deviation, yo of the mean Peak 1 . . 0.94 . . 6.283 . . 0.38 . . 0.10 . . 0.1691 . . 0.24 . . 2.38 Peak 2 1.50 8.119 0.20 0.04 0.225 1 0.06 0.45 Peak 3 5.21 20.421 0.46 0.04 0.575 3 0.19 0.55 of 0.94, 1.50 and 5.21. These results were obtained from 12 replicates taken on the same day over a period of about 5 h.It can be seen that the standard deviations of the retention times of the peaks vary between 0.04 and 0.1%. For the third peak, having a retention time of 20.421 min, the total spread of the results was about kO.9 s. The standard devi- ation of the peak width is somewhat greater but even so the total spread of peak 3 in time was h0.4 s at a mean value of 34.52 s. In Table VI the precision for the analysis of a mixture by normalisation of peak heights and peak areas is shown. The precision of analysis is slightly better for peak heights than peak areas, except for the last peak. Using peak-area analysis samples present at the level of 0.63, 7.49 and 91.9% could be measured, the total spread of values being *0.06, h0.14 and &0.16%, respectively. In Table VII the standard deviation for the analysis of the same mixture within days and between days is given.The standard deviation of retention time data is very similar between days and within days. This is true also for peak-height analysis. However, the standard deviation for peak-area analysis is significantly worse between days than within days, particularly for peaks eluted later in the chromatogram. TABLE VI PRECISION OF THE ANALYSIS OF A MIXTURE BY NORMALISATION OF PEAK HEIGHTS AND PEAK AREAS Parameter k’ value .. .. .. * . Analysis by peak heights : Mean . . .. . . .. . . Standard deviation . . .. .. Standard deviation, % of the mean Mean . . .. . . . . .. Standard deviation . . .. .. Standard deviation, yo of the mean Analysis by peak areas : . . .. .. .. .. .. .. Peak 1 0.94 1.937 0.046 5 2.46 0.633 0.032 5.071 Peak 2 1.50 16.491 0.121 0.736 7.486 0.072 0.97 Peak 3 5.21 81.574 0.148 0.18 91.884 0.082 3 0.09 Use of High Precision as an Alternative to Resolution If retention times can be determined with sufficient precision, retention data can be used to determine the composition of a mixture of two solutes that are completely unresolved. If the two solutes are run separately and their retention times accurately measured, the retention time of the composite peak of any mixture will fall between the two extremes for54 SCOTT : HIGHLIGHTS FROM CONTEMPORARY TABLE VII Analyst, Vol. 103 REPEATABILITY OF RETENTION TIMES, PEAR-HEIGHT AND PEAK-AREA MEASUREMENTS FOR TWELVE REPLICATE SAMPLES TAKEN OVER A 4-DAY PERIOD Parameter* k' value .... Retention time : Meanlmin . . .. T,, min . . .. Tt9 Y?. * .. .. T,/min . . .. T d i % .. .. T , . . .. .. Tt, %. * .. . . T , .. .. .. T,, % .. .. T , . . .. ,. Tt, % - .. . . Td . . .. .. T d , % .. .. Peak-height analysis : Mean . . .. .. Peak-area analysis : Mean . . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. * . .. .. .. .. .. Peak 1 0.94 6.286 0.009 3 0.15 0.0096 0.15 1.958 0.098 6 5.04 0.080 2 4.09 0.634 0 0.050 8 8.0 0.043 8 6.90 Peak 2 1.50 8.121 0.007 4 0.09 0.009 1 0.11 16.484 0.088 3 0.54 0.206 5 1.25 7.457 0.054 6 0.73 0.1549 2.08 Peak 3 5.21 20.929 0.018 6 0.08 0.024 6 0.12 81.558 0.145 3 0.18 0.226 4 0.25 91.890 0.062 6 0.07 0.156 0.17 * Standard deviation of the difference between two determinations on the same day: T , (actual), Ti,% Standard deviation of the difference between two determinations on different days: T , (actual), Tdr% (percentage).(percentage). the pure substances. Thus, a calibration graph can be obtained relating the retention time of the composite peak to the percentage composition of the mixture. Such a graph is shown in Fig. 21 for a mixture of nitrobenzene and deuteronitrobenzene. Nitrobenzene is eluted a t 8.927 min and deuteronitrobenzene at 9.061 min, the difference being about 8 s. As the percentage of deuteronitrobenzene increases the retention time of the composite peak increases from that of nitrobenzene to that of pure deuteronitrobenzene. The retention time of the composite peak can be calculated by means of the plate theory from the retention time of the individual peaks and the efficiency of each peak.However, in practice the peaks are often not symmetrical and as the rear of the first eluted peak combines with the front of the second eluted peak two different efficiency values have to be used. The efficiency of the first peak would be taken as equivalent to the second half of the first peak and the efficiency of the second peak would be taken as equivalent to the first half of the first peak. Amount of deuteronitrobenzene, % m/m Fig. 21. Graph of retention time difference against sample composition for different mixtures of nitrobenzene and deutero- nitrobenzene. The times of 8.927 and 9.061 min correspond to elution of nitro- benzene and deuteronitrobenzene, respec- tively.Janzcary , 19 78 ANALYTICAL LIQUID CHROMATOGRAPHY 55 Then, using the Gaussian function to describe the profile of the elution curve and with the aid of the computer, the profile of the composite curve can be calculated for different mixtures and the retention time a t the peak maximum determined.The method of calculation has been described in For Fig. 21 it should be pointed out that the curves are theoretical and the points were obtained experimentally. that 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Future Developments in Liquid Chromatography Any prediction on future developments in a field that is changing as rapidly as liquid chromatography can only be an extrapolation of present trends.A single innovative concept could change the whole pattern and course of the development of the technique. However, extrapolating the present progress would indicate further advances in column technology and columns of millions if not tens of millions of plates could become available in the not too distant future. Detectors of improved design to cope with such columns will also be pro- duced with hopefully greater sensitivity, wider linear dynamic range and improved stability. Unique packings specially designed for particular separation problems will no doubt be developed and the bonded phases presently available made more stable, thus providing higher column efficiencies. The nature of the solute - solvent interactions in liquid chromato- graphy are beginning to be understood and it is likely that the choice of phase systems for specific separations will, in the future, be arrived a t on a far more rational basis. The value of the liquid chromatograph-mass spectrometer system has yet to be established but its future looks very promising and the system will no doubt be developed to provide higher sensitivities and automatic operation. It is also possible that liquid chromatograph - Raman spectro- meter systems may be developed using the laser Raman spectrometer in pulsed form and with Fourier transform analysis. Such a system, however, requires significant advances in Raman spectroscopy before it could become commercially available. Probably the most exciting future for liquid chromatography lies in its development towards effective use in the separation of macromolecules and biological materials. The biologists and bio- chemists anxiously await the million-plate column that will be stable in the presence of electrolytes and over a range of pH to help them handle the difficult separation problems face them. References Linsen, B. G., Editor, “The Physical and Chemical Aspects of Adsorbents and Catalysts,” Academic Linsen, B. G., Editor, “The Physical and Chemical Aspects of Adsorbents and Catalysts,’’ Academic Scott, R. P. W., and Kucera, P., J . Clwomat. Sci., 1974, 12, 473. Scott, R. P. W., and Kucera, P., J . Chromat., 1976, 125, 215. Majors, R. E., Analyt. Chem., 1972, 44, 1722. Kirkland, J. J., J . Chromat. Sci., 1971, 9, 206. Scott, R. P. W., and Kucera, P., Analyt. Chem., 1973, 45, 749. Lawrence, J. G., and Scott, R. P. W., J . Chromat. Sci., 1970, 8, 619. Scott, R. P. W., and Kucera, P., J . Chromat. Sci., to be published. Scott, R. P. W., J . Chromat., 1976, 122, 35. Unger, K. K., Becker, N., and Roumeliotis, P., J . Chromat., 1976, 125, 116. Halasz, I., and Sebastian, J., Angew. Chem., Int. Edn, 1969, 8, 453. Scott, R. P. W., and Kucera, P., J . Chromat., in the press. Kirkland, J. J., J . Chromat., 1976, 125, 231. Scott, R. P. W., and Kucera, P., J . Chromat., 1976, 125, 251. Baldwin, M. A., and McLafferty, F. W., Org. Mass. Spectrom., 1973, 7, 1111. Arpino, P., Baldwin, M. A., and McLafferty, F. W., Biomed. Mass. Spectrum., 1974, I, 80. Arpino, P., Dawkins, B. G., and McLafferty, F. W., J . Chromat. Sci., 1974, 12, 574. Scott, R. P. W., Scott, C. G., Munroe, M., and Hess, J., “The Poisoned Patient: The Role of the Scott, R. P. W., Scott, C. G., Munroe, M., and Hess, J., J . Chromat., 1974, 99, 395. McFadden, W. M., Schwartz, K. L., and Evans, S , J . Chromat., 1976, 122, 389. Scott, R. P. W., and Reese, C., J . Chromat., 1977, 138, 203. Press, London and New York, 1970, p. 215. Press, London and New York, 1970, p. 229. Laboratory,’’ Elsevier, New York, 1974, p. 155.

ISSN:0003-2654    

DOI:10.1039/AN9780300037

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

56 Analyst, January, 1978, Vol. 103, pp. 56-67 Hybrid Methods of Analysis Plenary Lecture* Yu. A. Zolotov Vernadsky Institute of Geochemistry and Analytical Chemistry, Academy of Sciences, Moscow 117334, USSR Methods of separation and determination are usually combined and their combinations are diverse. The separation and determination stages are not often connected closely, but in a number of instances the properties of a separation product (concentrate) have a significant effect on the determina- tion. Such a combination is not a simple sum of separation and determination techniques; sometimes virtually new methods arise. An example is the combination of solvent extraction and flame atomic-absorption spectrophoto- metry (if the extract is introduced directly into the flame).There is a tendency to carry out the separation and determination in the same device (gas chromatography, high-performance liquid chromatography). In gas chromatography a group of analytical methods was developed that were hybrids of separation and determination techniques. New methods of such a type have been developed in this and in other laboratories. These methods include a combination of solvent extraction pre-concentration and spark-source mass spectrometry. Trace elements from a sample are collected on high-purity alumina after separation of the matrix by solvent extraction; the alumina is then arialysed by mass spectro- metry. Copper oxide has been used as a collector for the determination of trace amounts of noble metals. Other hybrid methods have been suggested that are based on combinations of solvent extraction with polarography and spectrochemical analysis.Organic chelate-forming reagents, which are free radicals, have been used in a combination of separation and electron-spin resonance spectroscopy; this allows the latter technique to be made a universal analytical technique and the sensitivity to be increased. High- performance liquid chromatography can be used for inorganic analysis in combination with solvent extraction. Some new reagents have been pro- posed for the group pre-concentration of trace elements. Keywords ; Hybrid methods of analysis ; pre-concentration ; physical methods of analysis The science of analytical chemistry is constantly changing, and the most important of these changes is the rapid increase in the application of physical techniques, especially spectro- scopic and nuclear methods.In this situation, it is desirable to emphasise the interdependence of physical and chemical methods. In particular, physical methods of analysis often benefit by combining them with chemical methods of separation and pre-concentration. Physical and partly physico-chemical methods are, in many instances, distinguished by speed, objectivity of the results and the possibility of automation. There are many physical methods of multi-element analysis, some of which possess high sensitivity and adequate accuracy and precision. However, there are no ideal techniques. Almost a general limita- tion of physical methods is the necessity to use standard reference materials; sometimes the preparation of a representative sample for a determination by physical methods is difficult ; the selectivity, sensitivity, precision and accuracy are frequently not adequate.On the other hand, analytical chemists have accumulated great experience in the preparation of samples for determination, particularly in separation and concentration techniques. The combination of separation and concentration with a subsequent determination by instru- mental techniques has proved to be one of the most promising trends in modern chemical analysis. Two groups of analytical techniques can be distinguished, involving the combination of either separation or concentration with determination.1 * Presented at the Fourth SAC Conference, Birmingham, July 17th to 22nd.1977.ZOLOTOV 57 In the techniques in the first group, the concentration and determination steps are not connected closely and more or less independent methods are used for each. The properties of a concentrate do not have a substantial effect on the determination and it is of little importance to the determination whether the concentration was carried out by, for example, ion exchange, solvent extraction with subsequent back-extraction or any other method. The analyst will have a concentrate in the form of an aqueous solution that reflects the historv of sample treatment only to a small extent and such a solution can be analysed by different techniques. I t is clear that combinations of such a type (which have been widely used for a long time) do not have any substantial peculiarities that need to be discussed in detail.However, sometimes the properties of the concentrate have a considerable effect on the determination and specific peculiarities of the combination arise in these instances as we are no longer dealing simply with the successive use of two methods. For example, whereas during the flame atomic-absorption analysis of aqueous solutions it is not very important which method was used to separate the elements being determined, during the atomic- absorption analysis of organic extracts important peculiarities arise, connected with the different viscosities of organic solvents in comparison with that of water, with their com- bustibility and with changes in the mechanism of processes in the flame.These peculiarities are so pronounced that it is necessary somehow to distinguish such methods from the usual ones. A similar situation arises in the combination of solvent extraction with photo- metric determination if the determination is carried out directly in the organic phase. In recent years, attempts have been made to combine an extraction separation with direct polarography of the extracts. It is obvious that the polarography of such organic solutions differs from the polarography of aqueous solutions. We can also speak about the very close combination of separation and determination in a different sense. There are examples of the combination of processes of separation and determination in one instrument such as in a gas chromatograph, and a similar combination is found in modern high-pressure liquid chromatographs.The need for some means of marking out the methods of the group under consideration is most clearly exemplified by gas chromatography, which can be treated as a method of separation but is also used for the identification of the components of a mixture and for their determination. The com- bined methods of the second group may be called hybrid methods2 The development of these methods reflects the general trend to combine distinct analytical operations. Laitinen3 wrote about this trend, “The classical approach to chemical analysis involves a series of distinct operations, including sampling, sample pre-treatment, adjust- ment of conditions, separation, measurement and data processing. While these distinct steps are still useful for didactic purposes, and often even in practice, there is a pronounced trend towards their consolidation in modern analytical methods.” For example, it is sometimes import- ant to carry out the concentration (separation) as rapidly as possible after sampling. In some instances, it is desirable in practice to combine these operations. We face such a situation when samples are too large in volume and it is difficult to transport them. Also, some samples are unstable during storage and their composition and properties vary. This applies, in particular, to natural waters. In the analysis of high-purity materials, the combination of the decomposition of a sample and separation and determination steps reduces systematic errors, especially when the operations or some of them are carried out in one small, closed ~ y s t e m .~ This makes it possible to eliminate errors due to the transfer of substances from one system to another. There are a number of stimuli to such an integration. Some Advantages of Preliminary Separation and Concentration It is useful to note some peculiarities of separation and especially pre-concentration methods that make them an important part of hybrid techniques5 Pre-concentration is used if the relative detection limit ensured by the analytical technique is higher than the concentration of trace elements in the sample. Concentration allows one to separate the matrix or the bulk of it and sometimes a number of interfering micro- components also. In a prepared concentrate, the relative concentration of trace elements58 ZOLOTOV: HYRRID METHODS Analyst, VoI.103 is usually higher than in the initial sample. Moreover, the possibility of increasing the size of the sample being analysed allows one to increase the absolute amounts of the elements to be determined. As a result, it is possible to reduce the detection limit of trace elements (sometimes very significantly, e.g., 100- or 1 000-fold). It is the main but not the sole reason for the wide use of concentration techniques. Preliminary concentration is virtually essential if trace elements are distributed inhomo- geneously in the material. In this instance, a representative sample must be large and it is difficult to analyse it directly, especially if the determination method requires a small sample, such as spark-source mass spectrometry or spectrochemical analysis.A classical example is the concentration of gold and other noble metals, which are distributed inhomogeneously in ores, rocks, etc., from very large samples of natural materials by the fire assay technique. In many other instances, pre-concentration with preliminary dissolution and the production of a small concentrate facilitates the preparation of a representative sample. Of course, homogeneity of the sample can also be achieved during other operations of its treatment. Concentration facilitates the calibration, especially if standard reference materials are lacking. It makes it possible to obtain identical concentrates during the analysis of materials of various compositions, for example concentrates on carbon powder in spectrochemical analysis. Reference samples are prepared as concentrates of such a type.In this instance, there is no need to have standard reference materials for all of the materials being analysed. Preliminary concentration with exhaustive removal of the matrix is desirable in the analysis of toxic, radioactive or very expensive materials. Moreover, it is convenient to add elements as internal standards, if necessary, during decomposition of the sample and concentration. Sometimes, pre-concentration enables one to increase the number of trace elements that can be determined by a selected technique or provides for the possibility of using this determination technique for a particular application. These advantages of separation and pre-concentration make them an important aspect of analysis.In spite of the progress made with sensitive instrumental methods of direct analysis, the significance of concentration has not diminished. On the contrary, its possibilities have been increased, particularly owing to new combinations with methods of determination. In the following sections we consider some established combinations of separation and pre-concentration techniques with some important methods of determination. Gas and Liquid Chromatography The most widely known methods of the type considered are gas chromatography in combination with various methods of detection, including mass spectrometry, and high- performance liquid chromatography. It is not necessary to describe these well known techniques in detail, but only to note that chromatographic separations can be combined with new detection systems.Thus, in addition to gas chromatography - mass spectrometry, a new method has been developed that combines high-performance liquid chromatography with mass spectrometry. Horning6 has used mass spectrometry with ionisation at atmo- spheric pressure, the system including a liquid chromatograph, a mass spectrometer and a computer. We have combined high-performance liquid chromatography, including the use of a spectrophotometric detector, with preliminary solvent extraction of metal chelates (diethyldithiocarbamates) . The use of high-performance liquid chromatography in inorganic analysis appears to be a very promising field for investigation.An atomic-absorption spectrophotometer has been used as a detector with gas and liquid chromatographs. This combination is suitable for the analysis of organometallic and inorganic compounds. Such a combination increases the possibilities of atomic-absorption spectrophotometry because it permits the determination of individual forms of elements. The combination of gas chromatography with atomic-absorption spectrophotometry was described for the analysis of volatile organometallic compounds, especially those of mercury and lead.'~* Recently, the combination of high-performance liquid chromatography with atomic-absorption spectrophotometry was used for this purpo~e,~ a common atomic-absorption spectrophotometer being combined with an ordinary high-pressure liquid chromatograph.January, 1978 OF ANALYSIS 59 Combination of Concentration and Spectrochemical Analysis There are some close combinations of concentration techniques with spectrochemical analysis.1° One uses pre-concentration when the direct analysis does not enable one to attain the necessary relative detection limit, as well as in other instances that have been considered above.It is necessary to separate the matrix if it hinders the determination of trace elements as a result of overlapping of spectral lines or an unfavourable excitation potential of the bulk element. It is useful to convert the samples into a single physico- chemical form in order to utilise a universal series of reference samples. In addition, the increase in the concentration of trace elements in the concentrate in comparison with the sample obviates the necessity to use unreliable reference samples with very low contents of the elements to be determined.Fluctuations in results because of an uneven distribution of trace elements in the sample decrease. The group concentration of all elements to be determined is advantageous in these methods. If the emission spectrum of the matrix elements is sufficiently simple and the physico- chemical properties of the matrix are favourable for the spectrographic determination of trace elements, it is not necessary to separate it completely and an enrichment suffices. In any event, the concentrate should be in a form convenient for direct excitation of the spectrum of trace components: they must be concentrated in a collector for small masses or on a small surface area of the electrode.What are the necessary properties of the collector (matrix of the concentrate) in spectro- chemical analysis? It should be a substance that is stable during storage, non-hygro- scopic, and with a simple emission spectrum of its component elements. The material of the collector must be easily obtained in a reproducible form as a result of a small number of simple operations. It is desirable that it should have a relatively low volatility, which favours the uniform supply of trace elements into the discharge zone. Carbon (graphite) powder satisfies these requirements. Many methods of chemical - spectrographic analysis based on the use of this collector have been developed. Methods elaborated in our laboratory (Table I) may serve as an example.Materials analysed High-purity silver Steel Aqueous solutions Sodium iodide Solutions after decomposition of ores and concen- trates Aluminium, gallium ; indium phosphide * Parts per lo9. TABLE I SOME CHEMICAL - SPECTROGRAPHIC METHODS Pre-concentration method Extraction of matrix Extraction of trace elements Extraction of trace elements Extraction of trace elements Extraction of trace elements Sorption on CuS and chelate resin Elements concentrated AI, As, Ba, Be, Bi, Ca, Cd, Co, Cr, Cu, Fe, Ga, In, Mg, Mn, Ni, Pb, Sn, Te, T1, V, Zn, Zr Bi, Cd, Cu, Pb, Sb, Sn, Zn Cd, Co, Cr, Cu, Fe, Mn, Mo, Ni, Pb, Ti, V Ag, Al, Bi, Ca, Cd, Co, Cr, Cu, Fe, Mg, Mn, Mo, Ni, Pb, Sb, Sn, Ti, Zn Ag, Au, Ir, Pd, Pt, Rh, Ru Ag, Au, Ir, Pd, Pt, Rh, Ru Co-precipitation on Bi,S3, Ag, Au, Cd, Co, Cr, Cu, Mn, Ni, Pb, Sn, Zn Bi(OH),, In& Detection limit, p.p.b.* Reference 1-20 11 10-1 000 12 13 - 1-100 14 10-1 000 15 1-100 16 1-100 17 There is a specific technique for concentration in spectrochemical analysis, namely the vaporisation method.This technique was developed by Zaidel et aZ.l* and Mandelstam et aL19 for application in the analysis of pure nuclear materials in two variants : vaporisation in air and under vacuum. The high melting-point powdered sample is placed in a carbon (graphite) crucible and heated with a high current in a small graphite furnace, clamped between graphite blocks on water-cooled copper electrodes. The vapours are condensed on60 ZOLOTOV: HYBRID METHODS Analyst, Vol.103 cooled graphite or metal caps, which then act as the electrode for the arc or spark during spectrographic determination. A double-stage vaporisation can be used for the analysis of samples of average volatility, for example silicon. A cooled graphite crucible serves as the receiver in the first stage. In the second stage, trace elements are evaporated from the crucible to the cap electrode. This method ensures a more complete removal of the matrix. In a number of studies, the volatile matrix, e.g., molybdenum or chromium oxides, was preliminarily converted into non-volatile carbides. Combination of Separation Methods with Atomic-absorption Spectrophotometric Determination The combination of preliminary separation and concentration with atomic-absorption spectrophotonietry is most frequently accomplished in the form of solvent extraction - flame atomic absorption with direct nebulisation of the extract into the flame.The advantage of such niethods in comparison with the determination of elements in aqueous solution is that lower detection limits and more rapid and simpler analyses are achieved. Operations such as evaporation of the extracts and back-extraction are avoided, which leads to a decrease in the correction for .the blank experiment during determination of abundant elements. Sometimes, when a flammable organic solvent is used, a combustible gas is not required. The selectivity of atomic-absorption determinations allows one to concentrate a group of trace elements into the extract. Therefore, group extraction is used together with the extraction of individual elements. The extraction of chelates of quinolin-8-01, /3-diketones, xanthates and especially dithiocarbamates is widely used for this purpose.Chloroform, carbon tetrachloride and benzene are not suitable as solvents for flame atomic-absorption spectrophotometry, 4-methylpentan-2-one and butyl acetate being more satisfactory. The system ammonium tetramethylenedithiocarbamate - 4-methylpentan-2-one is of great importance for group extraction of trace elements; many methods for the analysis of water, biological and other materials have been developed with the use of this system. Instead of ammonium tetramethylenedithiocarbamate, hexarnethyleneammonium hexa- metliylenedithiocarbamate (I) has been used successfully.The reagent is synthesised from hexamethyleneimine, which is a by-product of nylon manu- facture and is readily available.20 Several methods have been elaborated with the use of this reagent, for example for the determination of cobalt, copper, iron, manganese, nickel, lead and zinc in sea water.21 Some other extractants for the separation and concentration of elements before atoniic- absorption determination have been suggested in our laboratory. Dialkyltin dinitrates, e.g., dinonyl- and dioctyltin dinitrates (11), possess the unique power oE extracting highly charged 0x0 anions ; these extractants extract best triply charged phosphate and arsenate ions and worst (although satisfactorily) doubly charged selenite and selenate ions ; singly charged anions are poorly e ~ t r a c t e d .~ ~ ~ ~ ~ Such organotin extractants differ from all other known extractants. By using these extractants, procedures have been developed for the pre-concentration of phosphorus, arsenic and selenium. I I l l Triphenylphosphine (111) has been proposed for the selective pre-concentration of trace amounts of gold and silver followed by atomic-absorption determination of gold in minerals and cyanide solutions24 and silver in rocks, ores and minerals.25 The peculiarity of this reagent consists in its tendency to react with metal ions that are Pearson's weak acids.26 1-Pheriyl-3-methyl-4-benzoylpyrazo1-5-one (IV) can be used for the group concentration of some elements.27 It is also used for the selective extraction of some elements prior to atomic-January, 1978 OF ANALYSIS 61 absorption determination.28 More than 100 studies have been carried out with this reagent, and the subject has recently been revie~ed.~’ Diphenylthiourea (V) and hexamethylene- phenylthiourea have been suggested for the solvent-extraction concentration of elements that form chalcogenides, particularly silver and the platinum group metals.The reagents make it possible to concentrate noble metals and to separate them from non-ferrous metals. Diphenylthiourea has been used for the pre-concentration of trace amounts of silver in rocks, ores and concentrate^^^ and copper, silver and thallium in chemical reagents30; the elements were determined by atomic-absorption spectrophotometry in both instances. Devices for automatic extraction concentration have been suggested; in particular, the use of Technicon instruments allows 60 samples per hour to be handled.31 Another variant of automatic extraction concentration has been proposed, in combination with a six-channel atomic-ff uorescence spectr~photometer.~~ The so-called “injection” method for the atomic- absorption determination of small amounts of substances and for the analysis of concentrates is also A system for drying aerosols has been suggested a t Moscow Univer~ity,~~ making it possible to decrease the detection limits for elements and to analyse extracts containing chloroform, diethyl ether, benzene and other solvents that are considered as unsuitable for flame atomic-absorption analysis.Some methods for the atomic-absorption determination of trace elements in various materials have been developed in our laboratory (Table 11).The peculiarities of the combination of solvent extraction with flame atomic-absorption spectrophotometry have been considered in more detail in a monograph37 and in r e v i e ~ s . ~ * s ~ ~ TABLE I1 EXAMPLES OF ATOMIC-ABSORPTION DETERMINATIONS WITH PRELIMINARY EXTRACTIOK CONCENTRATION Elements Main extraction Detection limit, determined Materials analysed reagent p.p.m. Reference Rocks, ores, concentrates Diphenylthiourea 0.01 29 Metal halides (chemicals) Diphenylthiourea 0.001-0.1 30 Ag, Cu, T1 Rocks, ores, minerals Triphenylphosphine 0.01 26 Minerals, cyanide solutions Triphenylphosphine 1 (2 x lo3 pg ml-l) 24 Au Cu. Fe, Mo, V, Zn Titanium(1V) chloride Trioctylamine 0.02-0.2 36 Sb, Sn, Zn Ag Ag Bi, Cd, Cu, Pb, Steel Trioctylamine 0.1-1 36 Combinations of solvent-extraction concentration and atoinic-absorption determination with electrothermal atomisation can also be used, e g ., with determination in a graphite furnace. In this instance, the natures of the solvent, extractant and type of compound extracted are of importance for the sensitivity and precision of the determination. The elements to be determined after the direct introduction of extracts into the furnace must not form products that are too volatile and difficult to atomise during removal of the solvent and thermal decomposition of the extractants and extracted species. We have studied the possibility of determining gallium in the HGA-74 furnace of the Perkin-Elmer Model 503 instrument after its extraction with different extractants from hydrochloric acid solutions. Under the conditions chosen, gallium can be determined in extracts with better sensitivity than in hydrochloric acid solutions owing to the separation of excess of acid and macro- components of the sample.Solvent Extraction and Spark-source Mass Spectrometry Pre-concentration in a new combination with spark-source mass spectrometry is used in62 ZOLOTOV: HYBRID METHODS Analyst, Vol. 103 our laboratory. After separation of the matrix, trace elements are collected on the collector (high-purity alumina). For the separation of trace amounts, we have used solvent extrac- tion. In the determination of eight elements in steel,40 they were extracted with 4-methyl- pentan-Zone from iodide solution, then back-extracted with dilute sulphuric acid containing hydrogen peroxide.The solution obtained was evaporated in the presence of alumina. After calcination, the concentrate was analysed by using a mass spectrometer. Such a method has been elaborated for the determination of elements as impurities in high-purity sulphuric acid. Another variant has been suggested for the determination of trace amounts of noble metals in natural and industrial materials.41 The noble metals were extracted as complexes with hexamethylenephenylthiourea (VI) from a hydrochloric acid solution containing copper VI and other non-ferrous metals, the copper being simultaneously extracted as copper(1). The extract was evaporated and calcined, during which a concentrate on copper oxide was obtained that was then analysed.The detection limits of the platinum metals in the concentrate are 0.01-0.1 p.p.m. at a concentration coefficient of 50; the relative standard deviation is 0.18-0.38. Results for the analysis of copper are shown in Table 111. Copper does not interfere in the determination. TABLE I11 RESULTS FOR THE ANALYSIS OF COPPER (STANDARD REFERENCE SAMPLE) Relative Element Certificate Results of standard determined value, p.p.m. analysis, p.p.m. deviation Rh 1.6 1.36 0.27 Pd 16 14.6 0.22 I r 1.6 1.62 0.32 Pt 6.0 4.7 0.18 For spark-source mass spectrometry, the advantages of concentration mentioned above are of importance. A weighed sample of 10-20mg is taken for analysis by direct spark- source mass spectrometry, and it is difficult to make this sample representative.However, we can take a sample of 1 g or even larger for concentration, and the concentration method ensures that the sample becomes homogeneous. The 20 mg of alumina that are used for analysis are reasonably representative. A similar effect also occurs in standardisation. Standard reference samples of the numerous materials that may be analysed become unnecess- ary; it is sufficient to have samples of alumina that become a universal standard. Sorption Concentration and X-ray Fluorescence Determination In some instances, the combination of separation and pre-concentration with X-ray fluorescence determination is applied. This combination is especially important for the analysis of liquid samples.Usually, a suitable method has to be found for the preparation of solid samples. The most appropriate technique for this purpose is sorption of elements to be determined on a sorbent that can then be used for direct analysis by the X-ray fluorescence technique. In this instance, the concentrate has only to be pressed into a tablet. Suitable sorbents for this purpose include cellulose containing functional groups of high selectivity, for example 1-(2’-hydroxyphenylazo)-2-maphthol (VII), 4-(pyridy1)azoresorcinol (VIII)42943 and diamines (IX).44January, 19 78 OF ANALYSIS ,O H OH 63 V I I Vlll IX Leyden and co-workers 45-47 used the sorption of trace elements on silica gel or glass pellets Complex-forming groups immobilised on the with reagents immobilised by silylation.surface of pellets are shown here (X-XIII). E E > S i - (C H 2 ) 2 --N H (C H 3) -C 4s O ‘S- X XI XI I X l l l The diethyldithiocarbamate group ensures the sorption of cations and the diamine group the sorption of anions. Sorption of trace elements is sufficiently complete. If it is necessary to isolate trace amounts from a large volume, the column variant is preferable, although in this instance it is better to use glass balls but not silica gel. After the sorption, the glass or silica gel particles are mixed with an equal amount of powdered cellulose, pressed, and the tablet is analysed by the X-ray fluorescence technique. Elements can be extracted from very dilute solutions, e.g., selenium from 2 1 of water at the level of 10 pg ml-l. Solvent extraction is a less suitable concentration technique for combination with X-ray fluorescence analysis.However, an extraction method that is promising for this purpose is solvent extraction with fusible extractants at increased tem~erature.~~ In this instance, the concentrate (after cooling and separation from the solution) can immediately be pressed into a tablet and analysed. Of course, solvent extraction can be applied before the sorption as a preliminary step of separation and c~ncentration.~&~~ Separation of Metal Complexes with Stable Free Radicals and Electron-spin Resonance Spectroscopic Determination Very interesting possibilities are provided by the new combination of electron-spin resonance (ESR) spectroscopy and separation methods based on the use of complexes of metals with chelate-forming reagents that are stable free radicals.It is known that the direct use of ESR spectroscopy for the determination of element concentrations is based on paramagnetism of the metal ions themselves. This restricts considerably the possibilities of the technique because the number of metal ions that give ESR signals is relatively small, particularly for the common ions. However, the technique can be used more widely for analytical purposes if the metals are bound in complexes with organic complex-forming reagents that simultaneously act as stable free radicals. Such compounds can now be synthesised and their use will permit the development of a radio-64 ZOLOTOV: HYBRID METHODS Afialyst, Vol. 103 spectroscopic method of analysis based on the measurement of paramagnetism arising from an organic part of a complex molecule.This makes it possible to have a universal technique, independent of the nature of the metal. In addition, the determination of metals by using reagent-radicals is very promising because of the high sensitivity of the ESR method with respect to organic free radicals. The analytical utilisation of reagent-radicals requires the separation of excess of reagent from the complex as the ESR spectrum of the complex can overlap with the spectrum of the reagent itself. Even if the ESR spectrum of the complex is qualitatively different from the spectrum of the reagent, the excess of reagent causes a significant background and decreases the sensitivity of the metal determination. Solve-nt extraction can be used as a separation method.We have studied some reagen t-radicals ,52 e. g., potassium 2,2,6,6- t et rame t h ylpiperidine- 1 - oxy-4-xanthate (XIV). 0-c H3C CH3 A- XIV The reagent is readily soluble in water and therefore it is possible to extract metal com- plexes in conditions under which the excess of reagent remains in the aqueous phase. It has been found that zinc can be extracted and determined in the organic phase from the ESR signal of the complex with this xanthate. The detection limit is 0.01 pg ml-1, but the reagent and the complex are not sufficiently stable during the extraction. At present we are continuing to study such compounds. Solvent Extraction and Catalytic Determination of Metals Solvent extraction is combined with the catalytic determination of metals in order to increase the selectivity of catalytic n i e t h o d ~ .~ ~ , ~ * By selection of a suitable extraction system, the catalyst can be determined directly in the organic phase. The determination of molybdenum based on the catalytic oxidation of 1-naphthylamine with bromate after extraction of 8-quinolinatomolybdenum can be taken as an example; M vanadium, 2 x M chromium, 5 x M tungsten and M manganese, nickel, copper, cobalt, iron, magnesium, zinc, cadmium, aluminium and lead do not interfere. The method has been used for the determination of molybdenum in sea water. This combined technique has been called “extraction catalimetry.” Liquid Ion-selective Electrodes Based on Solvent-extraction Systems The use of solvent extraction with liquid ion-selective electrodes can serve as an example of the close integration of different fields.The achievements of solvent extraction theory and practice related to the choice of selective extractants and extraction conditions can be utilised for the development of such electrodes. Investigations carried out in our laboratory have shown that the choice of an optimum ion exchanger and the detection limit of the metal depend on constants related to the extraction of the metal. Knowing the behaviour of the metal in a common extraction system, it is possible to choose optimum conditions for its determination in a solution being analysed. Corresponding correlation equations have been developed. By using the extraction systems tetraphenylarsonium - organic solvent - complex metal- containing anion, we have developed ion-selective electrodes for the determination of goId(II1) and gold(1).In the first instance,55 tetrachloroaurate was used, which permits the determination of gold at concentrations in the range 3.2 x 10-2-1.5 x 10-7 g-ion 1-l. The detection limit depends on the nature of the organic solvent used and on the concentration of the ion exchanger in the membrane. In the second instance,56 dicyanoaurate was used;January, 1978 OF ANALYSIS 65 the electrode makes it possible to determine gold in cyanide solutions in the concentration range 3.2 x 10A2-3.2 x 10-6 g-ion 1-l. The length of the linear electrode function also depends on the nature of the solvent and the concentration of the ion exchanger in the membrane. The electrode potential does not depend on pH over the range 2.5-11.5.The interferences of CN-, Fe(CN)g-, Fe(CN)3,- and Ag(CN), were measured. The electrode developed has been used to determine gold in gold-plating baths and in technological solutions in the non-ferrous metals ind~stry.~' Other Methods There are, of course, many other combined methods that can be called hybrid methods. Combinations of separation and concentration, especially extractive, with spectrophoto- metric determination in the visible and ultraviolet regions are widely known. Individual concentrations or successive separations of several elements are usually applied and the matrix is very rarely separated in this instance. Most frequently, a reagent that gives a coloured complex with the element to be determined is used for concentration.However, two reagents can also be used: in the first stage, the most selective reagent is used for the separation and then, in the second stage, a reagent that perhaps is not selective but is most suitable from the point of view of photometry is introduced into the extract. Some extrac- tion photometric methods developed in our laboratory are listed in Table IV. TABLE IV SOME EXTRACTION - PHOTOMETRIC METHODS DEVELOPED AT THE VERNADSKY INSTITUTE Element determined Reagent Hg Astraflocsin FG Cd 1,4-Dimethyl-l,2,4-tri- azoline-( 3-azo) -4-NN- dimethylaniline Co Acid Monochrome Green S Nd 5,7-Dibromoquinolin-8-01 after extraction with l-phenyl-3-methyl- 4-benzoylpyrazol-5-one A1 Salicylal-2-aminophenol Au Malachite green, after extraction as quinolin- 8-thiolate Solvent Material analysed Benzene Zinc, cadmium, Mixture of Zinc sulphate, Butanol Gallium metal water, ores benzene and oxide, zinc metal cyclohexanone Chloroform Lithium niobate Tributyl Sodium iodide, phosphate Mohr's salt Chloroform Aluminium nitrate, nickel metal Detection limit, p.p.m.Reference 0.01-20 58 5-14 59 0.1-5 60 61 - 0.01-1* 62 0.01-0.1 63 * Fluorimetric determination. Polarography of extracts containing elements to be determined is another useful method. Before polarographic determination, a supporting electrolyte and solvent homogeniser, for example methanol, are introduced into the extract. Such a combination accelerates the analysis as it eliminates the necessity to decompose the extract or to carry out a back- extraction. In addition, the selectivity of polarographic determination is added to the selectivity of extraction concentration.For example, we have developed a method of determining an alloying indium additive in a complicated semiconductor, CdSiAs,.64 Indium is extracted at pH 1.2 with chloroform as a chelate with l-phenyl-3-rnethyl-4-benzoyl- pyrazol-5-one. Ammonium bromide in methanol has been used as a supporting electrolyte. Many papers on such work have been published; for example, Vydra in Czechoslovakia and Pyatnitsky in the USSR have carried out a series of such investigations. The first review of extraction-polaro- graphic methods was published by Kuz'min et aZ.65 in 1969; several other such reviews have also appeared.In this method, Cadmium and arsenic do not influence the extraction of indium. Electrochemical stripping analysis is also a hybrid method of analysis.66 ZOLOTOV: HYBRID METHODS Analyst, Vol. 103 elements to be determined are preliminarily concentrated electrochemically on a solid or mercury electrode, and then determined during anodic dissolution. “Double” concentration, as used by KarbainoP at the Tomsk Polytechnical Institute, is interesting. It combines extraction concentration with subsequent electrochemical concentration on an electrode. In this variant, both of the above-mentioned ways have been combined. Neutron-activation analysis with sub-stoicheiometric separation of radioactive elements can also be utilised.Conclusion All of the examples mentioned demonstrate that combined techniques of the type con- sidered are of irnportance in modern analytical chemistry, and have a promising future. 1. 2. 3. 4. 5. 6. 6. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. References Zolotov, Yu. A., Zh. Analit. Khim., 1977, 32, 2085. Zolotov, Yu. A., in “XI Mendeleev Congress on General and Applied Chemistry, Summaries of Laitinen, H. A., Analyt. Chem., 1976, 48, 2049. Tolg, G., Talanta, 1974, 21, 327. Zolotov, Yu. A., P y e Appl. Chem., in the press. Horning, E. C., in Parris, G. E., Blair, W. R., and Brinckman, F. E., Analyt. Chem., 1977, 49, 378. Coker, D. T., Analyt. Chem., 1975, 47, 386. Botre, C., Cocace, F., and Cozzani, R., Analyt. Lett., 1976, 9, 825.Zilberstein, Kh. I., Editor, “Spectrochemical Analysis of Pure Substances,” Adam Hilger, Bristol, 1977. Zolotov, Yu. A., Vanifatova, N. G., Chanysheva, T. A., and Yudelevich, I. G., Zh. Analit. Khim., 1977, 32, 317. Krivenkova, N. P., Pavlenko, L. I., Spivakov, B. Ya., Popova, I. A., Plotnikova, T. S., Shkinev, V. M., Kharlamov, I. P., and Zolotov, Yu. A., Zh. Analit. Khim., 1976, 31, 614. Zolotov, Yu. A., Sizonenko, N. T., Zolotovitskaya, 12. S., and Yakovenko, E. I., Zh. Analit. Khim., 1969, 24, 20. Pavlenko, L. I., Petrukhin, 0. M., Zolotov, Yu. A., Karyakin, A. V., Gavrilina, G. N., and Tumanova, I. E., Zh. Analit. Khim., 1974, 29, 933. Vorobjeva, G. A., Zolotov, Yu. A., Izosenkova, L. A., Karyakin, A. V., Pavlenko, L.I., Petrukhin, 0. M., Seryakova, I. V., Simonova, L. V., and Shevchenko, V. N., Zh. Analit. Khim., 1974,29,497. Myasoedova, G. V., Malofeeva, G. I., Shvoeva, 0. P., Illarionova, E. V., Savvin, S. B., Zolotov, Yu. A., Zh. Analit. Khim., 1977, 32, 645. Rudnev, N. A., Pavlenko, L. I., Malofeeva, G. I., and Simonova, L. V., Zh. Analit. Khim., 1969, 24, 1223. Zeidel, A. N., Kaliteevsky, N. I., Lipis, L. V., Chaika, M. P., and Belyaev, Yu. I., Zh. Analit. Khim., 1956, 11, 21. Mandelstam, S. L., Semenov, N. I., and Turovtseva, 2. M., Zh. Analit. Khim., 1956, 11, 9. Busev, A. I., Byr’ko, V. M., Tereshchenko, A. P., Novikova, N. N., Naidina, V. P., and Terent’ev, Tsalev, D. L., Alimarin, I. P., and Neiman, S. I., Zh. Analit. Khim., 1972, 27, 1223. Spivakov, B. Ya., Shkinev, V.M., and Zolotov, Yu. A., Zh. Analit. Khim., 1975, 30, 2182. Zolotov, Yu. A., Spivakov, B. Ya., and Shkinev, V. M., in “Second National Conference on Analytical Chemistry with International Participation,” September 20th-24th, 1976, Golden Sands-Varna, Bulgaria, p. 10. Serebryany, B. L., Fishkova, N. L., Petrukhin, 0. M., and Rakovsky, E. E., Zh. Analit. Khim., 1973, 28, 2333. Fishkova, N. L., and Petrukhin, 0. M.. Zh. Analit. Khim., 1973, 27, 645. Petrukhin, 0. M., Zolotov, Yu. A., and Izosenkova, L. A., Zh. Neorg. Khim., 1971, 16, 3285. Zolotov, Yu. A., and Kuf,’min, N. M., “Ekstraktsiya Metallov Atsilpirazolonami (Metal Extraction Akama, Y., Nakai, T., and Kawamura, F., Bunseki Kugaku, 1976, 25, 496. Vall, G. A., Usoltseva, M. V., Seryakova, I. V., and Zolotov, Yu.A., Zh. Analit. Khim., 1976, 31, Shaburova, V. P., Yudelevich, I. G., Seryakova, I. V., and Zolotov, Yu. A., Zh. Analit. Khim., Pierce, F . D., Cortatowski, M. J., Mecham, H. D., and Fraser, R. S., Analyt. Chem., 1975, 47, 1132. Jones, M., Kirkbright, G. F., Ranson, L., and West, T. S., Analytica Cham. Acta, 1973, 63, 210. Berndt, H., and Jackwerth, E., S@ctrochim. A&, 1975, 30B, 169. Tsalev, D. L., Tarasevich, N. I., and Alimarin, I. P., Zh. Analit. Khim., 1973, 28, 19. Orlova, V. A., Spivakov, B. Ya., Shkinev, V. M., Kirillova, T. I., Ivanova, V. A., Malyutina, T. M., Lectures and Papers, No. 5, Analytical Chemistry,” Nauka, Moscow, 1975, p. 14. International Symposium on Microchemical Techniques 1977, Davos, Switzer- land, May 22nd-27th, 1977, Abstracts,” 1977, p.13. P. B., Zh. Analit. Khim., 1970, 25, 665. with Acylpyrazolones), Nauka, Moscow, 1977. 27. 1976, 31, 255. and Zolotov, Yu. A., Zh. Analit. Khim., 1978, 33, No. 1.January, 1978 OF ANALYSIS 67 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. Spivakov, B. Ya., Lebedev, V. I., Shkinev, V. M., Krivenkova, N. P., Plotnikova, T. S., Kharlamov, Zolotov, Yu. A., and Kuz’min, N. M., “Ekstraktsionnoe Kontsentirovanie (Solvent Extraction Con- Zolotov, Yu. A., and Kuz’min, N. M. Zh. Analit. Khim., 1967, 22, 773. Kuz’min, N. M., Vlasov, V. S., Krasil’tschik, V. Z., and Lambrev, V. G., Zav. Lab., 1977, 43, 1. Zolotov, Yu. A., Shakhova, N. V., Kryuchkova, 0. I., Gronskaya, S.I., Spivakon, B. Ya., Ramendik, G. I., and Gutshin, V. N., Zh. Analit. Khim., in the press. Petrukhin, 0. M., Zolotov, Yu. A., Shevchenko, V. N., Kryuchkova, 0. I., Gronskaya, S. I., Gushchin, V. N., Ramendik, G. I., Dunina, V. V., and Rukhadze, E. G., Zh. Analit. Khim., in the press. I. P., and Zolotov, Yu. A., Zh. Analit. Khim., 1976, 31, 757. centration),” Khimiya, Moscow, 1971. Burda, P., and Lieser, K. H., Angew. Makromol. Chem., 1976, 50, 151. Burda, P., and Lieser, K. H., 2. Analyt. Chern., in the press. Smits, J ., and Grieken, van, R., in “International Symposium on Microchemical Techniques 1977, Davos, Switzerland, May 22-27, 1977, Abstracts,” 1977, p. 135. Leyden, D. E., and Luttrell, G. H., Analyt. Chem., 1975, 47, 1612. Leyden, D. E., Luttrell, G. H., Nonidez, W. K., and Werko, D. B., Analyt. Chem., 1976, 48, 67. Leyden, D. E., Luttrell, C . H., Sloan, A. E., and De Angelis, N. J., Analytica Chim. Ada, 1976, 84, Lobanov, F. I., in “X-Ray and Emission Spectral Methods of Analysis” (in Russian), Moscow, 1975, Zimmerman, J. B., Can. J . Sfiectrosc., 1973, 18, 147. Leyden, D. E., Nonidez, W. K., and Carr, P. W., Analyt. Chem., 1975, 47, 1449. Morris, A. W., Analytica Chim. Acta, 1968, 42, 397. Petrukhin, 0. M., Kurdyukova, N. A., Zolotov, Yu. A., Zhukov, V. V., and Marov, I . N., Zh. Analit. Khim., in the press. Otto, M., and Muller, H., Talanta, 1977, 24, 15. Werner, G., in “International Symposium on Microchemical Techniques 1977, Davos, Switzerland, May 22-27, 1977, Abstracts,” 1977, p. 59. Bychkov, A. S., Petrukhin, 0. M., Zarinski, V. A,, and Zolotov, Yu. A., Zh. Analit. Khim., 1975, 30, 2213. Bychkov, A. S., Petrukhin, 0. M., Zarinski, V. A,, Zolotov, Yu. A., Bakhtinova, L. V., and Shanina, G. G., Zh. Analit. Khim., 1976, 31, 2114. Zarinski, V. A., Bakhtinova, L. V., and Zolotov, Yu. A., Zh. Analit. Khim., in the press. Kish, P. P., Spivakov, B. Ya., Roman, V. V., and Zolotov, Yu. A., Zh. Analit. Khim., 1977,32, No. 10. Kish, P. P., Balog, I. S., Spivakov, B. Ya., and Zolotov, Yu. A., Zh. Analit. Khim., 1975, 31, 1114. Bagreev, V. V., and Zolotov, Yu. A,, Talanta, 1968, 15, 988. Sizonenko, N. T., and Zolotov, Yu. A., Zh. Analit. Khim., 1969, 24, 1341. Demina, L. A., Petrukhin, 0. M., Zolotov, Yu. A., and Serebryakova, G. V., Zh. Analit. Khim., 1972, 27, 1731. Demina, L. A., Petrukhin, 0. M., and Zolotov, Yu. A., i n “Vsesoyuznoe Sov. PO Analit. Kontr. Proizv. v Promyshl. Khim. Reaktivov i Osobo Chist. Veshchestv. Tezisy Dokl. 1-3, XII, 1970,” Inst. Khim. Reaktivov, MOSCOW, 1970, p. 20. Revenko, V. G., Bagreev, V. V., Zolotov, Yu. A., and Kopanskya, L. S., Zh. Analit. Khim., 1972 27, 187. Kuz’min, N. M., Zolotov, Yu. A., and Karbainov, Yu. A., Trudy Konm. Analit. Khim. Akad. Nauk, SSSR, 1969, 17, 288. Karbainov, Yu. A., Izu. Tomsk. Politekh. Inst., 1967, 164, 228. 97. p. 129.

ISSN:0003-2654    

DOI:10.1039/AN9780300056

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

68 Analyst, January, 1978, Vol. 103, pp. 68-71 Calcium Ion-selective Electrodes Based on Calcium Bis[di(p-I,1,3,3=tetra methyl butyl pheny1)- phosphate] Sensor and Trialkyl Phosphate Mediators G. J. Moody, N. S. Nassory and J. D. R. lhomas Chemistry Depavtment, University of Wales Institute of Science and Technology, Cardifl, CF 1 3N U, Wales Calcium ion-selective electrodes based on PVC matrices with membranes composed of calcium bis[di(p-l,l, 3,3-tetramethylbutylphenyl)phosphate] sensor and trialkyl phosphate solvent mediator are described and compared with electrodes of the same sensor with di-n-octyl phenylphosphonate mediator. Trialkyl phosphates of sufficiently high viscosity, for example, tri-n-butyl phosphate, tri-n-amyl phosphate, tri( 1,1,3,3-tetramethylbutyl) phosphate and tri-n-octyl phosphate make good solvent mediators and maintain the calcium ion selectivity of the sensor for the range of alkali metal and bivalent ions studied, namely, Naf, Kf, Mg2+, Sr2+, Ba2f, Mna+, Cu2+, Ni2+ and Zn2+.The trialkyl phosphate mediators provide commercially available alternatives to di-n-octyl phenylphosphonate, the best in the range examined being tri- n-amyl phosphate. Keywords : Ion-selective electrodes ; calcium ion-selective electrodes It has been suggested that the positions of maxima in solvent extraction graphs and corre- sponding minima in ion-selective electrode e.m.f. .- pH graphs in systems involving dialkyl- phosphoric acid are associated with the acidic nature of the proton and that introducing groups -of pronounced electrophilic character into the alkyl substituent would make the proton 1abile.l In calcium ion-selective electrode terms this suggestion means that the region of existence of the acid would be shifted to lower pH values, leading to increased selectivity towards calcium compared with hydrogen ions, and that the sensing sites within an ion-selective electrode membrane would be fully occupied at the optimum level for calcium ions over a wider pH range.A test of this hypothesis has shown1 that ion-selective electrodes using calcium bis [di(n-octylphenyl)phosphate] as a sensing component with di-n-octyl phenyl- phosphonate mediator have minima in their e.m.f, - pH graphs towards the acid region, where, of course, hydrogen ions contribute to the charge transport across the membrane. In addition to the wider pH range over which the response to calcium ions is free from interference from hydrogen ions1 these ion-selective electrodes have certain other advantages, such as a lesser susceptibility to interference from sodium ions.2 These differences in properties from those of calcium ion-selective electrodes based on calcium bis [didecylphos- phate] sensor with di-n-octyl phenylphosphonate mediator justify wider study.This paper is devoted to a study of the calcium ion-selective behaviour of poly(viny1 chloride) (PVC) matrix membranes containing calcium bis[di(p-I , 1,3,3-tetramethylbutylphenyl)phosphate] (calcium bis [di (isooctylphenyl)phosphate]) sensor and certain trialkyl phosphate mediators. Di-n-octyl phenylphosphonate solvent mediator was used as the reference.Experimental Electrodes Ion-selective electrodes with membranes containing the calcium liquid ion exchanger {solvent mediator plus calcium bis [di(~-1,1,3,3-tetramethylbutylphenyl)phosphate]~ trapped in PVC were prepared as previously de~cribed.~g~ The master membranes contained 0.36 g of the solvent mediator [one of di-n-octyl phenylphosphonate, triethyl phosphate, tri-n- propyl phosphate, tri-n-butyl phosphate, tri-n-amyl phosphate, tri( 1,1,3,3-tetramethylbutyl) phosphate (isooctyl phosphate) or tri-n-octyl phosphate] plus appropriate amounts (up to 0.036 g) of calcium bis[di(~-1,1,3,3-tetramethylbutylphenyl)phosphate] sensor in 0.17 g of PVC.MOODY, NASSORY AND THOMAS 69 For comparison, membranes were also prepared from Orion (92-20-02) calcium liquid ion exchanger (0.40 g of exchanger and 0.17 g of PVC) and from calcium bis[di(n-octyl- phenyl)phosphate] (0.06 g) plus di-n-octyl phenylphosphonate (0.36 g) in PVC (0.17 g).Reagents Di-n-octyl phenylphosphonate was synthesised as has been described previously.3J-7 Calcium bis [di(p-l,l,3,3-tetramethylbutylphenyl)phosphate] was synthesised by a modifica- tions of the procedure described by RfiiiEka et aZ.l and purified by ethylene glycol extractions, as described by Cattrall et aZ.9 All other materials were of laboratory-reagent grade. Physical Constants of Solvent Mediators dynamic viscosities at 25 "C by use of an Ostwald viscometer. Relative permittivities were determined at 25 "C by using a Calvert dielectric bridge and Procedure All e.m.f.measurements were made relative to a Corning, Model 476 109, ceramic plug type calomel reference electrode containing 4 M potassium chloride solution. A Beckman, Model 4500, digital pH/millivolt meter was used in conjunction with a Servoscribe, Model RE 4541, potentiometric recorder. Selectivity coefficients, kg:h, were determined by means of a mixed solution method,lOJ1 with a fixed level of interferent, B. Interference as a result of pH changes was assessed by the mixed solution method,1°-12 in which the calcium level was fixed and the pH varied. Results Ion-selective electrodes were assembled from membranes containing triethyl phosphate ( E = 12.7 at 25 "C) and tri-n-propyl phosphate ( E = 10.7 at 25 "C) as solvent mediators, but because the resulting general electrode qualities were poor, the data obtained are not included among those discussed in this paper.The poor electrode quality is attributed to the low viscosities (1.55 and 2.62 mN s m-2, respectively) of these solvent mediators resulting in electrodes of very limited lifetime and restricted range. For example, electrodes assembled from membranes of calcium bis [di(~-l,1,3,3-tetramethylbutylphenyl)phosphate] (0.025 g) and triethyl phosphate or tri-n-propyl phosphate in PVC were linear (with slopes of 32 and 27 mV per decade) only in the 10-4-3 x l O W 3 ~ and 10-5-3 x 1 0 - 3 ~ range, respectively, while the other electrodes examined gave much more extensive linear calibration ranges with near-Nernstian slopes. Details of the various electrodes assembled from membranes based on calcium bis [di- (~-1,1,3,3-tetramethylbutylphenyl)phosphate] sensors with di-n-octyl phen ylphosphonate, tri-n-butyl phosphate, tri-n-amyl phosphate, tri( 1,1,3,3-tetramethylbutyl) phosphate and tri-n-octyl phosphate mediators are sumniarised in Table I and the corresponding selectivity coefficients in Table 11. Discussion Table I shows that good calcium ion-selective electrodes result from membranes with the viscous trialkyl phosphate solvent mediators tri-n-butyl phosphate, tri-n-amyl phosphate, tri( 1,1,3,3-tetramethylbutyl) phosphate and tri-n-octyl phosphate when used with calcium bis [di(~-l,1,3,3-tetramethylbutylphenyl)phosphate] as sensor in PVC matrices.These membranes (VI-XXI) compare favourably with electrodes prepared from membranes based on the same sensor but with the more conventional di-n-octyl phenylphosphonate mediator (membranes 11-V) .Such comparability is of considerable practical interest as the trialkyl phosphates are more widely available commercially, but electrodes with membranes using tri-n-butyl phosphate and tri-n-octyl phosphate do not have the longer lifetimes associated with the other two trialkyl phosphates. The limited lifetimes are associated with premature shortening of the calibration range. The general favourable features of the electrodes with the trialkyl phosphate mediators extend to selectivity (Table 11), whereby the calcium ion selectivity in relation to a range of divalent metals compares with that for those electrodes with di-n-octyl phenylphosphonate70 Analyst, VoL.103 MOODY, NASSORY AND THOMAS : CALCIUM TABLE I PROPERTIES OF ION-SELECTIVE ELECTRODES WITH MEMBRANES OF CALCIUM BIS [DI(p-1,1,3,3-TETRAMETHYLBUTYLPHENYL)PHOSPHATE] SENSOR AND VARIOUS MEDIATORS I N PVC MATRICES Solvent mediator viscosity and relative permittivity a t 25 O C are quoted below the name of each solvent mediator. Liquid ion-exchanger system Solvent mediator r- Membrane Wiscositvl (Relative Mass of number I I1 I11 IV V VI VII VIII IX X XI XI1 XI11 XIV xv XVI XVII XVIII XIX xx XXI XXII AN s m-'aj permittivity) Orion 92-20-02 calcium liquid ion exchanger Di-n-octyl phenylphosphonate (16.7) (6.2) Tri-n-butyl phosphate (3.25) Tri-n-amyl phosphate (4.64) Tri-n-octyl phosphate (11.8) (8.3) (7.0) (5.1) Tri(l,l,3,3-tetramethylbutyl) phosphate (14.3) (5.0) Di-n-octyl phenylphosphonate (with calcium bis[di(p-n-octyl- pheny1)phosphatel sensor 1 (16.7) (6.2) sensorlg 0.40 0.007 5 0.015 0.025 0.036 0.007 5 0.015 0.025 0.036 0.007 5 0.015 0.025 0.036 0.007 5 0.015 0.025 0.036 0.007 5 0.015 0.025 0.036 0.036 Lower limit of detection for Caa+/ix 7.5 x 10-6 3.2 x 10-6 5.8 x 10-6 4.1 x 3.0 x 2.7 x 2.6 x 2.0 x 1 0 - 6 1.2 x 10-6 4.4 x 10-0 6.0 x 10-6 2.0 x 10-6 6.0 x 9.0 x 10-6 9.0 x 10-8 8.5 x lo-'' 1.1 x 10-5 4.5 x 10-0 6.5 x lo-' 5.6 x 6.0 x 10-6 2.7 x Electrode slope a t 25 OC/mV Per decade 33.5 30.5 30.0 30.5 30.5 30.5 30.0 31.0 30.5 30.5 31.5 31.0 31.0 29.0 30.5 31.0 30.0 30.5 31.5 31.0 31.5 30.5 Static response time/s 7- M M Operational CaCl, CaC1, lifetime 84 144 3 months 96 84 36 ii ] 3months 84 30 60 I lweek 30 30 60 30 24 7 weeks 60 48 ! } 3months 120 48 48 48 24 5; } 4weeks 24 24 42 48 1 3 months 52 60 48 72 120 pH range for 10-8 M CaCl, 6 A 8 .3 4.5-8.8 4.5-8.7 4.5-8.5 4.8-8.8 5.5-8.7 4.5-9.0 5.0-8.9 4.5-8.9 5.0-8.4 4.9-8.1 4.6-8.0 4.8-8.2 5.5-8.6 5.2-8.5 4.7-8.5 4.9-8.3 5.1-7.7 5 .O-8.9 5.3-8.1 5.2-8.2 4.6-8.2 mediator, the best being membrane XI1 containing 0.025 g of tri-n-amyl phosphate. The general selectivity sequence for bivalent ions is Ca>Zn >usually Mn>frequently Sr>the other metal ions in various orders. Calcium-ion selectivity in the presence of sodium and potassium ions also emphasises the scope of the trialkyl phosphates as solvent mediators (Table 11). mol dm-3 these ions did not affect the general nature of the calcium ion response graph, signifying no interference at this level from the alkali metal ions.The increased selectivity towards calcium compared with hydrogen ions shown by calcium bis [di(p-l, 1,3,3-tetrarnethylbutylphenyl)phosphate] sensor with di-n-octyl phenyl- phosphonate solvent mediator has been referred to above and extends to the mediator systems of this study wherein the electrode e m f . response to a given calcium ion level remains constant into lower pH ranges. Thus, Table I shows that the pH ranges for 10-3 mol dm-3 calcium chloride solution extend down to about pH 5 for the bis[di(+1,1,3,3- tetramethylbutylpheny1)phosphatel sensor systems (membranes 11-XXI) compared with about pH 6 for the sensing system of the Orion 92-20-02 liquid ion exchanger (membrane I).Varying the proportion of sensor to mediator shows that membranes with as little as 0.007 5 g of sensor to 0.36 g of mediator in the master membrane function well and there is no regular feature that suggests an optimum ratio of sensor t o mediator. There is, how- ever, a tendency for gel-like inclusions to be present at higher sensor to mediator ratios (Table I). It is helpful to practitioners that the exact amount of sensor used is not critical and electrodes with master membranes containing 0.0075 to 0.036 g of sensor to 0.36 g of solvent mediator all behave well. Apart from the fact that very low viscosity mediators give poor PVC-matrix membrane ion-selective electrodes, investigation of the viscosity and permittivity of solvent mediators does not lead to any regular conclusion on the quality of ion-selective electrodes.Concerning the sensor, the branched isooctyl chain, (1,1,3,3-tetramethylbutyl) is much more convenient than the straight n-octyl chain as the branched-chain phenol is easier At a concentration ofJanuary, 1978 ION-SELECTIVE ELECTRODES TABLE I1 ELECTRODE SELECTIVITY CHARACTERISTICS OF THE MEMBRANES Level of interferent chloride B = 5 x M for sodium and potassium and 5 x 10-4 M for bivalent ions. 71 Selectivity coefficient, k g i k r A 3 Membrane kPot RPot kPot kPot kPot Pot kPOt kPot kPot number Ca Na Ca I( Ca M g C a Sr c a ~ a k c e ~ n Ca Cu Ca Ni Ca Zn I 4.5 x 10-8 6.2 x 10-2 1.3 X lo-' 1.4 X 10-1 5.8 x lo-' 2.3 X lo-' 1.6 x lo-' 9.6 X lo-* * I1 2.6 x lo-' 2.1 x 3.3 x 10-2 2.6 x 2.6 x 6.7 x 2.6 X lo-* 2.0 X lo-* 4.8 X lo-' I11 1.2 x 10-2 9.3 x 10-3 4.5 x lo-* 5.5 x 2.4 x lo-* 6.7 x lo-' 4.1 x lo-* 2.6 X lo-' 4.1 X lo-' IV 2.2 x 10-8 1.8 x 10-2 4.5 x 8.1 x lo-* 4.1 x 1.2 x lo-' 4.8 x 2.9 x lo-' 5.6 X lo-' V 1.7 x 10-2 1.8 x lo-* 2.1 x 4.1 X 9.0 x 4.0 x lo-* 1.4 x lo-' 1.3 X 3.0 X lo-' 3.3 x 10-8 2.4 x lo-* 4.8 x 10-2 6.4 x 2.9 x lo-* 7.6 x 3.1 X 3.8 X lo-' 4.3 X lo-' 2.4 x 10-2 1.6 x 10-2 4.8 X lo-* 7.2 X lo-* 3.2 x lo-* 9.1 X 6.9 X lo-' 3.3 X 10-1 3.4 X lo-' VI 1.6 x 10-8 1.3 x 10-e 3.3 X lo-* 6.0 X 3.6 X lo-' 1.1 X lo-' 2.7 X lo-' 1.9 X lo-' 2.3 X lo-' VII 2.4 x 10-2 1.9 x 4.3 X lo-* 3.6 X lo-' 3.0 x lo-' 1.2 X lo-' 5.5 X lo-* 3.1 X 4.3 X lo-' VIII IX 2.1 x 10-2 1.7 x 10-2 4.3 X 10-8 6.7 X lo-' 5.5 X lo-* 1.0 X lo-' 2.9 X lo-' 3 6 X 4.3 X 10-1 X X I 9.9 x 10-8 8.9 x 10-8 3.1 x 5.5 X 2.0 X lo-' 6.9 X lo-' 2.1 X lo-' 1.8 X lo-' 2.2 X lo-' 2.1 x 10-8 2.2 x 10-8 6.2 x 10-2 9.1 X 10-2 4.3 x 2.2 x 10-1 8.6 x lo-' 8.2 x lo-' 4.8 X 10-1 XI1 XI11 9.7 x 10-2 5.6 x 1.8 x lo-' 1.2 X lo-' 8.1 X 5.6 X lo-' 2.0 X lo-' 6.9 X lo-* * 1.2 x 10-1 5.6 x 10-2 1.5 x 10-1 1.1 x 10-1 9.6 X 3.8 X 10-l 9.6 X lo-* 6.9 X lo-* 1.2 x 10-1 7.5 x 10-3 1.3 x 10-1 1.2 x 10-1 6.7 x 10V 4.8 x lo-' 1.6 x lo-' 6.9 X 10-' XIV xv 4.3 x 10-8 3.5 x lo-* 1.9 x 10-1 1.4 x lo-' 7.7 x lo-* 4.6 X lo-' 1.8 x lo-' 1.0 X lo-' XVI XVII XVIII XIX xx XXI 2.5 10-8 1.6 x 10-2 5.3 x 10-2 8.1 x lo-* 3.8 x lo-' 6.7 x 6.0 x 3.8 x lo-* 3.6 x 10-l I 2.6 10-8 2.0 x 10-3 8.1 x lo-* 1.1 x 10-1 3.6 x 1 0 - 2 4.8 x lo-% 4.3 x lo-' 6.2 x 4.8 x lo-' 3.0 x 10-8 2.6 x 10-8 4.3 x 6.0 x lo-* 3.3 x 5.5 x lo-$ 2.9 x lo-' 3.2 X lo-' 3.6 x 10-1 2.6 x 10-2 2.2 x 10-0 4.5 X 6.9 X lo-' 2.9 X lo-' 9.1 X 6.0 X lo-' 4.3 X lo-' 5.1 X 10-' 2.0 x 10-P 1.9 x 10-2 5.3 x lo-% 6.4 x 3.3 x 6.9 x lo-' 4.1 x lo-' 3.6 x lo-' 5.3 x lo-' XXII 1.2 10-a 9.3 x lo-* 6.2 x 10-2 9.0 x lo-* 4.1 x lo-' 6.2 x lo-* 3.8 x 10-a 2.0 x lo-' 1.2 x 10-l * k;:in could not be determined because the calibration graph for calcium in the presence of zinc ions did not, at any stage, coincide with the normal calcium-ion calibration.to obtain commercially as a suitable starting material for the calcium bis[di(p-l,l13,3-tetra- methylbutylphenyl)phosphate] sensor.8 The n-octylphenyl starting material for calcium bis [di(n-octylphenyl)phosphate] has to be synthesised and considerably lengthens pro- ceedings8 In any case, there is little to be gained by using an n-octyl chain sensor as its general performance in an ion-selective electrode membrane is similar to that of membranes with the branched isooctyl chain sensor, as illustrated here by membranes XXII and V.Conclusion Viscous trialkyl phosphates, such as tri-n-butyl phosphate, tri-n-amyl phosphate, tri- (1,1,3,3-tetramethylbutyl) phosphate and tri-n-octyl phosphate, can be used as solvent mediators in calcium ion-selective electrode membranes with calcium bis [di($- 1,1,3-3-tetra- methylbutylphenyl)phosphate] sensor instead of di-n-octyl phenylphosphonate. Tri-n- amyl phosphate is the best choice, especially membrane XII. The authors thank the University of Technology, Baghdad, Iraq, for paid leave of absence (to N.S.N.). 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References RfiiiCka, J., Hansen, E. H., and Tjell, J. C., Analytica Chim. Acta, 1973, 67, 155. Craggs, A., Moody, G. J., and Thomas, J. D. R., to be published. Moody, G. J., Oke, R. B., and Thomas, J. D. R., Analyst, 1970, 95, 910. Craggs, A., Moody, G. J., and Thomas, J. D. R., J . Chem. Educ., 1974, 51, 541. Griffiths, G. H., Moody, G. J., and Thomas, J. D. R., J . Inorg. Nucl. Chem., 1972, 34, 3043. Craggs, A., Keil, L., Moody, G. J., and Thomas, J. D. R., J . Inorg. Nucl. Chem., 1975, 37, 677. Petrov, K. A., Bliznyuk, N. K., and Maklyaev, F. L., Zh. Obschch. Khim., 1959, 29, 3403. Craggs, A., Delduca, P. G., Keil, L., Key, B. J., Moody, G. J.. and Thomas, J. D. R., J . Inorg. Nucl. Cattrall, R. W., Drew, D. M., and Hamilton, I. C., Analytica Chim. Acta, 1975, 76, 269. Moody, G. J., and Thomas, J . D. R., Lab. Pract., 1971, 20, 307. Moody, G. J., and Thomas, J . D. R., TalaPzta, 1972, 19, 623. Moody, G. J., and Thomas, J . D. R., Sel. A . Rev. Analyt. Sci., 1973, 3, 59. Chem., submitted for publication. Received May 9th. 1977 Accepted August 22nd. 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300068

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

72 Analyst, January, 1978, Vol. 103, pp. 72-78 Automatic Determination of the Specific Gravity of Alcohol and Water Mixtures Part I. W. Bunting" and P. B. Stockwell Development and Evaluation of a Simple Sensing System Department of Industry, SE19NQ Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, The design of a simple sensing system to determine the specific gravity of liquids is described. The evaluation of a prototype system to measure the specific gravity of alcohol - water mixtures is discussed. The system is fully automatic and data processing is achieved in an off-line manner, Further possible areas of application are discussed. Keywords : Specific gravity ; automatic determination ; data processing Automatic analysis has developed rapidly since the introduction by Skeggs in 1957 of the continuous-flow automatic analysers that revolutionised clinical chemistry ; they also had a major effect on the work of the Laboratory of the Government Chemist.Early attempts at automation involved application of the basic AutoAnalyzer approaches directly to the work of this lab~ratoryl-~ but in 1969 a group was set up with direct responsibility to realise the full economic and scientific benefits of automation. This group has automated a wide range of analyses and particularly has pioneered the use of hybrid techniques, merging both the philosophies of automation and different analytical techniques. For example, discrete- and continuous-flow principles have been coupled together4 and gas chromatography has been coupled to specific colorimetric detection systems5 A potential area of automation at this laboratory relates to the analysis of wines and spirits, for which the alcohol content, and hence the excise duty, must be determined.The number of these analyses is large and high precision is required; automatic analysis is thus desirable on both counts. The statutory analytical procedures6 require that samples are distilled to remove the alcohol, the alcoholic distillate is diluted with water to the original volume of the spirit and the specific gravity is measured by weighing in a specific gravity bottle. The physical characteristics of alcohol - water mixtures, coupled to the high duty rates on alcohol, call for precise weighing for the measurement of specific gravity.An automatic balance, or even an electronic balance, with the required precision is not available and alternative procedures had to be considered. Automatic chemical procedures were successfully developed for the analysis of beerl-3 and a system complete with data processing has been in routine operation for a number of years. The precision required for spirit analysis is, however, greater than that for beer analysis and these chemical methods are not sufficiently precise. An automatic rapid measurement of specific gravity after distillation would be suitable as a screening procedure and would also satisfy the statutory requirements of analysis if the desired precision was obtained. The specification of such an automatic system includes a capability to analyse samples a t a rate of 20 samples h-l using a sample size of approximately 30 ml and to measure the specific gravity to a precision of 0.00005 at 20 "C.This paper describes an automatic system for the measurement of specific gravity designed and built in this laboratory. Other methods available commercially or published prototype systems were unsuitable for spirit determinations tor various reasons. A popular method reported as long ago as 1913 by Lamb and Lees uses a magnetically controlled float to deter- mine the density of the liquid. Many other investigators have produced similar systems, further developing and refining the technique.lO-ls However, for precise measurements the * Present address : The Radiochemical Centre, Amersham, Buckinghamshire, HP7 9LL.Crown Copyright.BUNTING AND STOCKWELL 73 final volume required for each measurement was greater than the volume available from the prescribed distillation procedure. More recent work described by Haynes et aZ.19920 uses the magnetic suspension principle to determine the density of materials of cryogenic interest. Density gradient,21 oscillation,22 radiation23 and viscodensimetric methods24 were also re- viewed but were unsuitable because large volumes were required or long analysis times were necessary. The principle of operation of the approach selected was to measure the deflection of a totally immersed float connected to a stainless-steel cantilever suspension. A polyethylene float coupled with an optical sensing system has been found to operate with acceptable precision when colourless solutions are analysed.However, in practice cloudy solutions are often formed with the distillation procedure and these lead to erroneous instrument responses. The technique also has limitations when solutions of varying colour intensity are to be analysed. The preferred system uses a pair of inductive proximity sensors t o detect the deflection of a hollow metal float. The output from these sensors arranged in a Wheatstone bridge circuit can be calibrated directly in terms of specific gravity by using standard solutions of known specific gravity. The sensing system described is the subject of a patent applica- t i ~ n . ~ ~ Design of Apparatus Fig. 1 shows a schematic diagram of the prototype system that has been evaluated for the measurement of the specific gravity of alcoholic distillate.It consists of four component modules: (i) a measurement cell; (ii) a measurement system and associated data output; (iii) a sample turntable complete with sample transfer facilities maintained at a constant temperature; and (iu) an electronic control system. In addition to these hardware modules, a computer program is required to analyse the data and prepare a suitable report. This program, written in Fortran, is designed to operate on the Control Data Corporation CDC 6600 computer at the London University Computer Services Bureau. Data are transferred to the Bureau through a Remote Batch Terminal. Fig. 1. Schematic diagram of the modified automatic relative density system and data handling equipment.Measurement Cell The measurement cell is the heart of the apparatus; a detailed diagram is shown in Fig. 2. A hollow metal float (A) of precisely determined displacement is situated at one end of the cell. The float is constructed from two pieces of brass, which are pressed in a die to form cylinders closed at one end; the halves are then joined with solder, to form a sealed cylindrical float (diameter 1 cm x 0.6cm long). Excess of solder is removed with a file and the float polished to a mirror finish. One end of a 6-cm length of stainless-steel wire (B) is flash gold plated and soldered to the float. The complete float and wire are again polished to remove any imperfections and electroplated with gold. The mass of the float is adjusted so that the support wire remains horizontal when the cell is filled with water at 20 "C.The free end of the wire is silver soldered carefully to a short length of stainless-steel capillary tube, which74 BUNTING AND STOCKWELL: AUTOMATIC DETERMINATION OF THE Analyst, VoZ. 103 is then mounted into the wall of the density cell with an adjustable retaining screw. The inductive sensors (C and D) are mounted vertically above and below the float in the density cell. G E F Fig. 2. Specific gravity measuring cell: A, hollow metal float; B, stainless-steel wire; C and D, transduc,ers; E, inlet tube; F, exit tube; G, overflow tube; and H, tube for electrical connections. The cell itself is constructed from 18/80 non-magnetic steel; all faces are polished to a mirror finish. Glass panels are mounted on both sides with Sealomastic gaskets and retaining screws to give a quick visual indication of the cell performance.Readily available models of the position-sensing transducers absorb alcohol from the solutions and swell, thus producing an erratic and variable performance. To avoid this effect each sensor is encapsulated in stainless steel. The height above the float can be adjusted by means of a screw thread with a locking adjustment. The electrical leads for each sensor are brought through the top of the cell through the tube (H) and the bottom transducer is sealed with Sealomastic to protect the cell from corrosion. The density cell is placed in a thermostatically controlled water- bath, the water for which is pumped from the large bath which accommodates the sample turntable.The cell has been designed so that when it is filled air bubbles are not trapped either in the cell or on the float. It is filled through a capillary tube (E) at the top of the cell, which directs the flow of liquid directly on to the float. Grease or dirt retained in the cell or on the float can have a deleterious effect on the system and thorough cleaning with a detergent solution on a regular basis is essential for precise measurements. The range and precision of the cell operation can be adjusted in three ways, by moving the relative position of the transducers, by altering the mass of the float or by changing the physical characteristics of the suspension wire. The construction materials are specifically chosen so that water or alcohol is not absorbed or desorbed during the analysis.Temperature Control of Sampler Unit In this prototype a large constant-temperature water-bath that maintains the measurement cell and the samples at 20 -+ 0.1 "C is used. A turntable constructed to hold duplicate sample tubes fully immersed in the water is mounted within the bath. Samples are equilibrated and then the turntable control mechanism presents each tube in turn to the sampling probe. The pneumatically operated probe assembly moves in both a vertical and a horizontal plane, in order to pierce the seal of each of the duplicate sample tubes. The probe is a stainless- steel syringe tube sharpened to a point, which punctures the Mylar film seal on the top of the sample tubes. These seals prevent loss of alcohol by evaporation.A peristaltic pump transfers the sample through the probe assembly into the measurement cell. This type of pump provides a positive seal against sample loss by a siphoning action and prevents inter- sample contamination. After measurement of the float displacement, the cell is drained quickly by using a syringe pump. Precise control of the temperature is mandatory for determination of specific gravity.January, 1978 SPECIFIC GRAVITY OF ALCOHOL AND WATER MIXTURES 75 Measurement System Two similar position-sensitive transducers (Type T6, available from Sangamo Weston Ltd.) are mounted at 180" to each other in a push-pull configuration, and excited to a fixed oscillator frequency, nominally 75 kHz. The bridge output is connected through a trans- ducer amplifier to a digital voltmeter and data transfer unit (Solartron) in order to produce data in the form of a paper tape.The gaps between the float and the transducers are adjusted for the range of density required and the bridge is electrically balanced with the float surrounded by water. The data handling system is so arranged that when the measure- ment system is triggered five independent readings are taken of the float position, measured by the out of balance output of the transducers. Control System to control the prototype system. A simple timed sequence, using a cam timer in association with interlocked relays, is used Method of Operation Eight standards covering the range of specific gravity 0.970 00-1.000 00 are prepared. Initial calibration by volume of these standards is obtained by mixing known volumes of absolute alcohol with water in 250-ml calibrated flasks.The precise specific gravity of the solutions is then determined by using the statutory specific gravity bottle method, taking great care to weigh to constant mass. The precision of the measurement system described in this paper is only as reliable as the calibration standards. Alcohol distillates prepared in the conventional manner and standard alcohol - water solutions are loaded into duplicate tubes and sealed with Mylar film. The tubes are then loaded in sequence on the turntable and immersed in the water-bath to equilibrate. Standards occupy the first eight positions and sample solutions occupy the remaining positions.Standard solutions are also interspersed between samples in order to monitor the performance of the instrument. A pre-printed control sheet, an example of which is shown in Fig. 3, is completed by the analyst when the tubes are loaded on the turntable. Data from this sheet are punched on to computer cards and form one input stream for the computer data handling system. The operator can specify a number of options, which is useful in the initial setting up of the system and as a check on the performance of the instrument. After a title line that includes the run date, the operator can specify whether the calibration data should be plotted or whether the full statistics of the calibration need to be printed. These options are identified using a simple 0 or 1 notation.For each analysis four parameters are required on the control sheet. Sample identity is specified by a single digit field: a standard is specified by 0, samples by a 1, and any error parameter recorded on the paper tape by a 9, which allows any erroneous data to be ignored by the computer software. Sample or standard numbers are identified; six letters or figures can be used in the identification. A single alphanumeric field allows various post-programming options that are available according to the sample origin; for example, a W will provide results in the form of proof spirit strength. A final field of ten digits allows the declaration of a sample or the known value of a standard to be entered in floating point form. Use of control sheets of this format provides a simple method of controlling the operation of automatic systems, a number of applications of which have been made in this laboratory, for example, amino acid analysis2s and the determination of the original gravity of beer.8 When temperature equilibrium is attained the automatic cycling of the specific gravity instrument is started.The first tube is located under the probe and the measurement cell is washed with sample from it. The cell is re-washed with sample from the duplicate tube and then completely filled with the sample (or standard) under examination. After a pre-set time the electronic measurement system is triggered and the relative displacement of the float is recorded five times on to punch paper tape. The cell is then emptied, the next tube located and the sequence of operations repeated. The cycle is repeated until all samples and standards have been analysed. On completion of the analyses the cell is washed with Decon solution and the cell left in a standby state filled with water.76 BUNTING AND STOCKWELL : AUTOMATIC DETERMINATION OF THE Analyst, VoZ.103 SPEC I F IC GRAVl TY DETERMINATION * *AG RAV FORBE E RS-CY=I * * TITLE ................................................................................................................. (ccl-80) PLOT OPTION ( 1 or 0 ) - (ccl) PRINT OPTION (1 or 0) - (cc2) N.B. IF PLOT OPTION = 1 - USE DISSPLA VERSION OF JOB DECK***** * * * * * Sample No. Treatment type (W, B, or X I ""N.B. - W gives Proof Spirit on output B gives Gravity Lost on output X gives Extract Gravity (1000 notation) on output Fig.3. Pre-printed control sheet. Data Processing The paper tape and cards form the input streams for the data processing software, and are transmitted by the EAL Remote Batch Terminal by the telephone-switched network to the London University Computer Services Bureau. The software averages the five data readings, correlates the readings for samples and standards and sorts the values for each into separate files. The values for the standard solutions are used to calculate a calibration line using a cubic regression technique to provide the best statistical fit of the data. The average displace- ment value for the sample solutions is then referred to the calculated calibration graph in order to determine the specific gravity.This value can either be printed as specific gravity or referred to additional tables and expressed directly as proof spirit. On completion of the calculations a printed report is generated, sorting the solutions into samples and standards using the options specified on the control sheet by the operator. The format of these output reports can be tailored to the particular application.January, 1978 SPECIFIC GRAVITY OF ALCOHOL AND WATER MIXTURES Results and Discussion 77 A prototype instrument has been constructed that can analyse 20 samples per hour. It requires 30 ml of sample, most of which is used to wash out the cell and associated pipework. The system has been evaluated using ethanol - water solutions of known specific gravity and a calibration typically shows a standard deviation of less than 0.00005 for relative densities between 0.97000 and 1 .OOO 00.For alcohol determinations (on distillates from spirits) a precision of better than ,t0.05% proof spirit at 70% proof spirit (ten replicates) is obtained. Table I shows a comparison of relative densities obtained by the automatic method and by the manual method by specific gravity bottle for distillates from various alcoholic preparations. The automatic system eliminates interpreter error and provides repeatable results over long periods ; the present system, however, requires direct access to computing facilities for result calculation and reporting. Recent electronic developments can avoid this criterion although a t this laboratory the use of a central computer system has been justified on numerous other counts.TABLE I COMPARISON OF THE SPECIFIC GRAVITIES OBTAINED BY THE AUTOMATIC METHOD AND THE MANUAL METHOD BY SPECIFIC GRAVITY BOTTLE, FOR DISTILLATES FROM VARIOUS ALCOHOLIC PREPARATIONS Automatic Manual Automatic Manual method method method method 0.993 10 0.993 21 0.978 48 0.978 49 0.989 84 0.989 93 0.978 00 0.978 08 0.986 81 0.986 78 0.978 03 0.978 06 0.983 92 0.983 91 0.978 14 0.978 22 0.981 03 0.981 01 0.977 97 0.978 03 Experience with the prototype system has proved that the sensor is suitable for the task in hand and has generated a design specification for an engineered system capable of providing a rapid measurement system. This instrument, which is at present under construction, wiii be linked to the laboratory’s in-house computer system and will provide analytical results directly to the operator on completion of each analysis, so overcoming the delays caused by off-line operation.For example, the prototype system has been used to measure the density of cows’ milk (on which basic payment is paid to farmers), of iron(II1) chloride solution (used in the printing industry for etching) and beer. For these applications care must be taken to provide suitable standards and proper wash-out characteristics, particularly for corrosive solutions such as iron( 111) chloride. Tables I1 and I11 show results for iron(II1) chloride solution and for cows’ milk as measured on the prototype specific gravity instrument and by manual measurements. The principle of the measurement cell is simple and has a variety of applications.TABLE I I SPECIFIC GRAVITY OF IRON( 111) CHLORIDE SOLUTION Comparison of calculated results with those obtained by use of the automatic instrument. Specific gravity A I I Found by Concentration, % Calculated instrument Error 1 1.005 1.006 + 1 x 10-3 2 1.009 1.011 + 2 x 10-3 3 1.014 1.016 + 2 x 10-3 4 1.018 1.019 + 1 x 10-378 BUNTING AND STOCKWELL TABLE 111 COMPARISON OF RESULTS OBTAINED FOR SPECIFIC GRAVITY OF cows’ MILK USING MANUAL HYDROMETER READINGS AND THE AUTOMATIC SPECIFIC GRAVITY INSTRUMENT Specific gravity r- Sample A B C D E F Value from meter 1.030 8 1.030 9 1.030 7 1.028 6 1.028 4 1.028 8 Corrected value from meter 1.028 9 1.029 0 1.028 8 1.026 7 1.026 5 1.026 9 (-0.001 9)* I > Hydrometer value Analyst A Analyst B 1.028 8 1.028 7 1.029 1 1.029 0 1.029 0 1.028 9 1.026 5 1.026 5 1.026 5 1.026 6 1.027 1 1.027 0 A \ * The correction factor was introduced because a primary standard was not used between manual and automatic methods.The authors thank Mr. I. Telford for developing the computer software used for the evaluation of the prototype system. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. References Sawyer, R., and Dixon, E. J., Analyst, 1968, 93, 669. Sawyer, R., and Dixon, E. J., Analyst, 1968, 93, 680. Sawyer, R., Dixon, E. J., Lidzey, R. G., and Stockwell, P. B., AnaZyst, 1970, 95, 957. Tucker, K. B. E., Sawyer, R., and Stockwell, P. B., Analyst, 1970, 95, 730. Lidzey. R. G., and Stockwell, P. B., Analyst, 1974, 99, 749. “The Spirit Regulations,” SI 1952 No. 2229, HM Stationery Office, London. Sawyer, R., Dixon, E. J., and Johnson, E., Analyst, 1969, 94, 1010. “Report of the Government Chemist, 1972,” HM Stationery Office, London, 1973, p. 141. Lamb, A. B., and Lee, R. E., J . Am. Chem. Soc., 1.913, 35, 1668. Hall, N. F., and Jones, T. O., J . Am. Chem. Soc., 1936, 58, 1915. Richards, A. R., Ind. Ertgng Chem., Analyt. Edn, 1942, 14, 595. Geffcken, W., Beckmann, C., and Krus, A., 2. Phys. Chem., 1933, B20, 398. Tsuki, T., Diss. Abstr., 1958, 18, 423. Hargens, C. W., Rev. Scient. Instrum., 1957, 28, 921. Beams, J. W., and Clarke, A. M., Rev. Scient. Instrum., 1961, 33, 735. Millery, F. J., Jr., Rev. Scient. Instrum., 1967, 38, 1444. Clarke, A. L., J . Scient. Instrum., 1967, 44, 384. Grigor, A. F., and Steele, W. A., Rev. Scient. Instrum., 1966, 37, 54. Haynes, W. M., Hiza, M. J., and Frederick, W. V., Rev. Scient. Instrum., 1976, 47, 1237. Haynes, W. M., Rev. Scient. Instrum., 1977, 48, 40. Pelsmaekers, J., and Amelinckx, S., Rev. Scient. Instrum., 1961, 32, 830. Kratky, O., Leopold, A., and Stabinger, H., 2. Angew. Phys., 1969, 27, 273. Tabor, P., and Lakosi, L., Acta Imeko, 1964, 435. Gordon, M., Hope, C. S., Loan, L. D., and Ryong-Joon Roe, Proc. R. SOL, A , 1960, 258, 235. Bunting, W., UK Patent Application No. 4846/71. Bunting, W., Morley, F., Telford, I., and Stockwell, P. B., Analyst, 1975, 100, 369. Received July 7th, 1977 Accepted August llth, 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300072

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, January, 1978, Vol. 103, $9. 79-83 79 Rapid Spectrophotometric Determination of Copper in Steel and Non-ferrous Alloys After Synergistic Extraction P. S. Patil and V. M. Shinde Department of Chemistry, Shivaji University, Kolhapur 41 6004, India A simple and rapid procedure is described for the separation and deter- mination of copper in steel and alloys comprising extraction of the copper into chloroform with isonitrosoacetylacetone (H-INAA) in combination with pyridine from an aqueous solution of pH 6.2. The extracted species absorbs at 450nm; the molar absorptivity is 2.44 x 1031mol-1cm-1 and Sandell's sensitivity is 2.6 x pg cm-2. Beer's law is obeyed over the range 20-125 pg of copper per 10 ml of organic phase. The complex is stable for 48 h and the extracted species is probably Cu(INAA),.2C6H,N.Copper is separated from associated elements. Results of the analysis of synthetic mixtures and standard samples are reported. Keywords : Copper determination ; steel analysis ; alloy analysis ; spectro- photometry ; synergistic copper extraction Isonitrosoacetylacetone (H-INAA) has been used for the spectrophotometric determination of palladium,l iron2 and ruthenium.3 We propose a simple method for the synergistic extraction of copper(I1) from an aqueous solution of pH 6.2 by using H-INAA in combina- tion with pyridine. The extracted red - brown species is well suited for the spectrophoto- metric determination of copper at 450 nm. Experimental Apparatus pH values were measured on a Radiometer pH meter. Absorbance measurements were made on a Zeiss spectrophotometer using 1-cm silica cells.Reagents Copper stock solution. A stock solution of copper was prepared by dissolving 9.8g of analytical-reagent grade copper sulphate pentahydrate in 500 ml of distilled water containing 1% m/V of sulphuric acid. The copper content of the stock solution was determined by titration with 0.01 M EDTA solution using 1-(2-pyridylazo)-2-naphthol as indi~ator.~ Solu- tions of lower concentration were prepared by appropriate dilution of the stock solution. Isonitrosoacetylacetone soZution, 0.038 M. Isonitrosoacetylacetone (H-INAA) was prepared by the method given by Wel~her.~ A suspension of 50 g of acetylacetone in 500 ml of 7% mlV sulphuric acid was prepared and a solution of 35 g of sodium nitrite in 150 ml of water was added in a thin stream with shaking.The mixture was extracted with ether, the extract washed with water and the ether distilled off, when the reagent separated as a crystalline mass. It was re-crystaIlised from ethyl acetate (m.p. 73 "C). A 0.038 M solution was prepared by dissolving 0.5 g of the reagent in 100 ml of doubly distilled water. Pyridine solution, 0.063 M. Sodium hydroxide solution, 1 M. Sul$huric acid, 1 M. All other reagents used were of analytical-reagent grade. Prepare a 0.5% Y/V aqueous solution of pyridine. Extraction Procedure To a 1-ml aliquot of sample solution containing between 20 and 125 p,g of copper were added 2 ml of pyridine solution and 3 ml of H-INAA solution. The solution was diluted to80 PATIL AND SHINDE : RAPID SPECTROPHOTOMETRIC DETERMINATION Analyst , Vol.103 20 ml, the pH of the solution adjusted to 6.2 by addition of sulphuric acid or sodium hydroxide solution and the volume made up to 25 ml. The solution was transferred into a 100-ml separating funnel and shaken with 10 ml of chloroform for 5 min. After separation of the two phases, the organic phase was transferred into a 1-cm cell and the absorbance of the red- brown complex measured at 450 nm against a reagent blank prepared in the same manner. The copper content was read from a calibration graph. Results and Discussion Spectral Characteristics and Extraction Conditions The absorption spectrum of the copper - INAA - pyridine complex extracted into chloro- form at pH 6.2 is shown in Fig. 1. In the absence of pyridine the complex was not extracted.Addition of pyridine, however, ensures a quantitative and rapid extraction of copper. Pyri- dine thus exerts a synergistic effect on the extraction of copper - INAA complex from an acidic solution. The extraction of copper was studied at various pH values. The results in Fig. 2 indicate that the optimum pH for quantitative extraction of the copper - INAA - pyridine complex is 6.2-6.6. The pH was adjusted with sodium hydroxide solution or sulphuric acid. The spectrum of the extracted complex has a maximum absorbance at 450nm and the molar absorptivity of the complex in chloroform is 2.44 x lo3 1 mol-l cm-l, calculated on the basis of total copper. 0.4 0.3 Q) c e 0.2 f2 2 O.' t 0- 400 450 500 650 600 Wavelength/nm Fig.1. Absorption spectrum of copper - INAA - pyridine complex against a reagent blank. [Cu] = M ; [pyridine] = 6.3 x 1 0 - 2 ~ ; pH 6.2. 1.67 X ~ O - ' M ; [H-INAA] = 3.8 x lo-' 0.4 0.3 W E -ft 2 0.2 2 0.1 0: 5 6 7 8 9 PH Fig. 2. Extraction as a function of pH. The calibration graph gave a straight line in the range 20-125 pg of copper per 10 ml of organic phase, indicating that Beer's law is obeyed at 450 nm. The concentrations of H-INAA and pyridine were varied from 0.1 to 0.5%. The results showed that 3 ml of 0.5% (0.038 M) H-INAA and 2 ml of 0.5% (0.063 M) pyridine solution were required for quantitative extraction and colour development of 20-125 pg of copper in 10 ml of chloroform. Excess of the reagent had no effect on the intensity of the colour. The absorbance of the complex in chloroform solution measured a t different time intervals between 10 min and 72 h did not show any significant change up to 48 h; after 48 h the absorbance decreased by 4%. Variation of shaking time showed that a minimum of 5-min shaking was sufficient for complete extraction of the complex.Prolonged shaking had no adverse effect on the extraction of the complex. The extraction coefficients (Table I) of the copper - INAA - pyridine complex between organic solvents and an aqueous solution of pH 6.2 showed partial extractions into benzene, toluene and amyl alcohol. The complex is insoluble in carbon tetrachloride, ethyl acetate, 4-methylpentan-2-one and butan-2-one. The only effective solvent tested was chloroform.January, 1978 OF COPPER I N STEEL AND NON-FERROUS ALLOYS TABLE I 0.05 0.1 - I I I 81 - I I I I EXTRACTION COEFFICIENTS OF COPPER - INAA - PYRIDINE ADDUCT BETWEEN ORGANIC SOLVENTS AND AQUEOUS SOLUTION OF PH 6.2 Solvent Carbon tetrachloride .. .. Benzene . . .. .. .. Toluene . . .. .. .. Ethyl acetate . . .. .. Amy1 alcohol . . .. .. 4-Methylpentan-2-one . . .. Butan-Bone . . .. .. Chloroform . . .. .. Extraction coefficient 0.0 0.9 0.8 0.0 1.4 0.0 0.0 co Copper extracted into organic solvent, yo 0.0 27.9 24.6 0.0 36.3 0.0 0.0 >99.9 Effect of Foreign Ions Various amounts of foreign ions were added to a fixed amount of copper (50 pg) in 25 ml of solution and the above procedure was applied. There was no interference (as shown by less than 1% deviation) from 10 mg each of zinc(II), cadmium(II), vanadium(V), tungsten- (VI), selenium(IV), manganese(II), iridium(III), fluoride, phosphate, bromide and nitrate ; 7.5 mg each of molybdenum(V1) and uranium(V1); 5 mg each of silver(1) and thorium(1V); 2.5 mg each of thallium(II1) and tartrate; and 1 mg each of mercury(II), lead(II), aluminium- (111), gold(III), osmium(VII1) and platinum(1V) ; 1 mg of bismuth(II1) can be masked with tartrate.Trace amounts of chromium (50 pg), iron (100 pg) and nickel (200 pg) did not interfere. Cobalt, tin(I1) , tellurium( IV) and palladium( 11) interfered. Complexing anions such as citrate, oxalate, thiourea, EDTA, ascorbate and thiocyanate interfered severely and must be absent. Nature of Extracted Species The composition of the mixed complex was determined by Job's continuous variation For Job's method, the colour of the complementary mixtures of equimolar solutions of the metal and the reagent was developed at pH 6.2 and the absorbance was measured at 450 nm.The maxima (Figs. 3 and 4) indicate that the copper to INAA ratio as well as the copper to pyridine ratio in the adduct is 1 : 2. Hence the probable composition of the extractable species is Cu(INAA),.ZC,H,N. The extraction reaction, therefore, can be expressed as CU(H,O)6'. + 2 H-INAA + CU(INAA)~.~H,O + 2H+ + 4H2O Cu(INAA),.2H20 + 2C,H,N + Cu(INAA),.C,H,N + 2H2082 PATIL AND SHINDE : RAPID SPECTROPHOTOMETRIC DETERMINATION Analyst, VoZ. 103 Pyridine thus exerts a synergistic effect on the extraction of the copper - INAA complex into chloroform. The 1 : 2 ratio of the metal to INAA and of the metal to pyridine has also been confirmed by the slope ratio method.Reproducibility and Sensitivity In a test of reproducibility of the method, six samples each containing 60, 100 and 120 pg of copper were prepared according to the general procedure and their absorbances measured at 450 nm. The standard deviation of the absorbance was found to be 0.004 for 100 pg of copper. The coefficient of variation was 1.16%. The results in Table I1 show that the method provides good precision. Sandell's spectrophotonietric sensitivity was found to be 2.6 x 10-2 pg cM2. TABLE I1 REPRODUCIBILITY OF THE METHOD Amount of Mean absorbance Standard Coefficient of copper/pg (6 determinations) deviation variation, yo 60 0.23 0.004 1.94 100 0.385 0.004 1.16 120 0.46 0.004 0.97 Detection and Determination of Copper in Synthetic Mixtures Synthetic mixtures containing either manganese, nickel, chromium, molybdenum and iron, or nickel, manganese, iron, bismuth and zinc or nickel, manganese, iron and aluminium in addition to copper were analysed by the proposed method.The results given in Table 111 show that the detection and determination of copper is possible in the presence of these metal ions. If the iron content is greater than 100 pg, it is removed from the mixture by selective extraction* with 5% trioctylamine from a solution 0.75 M in hydrochloric acid (the procedure is described below) and copper is deterniined in the aqueous extract as described in the general procedure. TABLE 111 ANALYSIS OF SYNTHETIC MIXTURES All analyses were carried out in triplicate.Copper Relative recovered , error, Composition of mixture and amounts takexilmg % % Cu, 0.1; Mn, 5; Ni, 0.2; Cr, 0.05; Mo, 5 ; Fe, 6 . . . . 98.7 1.3 Cu, 0.1; Pb, 1; Zn, 5 . . .. .. .. .. . . 99.4 0.6 Cu, 0.1; Ni, 0.2; Mn, 6 ; Fe, 0.1; Bi, 0.1; Zn, 5 .. . . 98.7 1.3 Cu, 0.1; Ni, 0.2; Mn, 6 ; Fe, 0.1; Al, 1 .. .. . . 98.5 1.5 Practical Applications Determination of Copper in Steel Dissolve 500 mg of 33d cast iron steel (NBS certified sample) in about 10 ml of aqua regia, evaporate nearly to dryness and then dilute to 100 ml. Take a l-ml aliquot of the solution containing 4.5 mg of iron and 100 pg of copper and add enough hydrochloric acid (0.0 ml of concentrated hydrochloric acid) to give a concentration of 0.75 M in a total volume of 10 ml.Extract the solution in a separating funnel by shaking for 3 min with three 6-ml portions of 5% trioctylamine in benzene. This extraction removes the iron in the organic phase while keeping the copper completely in the aqueous phase. Determine copper in the aqueous phase by the proposed method. Determination of Copper in Brass, Nickel Silver and Aluminium Alloy Dissolve about 1 g of brass, nickel silver or aluminium alloy as described elsewhere8 andJanzGary , I 9 78 determine copper by the proposed method in an aliquot of the solution. analysis of some standard samples are reported in Table IV. OF COPPER I N STEEL AND NON-FERROUS ALLOYS 83 Results for the TABLE IV Sample 33d steel (NBS sample) 4 1 brass (NML* sample) Nickel silver Aluminium alloy ANALYSIS OF STANDARD SAMPLES Recovery of other metals Composition, Copper content, % Relative A 3 elements other than error, Amount Relative copper, % Declared Found % Metal found, yo error, yo C, 2.30; Si, 1.63; Mn, 0.63; Ni, 2.38; Cr, 0.32; Mo, 0.48; remainder iron, 90.30 1.64 1.52 1.3 Iron 90.2 0.11 Pb, 2.35; Zn, 40.65; Fe, 0.009 66.9 56.5 0.7 Zinc 40.4 0.6 Ni, 18.20; Mn, 0.27; - - - Bi, 0.6; remainder Zn 66.1 54.75 0.6 Fe, 0.031 ; remainder Al 4.0 3.95 1.2 Ni, 2.12; Mn, 1.7; - - - * National Metallurgical Laboratory.Conclusion The proposed method is simple, rapid and free from interference from a large number of elements. The method effects the separation of copper from nickel, chromium, manganese, €ead, zinc, molybdenum, bismuth and aluminium in a single extraction with chloroform and its spectrophotometric determination in the solution. The results are accurate to within 50.5%. The method is rapid, the time required being of the order of 30-60 min. The wide applicability of the method is shown by the satisfactory analysis of a variety of samples. The authors thank the university authorities for providing financial assistance. 1. 2. 3. 4. 6. 6. 7. 8. References Talwar, U. B., and Haldar, B. C., Analyt. Chem., 1966, 38, 1929. Talwar, U. B., and Haldar, B. C., J . Indian Chew. SOC., 1972, 49, 785. Yeole, V. V., and Shinde, V. M. 2. Analyt. Chem., in the press. Welcher, F. J., “The Analytical Uses of EDT+;” Van Nostrand, New York, 1962, p. 242. Welcher, F. J., “Organic Analytical Reagents, Volume 3, Van Nostrand, New York, 1966. Irving, H., and Pierce, T. B., J . Chem. SOC., 1969, 2565. Job, P., Annali Chim., 1928, 9, 113. Patil, P. S., and Shinde, V. M., Mikrochim. Acta, in the press. Received May 16th, 1977 Accepted July 12th, 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300079

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

84 Analyst, Janwary, 1978, Vol. 103, p p . 84-92 Preparation of Trace Element Reference Materials by a Co-precipitated Gel Technique Alan R. Date Institute of Geological Sciences, Analytical and Ceramics Unit, 64/78 Gray’s Inn Road, London, WClX 8NG A method is described for the preparation of trace element reference materials by a co-precipitated gel technique. The technique is applied to the manu- facture of multi-element calibration standards for the analysis of geological materials by direct-reading emission spectrometry. The quality of gel reference material is considered in terms of homogeneity, major element and trace element accuracy and purity. Results are presented for the analysis of the gels by a variety of techniques, and for the analysis of international standard geological materials by direct-reading emission spectrometry with gel reference material calibration.Refevence material preparation ; geochemical trace elements ; GO- precipitated gel technique Keywords Powder reference materials that have been prepared by conventional techniques for use in silicate analysis by emission spectroscopy suffer from several disadvantages. The prepara- tion of glass usually leads to some loss of volatile trace elements during fusionlJ while solid dilution involves mixing errors and is limited by the availability of suitable starting materials. Both techniques often result in reference materials the matrices of which are incompatible with the samples for analysis. Geochemical research , in particular, involves the analysis of a wide variety of natural materials that have matrices differing in both chemical compo- sition and mineralogy.Although still not ideal, the method described below seeks to overcome some of the disadvantages inherent in reference materials prepared by conventional techniques. It is a development of the co-precipitated gel technique,3-6 in this instance involving the introduction of up to 30 elements in the same reference material, Development of the Method The technique was developed for multi-element trace element reference materials in order to satisfy a demand for large amounts of calibration standards for use in the analysis of geological materials by direct-reading emission spectrometry. The first such reference materials contained only those elements forming stable soluble nitrates, and the remainder were prepared by liquid - solid dilution additions to a gel matrix reference material.The technique was later extended to include elements having amphoteric properties by addition in sodium hydroxide solution. It was found that at the low concentrations required for most elements (it?., less than 1000 p.p.m.) there was no evidence of insoluble species being formed immediately prior to gelling. In this respect the only problem encountered was in trying to incorporate 2% of titanium(1V) oxide and 20% of magnesium oxide in the same major element reference material. A correction is usually applied to silicon dioxide to allow for a significant contribution in mass from the trace elements. This contribution is determined by assuming the formation of an oxide having the element in its highest oxidation state.This hypothesis was tested for 10-g amounts of reference materials, which were found to fall within 0.01 g of the calculated mass. Although some loss occurred in the preparation of 50-g batches of reference material, results presented for analysis of the products by several methods suggest that the composi- tional accuracy and precision are good. The addition of some elements in solution in sodium or potassium hydroxide involves the presence of sodium or potassium oxide in the reference material matrix, and a correction is made (in this instance to sodium oxide) to allow for the contribution from the hydroxide solution. Reference Material Preparation Major Element Reference Materials The main field of application of the technique was to the preparation of reference materialsDATE 85 having a range of silicon dioxide concentrations from 30 to SOYo, and it was therefore convenient to use silicon as the major gel-forming agent.Tetraethyl orthosilicate, available from Monsanto Chemicals Ltd., was used as a source of silicon dioxide. Standardisation was effected by gelling, igniting and weighing three weighed aliquots following a 1 + 1 dilution with ethanol. The major elements sodium, magnesium, potassium, calcium, iron and aluminium were incorporated by dissolving sodium carbonate, magnesium oxide, potassium carbonate, calcium carbonate, iron sponge (Johnson Mat they, Specpure) and aluminium powder (Koch-Light, 99.98%) in dilute nitric acid.Manganese, as both major and trace element, was incorporated by dissolving manganese(1V) oxide (Johnson Matthey, Specpure, standardised by a conventional titrimetric technique) in dilute nitric acid, with the aid of a small amount of hydrogen peroxide (100 volume). Titanium was incorporated by dissolving ammonium titanyl oxalate (Johnson Matthey, Specpure, crystalline, soluble in water, specially prepared to order) in water with a small amount of oxalic acid in order to reduce the risk of premature precipitation. Standardisation was effected by igniting the compound to the oxide and then weighing the residue. Trace Element Stock Solutions The range of trace element stock solutions was determined by the compatibility of their concentration ranges in the reference materials, the availability of suitable high-purity starting materials and the choice of a suitable solvent.The following groups were identified. Elements having compounds soluble in water Solutions for rubidium, silver, chromium, uranium and thorium were prepared using rubidium nitrate, silver nitrate (Johnson Matthey, Specpure), ammonium dichromate, uranyl nitrate hexahydrate and thorium(1V) nitrate hexahydrate (Hopkin & Williams, AnalaR) . Elements and compounds soluble in dilute nitric acid Solutions were prepared for lithium, cobalt, nickel, copper, zinc, strontium, yttrium, palladium, cadmium, indium, caesium, barium, lanthanum, cerium, lead and bismuth using lithium carbonate, cobalt sponge, nickel sponge, copper( 11) oxide, zinc oxide, strontium carbonate, yttrium oxide, palladium sponge (Johnson Matthey, Specpure), cadmium powder, indium powder (Koch-Light, high purity), caesium(1) carbonate, barium carbonate, lanthanum oxide, cerium( IV) oxide, lead(I1) oxide (Johnson Matthey, Specpure) and bismuth powder (Koch-Light, high purity).Elements and compounds sol.uble in concentrated nitric acid oxide (Johnson Matthey, Specpure) in hot concentrated nitric acid (sp. gr. 1.42). Elements soluble in aqua regia Reference materials for gold were prepared by using a stock solution obtained by dissolving gold sponge (Johnson Matthey, Specpure) in hot aqua regia. The concentration of chloride ion was reduced by repeated evaporation with nitric acid. Elements and compounds soluble in sodium hydroxide solution Solutions for vanadium, germanium, molybdenum, tin and tungsten were prepared by dissolving germanium powder and tin powder (Koch-Light, high purity), vanadium(V) oxide, molybdenum(V1) oxide and tungsten(V1) oxide (Johnson Matthey, Specpure) in sodium hydroxide solution (BDH, standardised solution). The standardised sodium hydroxide solution was transferred quantitatively into a PTFE beaker (500ml) and evaporated to small volume.The remaining solution was cooled and weighed amounts of the compounds then added. For solutions containing tin and germanium, hydrogen peroxide was added drop by drop to aid dissolution. In this instance the stock solution for gallium was prepared by dissolving gallium(II1) Zirconium stock solution Zirconium was added in the form of zirconyl nitrate solution (Johnson Matthey, Specpure),86 DATE : PREPARATION OF TRACE ELEMENT REFERENCE Analyst, VoZ.103 containing 5% m/m of zirconium, either by weighing the solution direct, for higher con- centrations, or by diluting a weighed aliquot, for lower concentrations. The solution was standardised by drying a weighed aliquot, igniting the residue and weighing the ignited oxide (at 1000 "C for 1 h). Niobium stock solution The solution was prepared by dissolving niobium(V) oxide (Johnson Matthey, Specpure) in a mixture of hydro- fluoric and nitric acids overnight in a PTFE-lined steel bomb at 120 "C. The resulting solution was transferred to a platinum dish and evaporated to small volume. It was then transferred to a polythene centrifuge tube, 1 g of ammonium nitrate added and the hydrated oxides were precipitated with ammonia solution using thymol blue as the indicator.The precipitate was washed three times with ammoniacal ammonium nitrate solution, dissolved in oxalic acid solution and made up to volume in a calibrated flask. Standardisation was effected by drying and igniting (at 1000 "C for 1 h) a known volume of the stock solution. Niobium was added as its oxalate complex in oxalic acid solution. Procedure The method described below applies to the preparation of a 50-g batch of reference material, although the same technique has been used in order to prepare standards for gold in batches weighing up to 500 g. Weigh the major element compounds, excluding those for silicon and titanium(1V) oxide and manganese, directly into a 1-1 PTFE beaker.When several reference materials are to be prepared, for reasons of economy 500-ml beakers can be used for the dissolution and the gelling subsequently carried out in the large beaker. Add 200 ml of de-ionised water to the beaker, followed by the calculated volume of nitric acid (sp. gr. 1.42), in an excess of 10% and added carefully in 5-ml portions to avoid excessive effervescence. When the readily soluble compounds have dissolved, warm the beaker gently on a hot-plate until the evolution of brown fumes has subsided. Dissolution of iron and aluminium can usually be completed by leaving the covered beaker on gentle heat overnight. When manganese is to form a major constituent its solution can be added at this point. Next, remove the cover-glass and evaporate the solution until a skin begins to form, then add de-ionised water (20 ml) and repeat the evaporation several times in order to reduce the acid concentration without reducing the volume to the point at which insoluble iron( 111) salts are precipitated.Cool the solution and add the required trace element solutions by pipette. When a large volume of trace element solution is involved, the nitrate solutions can be added first and evaporation continued to reduce the volume. Combine the various solutions by addition in the following order : trace element nitrate solutions ; trace element hydroxide solutions; major and trace element oxalate solutions; tetraethyl orthosilicate plus ethanol washings ; and finally, sufficient ethanol to give complete miscibility of the aqueous and ethanolic solutions.Stir the solution with a PTFE rod and add 20-50 ml of ammonia solution (sp. gr. 0.88) to promote rapid gelling. In the presence of other constituents the tetraethyl orthosilicate hydrolyses in a few seconds. The volume of ammonia solution required depends on the final acid concentration of the solution, which must be reduced as much as possible in order to avoid excess of liquid reducing the solid gel to a slurry. Remove the stirring rod and cover the beaker overnight to ensure that gelling is complete. Remove the cover and dry the gel in an oven a t 80 "C until there is no possibility of loss by sputtering, whereupon the temperature can be raised to 120 "C. When the gel is dry, transfer it to a platinum dish and heat gently over a Meker burner or in a furnace to decompose the nitrates.Transfer the gel to a Tema mill (agate) and grind it for 5 min at slow speed. Then return the powder to the platinum dish and ignite it at a temperature determined by experimentation for each gel composition so that slight sintering occurs and the hygroscopic nature of the gel is largely (or completely) destroyed. Finally, return the gel to the Tema mill, grind it for 10 min and place it in a bottle. Analytical-grade reagents are used throughout.January, 1978 MATERIALS BY A CO-PRECIPITATED GEL TECHNIQUE 8 7 Reference Material Quality Control Reference materials prepared by means of the technique described in this paper have been produced, where possible, from high-purity starting materials (such as Johnson Matthey Specpure compounds) and using laboratory practice commensurate with their intended primary use as calibration reference materials for the determination of trace and minor elements by d.c.arc direct-reading emission spectrometry. Data presented in this paper for previously analysed international geochemical reference materials therefore reflect the precision and accuracy that may be expected when using the d.c. arc excitation technique. For this reason the gels have also been analysed by independent analytical methods, including atomic-absorption spectrophotometry, emission spectrography, instrumental neutron-activa- tion analysis and colorimetry. Both classes of data are used in assessing “gel” compositional accuracy, which may be considered in terms of homogeneity, major element accuracy, trace element accuracy and purity. Homogeneity In a previous application of the co-precipitated gel technique,6 tests for major element homogeneity were carried out by using radiotracer techniques. The results showed that, within the limits of experimental error, the gels were homogeneous for a 30-mg mass of sample even before they were ground and mixed.In the present study, 100-mg sample masses were mixed with spectroscopic buffer, and three pellets each containing 15 mg of gel reference material were prepared and arced for each reference concentration in a single calibration. The gel grinding stage is retained to minimise variations in the physical properties of the gel, and to reduce the lightly sintered material to a fine powder suitable for weighing.Although standard analytical techniques normally require larger amounts of sample than those taken for analysis by emission spectroscopy, the results presented below for major and trace element accuracy nevertheless constitute evidence of homogeneity. Additional evidence is provided in the form of statistical data on a reference material that is used to monitor calibration stability in analysis by direct-reading emission spectrometry. Direct-reader calibration is carried out at the beginning of each week, and reference materials containing the total element range of the instrument are arced a t intervals during the week. Statistical data for a number of elements in reference material IAB-3 are presented in Table I, and a direct comparison can be made with international standard rocks having similar concentration levels that were arced on the same days.There is no significant difference between the two groups and gel reference material homogeneity can be considered good a t the 100-mg level with a 15-mg sub-sample mass. TABLE I COMPARISON OF PRECISION FOR SEVERAL ELEMENTS IN REFERENCE MATERIAL IAB-3 WITH SIMILAR CONCENTRATION LEVELS IN INTERNATIONAL STANDARD ROCKS Ten determinations in 6 months. Reference material Calibration concen- Element range tration MgO .. 0.2-20 5.0 A1203 .. 3.0-30 16.0 CaO . . 0.2-20 5.0 TiO, . 0.03-5.0 0.5 Fe,O, .. 0.2-20 6.0 Major element data, yo- K2O .. 0.1-10 2.0 Trace element data, 9.p.m.- Li .. . . 0.5-50 60 v .. .. 10-1000 100 Co .... 6-500 50 CU .. .. 6-500 50 Ga ,. . . 0.5-50 10 Rb .. . . 6-600 60 Direct- reader mean 5.90 1.60 6.38 0.49 4.69 15.4 48.7 50.3 66.9 8.9 53.6 106 Relative Inter- standard national deviation, standard % rock 12.3 BCR-1 27.1 G-2 11.1 AGV-1 10.8 BCR-1 14.2 G-2 15.3 AGV-1 10.4 G-2 37.1 AGV-1 25.5 BCR-1 9.0 AGV-1 13.0 G-2 11.6 BCR-1 Recom- mended value’ 3.46 2.89 6.92 0.50 6.76 15.4 34.8 125 38 69.7 22.9 46.6 Direct- reader mean 3.80 2.74 7.63 0.57 5.91 15.2 Relative standard deviation % 10.6 16.3 16.6 11.2 10.4 16.5 33.4 12.9 112 23.6 38.3 28.8 67.3 13.4 17.9 11.2 63.9 18.988 DATE: PREPARATION OF TRJQCE ELEMENT REFERENCE Analyst, Vol. 103 Major Element Accuracy Reference materials that provided variation in content of the major elements were prepared primarily to facilitate matrix correction in analysis by direct-reading emission spectrometry.This does not imply that tests for gel accuracy should be any less strict than for trace elements. Previous workers in this field5y6 have shown that gel yields close to 100% will result in accurate major element compositions. In the present study only two major element reference materials have been independently analysed (Table 11). TABLE I1 MAJOR ELEMENTS IN MATRIX RE:FERENCE MATERIALS C AND I Reference material C I -l Oxide Expected, ya Found,* yo Found,? (yo SiO, . . . . 60.0 59.98 59.8 Al,o, .. .. 21.0 21.15 20.9 TiO, , . . . 2.0 1.72 1.71 Fe,O, . . . . 2.0 2.01 2.05 MgO .. .. 1.0 1.02 1.12 CaO .. .. 10.0 10.14 10.2 Na,O . . . . 2.0 2.04 1.9 K,O .. . . 2.0 1.98 2.1 Reference material I Expected, % Found,* % Found,$ yo 16.0 16.02 15.7 f h -7 65.0 64.32 3 0.5 0.52 <0.33 5.0 4.90 5.48 5.0 5.08 3 5.0 5.07 3 1.5 1.52 1.36 2.0 1.96 2.04 * Analysis by classical methods, Dawn Hutchison, G,eochemical Division, IGS.t Analysis by X-ray fluorescence spectrometry, G. R. Lachance, Geological Survey of Canada. $ Instrumental neutron-activation analysis, J . Herrington, AWRE, Aldermaston. S Not determined. In addition, results are presented in Table I11 for the determination of titanium(1V) oxide and calcium oxide in a series of international geochemical reference materials by means of direct-reading emission spectrometry that is based on calibration using standards prepared by the proposed method. Major element accuracy is considered to be satisfactory in the present context, particularly for titanium( IV) oxide, which is normally a difficult element to incorporate in gel reference materials.TABLE I11 DETERMINATION OF TITANIUM (Iv) OXIDE AND CALCIUM OXIDE IN INTERNATIONAL STANDARD ROCKS BY DIRECT READER WITH GEL REFERENCE MATERIAL CALIBRATION Ten determinations in 5 months. Titanium(1V) oxide Calcium oxide Standard rock G-2 .. .. GSP-1 . . AGV-1 . . BCR-1 , . DTS-1 . . PCC-1 . . SY-1 .. .. SY-2 .. .. SY-3 . . . . MRG-1 . . Published value, % 0.507 0.667 1.047 0.013? 0.0157 0.4g7 0.158 0.158 3.6g8 2.207 Direct- reader mean, yo 0.57 0.73 0.98 1.94 nd* nd 0.47 0.13 0.13 t * nd = not detected (<0.03%). t Overflow (>3.0y0). Relative standard deviation, :& 10.5 12.3 11.2 11.9 c - 12.8 15.4 15.4 - f Published value, yo 1.947 4.907 6.927 0.157 0.517 8.038 8.308 14.6Ss 2.027 10.27 Direct- reader mean, yo 2.13 2.23 5.97 7.63 0.04 0.35 8.79 8.82 8.62 16.0 Relative7 standard deviation,l% 5.0 3.9 7.9 11.2 39.1 33.3 12.7 9.2 8.5 12.9 Trace Element Accuracy Trace element reference materials prepared by the proposed method have been analysed by use of several independent analytical techniques and results are presented in Tables IV and V for two series of calibration standards used in analysis by direct-reading emission spectrometry.January, 1978 MATERIALS BY A CO-PRECIPITATED GEL TECHNIQUE 89 Table IV covers the results obtained for a series of volatile element reference materials. Compositional accuracy is considered to be good, even for molybdenum and tin, elements incorporated into the reference material by addition in sodium hydroxide solution.TABLE I V ANALYSIS OF MATRIX I SERIES A TRACE ELEMENT REFERENCE MATERIALS BY SEVERAL TECHNIQUES IA5 IA4 IA3 IA2 ---- Expected, Found, Expected, Found, Expected, Found, Expected, Found, Element p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. Li .. .. 2 * 2 10 * 8 50 *37 200 *193 cu .. .. 2 * 4 10 *13 50 * 53 200 *196 t177 $<30 50 $<30 200 3192 Ga .. .. 1 *11 Rb .. . . 2 lI M o . . .. 2 *11 t 4 $< 10 t l $< 10 § 3 § 7 25 $ 22 10 *11 50 * 53 t 43 $ 31 8 54 f l28.5 20 * 23 100 500 t492 500 *510 Ag .. .. 1 * 0.8 2 1;: 200 *216 50 1;; 50 Sn .. .. 20 Pb .. .. 20 t25 t51 7262 7522 * Atomic-absorption spectrophotometry, P. T. S. Sandon, Geochemical Division, IGS. t Optical-emission spectrography, B.A. T. Tait, Geochemical Division, IGS. 3 Instrumental neutron-activation analysis, J. Herrington, AWRE, Aldermaston. Colorimetry by visual comparison, M. Shah, Geochemical Division, IGS. 7 Not determined. t 49 10 t 7 $< 10 IZl w- Expected, Found p.p.m. p.p.m. 1000 * 990 1000 * 990 t>400 100 $ 92 1000 $1118 100 * 105 t 94 $ 87 100 100 * 84 1000 71015 2 000 *1950 $ > l o o 0 Table V contains results obtained for part of a more recent direct-reader calibration standard series, in this instance incorporating mainly involatile elements with lithium, zinc and lead at high concentration levels. In this instance vanadium was incorporated into the reference material by addition in sodium hydroxide solution. TABLE V ANALYSIS OF DR SERIES TRACE ELEMENT REFERENCE MATERIALS BY THREE TECHNIQUES Element Li ..v .. Cr .. Mn . . co .. Ni .. Zn .. Y .. Ba .. La .. Pb .. DR09 - Expected, Found, p.p.m. p.p.m. .. 50 .. 20 ; ;; i < 50 .. 50 - j 54 .. 200 *200 t190 $190 -. 20 * 15 ,. 50 .. 100 .. 20 .. 200 .. 20 .. 100 DR07 M Expected, Found, p.p.m. p.p.m. 100 * 81 50 t 47 100 i1:; 500 *530 t600 $520 60 * 48 t 44 100 200 *210 $290 50 500 50 200 1555: 52;: DR05 - Expected, Found, p.p.m. p.p.m. 200 * 200 100 t 80 : 120 200 $ 180 1000 *lo70 t 950 $ 980 100 * 105 t 100 $ 110 200 * 210 500 * 500 $ 560 100 100 $ 110 500 * 500 1000 110;; DR03 - Expected, Found, p.p.m. p.p.m. 500 * 510 200 t 200 $ 290 500 $ 460 2000 *2050 t 2 150 $2 000 200 * 200 t 220 $ 220 500 * 540 1000 *lo50 $1 000 200 $ 220 1000 * 990 200 2000 12% DROl - Expected, Found, p.p.m.p.p.m. 1000 * 990 500 t 590 $ 590 1000 900 5000 *5200 500 540 433 $ 510 1000 *lo70 2000 82000 $1 900 500 t 440 5000 $5200 500 $ 530 2000 *2010 1 4 $00 * Atomic-absorption spectrophotometry, P. T. S Sandon, Geochemical Division, IGS t Optical-emission spectrography, B. A. R. Tait, Geochemical Division, IGS. $ Instrumental neutron-activation analysis, J. Herrington, AWRE, Aldermaston Not determined. A further assessment of trace element compositional accuracy can be made by considering the results obtained by use of direct-reading emission spectrometry for readily analysed elements in a series of international natural reference materials, including rocks and geo- chemical prospecting samples. Results obtained for lithium, copper and bismuth are presented in Tables VI and VII.Purity The application of co-precipitated gel reference materials in ultra-trace element analytical fields would obviously require more stringent laboratory practice than that described in the90 PREPARATION O F TRACE ELEMENT REFERENCE TABLE VI DETERMINATION OF LITHIUM AND COPPER IN INTERNATIONAL STANDARD ROCKS BY Analyst, Vol. 103 DIRECT READER WITH GEL REFERENCE MATERIAL CALIBRATION Standard Published G-2 . . 34.B7 GSP-I" .. 32.17 AGV-1 . , lz7 BCR-1 . . 12.8' DTS-1 . . 27 rock value, p.p.m. Pcc-i . . 27 s y - i .. .. 1217 SY-2 .. .. 868 MRG-1 .. 48 SY-3 . . .. 8ij8 Ten determinations in 5 months. Lithium Direct- reader mean, p.p.m. 33.4 31.3 11.8 15.8 -L 2.03 1.22 130 95.8 96.9 4.8 Copper 1 Relative: standard deviation, yo 12.9 8.5 16.0 16.1 20.9 41.2 13.0 18.4 13.1 20.5 Published value, p.p.m.11.77 33.37 59.77 18.4l 7.07 11.3' 237 78 1 88 135" Direct- reader mean, p.p.m. 11.3 37.4 67.3 21.5 9.4 14.4 24.1 6.7 16.0 149 1 Relative standard deviation, yo 24.0 12.0 13.4 15.1 27.1 76.7 33.4 24.4 32.5 6.7 TABLE VII DETERMINATION OF BISMUTH I N USGS GEOCHEMICAL EXPLORATION SAMPLES BY DIRECT READER WITH GEL REFERENCE MATERIAL CALIBRATION Ten determinations in 5 months. Published Standard median data, Direct-reader sample p.p.m.@ mean, p.p.m. GXR-1 .. .. * 1576 1 004 t 1000 $1000 t < 10 < 10 15 GXR-2 .. .. * 6.5 6.0 GXR-3 .. .. * 18.6 20 GXR-4 .. .. * 25 19 GXR-6 .. .. : 7 2.9 t < 10 Relative standard deviation, % 6.7 30 136 19 99 GXR-6 .... * 11 t < 10 * Atomic-absorption spectrophotometry. $ Colorimetry. Emission spectrography. 3.1 95 present context, and more careful consideration of both non-stoicheiometry in the starting materials and possible sources of contamination. In the present context only two serious contaminants have been identified, namely tin and lead. A particularly high level of tin contamination. is shown by the intermediate rock matrix reference material PG (Table VIII) when analysed by emission spectrography using a visual comparison technique of plate evaluation. This reference material was prepared at an early stage in the development of the method when only small amounts were required. Tetra- ethyl orthosilicate was purchased in 1-gallon amounts and delivery was made in tinplate cans.Although its transfer to glass storage bottles was effected immediately on receipt, it was found that slight acidity in the tetraethyl orthosilicate (0.01% of hydrochloric acid) was sufficient to cause serious contamination. Subsequent reference materials were prepared from tetraethyl orthosilicate purchased in 25-kg drums, in which the commodity is normally shipped and stored, and transferred on receipt to glass. No contamination is apparent for the elements of interest in this study; the 25-kg drums are lined with polythene and consideration may have to be given in other applications to purification of the organic silicate by, possibly, a distillation process. Although contamination by tin is dependent on the silicon dioxide content of the referenceJanuary, 1978 MATERIALS BY A CO-PRECIPITATED GEL TECHNIaUE TABLE VIII 91 MATRIX REFERENCE MATERIAL PG Analysis by emission spectrography (visual comparison) by P.Greenwood, Geochemical Division, IGS. Element Be .. B .. v .. Cr . , Mn . . co .. Ni .. cu .. Zn .. Y .. Concen- tration expected, p.p.m. .. 0 .. 0 .. 0 .. 0 .. 0 .. 0 .. 0 .. 0 .. 0 .. 0 Detection limit, p.p.m. 2 6 6 32 10 2 1 1 100 2 Concentration found, p.p.m. Element Not detected Zr Not detected Nb Not detected Mo Not detected In Not detected Sn Very slight trace Ba 1, 1 W Not detected Pb Not detected Bi Not detected Ag Concen- tration expected, p.p.m. 0 0 0 0 0 0 0 0 0 0 Detection limit, p.p.m. 56 66 1 0.1 2 1 100 320 1 2 Concentration found, p.p.m. Not detected Not detected Not detected Not detected Very slight trace 56, 42 Not detected Not detected Not detected Not detected material, lead contamination is a different matter.Results are presented in Table IX for a series of trace element reference materials in a nominally common matrix (a slight correction is made to the silicon dioxide content to allow for the trace element contribution) analysed by emission spectrography at the Analytical Chemistry Section of the Geological Survey of Canada. Contamination by lead is apparent in matrix C and reference material CWO9, but is absent from CW06, CW08, CWlO and intermediate rock matrix PG (Table VIII). The source of this irregular contamination, which cannot be related to any particular major element component of the reference material, nor to variation in reagent use in gel prepara- tion, is thought to be dust in the IGS laboratories in Central London, which has been analysed by atomic-absorption spectrophotometry and found to contain up to 2000 p.p.m.m/m of lead. TABLE IX LEAD CONTAMINATION IN MATRIX C SERIES W REFERENCE MATERIALS Analysis by means of emission spectrography, using globule arc fractional distillation, by C. F. Meeds, K. A. Church and W. H. Champ, Analytical Chemistry Section, Geological Survey of Canada. Matrix C CW06 CW08 cwo9 cwlo Expected, Found, Expected, Found, Expected, Found, Expected, Found, -, Element p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m, p.p.m. p.p.m. Ag .. 0 0.069 2 2.a 0.6 0.46 0.2 0.29 0.1 0.20 0.097 2.3 0.63 0.27 0.20 Sa .. 0 <0.6 20 13 6 4.8 a 2.4 1 1.2 ---- 16 4.4 2.0 1.1 Pb ..0 23 20 17 6 8.2 2 19 1 3.0 13 20 7.6 20 3.8 Bi .. 0 <O.B 20 14 6 2.3 2 0.89 1 <0.6 12 2.8 0.93 The possibility of contamination being introduced at the gel drying stage, when the PTFE beaker is exposed to the atmosphere for up to 48 h, was tested by ensuring that drying took place with the beaker covered by a paper tissue. Lead contamination was effectively eliminated in this way. The possibility that the PTFE had influenced preferential enrich- ment of lead was not investigated. Conclusions Results have been presented to show that co-precipitated gel reference materials can be used as calibration standards in the trace element analysis of geochemical materials. In the proposed method of preparation silicon forms the major gel-forming agent, although the other geochemically significant major elements, aluminium, magnesium, calcium and iron, also play a part in gel structure.92 DATE The main advantage of the technique lies in the production of significantly large amounts (50-500 g) of reliable calibration reference materials.The disadvantages associated with glass preparation and solid - solid mixing are avoided; the reference materials approximate closely in both physical and chemical characteristics to the samples for analysis, with a consequent improvement in analytical accuracy. The lightly sintered gels are particularly suited to calibration in physical methods of analysis. They can be applied to chemical methods when due consideration is given to their behaviour under the conditions chosen for sample decomposition. The possibility of incorporating a selected wide range of elements at known fixed levels of concentration in a solid matrix is of great value in assessing the performance of new analytical procedures and in comparing several independent analytical techniques. The author thanks Mr. Sydney Abbey of the Analytical Chemistry Section, Geological Survey of Canada, Dr. J. Herrington of the Atomic Weapons Research Establishment, Aldermaston, and colleagues in the Analytical and Ceramics Unit, Geochemical Division, Institute of Geological Sciences, who are responsible for analyses carried out in support of this publication. Particular thanks are due to Mr. P. J. Moore, Chief Chemist, Analytical and Ceramics Unit, Institute of Geological Sciences, for critically reading the manuscript, and to Dr. A. W. Woodland, Director, Institute of Geological Sciences, for allowing publication of the paper. References 1. 2. 3. 4. 5. 6. 7. 8. Schairer, J. F., and Bowen, N. L., Am. J . Sci., 1955, 253, 681. Schairer, J. F., and Bowen, N. L., Am. J. Sci., 1956, 254, 129. Roy, R., J. Am. Ceram. Soc., 1956, 39, 145. Hamilton, D. L., and MacKenzie, W. S., J. Petrology, 1960, 1. 56. Luth, W. C., and Ingamells, C. O., Amer. Min., 1965, 50, 255. Hamilton, D. L., and Henderson, C. M. B., Min. Mag., 1968, 36, 832. Flanagan, F. J., Geochim. Cosmochim. Acta, 1973, 37, 1189. Abbey, S., Gillieson, A. H., and Perrault, G., “A Report on the Collaborative Analysis of Three Canadian Rock Samples for Use as Certified Reference Materials,” Canadian Certified Reference Materials Project, Ottawa, 1975; Can. J. Spectrosc., 1975, 20, 113. Allcott, G. H., and Lakin, A. W., “Statistical Summary of Geochemical Data Furnished by 85 Laboratories for Six Geochemical Exploration Reference Samples,” US Geological Survey Open File Report, Denver, 1974. Received May 6th, 1977 Accepted August 12th, 1977 9.

ISSN:0003-2654    

DOI:10.1039/AN9780300084

出版商:RSC    

年代:1978

数据来源: RSC