ISSN:0003-2654    

DOI:10.1039/AN97499FX041

出版商:RSC    

年代:1974

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97499BX043

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

iv THE ANALYST [November, 1974THE ANALYSTEDITORIAL ADVISORY BOARDChairman: H. J. Cluley (Wembley)*L. S. Bark (Solford)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)"R. M. Dagnall (Honfingdon)E. A. M. F. Dahmen (The Netherlands)*J. B. Dawson (Leeds)A. C. Docherty (Billingham)D. Dyrssen (Sweden)*W. T. Elwell (Birmingham)*D. C. Garratt (London)J. Hoste (Belgium)D. N. Hume (U.S.A.)H. M. N. H. Irving (Leeds)M. T. Kelley (U.S.A.)*J. A. Hunter (Edinburgh)W. Kemula (Poland)*G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)*J. M. Ottaway (Glasgow)*G. E. Penketh (Billingham)S. A. Price (Tadworth)D. 1. Rees (London)E. B. Sandell (U.S.A.)*R. Sawyer (London)A. A.Smales, O.B.E. (Harwell)H. E. Stagg (Manchester)E. Stahl (Germany)A. Walsh (Australia)T. S. West (London)P. Zuman (U.S.A.)*A. Townshend (Birmingham)* Members of the Board serving on the Executive Committee.NOTICE TO SUBSCRIBERS(other than Members of the Society)Subscriptions for The Analyst, Analytical Abstracts and Proceedings should beThe Chemical Society, Publications Sales Ofnce,Blackhorse Road, Letchworth, Herts.Rates for 1974sent to:(a) The Analyst, Analytical Abstracts, and Proceedings, with indexes . . . . €37.00(b) The Analyst, Analytical Abstracts printed on one side of the paper (withoutindex), and Proceedings . . . . . . . . . . . . . . €38.00(c) The Analyst, Analytical Abstracts printed on one side of the paper (withindex), and Proceedings .. . . . . . . . . . . . . a €45.00The Analyst and Analytical Abstracts without Proceedings-(d) The Analyst and Analytical Abstracts, with indexes . . . . . . . . €34.00(e) The Analyst, and Analytical Abstracts printed on one side of the paper (withoutindex) . . . . . . . . . . . . . . . . . . €35.00(f) The Analyst, and Analytical Abstracts printed on one side of the paper (withindex) . . . . . . . . . . . . . . . . . . €42-00(Subscriptions are NOT accepted for The Analyst and/or for Proceedings aloneV November, 19741 THE ANALYST HUMPHRY DAVY'SBRILLIANT INVENTIONnIt is well known that Sir Humphry had quite a few laughs finding a use fornitrous oxide. There would have been no problem, however, finding a gooduse for 'Pronalys' analytical reagents because they are exceptionally purechemicals ideal for use in the most exacting analytical procedures.Comprehensive specifications are available for every product and these,with their constant composition, ensure consistency of results.But as 'Pronalys' wasn't around at the time, he invented a miner'ssafety-lamp instead.1IIIII I~ = = I I I ~ I ~ I ~ ~ ~ I I I I = ~ I ~To:May & Baker Ltd Dagenham Essex RM10 7XSPlease send me further information on 'Pronalys' highpurity analytical reagentsName ............................................................................................................ III Address ...................................................................................................... I- r i r r r r r r r r r r r r r r r r r ~ r r m r r l.......................................................................................................................'Pronalys' is a trade markdaMay&Baker I Rh6ne-Poulenc Groupof CompanieSUMMARIES OF PAPERS IN THIS ISSUE [November, 1974Summaries of Papers in this IssueThe Use of Precipitate Based Silicone Rubber Ion- selectiveElectrodes and Silicone Rubber Based Graphite Voltam-metric Electrodes in Continuous AnalysisA ReviewSUMMARY OF CONTENTSIntroductionBehaviour of the electrodes in flowing systemspX electrodespE electrodesReference electrodesMeasuring cells and measuring methodsApplications in continuous monitoringContinuous monitoring of the chloride content of natural waterDetermination of cyanide content of industrial sewage effluentsDetermination of dissolution rate of drugs from pharmaceuticalApplication of the silicone rubber based graphite electrode to in vivoChromatovoltammetric detector cell containing silicone rubber basedApplication of the injection techniquepreparationsmeasurementselectrodeConclusionREPRINTS of this Review paper can be obtained from The Society for AnalyticalChemistry, Book Department, 9/10 Savile Row, London, W1X lAF, at 50pper copy (with a 25 per cent.discount for 4 or more copies), post free.A remittance for the correct amount, made out to The Society forAnalytical Chemistry, should accompany every order; these reprints are notavailable through Trade Agents.2s.FEHaR, G. NAGY, K. ToTH and E. PUNGORInstitute for General and Analytical Chemistry, Technical University, Budapest,Hungary.Analyst, 1974, 99, 699-708.A Simple Modification to a Flame-ionisation Detector LinearAmplifier for Logarithmic Display of Gas Chromatogramsand its UseWith the addition of a single diode, a linear electrometer amplifier for gaschromatography with flame-ionisation detection has been modified so as toenable it to respond to ionisation currents, over the range 10-l2 to A, in amanner closely approaching logarithmic proportionality. Some chromato-grams obtained with this amplifier are included and attention is drawn to theadvantages of logarithmic display.D. A. COLLINSWest Midland Forensic Science Laboratory, Priory House, Gooch Street North,Birmingham, B5 6QQ.Analyst, 1974, 99, 709-716November, 19741 SUMMARIES OF PAPERS IN THIS ISSUEA Convenient Method for Collecting Gas-chromatographicEffluents for Micro- scale Infrared SpectrophotometricAnalysis in the Gas PhaseA method for the collection of fractions of low relative molecular massissuing from a gas chromatograph followed by identification using infraredspectrophotometry is described.The fractions are transferred to a micro-scale gas cell by a simple cryogenic technique, which requires minimumhandling of the separated fractions. Good infrared spectra were obtained fora sample size of 30 to 100 pg.K. HARALD NORINArrhenius Laboratory, Department of Analytical Chemistry, University of Stock-holm, S-104 05 Stockholm, Sweden.Analyst, 1974, 99, 717-723.viiAntioxidant Analysis Incorporating a Thin-layer ChromatographicSeparation ProcedureThe method described consists of three distinct and separate steps.Initially, a thin-layer chromatographic technique, involving the use of mixedsilica gel - aluminium oxide plates, is used to separate and identify the ex-tracted antioxidants.This step is followed by confirmation of the suspectedidentity and determination of the antioxidant by using infrared and ultra-violet spectroscopy.This method is considered applicable, over a range of concentrations, tothe analysis of both antioxidants and ultraviolet screening compoundspossessing ultraviolet absorbing properties, in polyolefinic polymers such aspolyethylene, polypropylene, poly(4-methylpentene) , etc.I t can be used todetect such additives down to a level of 200 pg g-l, and possibly even less, in agiven polymer.In addition, the procedure has been successfully applied to the separationof isomeric antioxidants and to the identification of antioxidants present inlubricating oils.DAVID T. MILESAustralian Post Office Research Laboratories, 59 Little Collins Street, Melbourne3000, Australia.Analyst, 1974, 99, 724-728.Partly Quenched, Synchronously Excited FluorescenceEmission Spectra in the Characterisation of Complex MixturesNon-linear quenching of mixed fluorescences a t low concentrations ofquencher, and linear variation at increased concentrations, depend on theStern-Volmer coefficients and on the proportions of the components in amanner that can be predicted from the Stern - Volmer equation. The varia-tion with quencher concentration of emission intensities synchronouslyexcited from mixtures of benzo[a]pyrene and perylene at intervals of 45 nmis quantitatively interpreted in this way.A variety of examples is used to show that in the qualitative charac-terisation of complex mixtures, the specificity of synchronously excited spectrais enhanced by the use of varyingly quenched conditions. The high resolutionafforded by the synchronous excitation technique readily demonstrates thewidely differing relative sensitivities of components of fluorescent mixturesto quenching effects; and reveals new emissions that are obscured undernon-quenching conditions.J. B. F. LLOYDWest Midland Forensic Science Laboratory, Priory House, Gooch Street North,Birmingham, B5 6QQ.Analyst, 1974, 99, 729-738

ISSN:0003-2654    

DOI:10.1039/AN97499FP123

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

X SUMMARIES OF PAPERS I N THIS ISSUE [November, 1974A Kinetic Theory of Atomisation for Non-flameAtomic-absorption Spectrometry with a Graphite Furnace.The Kinetics and Mechanism of Atomisation for CopperA kinetic approach has been made to the atomisation process in non-flame atomic-absorption spectrometry. Time - absorbance profiles for thedetermination of copper, using a graphite furnace, have been investigatedin the temperature range 1720 to 2220 K and a rate equation derived thatdescribes the variation of the amount of copper atoms in the furnace with time.It has been shown that a slow first-order reaction involving reduction of copperoxide by carbon followed by the rapid vaporisation of the copper formed isthe most probable reaction mechanism. The greater sensitivity achieved inthe detqmination of copper using a tantalum-lined graphite furnace hasbeen attributed in part to the greater rate of reduction of copper oxide bytantalum.C.W. FULLERTioxide International Limited, Billingham, Cleveland.Analyst, 1974, 99, 739-744.Determination of Soap in Refined Vegetable Oils byAtomic-absorption SpectrophotometrySoap has been determined, as sodium, in alkali-refined vegetable oils byatomic-absorption spectrophotometry. The oil is treated with absoluteethanol, the mixture dissolved in ethyl methyl ketone and the solution thenaspirated. Oil solutions have been compared with standards containingvirgin olive oil and known amounts of sodium oleate. Concentrations ofsodium oleate in the range 3 to 1000 p.p.m.in oil show a linear absorption.The method is rapid and accurate and can also be applied to the detection ofadulteration of virgin olive oil with refined olive or other vegetable oils.DINA GEGIOUResearch Department, State Chemical Laboratories, 16 A. Tsoha Street, Athens,Greece.Analyst, 1974, 99, 745-748.Automatic Analysis of Trace Amountsof 2-Furfuraldehyde in Gas OilThe design and development of an automatic instrument for the identifi-cation and determination of 2-furfuraldehyde (0 to 6.0 mg kg-I) in gas oil isdescribed. The principle of operation of the instrument, which is in routineuse, involves a gas-chromatographic separation followed by a colorimetricdetermination.R. G. LIDZEY and P. B. STOCKWELLDepartment of Industry, Laboratory of the Government Chemist, Cornwall House,Stamford Street, London, SE1 9NQ.Analyst, 1974, 99, 749-754.An Improved Method for the Colorimetric Assay ofThiacetazone in Pharmaceutical PreparationsAn improved colorimetric procedure for the determination of thiacetazone(tibione) has been developed. It is based on the complete hydrolysis oftibione in a concentrated alkaline medium, after which the hydrolytic productis subjected to diazotisation and coupling with a-naphthylethylenediaminedihydrochloride.The procedure suggested in this paper eliminates somedisadvantages of the original method and affords reproducible assay resultsfor tibione in galenical preparations.JACOB S . SHOHETDepartment of Inorganic and Analytical Chemistry, Hebrew University of Jerusalem,Jerusalem, Israel.Analyst, 1974, 99, 755-758November, 19741 SUMMARIES OF PAPERS I N THIS ISSUEStudies on a Specific Colour Reaction for the Determination ofTestosteroneThe chromogen resulting from preliminary treatment of testosterone withconcentrated sulphuric acid gives a stable green colour on further treatmentwith a saturated aqueous solution of picric acid.The absorbance graph of thegreen colour shows two maxima, at 470 and 640nni, and obeys Beer’s law.Under the specified conditions, absorbances at the two maxima have a fairlyconstant ratio (2.80 to 2-96), which is characteristic for testosterone and itsesters in amounts ranging from 125 to 250 pg. Spectral and structural studiesinvolving a large number of steroids show that the colour reaction isspecific for testosterone.EMIL FAHMY, DAWOUD A.YASSAResearch and Control Department, Sociktk Misr pour 1’Industrie Pharmaceutique,92 El Mataria Street, Post El Zeitoun, Cairo, Egypt.and NAG1 WAHBABiochemistry Department, Faculty of Medicine, Ain Shams University, Cairo, Egypt.Analyst, 1974, 99, 759-764.xiAn Absolute Galvanic Detector for Nitrogen DioxideThe performance of a modified Hersch and Deuringer galvanic detectorfor nitrogen dioxide was examined. The results indicate that the response iscoulometric at low flow-rates and low concentrations on the basis of 1 Faraday(96 487 C ) per mole of nitrogen dioxide. This response is explained in termsof the electrochemical reduction of nitrogen dioxide to the nitrite ion ornitrous acid.The cell can thus differentiate between nitrogen dioxide andnitrous acid vapour, which might be present in humid air. The detectorhas an immediate application for calibrating low-concentration mixturescontaining nitrogen dioxide, but further work is necessary to investigatethe reliability of the selective scrubbers required for monitoring nitrogendioxide in polluted, ambient air.J. D. ALLENBritish Gas Corporation, Research and Development Division, Watson House,Peterborough Road, London, SW6 3HN.Analyst, 1974, 99, 765-770.The Determination of Mercury( 11) by RadiochemicalReplacement with Silver-llOmMercury(I1) in aqueous solution can be determined by shaking with adilute solution of silver- 11 Om dibutyldithiocarbamate in chloroform, andcounting the silver-1 lorn extracted into the aqueous layer. Alternatively, thereagent in chloroform can be shaken with up to fifty times its volume ofwater, and the mercury present measured by the decrease in specific activityof the chloroform layer.The method is rapid, has a minimum sensitivityof 1 ng ~ m - ~ of mercury and a precision of f5 per cent. Calibration graphsare linear, and the technique would be suitable for use in the field, in orderto avoid losses of mercury during the storage or transportation of aqueoussamples. Interference by chloride ions can be serious, but is completelyovercome by adding potassium cyanide. Gold and palladium are the onlymetals that are likely to interfere.H.J. M. BOWENChemistry Department, Reading University, Whiteknights, Reading, Berkshire.Analyst, 1974, 99, 771-773xii SUMMARIES OF PAPERS I N THIS ISSUEThe Control Analysis of Boron by Measurement of the Transmissionof Radioisotope Source NeutronsBoron at levels of between 0.1 and 10 per cent. has been determined non-destructively in bulk solids such as boron - aluminium alloys by measurementof the intensity of a transmitted neutron beam. The neutrons were producedfrom a 300-mCi 241Am - Be source and employed in an automatic instrumentalsystem developed for use in a control laboratory at a rate of up to thirty-sixsamples per hour. The accuracy of such a technique has been shown to begood and the precision is better than -& 1 per cent.The method is best appliedto the determination of boron in matrices that have low thermal neutronabsorption or scattering cross-sections.T. B. PIERCE, C. R. BOSWELL and P. F. PECKApplied Chemistry Division, Atomic Energy Research Establishment, Harwell,Didcot, Oxfordshire, OX11 ORA.Analyst, 1974, 99, 774-781.[November, 1974Bureau ofAnalysedSamples Ltd.announce the introduction of a newrange of BCIRA/BAS spectroscopicsetting-up samples for low, mediumand high phosphorus cast irons.Full details are available on request.Newham Hall, Newby,Middlesbrough, ClevelandTS8 9EATelephone: Middlesbrough 37216BOOKSM 0 N OG RAPH SREPRINTSorders for all publications ofthe Society (except journals)should be sent direct or througha bookseller to-THE SOCIETY FORANALYTICAL CHEMISTRYBook Department9/10 Savile Row,London, WIX IANovember, 19741 THE ANALYST xiiiOfficial,Standardised andRecommendedMethods of AnalysisSECOND EDITION (1 973)aCompiled and Edited forTHE ANALYTICAL METHODS COMMITTEEofThe Society for Analytical ChemistrybyN. W. Hanson, BSc, PhD, FRICThe first part of this book (500 pages) gives in full workingdetail all the methods developed under the aegis of the Society'sAnalytical Methods Committee in the past 40 years.The second part (358 pages) comprises a newly revisedbibliography of standard and recommended methods publishedby official bodies throughout the world.There is an extensive index (37 pages).Published: August 1974Pp. xxiv + 897 f 17.00 ; U.S. $42.50ISBN 0 85990 704 XObtainable from The Society for Analytical Chemistry,Book Department, 9-1 0 Savile Row, London W1 X 1 AFMembers of the Chemical Society are entitled to buy one copy fortheir own personal use at the special price of f14.50 (U.S. $36.50)

ISSN:0003-2654    

DOI:10.1039/AN97499BP129

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

NOVEMBER, 1974 THE ANALYST Editorial Vol. 99, No. 11 84 The Future of The Analyst MANY readers of The Analyst will be aware of moves in recent years towards amalgamation of those learned Societies in the United Kingdom whose prime interest is in Chemistry. Three such bodies, The Chemical Society, The Royal Institute of Chemistry and the Faraday Society, joined together in January 1972 to form a new enlarged Chemical Society (CS), which has continued to carry out the Scientific functions of the three former Societies. The Society for Analytical Chemistry (SAC) also entered into an association with the new CS in 1972 but on a provisional basis, with the option of either making the amalgamation permanent as from 1975, or severing the temporary association and continuing in its former way as a separate, independent Society.During this period of trial amalgamation, the SAC has in part functioned as the Analytical Division of the new CS, but it has also kept its identity as SAC and has maintained full control of SAC publications, including The Analyst. At a recent Extraordinary General Meeting, members of the SAC voted, by the constitu- tionally required majority, in favour of full amalgamation with the CS as from January 1975. This means that the SAC will now be wound up after 100 years of existence, and that the publication of its journals will become the responsibility of the CS, of which the SAC will now become permanently a part. It should be emphasised that the amalgamation agreement ensures the continued pub- lications of the SAC journals, including The Analyst.Moreover, the SAC, in its new analytical r81e within the CS, will continue to be closely associated with its former journals when published by CS. The purpose of this Editorial is to assure readers of the continuation of The Analyst, and also to outline various policy decisions as to its future recently taken by The Analyst Executive Committee. In broad terms, the future policy of The Analyst remains unchanged, namely to publish original and review papers of high standard on all aspects of the analytical sciences. In more detail, however, the aspirations of authors and the requirements of readers must be taken into account. One aspect of a primary journal of concern to both authors and readers is speed of publication, a matter partly under the control of authors themselves.It is sadly the case that some authors take an excessive time to deal with queries and suggestions made by referees. The Analyst has recently adopted a policy of more vigorous pursuit of dilatory authors, and in particular of rejecting papers which have been awaiting the authors’ attention for the still generous period of 1 year. This step has helped to reduce the risk of publishing papers at a time when they may have become out of date, and has appreciably improved the average time interval between submission of a manuscript and its publication in The AutaZyst. The median time to publication is now 32 weeks, a figure which compares favourably with other primary journals in analytical and chemical fields. Many papers are published more rapidly; since January 1973, some 25 per cent.of papers in The Analyst have appeared in 26 weeks or less from the date of receipt. With a dual refereeing system, which we deem essential as a means of maintaining a high standard of papers, and with current processes for production of the journal, the present speed of publication is about the minimum reasonably attainable. In this context, it is perhaps appropriate to remind potential authors that for some years The Analyst has offered facilities for rapid publication of communications, where the nature or significance of the subject matter warrants publication with minimal delay. These communications are intended for brief descriptions of work that has progressed to a stage at which it is likely to be of immediate value to workers faced with similar problems, and should not be simple claims for priority.Such communications are not subjected to the usual formal refereeing and publication within 6 to 8 weeks can normally be guaranteed. Intending authors in countries other than the United Kingdom are also reminded that The Analyst has Regional Advisory Editors located in ten different countries, and who are 697698 EDITORIAL listed on the inside cover page of The Analyst. Their function is to provide a channel of communication to The Analyst from their particular geographical locality, and to advise potential authors in that locality with regard to submission of manuscripts to the journal. A further factor of concern to readers and authors is the coverage of the journal, that is to say the extent to which the broad scope of analytical topics is reflected in the subject matter of the papers published.This is an ever-present problem, as analytical knowledge and techniques continue to broaden and break into new areas, a problem which The Analyst has sought to tackle in part by commissioning review articles on new or rapidly developing analytical topics. In the belief that such review papers are a valuable feature of the journal, it is proposed to commission and publish such papers at higher frequency than in the recent past. Another aspect of coverage by The Analyst that has recently been under review is the proportion of original papers of a theoretical or fundamental nature. There appears to be in some circles a mistaken impression that coverage by The Analyst is biased towards papers of a directly practical nature.Whilst we intend to continue our policy of publishing papers of direct interest to the practising analyst, papers concerned with the fundamental or theoreti- cal principles of analytical methods or techniques have also long been welcomed. Only with such contributions can our understanding of analytical systems be improved and new analytical approaches or techniques developed. It is our belief that an international analytical journal such as The Analyst has a positive duty to attract and publish such papers (consider, for example, the subsequent analytical significance of Schwarzenbach’s original papers on the metal-complexing properties of aminocarboxylic acids), and we would welcome a greater inflow of such papers to The Analyst. Finally, looking more deeply into the future, one must recognise that developments are likely to occur in methods of transmitting information. Even if, for many years to come, our medium of communication continues to be the printed page, there are suggestions that primary journals may move in the direction of printing only summaries of papers, with the full texts retrievable on request. While a careful watch on such developments will be main- tained, one must equally recognise that the analyst is one for whom the minutiae of procedures and equipment can be vitally important, and we envisage the need to continue to publish papers in full in the traditional manner for the immediately foreseeable future. We hope that these comments will assure readers and authors of the continuity of The Analyst as an international analytical journal, and of our endeavour to maintain, and if possible improve, the standards of service currently offered by the journal. H. J. CLULEY Chairman, Analyst Executive Committee

ISSN:0003-2654    

DOI:10.1039/AN9749900697

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, pp. 699-708 699 The Use of Precipitate Based Silicone Rubber Ion-selective Electrodes and Silicone Rubber Based Graphite Voltammetric Electrodes in Continuous Analysis A Review* By 2s. FEHgRt. G. NAGYT, K. TOTH AND E. PUNGOR (Institute foy General and Analytical Chemistry, Technical University, Budapest, Hungary) SUMMARY OF CONTENTS Introduction Behaviour of the electrodes in flowing systems pX electrodes pE electrodes Reference electrodes Measuring cells and measuring methods Applications in continuous monitoring Continuous monitoring of the chloride content of natural water Determination of cyanide content of industrial sewage effluents Determination of dissolution rate of drugs from pharmaceutical Application of the silicone rubber based graphite electrode to in vivo Chromatovoltammetric detector cell containing silicone rubber based preparations measurements electrode Application of the injection technique Conclusion INTRODUCTION THE automation of analytical processes is an important development in analytical chemistry and can be carried out either in stages or continuously.The former technique is generally used for titrations or with direct signal transformation after a selective reaction, and can be carried out either in stationary or streaming solutions. Continuous analysis is mainly performed in streaming solutions. Spectrophotometric methods have been the most commonly used concentration signal transformer techniques, but in the last decade electroanalytical methods have also gained importance, the pioneer work having been carried out by Novak,l Jordan, Javick and Ranz,t Blaedel and ~o-workers,~-~ Lighte and Fleet and Ho.' Many electroanalytical techniques and sensors are excellent for continuous analysis.In this paper the application in streaming solutions of precipitate based pX electrodes and a new sensor, the silicone rubber based graphite voltammetric electrode, are discussed. First, however, it is appropriate to review the essential features of electrodes, measuring cells and measuring methods. BEHAVIOUR OF THE ELECTRODES IN FLOWING SYSTEMS p x ELECTRODES- The precipitate based pX electrodes are excellent sensors for individual ion activities. They can be used favourably for signal transformation in continuous analysis also according to the following equation*: * Reprints of this paper will be available shortly.t Formerly at EGYT Pharmacochemical Works, Budapest, Hungary. Q SAC and the authors. For details see summaries in advertisement pages.700 FEHBR et aZ.: USE OF SILICONE RUBBER ION-SELECTIVE AND [Andyst, Vol. 99 where E is the electrode potential; Ed is the normal potential of the electrode; ai and ak are the activities of the primary and interfering ions; Kik is the selectivity constant (Ki,i is unity) ; k is the number of ions taking part in the electrode response; and z, is the valency of the primary ion. The selectivity constants of precipitate based pX electrodes can be calculated theoretically and measured potentiometrically.* From equation (1) it follows that if the value of Kik is small, then the effect of the presence of ion k on the electrode response is negligible.From the small Kik values shown in Table I, it is apparent that the precipitate based pX electrodes have excellent selectivity to the primary ion i in the presence of many other ions. Table I also shows that the cal- culated and the potentiometrically measured selectivity constants of pX electrodes are comparable. The potentials of the solid-state electrodes are generally not influenced by the flow-rate of the solutions when a simple flow-through tube cell and a 10-1 M concentration of the ionic strength adjusting agent are used, which may be explained by the rigid electrode construction and the fact that the electrode response is a result of a fast surface ion-exchange equilibrium reaction.However, the interpretation of the behaviour of this type of electrode near the detection limit is open to doubt. The dynamic response of precipitate based pX electrodes is of importance in continuous analysis. In this respect the response times of various precipitate based pX electrodes differ, but are always within a few tens or a few hundreds of milliseconds if the concentration changes around the electrodes are carried out relatively r a ~ i d l y . ~ It was also found that the response time is dependent to a great extent on the direction of the concentration change introduced at the surface of the pX electrode and that it does not depend on the thickness of the membrane layer (0-9 to 2.0 mm). As examples, the dynamic behaviour of precipitate based copper and iodide-selective electrodes is shown (Figs.1 and 2): On the basis of the experimental results obtained it was concluded that: different factors determine the potential veysus time curves of pX electrodes with increase and decrease in concentration; the response time is determined by electrode surface reactions, for example, desolvation of the ion in the boundary phase of the electrode, or the rate of ion-exchange reactions; the response time of pX electrodes allows their application in the monitoring of ion activities and in following certain chemical reactions. pE ELECTRODES- The silicone rubber based graphite electrodes can be used as redox electrodes or as voltammetric indicator electrodes ; with the former, the electrode potential is almost independ- ent of flow-rate.However, if used as voltammetric sensors, the voltammetric signal will become flow-rate dependent, because the signal is a result of transport processes. In stream- ing solutions, the electrode reaction, i e . , the signal, is controlled by convective diffusion, the sum of the convection and molecular diffusion, which has been theoretically investigated by Russian and Japanese worker^.^^-^* Correlations have been given for the relationships between the voltammetric current and the flow-rate, construction of the measuring cell and the shape and size of the electrodes. The appropriate equations valid for the main types of electrodes are shown in Table 11. REFERENCE ELECTRODES- Because problems may arise with streaming solutions, the appropriate selection of a reference electrode is generally very important for both potentiometric and voltammetric * For the response time study shown in Fig.1, about 300 ms were required to achieve a constant potential after switching the solutions, and in order to observe the first part of the curve in more detail the process was followed only for 100 ms.TABLE I 20 ms \ * 1 o - ~ Ion c1- Br- SCN- OH- Cr0,'- co3a- [ FeCN,] 4- I- SELECTIVITY OF THE HALIDE AND SULPHIDE ION-SENSITIVE ELECTRODES Chloride Bromide -7 -7 Kik calculated Kik measured I(ik calculated Kih measured 1 2.0 x 10-3 6.0 x 10-3 1 1-5 x loo 5.2 x 10-6 4.5 x 10-6 2.5 x 10-7 1.1 x 10-7 6.3 x 10-6 4-6 x 10-6 3.1 x 10-7 1.0 x 10-7 1.3 x 10-4 0.5 x 10-4 6.3 x 10-7 3.1 x 10-7 3.3 x 10-4 2.0 x 10-4 1.6 x 10-6 1.2 x 10-6 Iodide -' Kik calculated Kik measured 9.6 x 10-7 3.7 x 10-7 2.0 x 10-4 1.8 x 10-4 3.0 x 10-4 2.4 x 1 0 - 4 1.0 x 10-8 0.9 x 10-8 5-0 x 10-1' 6.6 X lo-" 1.2 x 10-10 0.2 x 10-10 3.2 x 2.6 x 2.4 x 3.5 X 1 Sulphide Kik measured I<ik calculated* c 7 6-3 x 10-15 3.2 to 8.0 x 10-15 8.0 x 10-13 7.4 to 13.8 x 10-13 6.3 x 10-17 1.9 to 10.0 x 10-17 2.5 x 10-1s 8.1 to 31.0 x 8.0 x 10-17 3.5 to 35.5 x 10-10 1.6 x 2-5 to 14.1 x 6.3 x lo-'' 1.0 x 10-22 * Ranges of calculated selectivity constants are given because different solubility values are recorded in the literature for the silver sulphide precipitate.Time - Fig. 1. Dynamic behaviour of a copper-selective electrode. AC = 10-3 to 10-a M J Time - Fig. 2. Dynamic behaviour of an iodide-selective electrode.AC = t o 1 0 - 2 ~ Y M t+ M 0 c3 w 0 U M cn z 0 0 $ c 0 d v,702 [Awdyst, Vol. 99 measurements. For example, in voltammetry the shift of the potential of the reference electrode causes an alteration in the potential of the indicator electrode, while in potentiometry the liquid junction potential or the change in this potential introduces errors. These errors can be avoided if the internal filling solution of the reference electrode streams into the sample at a low flow-rate (this technique is used mainly in potentiometry) and if the liquid phase of the reference electrode flows but with the reference cell separated from the ample.^ With these modifications contamination of the reference electrode is eliminated and a well defined solution interface is established.FEHER et al.: USE OF SILICONE RUBBER ION-SELECTIVE AND TABLE I1 LIMITING CURRENT EQUATIONS FOR VARIOUS TYPES OF VOLTAMMETRIC ELECTRODES Electrode shape Equation Rotating disc Planar Tubular Conical Disc Spherical iL = limiting current; n = number of electrons taking part in the electro- F = Faraday number; A = geometrical surface of the indicator electrode; co = concentration of the electroactive D = diffusion coefficient of the electroactive component; Y = kinematic viscosity of the solution; u = flow-rate; w = rotation speed; b, h, R, x , I. and a are the characteristics of the electrodes. iL = knFAc, DB v - i d iL = knFbhsc, DB v-t d iL = knFR; x b o D3 us ZL = knFAL-ic, D3 v-B uj iL = knFALic, DS v-i ut iL = knFab, D) u) chemical reaction; component in the streaming solution; The most commonly used reference electrodes are the electrodes of the second kind (calomel, silver - silver chloride, etc.).Electrodes of the first kind and redox electrodes, and, in potentiometry, other pX electrodes, have also been successfully employed. By the appro- priate selection of the reference electrode, cells with or without transference are used for the measurements. If the latter are used, then the concentration of the ion to which the reference electrode is reversible should be kept constant. MEASURING CELLS AND MEASURING METHODS Different types of detector cells have been used in continuous analysis.* There are several main requirements for measuring cells that are to be used in continuous analysis. First, the hold-up time and the volume should be kept to a minimum.A very important advantage is gained if the handling of the cell is free from complications, that is, if the electrodes can be replaced during operation and the cell readily cleaned. All the detector units should be arranged such that the whole is mechanically stable and of a size that is appropriate to the volume of the sample solution. In most instances it would be useful if the cell were made of a transparent material, thus enabling measurements to be followed visually. Above all, the voltammetric kind of measuring cell must also have a low electric resistance. In our Institute the measurements in streaming solutions are mostly carried out by using simple flow-through tube cells. The methods that we have used for continuous analysis are as follows.1. Monitoring the concentration level continuously in a flow-through cell. 2. The injection technique (discussed in detail later), which can be divided into the * Obtainable from Corning Inc., Medfield, Mass., U.S.A. (capillary sodium electrode, Model NAS- 1 1-18) and Foxboro Ltd., Redhill, Surrey.November, 19741 GRAPHITE VOLTAMMETRIC ELECTRODES IN CONTINUOUS ANALYSIS 703 following groups : the electroactive component (sample solution) is injected into the streaming blank solution ; the non-electroactive sample solution is injected into the streaming electroactive blank solution ; and the electroactive reagent is injected into the streaming solution to be analysed. In continuous monitoring, the signal due to the concentration must be evaluated either by calibration with standards, or by using the standard addition technique.15 The measuring system should also be checked frequently because of the possible drift of the electrode function and the change of slope of the electrode calibration graph.However, the injection techniques offer some advantages over continuous monitoring in that : (i) the zero level is controlled continuously, provided that the blank solution does not (ii) the regular alternate injection of samples and standards permits control of the slope (iii) calibration and the measurement can be carried out under the same conditions. react with the sample; of the calibration graph ; APPLICATIONS I N CONTINUOUS MONITORING Examples are given that demonstrate the possible applications of ion-selective and voltammetric electrodes.CONTINUOUS MONITORING OF THE CHLORIDE CONTENT OF NATURAL WATER- A flow-through tube cell with a pC1 indicator electrode and a double junction S.C.E. were used, the latter in order to avoid contamination of the sample with chloride. Sufficiently good results were obtained if only the indicator electrode was incorporated into the flow- through cell and the reference electrode placed into a vessel containing the outflow of sample solution. The calibration of the electrode was carried out with standard solutions that had almost the same composition as the sample. The standard solutions with higher chloride con- centrations were prepared from a sample solution with the lowest possible chloride content, which was determined coulometrically, by adding to it appropriate amounts of a concentrated chloride solution. Frequent re-calibration using natural water samples showed the average deviation of the measured values to be less than 5 per cent.of the appropriate chloride concentration. DETERMINATION OF CYANIDE CONTENT OF INDUSTRIAL SEWAGE EFFLUENTS- As reported earlier,16 the silicone rubber based iodide-selective membrane electrode is suitable for the determination of cyanide activity. By using a flow-through detector cell in which the indicator electrode was an iodide-selective electrode, the monitoring of the cyanide content of industrial sewage has been achieved. As the electrodes respond only to the dissociated cyanide ion, the pH of the sample was continuously checked and adjusted to 11 before it entered the electrode area.The evaluation of results was accomplished by using a calibration graph and with the continuous monitoring system the cyanide concentration of a sewage could be determined in a range as low as to M. DETERMINATION OF DISSOLUTION RATE OF DRUGS FROM PHARMACEUTICAL PREPARATIONS- The dissolution rate of drugs from solid pharmaceutical preparations (tablets and dragkes) is a most important parameter because the dissolution process plays an important r61e in drug absorption, which is especially true if the dissolution is slow, and the process will be rate determining in the consecutive reactions in the living organism. The dissolution process can be followed in some instances by employing pX electrodes, but a voltammetric sensor has a wider scope. The former measure the change in the inorganic content of the drug dissolved from the preparation, while the voltammetric sensor measures the organic compound to which the pharmacological effect is usually attributed.The silicone rubber based graphite electrode is an appropriate voltammetric sensor and its use is favoured because its anodic oxidation range is extensive and permits the voltam- metric determination of many organic compounds.17 This electrode has almost no memory704 [AnaZyst, Vol. 99 effect and repeated or continuous measurements can therefore be carried out without renewing the electrode surf ace. The recirculating dissolution measuring arrangement with a silicone rubber based graphite electrode and silver - silver chloride reference electrode measuring ce1P is shown in Fig.3. FEHBR et aZ.: USE OF SILICONE RUBBER ION-SELECTIVE AND 1 Fig. 3. Measuring set-up for investigation of the dissolution of drugs from pharmaceutical preparations. T, thermometer; K, stirrer; P, peristaltic pump; El, silicone rubber based graphite electrode; and E,, silver - silver chloride reference electrode A peristaltic pump continuously samples the dissolution system in the thermostatically controlled dissolution vessel during the measurement. A suitable constant potential is applied between the two electrodes and the voltammetric current veysus time curve recorded con- tinuously with the aid of a polarograph. First, a base-line is recorded, then the pharma- ceutical product is introduced into the vessel and, as the dissolution proceeds, the value of the voltammetric current corresponds to the momentary concentration of the drug in the solution.As a simde linear correlation has been found between the voltammetric current and the concenwariun ui L I I ~ wxuuacuve curriyuuiiu 111 LIIC ~ L I C U I U ~ ~ u ~ u L ~ u ~ ~ , LUG LUI vc 1G:LuIuCU can be taken as an integrated dissolution curve (Fig. 4). 1 / 1 min I Time - t o Fig. 4. Dissolution curve of an amidopyrine-contain- ing (Germicid) tablet (1) and a promethazine-containing (Pipolphen) dragde (2). Solution used for dissolu- tion: 10-1 M in NaCl and M in HC1. t = 37 "C; Ul = +0*8 V; U, = -+ 0.7 VNovember, 19741 GRAPHITE VOLTAMMETRIC ELECTRODES IN CONTINUOUS ANALYSIS 705 In curve (2) it can clearly be seen that the current is a function of the square root of t, while in curve (1) an almost linear relationship has been found between the two parameters, thus indicating that the promethazine is contained in a non-disintegrating type of dragde, while the amidopyrine is in a disintegrating type.APPLICATION OF THE SILICONE RUBBER BASED GRAPHITE ELECTRODE TO in vivo The silicone rubber based graphite electrode has been used continuously for several hours in flowing systems without renewal of the electrode surface, thus suggesting its use for in vivo measurements in which the flow-rate of the blood circulation and the distribution of drugs (in living organisms) could be studied. The effectiveness of this measuring technique was proved by experiments on drugged test animals (Figs.5 and 6). MEASUREMENTS"- ti 1 ti Time - ti Fig. 5. Current - time curves recorded in artery after the administration of amidopyrine. Amounts of drug injected 0.4, 0-S and 1-2 mg kg-l. U = +O-8 V; ti = time of injection In Figs. 5 and 6 the change in the current intensity with time can be seen as an effect of ascorbic acid or amidopyrine when injected into the rear limb of the narcotised cat test animal. The measuring cell was incorporated into the artery and the vein, respectively. t u C 2? 0 3 I I 2 min * Time -+ ti Fig. 6. Current - time curve recorded in vein after the administration of amidopyrine. Amount of drug injected 15 mg kg-l. U = +0.8 V; ti = time of injection The current recorded was found to vary about a constant average value, which corre- sponded to the actual drug level in the blood. This variation is due to changes in the pulse rate of the blood according to the heart beats.Various sympathomimetic drugs administered intravenously increased the amplitude of the variations, for example, to about three to five-706 [Analyst, Vol. 99 fold by 1 pg kg-1 of noradrenaline (Fig. 7 ) ; a technique for the detection of these drugs is thus offered. FEHBR et al.: USE OF SILICONE RUBBER ION-SELECTIVE AND Time - U = $0.9 V ti Fig. 7. Effect of noradrenaline on the voltammetric current recorded in a vein. Amount of drug injected 1 pg kg-l. CHROMATOVOLTAMMETRIC DETECTOR CELL CONTAINING SILICONE RUBBER BASED GRAPHITE ELECTRODE~O- A voltammetric cell of a special kind was fabricated in which a turbulent flow was achieved by pumping the solution to be measured with a powerful pump through a tube with a narrow jet against the sensing surface of the silicone rubber based graphite indicator electrode.The enhanced convection effect obtained resulted in a higher sensitivity of the voltammetric measurements. This cell was used as a chromatovoltammetric detector operating at a constant potential, and when connected to a suitable chromatographic column the separation and detection of various purine derivatives (adenine, guanine, xanthine, and hypoxanthine) was accomplished ; the lower limit of detection for these compounds was found to be 2 x 10-10 mol. The sensi- tivity of the detector was about the same as that of ultraviolet detectors, but it has the advantage of selective detection of purines in the presence of pyrimidines.APPLICATION OF THE INJECTION TECHNIQUE- As has been pointed out earlier, the principle of this method21 is that a small volume of the test solution is injected into the blank solution that is streaming at a constant flow-rate (Fig. 8). After mixing, the test solution enters the flow-through cell containing the indicator and reference electrodes. If a voltammetric electrode is used, then the change in the current intensity is recorded at a constant voltage corresponding to the limiting current range of the component to be measured. A peak-type signal is obtained. M I P Fig. S. Measuring set-up for continuous voltammetric analysis of various compounds using the injection technique.C, container; P, peristaltic pump; J, injection unit; T, thermostat; R, stirrer; D, detector cell; M, measuring instrument; and IP, integrator unit. N, north; and S, southNovember, 19741 GRAPHITE VOLTAMMETRIC ELECTRODES IN CONTINUOUS ANALYSIS 707 The amount of electroactive material injected can be determined by integrating the current or potential with respect to time, which corresponds to the area under the peak. The linear relationship between the area under the peak and the amount of the electro- active material injected at constant flow-rate was confirmed experimentally. With this direct technique it is possible to determine many electroactive materials, some examples of which, with appropriate potentials and supporting electrolytes, are given in Table 111.TABLE I11 DETERMINATION OF SOME PHARMACEUTICALS BY THE VOLTAMMETRIC TECHNIQUE Material K,CFe(CN) a1 H ydroquinone Variamine blue Ascorbic acid Isoprenaline Chlorpromazine Diethazine Amidop yrine Morphine Ethylmorphine Papaverine Supporting electrolyte 10-1 M KC1 10-1 M KC1 10-1 M KC1; M HC1 10-1 M KCl; M HC1 10-1 M KC1 10-1 M KC1; lo-, M HC1 10-l M KC1; lo-, M HC1 10-1 M KC1 4 N H,SO, 4 N H,SO, 4 N H,SO, Applied potential/V + 0.6 +0.7 + 0.8 + 0.8 + 0.7 + 0.7 + 1.0 + 1.0 + 1.0 + 1.0 Conversely, by use of a suitable electroactive reagent, electroinactive samples can also be analysed by the injection technique. Such determinations can be carried out in two ways, depending on the concentration level of the electroinactive sample. In the lower concen- tration ranges, it is an advantage for the electroinactive sample to be flowing continuously, but in the higher concentration ranges the measurement is more convenient if the electro- inactive sample is injected.22 CONCLUSION This paper shows that the development of ion-selective and voltammetric sensors has opened further possibilities in continuous analysis, which requires rigid electrode construction, stable and reproducible signals and suitable calibration.The injection technique has special advantages over direct continuous monitoring because it eliminates the problems of the zero signal level and simplifies calibration. The Aow-rate is an important parameter in voltammetric measurements as the current corresponds to the mass transport and if kept constant good results can be obtained.It can be expected that development of a wider range of reliable ion-selective and voltam- metric sensors will extend the scope of such sensors in continuous analysis. This paper is dedicated to Professor Dr. DDr. h. c. M. K. Zacherl in honour of his 70th birthday. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. REFERENCES Novak, J. V. A., “Progress in Polarography,” Volume 2, Interscience Publishers, New York, Jordan, J., Javick, R. A., and Ranz, E. W., J . Amer. Chem. SOC., 1958, 80, 3846. Rlaedel, W. J., and Strohl, J. H., Analyt. Chem., 1961, 33, 1631. Blaedel, W. J., Olson, C. L., and Sharma, L. R., Ibid., 1963, 35, 2100. Blaedel, W. J., and Boyer, S. L., Ibid., 1971, 43, 1538. Light, T. S., in Durst, R. A., Editor, “Ion Selective Electrodes,” National Bureau of Standards Special Publication No. 3 14, Washington, 1969,, Fleet, B., and Ho, A. Y . W., in Pungor, E., Editor, Ion Selective Electrodes,” Akad6miai Kiadb, Pungor, E., and T6th, K., Analyst, 1970, 95, 625. T6th, K., GavallCr, I., and Pungor, E., Analytica Chim. Acta, 1971, 57, 131. Levich, V. G., “Physicochemical Hydrodynamics,” Prentice-Hall, Inc., Englewood Cliffs, New Matsuda, H., J . Electroanalyt. Chern., 1968, 16, 153. -, Ibid., 1969, 21, 433. 1962. Budapest, 1972, p. 17. Jersey, 1962.708 F E H ~ R , NAGY, T ~ T H AND PUNGOR 13. 14. - - , Ibid., 1971, 30, 261. 15. 16. 17. 18. 19. 20. 21. 22. Matsuda, H., and Yamada, J., Ibid., 1971, 30, 271. Branh, M. J. D., and Rechnitz, G. A. Analyt. Chem., 1970, 42, 1172. T6th, K., and Pungor, E., Analytica Chim. Acta, 1970, 51, 221. Pungor, E., and Szepesvh-y, G., Ibid., 1968, 43, 289. Pungor, E., FehCr, Zs., and Nagy, G., Ibid., 1970, 51, 417. Fehhr, Zs., Nagy, G., and Pungor, E., Hung. Sci. Instrum., 1973, 26, 15. Vhradi, M., FehCr, Zs., and Pungor, E., J . Chromat., 1974, 90, 259. Nagy, G., Fehdr, Zs., and Pungor, E., Analylica Chim. Acta, 1970, 52, 47. FehCr, Zs., and Pungor, E., Analytica Chim. Acta, in the press. Received February 28th, 1974 Accepted June 5th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900699

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, $9. 709-716 709 A Simple Modification to a Flame-ionisation Detector Linear Amplifier for Logarithmic Display of Gas Chromatograms and its Use By D. A. COLLINS (West Midland Forensic Science Laboratory, Priory House, Gooch Street North, Birmingham, B6 6QQ) With the addition of a single diode, a linear electrometer amplifier for gas chromatography with flame-ionisation detection has been modified so as to enable it to respond to ionisation currents, over the range 10-la to 10-8 A, in a manner closely approaching logarithmic proportionality. Some chromato- grams obtained with this amplifier are included and attention is drawn to the advantages of logarithmic display. MUCH general qualitative and comparative work carried out by use of gas chromatography is performed on instruments fitted with flame-ionisation detectors. Usually, the detector current is amplified by a linear electrometer amplifier and is displayed on a potentiometric recorder.Complex mixtures are often dealt with either as actual samples or as the result of pyrolysis in pyrolysis - gas chromatography. The linearity and large dynamic range (seven decades or more) of the flame-ionisation detector are well known. It therefore responds to major, minor and trace components of complex mixtures in a linearly proportional manner. However, the use of linear amplification can result in the loss of pertinent information, because only two decades may be displayed, assuming that 1 per cent. of full-scale deflection is taken as the minimum detectable signal.Under these conditions, strong signals from major components go off scale or small signals from trace components fail to appear, depending on the sensitivity employed. Changing the sensitivity manually or automatically during the course of an analysis may alleviate the problem, but can lead to a rather confused display. Alternatively, a series of analyses can be made at different sensitivities, but this procedure consumes both extra time and sample. A useful approach is to record the chromatogram on a two-pen recorder, the two pens being connected in parallel to the amplifier output and having different sensitivities. However, even with a sensitivity ratio of l O : l , the dynamic range is extended to only three decades and the double chromatogram can assume a rather confused appearance.An entirely different solution to the problem is to use logarithmic amplification, with which the amplifier voltage output is proportional to the logarithm of the detector current. This method has received relatively little attention, perhaps because commercially built logarithmic amplifiers for gas chromatography are rarely encountered. The advantage of logarithmic display is that major, minor and trace components of complex mixtures can all be displayed in a single chromatogram, thus conserving both time and sample. In work on the analysis of fire accelerants in fire debris, it has been claimed that logarithmic presentation of the flame-ionisation detector current is superior to linear presentation.1 It has also been shown that quantitative work may readily be undertaken while using logarithmic amplifi- cation.Several circuits for logarithmic amplifiers intended to be used with flame-ionisation detectors have been described. That of Chen3 is purely logarithmic; those of Dewar and Maier4 and Gabriel and Morris5 include a logarithmic range as well as linear and integral ranges. The principles upon which these amplifiers operate are described in the respective papers, and some additional information on the Gabriel and Morris amplifier, especially with regard to the logarithmic range, appears elsewhere.6 The author has found that by analogy with the Gabriel and Morris amplifier, a simple modification can be made to a commercial Linear electrometer amplifier that gives the amplifier a useful, approximately logarithmic, range.@ SAC : Crown Copyright Reserved.710 COLLINS : MODIFICATION TO A FLAME-IONISATION DETECTOR LINEAR [Anb!&St, VOl. 99 EXPERIMENTAL A I’ PARATUS- Pye 104 gas chromatography equipment [W. G. Pye & Co. Ltd. (now Pye-Unicam Ltd.), Cambridge] was used with flame-ionisation detection ; temperature programming and Curie point pyrolysis facilities were available. The amplifier, which was originally part of a Model 4 chromatograph, was of the older type with the “zero/back off range 2/back off range 1” switch mounted on the rear panel. Its original battery-powered supply system for the detector had earlier been replaced by a mains unit from Pye. Potentiometric recorders (Leeds and North- rop, Honeywell, and Smiths Industries) were variously used.MODIFICATION OF THE AMPLIFIER- A Mullard BAV45 diode (Sasco Ltd., Crawley, Sussex) was soldered into place in parallel with the 9O-GQ resistor and the 10-pF capacitor designated R, and C,, respectively, in the circuit diagram of the amplifier.’ These components are the range setting resistor and the time constant determining capacitor for the linear attenuation settings x 1 to x 50, inclusive. I From backing off I v1 A I current source Attenuator switches I R25 w 500 200 100 50 20 10 5 2 1 R2 Fig. 1. Input circuitry of the modified amplifier, showing the logarithmic diode DL in parallel with the feedback resistor R, and capacitor C,. Diagram adapted from that given in the Pye 104 Manual (Fig. 104-9/3). R,, 9 GQ; R,, 10MR; R,, 90MQ; R,, 90 GO; R,, 900 MQ; and R,,, 2.2 MQ. C,, 10pF; C,, 33 pF; C,, 330 pF; C,,, 1 nF; and C,,, 10 nF The orientation of the diode with its anode towards the amplifier input was necessary in order to permit the required forward conduction to occur, with the detector jet having positive polarity.Part of the modified input circuitry of the amplifier is shown in Fig. 1.November, 19741 MATERIALS- AMPLIFIER FOR LOGARITHMIC DISPLAY OF GAS CHROMATOGRAMS 71 1 n-Pentane-Laboratory-reagent grade. n-Heptalze-Laboratory-reagent grade. n-Alkane reference mixture (C, to Cs> for gas - liquid chromatography. n-Alkane reference mixture (C, to CI2) for gas - liquid chromatography. Aviation spirit-Supplier not known. Polyethylene (ICI Alkathene). Three solutions containing 1.0, 0.01 and 0.0001 per cent.V/V of n-pentane in n-heptane A mixture of approximately equal volumes of the C, to C, and C, to C , , were prepared. n-alkanes was also prepared. GAS CHROMATOGRAPHY- The following columns were used: (i) glass, 1.5 m x 6 mm, packed with 80 to 120-mesh Celite coated with 10 per cent. SE30 silicone gum; (ii) glass, 1.5 m x 3 mm, packed with 60 to 80-mesh hydrogenated carbon black coated with approximately 1 per cent. OV-101 silicone oil (this type of column is a recent development8 and was made available by Dr. J. B. F. Lloyd); and (iii) stainless steel open tubular, 47 rn x 0.25 mm, coated with OV-101 silicone oil. A stream splitter was fitted to the injection port and a detector flushing facility was also used with this column. Details of the various chromatographic conditions are given in the figure captions.PROCEDURE- When the chromatographic conditions had been established, the recorder input was short-circuited, the pen was set on to the zero line of the chart-paper and the recorder was then connected to the amplifier. Zero setting of the amplifier and backing off of the detector standing current were accomplished in the normal manner.9 RESULTS AND DISCUSSION This provided a test of the logarithmic characteristic of the modified amplifier and verified that an acceptably logarithmic response was being obtained that was adequate, at least for the qualitative work for which it was intended. Estimates of the detector current represented by the extent of the deflection of the recorder pen were made in the following way.For any linear attenuation setting, Y , the ionisation current, I , at full-scale deflection is given bylo therefore a peak of height h per cent. of full scale at this attenuation represents a maximum current, Imax., where In Fig. 2 the response of the system to a series of injections of n-pentane is shown. I = Y (measured in amperes) h 100 I-=. = - x 10-12 Y A series of n-pentane injections at suitable linear attenuations showed that under the con- ditions used for Fig. 2, the following relationship was observed: 10-zA is the maximum current from an injection of 2.4 x 105-2 pl of n-pentane. A current scale, calibrated on this basis, has been added to Fig. 2. The attenuation settings available on the unmodified amplifier provide for full-scale deflection at currents of This modification t o the amplifier permits the whole of this range to be displayed logarithmically across the chart width; in fact, more than six decades can be displayed.The incorporation of this logarithmic modification into the amplifier does make the linear attenuation settings x 1 to x 50, inclusive, unobtainable, but it is the author's ex- perience that these settings are used rarely, if at all. Switching over the attenuation settings >( 1 to x 50 is effected at the amplifier output, and advantage can be taken of this to adjust the linear size of the basic logarithmic display in the modified amplifier. A (minimum) to 5 x lo-' A (maximum).712 COLLINS : MODIFICATION TO A FLAME-IONISATION DETECTOR LINEAR [Analyst, Vol.99 100 I 1 Volume of n-pentane injected/pl Fig. 2. Graph of n-pentane peak height v e w m volume of n-pentane injected, demonstrating the approximately logarithmic characteristic of the modified amplifier. Attenuator setting x 10. Chromatograph, Pye 104; column, 10 per cent. SE30 silicone gum on 80 to 120-mesh Celite (1.5 m x 6 mm); temperature, 80 " C ; carrier gas, nitrogen; flow-rate, 50 ml min-l; flame- ionisation detector, hydrogen 50 ml min-l, air 500 ml min-1; 1-pl and 0.1-pl injections of n-pentane and of suitable dilutions of n-pentane with n-heptane SPECIMEN CHROMATOGRAMS- Three specimen logarithmic chromatograms were obtained in order to illustrate the use of the modified amplifier. For comparison, conventional linear chromatograms of the same samples under the same conditions were also oht ained.Chromatograms of a mixture of C, to C,, n-alkanes are shown in Fig. 3 (logarithmic) and Fig. 4 (linear). Nine components are revealed in the linear trace, comprising the eight n-alkanes together with an impurity that occurred between n-heptane and n-octane; there is great disparity between the respective peak heights, such that while the n-octane peak is 90 per cent. of full scale in height, the n-pentane and n-hexane peaks are only just detectable in the chromatogram. In contrast, the logarithmic trace shows far less difference in peak I cfj c5 I 20 15 70 5 0 - Retention time/minutes Fig. 3. Logarithmic chromatogram of mixture of C, to C,, n-alkanes. Chromatograph and conditions as for Fig. 2 except temperature programmed, commencing a t 30 "C and rising at 8 "C min-'; 0.2 p1 of mixture injectedNovember, 19741 AMPLIFIER FOR LOGARITHMIC DISPLAY OF GAS CHROMATOGRAMS 713 heights, which is advantageous for retention indexing purposes, and, because of the much larger dynamic range that can be displayed, it could easily be obtained by an inexperienced operator who was undecided about the volume of mixture to inject or the sensitivity needed.c12 20 15 10 5 I - Retention time/minutes Fig. 4. Linear chromatogram of mixture of C, to C,, n-alkanes. Attenuator setting x lo5. Conditions as for Fig. 3 The impurity occurring between n-heptane and n-octane is prominently shown in the logarith- mic chromatogram and a number of other trace components are revealed; in all, twenty-four components were detected in the mixture.Increasing column bleed, consequent upon temperature programming, is responsible for the rising base-line of the logarithmic display ; 40 30 20 10 0 - Retention time/minutes Fig. 5. Logarithmic chromatogram of polyethylene pyrolysis products. Attenuator setting x 5 . Chromatograph, Pye 104; column, 1 per cent. OV-101 silicone oil on 60 to SO-mesh hydrogenated carbon black (1.5 m x 3 mm) ; temperature programmed, commencing a t 50 "C, rising a t 12 "C min-l t o 300 "C and continuing isothermally therefrom; carrier gas, nitrogen; flow-rate, 20 ml min-l; flame- ionisation detector, hydrogen 20 ml min-l, air 200 ml min-l; 0.5 mg of polyethylene pyrolysed a t 770 "C for 10 s714 COLLINS MODIFICATION TO A FLAME-IONISATION DETECTOR LINEAR [Andyd, VOl.99 it is emphasised by the much increased sensitivity at the lower part of the chromatogram compared with that in the corresponding linear chromatogram. Figs. 5 and 6 show logarithmic and linear chromatograms, respectively, of the products from the pyrolysis of polyethylene. In the linear chromatogram the sensitivity employed was such as to show the trace components clearly and it has resulted in the peaks from the major components going well off scale. However, in the logarithmic chromatogram the trace components are well shown and yet not a single peak in the chromatogram is off scale. - a 40 30 20 10 0 - Retention time/minutes Fig. 6. Linear chromatogram of polyethylene pyrolysis products. Attenuator setting x 600. Conditions as for Fig.5 Similar characteristics are evident in chromatograms of aviation spirit shown in Figs. 7 In this case the logarithmic chromatogram reveals some trace (logarithmic) and 8 (linear). components that cannot be seen in the linear chromatogram. CONCLUSIONS Although the chromatograms given here demonstrate how well trace components can be displayed among major components, this does depend upon column performance, thus emphasising the prime importance of the resolution of the column. The trace components in Figs. 5 and 7 are displayed more clearly than those in Fig. 3 because they were obtained on columns with superior resolution. Because of the much increased sensitivity towards the bottom of a logarithmic display, tailing, peak overlap and column-bleed effects are emphasised compared with linear display, and the illusion can be created of inferior column performance.In order to make the best use of logarithmic display, it would appear to be necessary to use high-resolution columns with low column bleed; no doubt the use of twin-column tech- niques with opposed detector outputs would counter the rising base-line that is evident in temperature-programmed chromatograms. It is therefore suggested that with good columns there is considerable advantage in logarithmically displaying the output of the flame-ionisation detector in work on complex mixtures, in pyrolysis - gas chromatography and for general screening and investigations. Furthermore, these advantages can be obtained with a simple modification to standard equipment, and it seems likely that this type of modification could be made inexpensively to many linear electrometer amplifiers. In particular, the newer models of the Pye-Unicam 104 ionisation amplifier can be modified in a similar way by fitting the BAV45 diode across the 91-GQ feedback resistor designated R15.The author thanks Mullard Ltd. for drawing his attention to the BAV45 diode. Advice on the preparation of this paper from Dr. Lloyd and from Mr. R. &I. Mitchell is gratefully acknowledged.November, 19741 AMPLIFIER FOR LOGARITHMIC DISPLAY OF GAS CHROMATOGRAMS L 715 0 I I I Retention time/minutes - Fig. 7. Logarithmic chromatogram of aviation spirit. Attenuator setting x 5. Chromatograph, Pye 104; column, open tubular coated with OV-101 silicone oil (47 m x 0.25 mm); temperature programmed, commencing a t 30 "C, rising a t 3 "C min-l; carrier gas, nitrogen; flow-rate, 1 ml min-1; flame-ionisation detector, hydrogen 25 ml min-l, air 250 ml min-l. A booster supply of nitrogen was introduced between the column and the detector, flow-rate 25 ml min-l. 0.2 p1 of aviation spirit injected, splitting ratio 0.02 0 I h I I 10 20 30 Retention time/minutes - 40 Fig. 8. Linear chromatogram of aviation spirit. Attenuator setting x 200. Conditions as for Fig. 7716 COLLINS 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Chisum, W. J . , and Elzerman, T. R., J . Fovens. Sci., 1972, 17, 280. Wade, R. L., and Cram, S. P., Analyt. Chenz., 1969, 41, 893. Chen, K. A., Ibid., 1968, 40, 1171. Dewar, R. A., and Maier, V. E., J . Chronzat., 1964, 15, 461. Gabriel, W. P., and Morris, R. A., J . Scient. Instrum., 1966, 43, 104. Olsen, G. H., “Electronics-A General Introduction for the Non-specialist,” First Edition, Pye 104, Model 4, Gas Chromatograph, Operating Manual, Fig. 104-9/3, W. G. Pye & Co. Ltd., DiCorcia, A., and Bruner, F., Analyt. Chem., 1971, 43, 1634. Pye 104, Model 4, Gas Chromatograph, Operating Manual 104-9, W. G. Pye & Co. Ltd., Cambridge, Ibid., p. 2. Butterworths, London, 1968, p. 388. Cambridge. p. 5. Received Apvil llth, 1974 Accepted June 3rd, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900709

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Analyst, November, 1974, Vol. 99, pp. 717-723 717 A Convenient Method for Collecting Gas-chromatographic Effluents for Micro-scale Infrared Spectrophotometric Analysis in the Gas Phase BY K. HARALD NORIN (Arrhenius Laboratory, Department of Analytical Chemistry, Univevsity of Stockholm, S-104 05 Stockholm, Sweden) A method for the collection of fractions of low relative molecular mass issuing from a gas chromatograph followed by identification using infrared spectrophotometry is described. The fractions are transferred to a micro- scale gas cell by a simple cryogenic technique, which requires minimum handling of the separated fractions. Good infrared spectra were obtained for a sample size of 30 to 100 pg. GREAT interest has been shown in the combination gas chromatography - infrared spectro- photometry, as a study of the literature reveals.The first publications dealing with this combination appeared in the late 1950s and early 1960s. Interest was concentrated mostly on relatively high-boiling compounds, which were analysed in liquid micro cell^,^-^ and in potassium bromide tablets or potassium bromide micro-tablet~.~~6-~ Good sensitivity and reproducibility were achieved, primarily with the potassium bromide tablets and potassium bromide micro-tablets. In the middle 1960s, interest was transferred to more rapidly scan- ning instruments in order to avoid the time-consuming and difficult transference of the gas-chromatographic fractions to the infrared spectrophotometer. The aim was to be able to obtain an infrared spectrum of the gas-chromatographic fraction in the carrier-gas stream.Many different types of infrared spectrophotometers were designed. Some instruments had rapid scanning p r o p e r t i e ~ ~ t ~ ~ but often with insufficient wavelength range, while others had moderate scanning speeds11-15 but usually poor resolution. At the same time, a number of effective gas cells for infrared spectrophotometric purposes were developed, as it had been observed that with an effective gas cell, ie., with a high ratio of length to volume (LIT/), and a slow-speed diff raction-grating instrument, infrared spectra with good resolution and high sensitivity could be obtained, This new type of so-called light-pipe gas cell was designed so that its length in relation to volume was the maximum possible, based on the fact that the amount of sample required for infrared spectrophotometry is determined by the molar absorptivity of the compound, and also the other terms in Beer’s law: A = alc, where A is the absorbance, a the molar absorptivity, l the length of the cell in centimetres and c the concentration in moles per litre.In order that sufficient energy may pass through the gas cell to give an infrared spectrum, the cross-sectional area of the cell must be at least 0.1 to 1 cm2. The molar extinction coeffi- cient is a function of the compound and of the wavelength. If this coefficient has values for a of about 1, 10 and 100 1 moll-l cm-l it can be considered to be weak, medium and strong, respectively. In infrared spectroscopy, it is not practical to operate with a lower absorbance than 0-005.Thus it is accepted that a high L/V value gives an effective gas cell for the analysis of small amounts of sample.16 Users of the technique involving the recording of the infrared spectrum of the gas- chromatographic fraction in the gas phase may find to their disadvantage that the infrared spectra of samples in the gas phase differ to some extent from those in the liquid or solid phases. However, the general pattern is the same, and the degree of difference depends on the nature of the sample and the physical and chemical environment.17-19 Q SAC and the author.718 NORIN COLLECTING GC EFFLUENTS FOR MICRO-SCALE [Analyst, VOl. 99 EXPERIMENTAL APPARATUS AND SAMPLES- A number of solutions containing various components with different boiling-points and vapour pressures (see Fig.1) were gas chromatographed and the separated components Fig. 1. Graphs of P = f(T) for the compounds listed in Table I1 (T measured in kelvin) A, n-butanol; B, isobutyl acetate; C, dioxan; D, isopropyl acetate; E, benzene; F, ethyl acetate; G, hexane; H, acetone; I, diethyl ether; and J, propylene oxide identified by infrared spectrophotometry. For their separation a Perkin-Elmer 116 (thermistor) gas chromatograph was used. The aluminium column, 2 m long and of 4 mm i.d., was filled with 20 per cent. m/m SE-30 on Chromosorb W, 60 to 80 mesh, and helium was used as carrier gas. For details of the conditions, see Table I. TABLE I GAS-CHROMATOGRAPHIC CONDITIONS Column 2 m long and 4 mm i.d. packed with 20 per cent.SE-30 on 60 to 80-mesh Chromosorb W Column temperature/'C Carrier gas Flow-rate of gaslrnl min-l Test 1 . . . . 72 Helium 45 Test 2 . . . . 93 Helium 25 A higher temperature (93 "C) was used for test 2 so as to avoid retention times that were too long owing to the slow rate of flow of carrier gas. For infrared spectrophotometric determinations, a Perkin-Elmer diff raction-grating spectrophotometer was used, together with a Wilks micro-scale gas cell. This cell was gilded at the inner walls so as to reduce the absorption of the materials on the walls and the gas cell was fitted with a potassium bromide window at both ends. The gilded inner walls caused multiple reflection, resulting in a higher L/V value. The cell, which had a cross-sectional area of 0.1 cm2 and length of 6 cm, i e ., L/V value of 10, was mounted in a mirror-focused system consisting of one plane mirror and three concave mirrors (see Fig. 2). Transmission through the optical system was measured and found to be approximately 30 per cent.November, 19741 INFRARED SPECTROPHOTOMETRIC ANALYSIS IN THE GAS PHASE mor From the light source To the ----c monochromator - 7 19 From the light source -t.------- Fig. 2. (a), Optical system for the micro-scale gas cell. The capillary is inserted into the gas cell, which is placed between the mirrors M, and M, and the gas cell is sealed. MI is a plane mirror, and M,, M, and M, are concave mirrors. (b), Sealed gas cell with the capillary sealed and inserted into the inlet tube of the gas cell. Surrounding the gas cell is an oven, U,, which maintains the temperature of the cell a t 150 "C.The oven U, is placed round the interlock of the capillary so as to improve the vaporisation of the sample COLLECTION OF FRACTIONS FROM THE GAS CHROMATOGRAPH- Several methods for trapping gas-chromatographic fractions have been described. The two most frequently used methods are: trapping of the fractions in a small column containing gas-chromatographic column filling material,3p4~20-22 and a cryogenic trapping technique.5J~~~-~6 Both methods are well known and have been tested in a number of different investigations. In the present investigation, the trapping system used consisted of a metal capillary, which could be attached to the inlet tube of the microcell, a three-way valve connected to the gas chromatograph and two PTFE stoppers (see Fig.3). The dimensions of the capillary, length 40 mm and i d . 0.45 mm, were the optimum for the rate of flow of carrier gas and the amount of sample to be frozen out. When a desired fraction issued from the gas chromato- graph, the three-way valve was turned so that the carrier gas and the fraction passed through the capillary, which was immersed in liquid nitrogen contained in a Dewar flask. After the fraction had been taken up in the capillary, the three-way valve was turned back to its original position, and the capillary sealed with a PTFE plug (see Fig. 3). Moisture was removed from the outside of the capillary, which was then inserted in the micro-scale gas cell. The gas cell was heated to approximately 150 "C by means of an oven, U,, which surrounded the cell (see Fig.2). Exhaust of the three-way valve capillary From gas Fig. 3. (a), Heated three-way valve through which the gas-chromatographic effluents issue; (b), sealing of the capillary720 [Analyst, Vol. 99 Prior to connecting the capillary to the inlet tube B of the gas cell (see Fig. 2), outlet tube A was sealed with a PTFE stopper. A second oven, U,, heated to 150 "C, was placed around the interlock connection of the capillary [see Fig. 2 (b)]. Because of the high tem- perature of the gas cell and extra heating with oven U,, the sample vaporised rapidly. About 1 minute after connecting the capillary recording of the spectrum was begun. As the amount of sample was small (35 to 100 pg) and the transmission low, the scanning speed could be kept slow, thus giving as good a spectrum as possible; the time required to record a complete spectrum from 2500 to 15 000 nm was about 15 minutes.Each mixture consisted of a solution of three compounds, each at a concentration of 5 per cent. V / V ; 1 p1 of solution contained 35 to 50 pg of each compound. Thus, on injecting 1 to 2 p1 of solution the peaks obtained in the gas chromatogram represented amounts of sample of TABLE I1 NORIN : COLLECTING GC EFFLUENTS FOR MICRO-SCALE Four mixtures of the compounds shown in Table I1 were separated. 35 to 1oopg. INVESTIGATION OF THE TRAPPING EFFICIENCY OF THE COLLECTION SYSTEM Maximum yield* Minimum yield* Substance Acetone . . . . . . Ethyl acetate . . . . Isopropyl acetate .. . . Diethyl ether . . . . Dioxan . . . . . . Isobutyl acetate . . . . n-Butanol . . . . . . n-Butyl acetate . . . . Propylene oxide . . . . n-Hexane . . . . . . Benzene . . .. ,. v Flow-rate 25 ml min-1 61-9 72.2 40.8 46.4 82-0 88.8 61.6 90.2 37.2 59.6 78.4 Flow-rate 45 ml min-1 51.9 69.9 50.5 69.1 87.7 73.3 55.5 92.2 47.7 66.6 76.3 r Flow-rate 25 ml min-l 52.7 62.9 34.9 23.7 68.7 66.0 49.4 86.6 36.2 33.8 60.7 Flow-rat e 45 ml min-I 44.3 50.9 49.7 42.5 68-1 59.0 46-1 63.7 36.7 59.7 72.0 * Yield calculated as per cent. of peak area (by gas chromatography). After recording the spectrum, the sample was removed from the gas cell by pumping with a vacuum pump connected to outlet tube B of the gas cell. In order to prevent dust from entering the gas cell, which could become attached to the walls and thus reduce the reflective properties, a 100-mm long steel column filled with molecular sieve was attached to inlet tube A.After pumping for 2 minutes, no detectable amount of sample remained. The cell was then ready for connection of the next capillary to record the infrared spectrum of a fresh sample. QUANTITATIVE EVALUATION OF THE COLLECTION METHOD- In order to make quantitative determinations and to test the efficiency of the freezing-out method, a gas-injection valve equipped with a loop was designed. The inlet and outlet of the loop were sealed, and the loop divided into two sections, which were connected by PTFE tubing to a two-way valve. The system was heated to a suitable temperature by electrical ovens (see Fig.4). When the fraction had been frozen-out in the capillary, the latter was connected to the two-way valve, which was kept closed. The capillary was then inserted into the part of the loop heated by oven U,. After about 1 minute, the gas valve was opened, and immediately afterwards the two-way valve was opened. The carrier gas now passed through the PTFE tubing and the capillary containing the vaporised sample. The sample was then transferred by the carrier gas to the injection block, and a new chromatogram was recorded, which could be compared with that obtained when freezing-out the sample. RESULTS AND DISCUSSION The limited volume (0.6 cm3) of the micro-scale gas cell made it impossible to connect the gas cell directly to a gas chromatograph with an ordinary packed column, as the volume of sample plzts carrier gas was too large for the gas cell.For the same reason, it was im- possible to collect the sample plus carrier gas in a gas syringe and transfer the contents toNovember, 19741 INFRARED SPECTROPHOTOMETRIC ANALYSIS IN THE GAS PHASE 721 L Loop of PTFE tube L Loop of PTFE tube Fig. 4. Gas injection valve. In Fig. 4 (a), the capillary is connected to the two-way valve and inserted into oven U,. Fig. 4 (b) shows the re-injection of the collected fraction the gas cell. The system described above was designed in order to separate the sample from the carrier, It may be of interest to estimate the smallest amount of substance that can be detected by the apparatus for the compounds described (see Table 111). The infrared spectra shown in Figs.5 (b) and 6 (b) do not represent minimum amounts of substance; the purpose here has been to demonstrate the usefulness of the method and the possibility of obtaining good spectra for interpretation. TABLE I11 APPROXIMATE MINIMUM SAMPLES FOR INFRARED SPECTROPHOTOMETRY ( A = 0.005) Weak Medium Strong band band band Concentration/mol 1-1 .. .. . . . . 0.005 0*0005 0.000 05 c x I* mol cm-2 5 0.5 0.05 Samples for 0.1-cm2 cell cross-section/pmol . . . . . . 0.5 0.05 0.005 Amount of substance of relative molecular mass 58 to 116/pg 30 to 60 3 t o 6 0.3 t o 0.6 * From Beer’s law: A = a x c x 1. In order to investigate the effect of varying the rate of flow of the carrier gas, two trial runs were carried out (see Table I). No great variations in the yields were noted and the differences found were not regular.The trapping efficiency of the condensing system depends on a number of factors, which include the temperature and the diameter of the trapping capillary, the flow-rate of the effluent gas emerging from the gas chromatograph, the time the sample remains in the trap and the vapour pressure of the sample under the trapping conditions employed. Many materials present problems because they form aerosols when they emerge from the exhaust of the gas chromatograph, particularly when drastic cooling conditions are used. The collec- tion of the gas-chromatographic effluents did not disturb the shape of the gas chromatogram, which can be seen in Figs. 5 (a) and 6 (a). The infrared spectra of the collected effluents [see Figs.5 (b) and 6 (b)] did not show all of the bands that appear in the liquid phase and there are many other differences between the gas-phase and condensed-phase spectra.722 6 5 4 > E 3 2 1 NORIN COLLECTING GC EFFLUENTS FOR MICRO-SCALE [Analyst, VOl. 99 (a) C I I I I 1 I 80 r 10 I I l l I I I I 1 1 2 1 0 8 6 4 2 0 4000 3000 2000 1750 1500 1250 1000 750 Time/minutes cm-1 Fig. 5 . (a), Gas chromatogram of a mixture of three compounds. Peaks: A, isobutyl acetate; B, collected fraction, dioxan; C , n-hexane (solvent) ; D, diethyl ether; and E, air. Conditions: column temperature, 93rC; cell volume, 0.6 ml; flow-rate of carrier gas, 25 ml min-l. (b), Infrared spectrum of 70 pg of dioxan When a molecule is exposed to infrared irradiation, a vibrational spectrum arises, which is determined by potential energy and nuclear configuration.The potential energy can be expressed, in terms of internal parameters, as displacement of inter-atomic distances from their equilibrium values. Molecules in the gas phase generally show little tendency to interact with other molecules. When molecules interact with each other in the condensed phase, the potential energy of the nuclei depends not only on the internal field of force in the molecule, but also on the configuration of other molecules relative to this particular molecule. Several other effects can be observed when molecules pass from the condensed phase to the gas phase. 6 5 4 > E 3 2 1 D i 80 1,o t- I l l I I I 1 I 12 10 8 6 4 .2 0 4000 3000 2000 1750 1500 1250 1000 750 Tim e/m i n'u tes cm-1 Fig.6. (a), Gas chromatogram of a mixture of three compounds. Peaks: A, collected fraction, (b), Infrared spectrum of 40 pg of isobutyl isobutyl acetate; B, dioxan; C, D and E, as in Fig. 5 (a). acetate The number of bands in the spectrum of a gas phase is generally less than that in the spectrum of a condensed phase. Bands observed in the spectrum of a condensed phase can be recognised in the spectrum of the corresponding gas phase, but are generally displaced with regard to frequency. However, this displacement is normally not more than 2 per cent. and is often only 1 per cent. The bands are normally displaced towards lower frequencies and the following can be observed: Vgas > Vsolution (non-polar solvent) > Vliquid > %,lidNovember, 19741 lNFRARED SPECTROPHOTOMETRIC ANALYSIS I N THE GAS PHASE 723 The most remarkable changes in the spectra when molecules pass from the condensed phase to the gas phase are the occurrence of discrete rotational fine structure and charac- teristic band contours.In the gas phase, the molecules are free to rotate and a rotational fine structure is found in the absorption bands. In small molecules, rotational energy levels are sufficiently separated to give rise to separate absorption bands. In larger molecules, the rotational energy levels are too close to each other to be resolved and absorption produced by rotational - vibrational bands is observed. The pressure of the gas gives rise to the important fine structure of absorption bands.With increased pressure the gas single lines are broadened owing to the increased number of collisions of molecules. A change of state is also accompanied by changes of intensity. Variations of more than a factor of five may occur when molecules pass from the gas phase to the condensed phase, but normally the change is moderately small. Significant changes in band shapes occur and are often accompanied by variations in intensities. Bands in the spectrum of the gas phase show typical PQR branch structure, depending on allowed vibrational and rotational energy transitions. The P and R branches depend on changes in the vibrational and rotational quantum numbers, the rotational quantum number decreasing for the P branch and increasing for the R branch. The Q branch depends on changes in the vibrational quantum number only and not in the rotational quantum number.The change in band shape when a molecule passes from a gas phase to a condensed phase is due to the inability of the molecule to pass through free rotation. As a result, the quantised rotational energy levels in the free molecule disappear as also does the associated rotational fine structure. Collision and other interactions between molecules have fairly small effects on the vibrational energy levels. Among other effects that appear when molecules pass from the gas phase to the con- densed phase are changes in the relative proportions of any rotational isomers that may be present. In some molecules internal rotation about one or more single bonds occurs. If the energy barrier, which restricts the rotation around the bond, is not large (as is usual in single bonds), rotational isomers may be completely converted from one form into another and a dynamic equilibrium will be achieved.The relative proportions between the species will depend on the physical environment. It has been found that the spectrum of a complex system in gas and liquid phases often becomes comparatively simple in a solid phase because 01 the 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. presence of a number of isomers. REFERENCES Haslam, J., Jeffs, A. R., and Willis, H. A., Analyst, 1961, 86, 44. Bergstedt, E., Chronzatographia, 1969, 2, 545. Fowlis, I. A., and Welti, D., Analyst, 1967, 92, 639. Howlett, M. D.D., and Welti, D., Ibid., 1966, 91, 291. Edwards, R. A., and Fajerson, I. S., Analyt. Chem., 1965, 37, 1630. Copier, H., and van der Maas, J . H., Spectrochiinica Acta, 1967, 23A, 2699. Curry, A. S., Read, J. F., Brown, C., and Jenkins, R. W., J . Chrounat., 1968, 38, 200. Gray, L. S., Metcalfe, L. D., and Leslie, S. W., “Fourth International Gas Chromatographic Krakow, B., Analyt. Chem., 1969, 41, 815. Low, M. J. D., and Freeman, S. K., Ibid., 1967, 39, 194. Brown, R. A., Kelliher, J. RI., Heigl, J. J., and Warren, C. W., Ibid., 1971, 43, 353. Bartz, A. &I., and Ruhl, H. D., Ibid., 1964, 36, 1892. Penzias, G. J., Ibid., 1973, 45, 892. Wales, R. J., Proc. SOC. Analyt. Chem., 1967, 4, 112. Wilks, P. A., jun., and Brown, 13. A., Analyt. Chem., 1964, 36, 1896. Littlewood, B., Chrornatographia, 1968, 1, 223. Herzberg, G., “Infrarcd and Raman Spectra of Polyatomic Molecules,” Van Nostrand, New York, Rao, C. N. R., “Chemical Xpplicatioii of Infrared Spectroscopy,” Academic Press, New York, 1963. Davies, M., Editor, “Infra-red Spectroscopy and Molecular Structure,” Elsevier Publishing Com- pany, Amsterdam, London arid New York, 1963. Amy, J . W., Chain, E. M., Baitinger, W. E., and McLafferty, F. W., AnaZyt. Chem., 1965, 37, 1265. Cartwright, M., and Heywood, A., Analyst, 1966, 91, 337. Widmark, K., and Widmark, G., ,4cta Chew. Scand., 1962, 16, 575. Snyder, R. E., J . Chvomat. Sci., 1971, 9, 638. Swoboda, P. A. T., Nature, Lond., 1963, 6, 31. Hoffman, R. L., and Silvera, A., jun., J . Gas Chromat., 1964, 2, 107. Symposium,” Michigan State University, Michigan, June, 1963. 1946. Received M a y 6th, 1974 Accepted July Sth, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900717

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

724 Ayzalyst, November, 1974, Vol. 99, pp. 724-728 Antioxidant Analysis Incorporating a Thin-layer Chromatographic Separation Procedure BY DAVID T. MILES j.4 ustralian Post Ofice Reseaych Laboratories, 59 Little Collins Street, Melbourne 3000, Australia) The method described consists of three distinct and separate steps. Initially, a thin-layer chromatographic technique, involving the use of mixed silica gel - aluniinium oxide plates, is used to separate and identify the ex- tracted antioxidants. This step is followed by confirmation of the suspected identity and determination of the antioxidant by using infrared and ultra- violet spectroscopy. This method is considered applicable, over a range of concentrations, to the analysis of both antioxidants and ultraviolet screening compounds possessing ultraviolet absorbing properties, in polyolefinic polymers such as polyethylene, polypropylene, poIy(4-methylpentene), etc.I t can be used t o detect such additives down to a level of 200 p g g-I, and possibly even less, in a given polymer. In addition, the procedure has been successfully applied to the separation of isomeric antioxidants and to the identification of antioxidants present in lubricating oils. CURRENT interest in the analysis of various polymers at the Australian Post Office Research Laboratories led to a decision to investigate methods available for the determination of antioxidants. Those applicable to the six antioxidants most frequently used in Australia for the production of polyolefins were of prime importance. It was immediately realised that numerous rnethod~l-~ existed for the identification of specific antioxidants and mixtures thereof.While accounts of some of these methods reported a subsequent quantitative deter- minati0n,~$~,6 none appeared to report a separation of the extracted antioxidants by thin- layer chromatography using 1 + 1 silica gel - aluminium oxide plates. Therefore, our studies were directed towards this aspect and eventually we developed a procedure that possessed improved antioxidant resolution and facilitated their identification. When linked with a step for the confirmation and quantitative determination of the antioxidants it formed the basis of the procedure reported in this paper. Users of this procedure will realise that new antioxidant compounds are constantly being encountered and conflicting conclusions may therefore be drawn. EXPERIMENTAL APPARATUS- All polymer extractions were performed with a standard Soxhlet apparatus.Any reputable thin-layer chromatographic equipment was considered to be suitable for Infrared spectra were obtained by using a Jasco IR-G diffraction grating spectrophoto- A comparable instrument would A Perkin-Elmer 124, double-beam, ultraviolet - visible spectrophotometer or a similar use in the separation step. meter covering the frequency range 4000 to 400 cm-l. suffice. unit was used for measurements. MATERIALS- Type E, mixed 1 : 1. reagent grade chemicals. Thin-layer plates were prepared from Merck Kieselgel G and Merck aluminium oxide, The solvents required, methanol, chloroform, and methylene chloride, were all analytical- 0 SAC and the author.MILES 725 Antioxidant reference standards with a concentration of 1000 mg 1-1 were prepared from pure samples of : A, 1,1,3-tris(2-methyl-4-hydroxy-5-t-butylphenyl)butane; B, 4,4’-thiobis(6- t-butyl-m-cresol) ; C, NN’-di-P-naphthyl-j5-phenylenediamine ; D, 2,2’-methylenebis[6-( 1- methylcyclohexyl)]-fcresol; E, 2,6-di-t-butyl-@-cresol; and F, 2,2’-thiobis(4-methy1-6-t- butylphenol).PROCEDURE- Cut the polymer specimen into small pieces and use accurately weighed portions of between 2 and 5 g, depending on the amount and type of antioxidant in the polymer. If these are not known use a maximum mass of 5 g. Extract the sample with 50 ml of methylene chloride or chloroform, by use of a Soxhlet extractor, for a period of 8 hours to ensure complete removal of the antioxidant.Cool and filter the extract if necessary; in isolated instances pigment particles will be observed to pass through the extraction thimble. Evaporate the cool, filtered extract nearly to dryness, immediately re-dissolve the residue in 500 p1 of methylene chloride and transfer the solution to a stoppered phial ready for the thin-layer chromatographic separation. Coat the thin-layer chromatographic plates with a slurry of aluminium oxide and silica gel (1 + l), containing calcium sulphate as binder, by use of the following technique. Dis- perse approximately 40 g of the powder mixture in 60 ml of distilled water and stir gently to ensure complete wetting of the powder and to assist the removal of any entrapped air.Allow the slurry to stand for about 5 minutes to thicken. Adjust the spreader to give a 300-pm coating to the glass plates in one steady pass. After gelling, load the plates horizontally into a thin-layer chromatographic plate rack and place the rack in an oven at 110 “C for 2 hours in order to activate the plates. Subsequently, store the rack, with the prepared plates in a vertical position, in a desiccator until required. Plates not used within 3 days should be re-activated in the oven for 15 to 30 minutes at 110 “C. Spot the solutions on to the plates in the usual manner by using a graduated spotting pipette and a template. Apply 5-pl aliquots of the reference standards and concentrated sample extract, such that the resulting spots are kept to a minimum size, e.g., 3 mm or less in diameter.This can be achieved by spotting several small aliquots, e.g., 2, 2 and 1 p1 in succession, while directing a current of air to the point of application. Place the spotted plate in an unlined developing tank that has been saturated at room temperature with chloroform vapour, chloroform being the eluting medium, and allow the solvent front to travel to a distance of 100 mm from the origin. Afterwards, dry the plate in air and locate the resolved component spots by spraying the plate with a 50 g 1-1 solution of iodine in methanol. As an alternative, the plate can be placed in a developing tank containing iodine vapour. In both instances the resolved spots appear as light yellow to red - brown areas on a near-white background.Confirm the identity of the antioxidant that has been indicated by thin-layer chromato- graphy by obtaining an infrared spectrum of a second sample extract, concentrated to the required level, and comparing it with the spectrum of a pure reference compound or by reference to a recognised atlas of spectra. When the initial extract contains components other than antioxidants or mixtures thereof, it may be necessary t o repeat the thin-layer chromatographic step. A streak of the concentrated extract, i.e., 30 p1 to 40 pl, is applied to the thin-layer plate together with a control spot. After elution, location of the area of interest is achieved as before with a solution of iodine in methanol, while the streaked portion of the plate is kept covered with aluminium foil.The foil is subsequently removed and the area indicated to contain the antioxidant is scraped off. This portion is then extracted with chloroform, filtered, the chloro- form evaporated down and the resulting concentrate applied to a potassium bromide disc in sufficient volume to yield an acceptable infrared spectrum following complete removal of the chloroform. The identity of the resolved component is then confirmed in the manner described above. When quantitative results are required it is advisable to remember that the normal concentration of antioxidant in a manufactured polymer ranges from 0-2 to 6 mg g-l. Assum- ing that the sample contains 2 mg g1 of an antioxidant, and taking 5 g for extraction, the final extract, adjusted to 50 ml, will contain 20 mg 1-l.In most instances the amount taken is adjusted to obtain a concentration of about 20 mg 1-l in the final extract.726 MILES : ANTIOXIDANT ANALYSIS INCORPORATING A [AnhdySt, VOl. 99 The final quantitative determination of the indicated antioxidant is obtained by use of ultraviolet spectrophotometry. Record the wavelengths at which the suitably diluted standard solutions show maximum absorbance in a 10-mm path length cell. Then, plot calibration graphs of concentration ‘ueysus absorbance for the range of concentrations from 0 to 50 mg 1-1 for each antioxidant studied at the characteristic wavelengths recorded in Table I. Alternatively, the required information can be obtained more conveniently from a table of A,,,. and absorbance per 10 mg 1-1 for each antioxidant.TABLE I CALIBRATION RESULTS FOR THE ANTIOXIDANTS STUDIED 2,6-Di-t-butyl-P-cresol a t 283 nm A I > Concentration/ mg 1-l Absorbance 10 0.112 20 0.227 30 0.330 40 0.434 50 0.535 4,4’-Thiobis(G-t-butyl- m-cresol) at 247 nm Concentration/ mg 1-1 Absorbance 10 0.380 20 0.800 30 1.190 40 1.580 60 1.920 2,2’-Methylenebis [6- (1 -methyl cyclohexyl)]-p-cresol at 287 nm Concentration/ n I 3 mg 1-1 Absorbance 10 0.115 20 0.232 30 0-360 40 0.475 50 0.600 NN’-Di- j3-naphthyl-p- phenylenediamine at 321 nm mg 1-l Absorbance 5 0-442 10 0.890 15 1.265 20 1.690 r A \ Concentration/ - L 1,1,3-Tris(2-rnethyl-4-hydroxy- 5-t-butylpheno1)butane at 281 nm Concentration / mg 1-l Absorbance 10 0.120 20 0.235 30 0.340 40 0.460 50 0-570 2,2’-Thiobis(4-methy1-6- t-butylphenol) at 296 nm I A > Concentration/ mg 1-’ Absorbance 10 0.242 20 0-460 0.702 30 40 0.915 60 1.1 18 From these results and a known mass of extracted polymer, the concentration, and hence the mass, of antioxidant in the original sample is calculated.DISCUSSION AND RESULTS An initial literature survey revealed that the majority of reported procedures involved the use of mixed solvents for elution on thin-layer plates used in the resolution of antioxidants. This probably resulted from necessity rather than choice. Mixed eluting solvents suffer from anomalous effects arising from the effect of varying humidity conditions on the evaporation rates of the components. These anomalies, in turn, affect the partitioning of the extracted antioxidants between the liquid and solid phases.Our experiments were therefore aimed at finding a suitable single solvent to replace the various mixtures advocated and they revealed that chloroform performs satisfactorily when used in conjunction with 1 + 1 silica gel - aluminium oxide thin-layer plates. The decision to investigate a mixed silica gel - aluminium oxide plate for the thin-layer chromatographic separation of the five antioxidants of interest resulted from the observations made when either simple silica gel or aluminium oxide plates were employed. Neither coating gave a completely satisfactory result ; either the resolved components were bunched near the solvent front or they were held back in the region of the origin. This bunching was attributed to the adsorptive forces of the coating.With silica gel, the adsorptive forces were too weak as a result of the inability of the sterically hindered phenolic antioxidants to achieve a silica gel - solute interaction by means of hydrogen bonding; hence, the components moved with the eluting solvent. With the simple aluminium oxide plates the resolved components were confined to an area near to the origin, probably because of a property of aluminium oxide whereby the liydroxyl groups of the antioxidant are superficially built into the crystal lattice of the oxide. It was consequently concluded that a 1 + 1 mixture of aluminium oxide and silica gel should provide a state midway between these two extremes. The results we achieved tended to support this conclusion. Experiments to determine the optimum thickness of the silica gel - aluminium oxideNovember, 19743 THIN-LAYER CHROMATOGRAPHIC SEPARATION PROCEDURE 727 layer indicated that a 300-pm coating gave the best resolution of sample extracts.Extracts in methylene chloride yielded fewer unwanted components, i e . , processing additives and degradation products, than those in chloroform. If the amount of antioxidant C is to be determined the analyst should be aware that chloroform will not completely extract this antioxidant from a polymer specimen. As antioxidant C is also far less soluble in methylene chloride at room temperature than at elevated temperatures it was necessary to heat extracts of this component nearly to the boiling-point prior to spotting. In the author's experience, when chloroform was replaced by methylene chloride in the initial extraction procedure, quantitative extractions of the order of 980 mg g-1 were obtained.The usual precautions for thin-layer chromatographic studies were observed, e.g., tightly fitting developing tank lids, saturation of the developing tank with solvent vapour, etc. Experiments with lined and unlined tanks showed that there was little difference between them, except perhaps a reduced elution time with a lined tank. Failure to take the above precautions did not necessarily mean that the plate concerned was wasted. If sufficient standards were spotted across the plate (a minimum of three was desirable) the variation in retention values could be recorded and an identification of the unknown obtained.In this event confirmation of the apparent identity was imperative. The migration distances of the extracted sample components on the thin-layer chromato- graphic plate were measured by determination of their individual R, values. Owing to the variations in R, values between plates, it was considered more reliable to relate the R, values of the components to the R, value of a reference standard, i.e., antioxidant E, The new ratio was known as the R, value' (see Table 11). TABLE I1 RETENTION VALUES OBSERVED FOR THE ANTIOXIDANTS STUDIED Observed R p and RT values of antioxidant- Antioxidant as a single substance as one of a mixture n f 1 -- RJ? H T RF R T A 0.25 0.25 0.25 0.25 B 0.43 0.44 0.51 0.52 C 0-87 0.89 0.90 0.92 D 0.95 0.97 0.95 0.97 E 0.98 1.00 0.98 1.00 F* - - - - * Reasons for the non-inclusion of antioxidant F are explained in the following paragraph.From Table I1 it will be noticed that the retention values varied according to whether the component of interest was present as a single substance or as part of a mixture. Those spots which did not correspond to any known antioxidant or its breakdown products were attributed to the processing aids mentioned earlier. Antioxidant A did not exhibit the behaviour under illustration; it performed in a similar manner to antioxidant D in this respect. Antioxidant F is not included because it has not been encountered in mixtures in our experience to date, and, from information received, it appears that this is most unlikely to occur. In addition to the determination of antioxidants in polyolefins, the determination of ultraviolet screening compounds4 in this type of polymer can be carried out in a similar manner.A prior separation step is necessary if the latter compounds are to be determined on the same sample as the antioxidants, as they are very strongly absorbing over the ultraviolet wavelength range and could interfere in any attempt to determine the amount of an antioxidant. This method should also prove useful for the determination of antioxidants in other classes of polymer. The procedure as reported above has, in addition, provided information regarding the identities of antioxidants, e.g., styrenated phenol, in various lubricating oils.6 CONCLUSIONS Mixed silica gel - aluminium oxide thin-layer plates eluted with chloroform can be used to improve the resolution and ease of identification of antioxidants extracted from polyolefins. Coupled with the confirmatory and quantitative steps, the reported thin-layer technique728 MILES forms the basis of a general procedure that can be used for determining antioxidants in polyolefins both qualitatively and quantitatively. The permission of the Australian Post Office to publish this paper is hereby acknowledged. 1. 2. 3. 4. 5 . 6. 7. REFEREKCES Slonaker, D. F., and Severs, D. C., Analyt. Chem., 1964, 36, 1130. Woggon, H., Kom, O., and Jehle, D., Nuhrung, 1966, 9, 496. Crompton, T. R., Eur. Polym. J., 1968, 4, 473. Crompton, T. R., “Chemical Analysis of Additives in Plastics,” Pergamon Press, Oxford, New Zubkova, N. D., Turskii, Yo. I., Genkina, V. I., and Klyacho, G. V., Khim. Tekhnol. Topl. Masel, Simpson, D., and Currell, B. R., Analyst, 1971, 96, 616. British Standard 2782: 1970, Part 5, in preparation. York, Toronto, Sydney and Braunschweig, 1971. 1964, No. 8, 60. Received October 15th, 1973 Amended A p i l 23vd, 1974 Accepted May 21st, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900724

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Amzlyst, November, 1974, Vol. 99, pp. 729-738 729 Partly Quenched, Synchronously Excited Fluorescence Emission Spectra in the Characterisation of Complex Mixtures BY J. B. F. LLOYD (West Midland Forensic Science Laboratory, Priory House, Gooch Street North, Birmingham, B5 6QQ) Non-linear quenching of mixed fluorescences at low concentrations of quencher, and linear variation a t increased concentrations, depend on the Stern - Volmer coefficients and on the proportions of the components in a manner that can be predicted from the Stern - Volmer equation. The varia- tion with quencher concentration of emission intensities synchronously excited from mixtures of benzo[a]pyrene and perylene at intervals of 46nm is quantitatively interpreted in this way. A variety of examples is used to show that in the qualitative charac- terisation of complex mixtures the specificity of synchronously excited spectra is enhanced by the use of varyingly quenched conditions.The high resolution afforded by the synchronous excitation technique readily demonstrates the widely differing relative sensitivities of components of fluorescent mixtures to quenching effects; and reveals new emissions that are obscured under non-quenching conditions. COMPLEX mixtures of fluorescent compounds often yield broad, poorly resolved spectra irrespective of excitation wavelength. By the synchronous excitation technique, with which the excitation wavelength is varied continuously with the plotted emission,lV2 spectra are obtained to which each fluorescent component may contribute in a resolved and distinguish- able way.The spectra serve as fluorescent fingerprints, for example, in forensic science,3 and, in conjunction with chromatography, in the analysis of complex mixtures4 Even so, seemingly well resolved emissions are sometimes composite and may envelop others of significance. In these circumstances more detail can be obtained from partially quenched spectra. The reduction in fluorescence emission caused by interaction between quenching and fluorescent molecules, in their ground and excited states, respectively, can generally be expressed by the Stern - Volmer relationship: f”(f)-l = K[Q] + 1, in which f” and f are fluorescence yields in the absence and presence, at concentrations [Q], of quencher, and K , the variation of f”(f)-l with respect to [Q], is the Stern -Volmer coefficient.When the efficiency of quenching encounters approaches unity, K = kqTO, where k , approaches the diffusion-controlled bimolecular rate constant and TO is the unquenched lifetime. Quenching effects, therefore, are dependent on fluorescence lifetimes, which vary extensively between different compounds. For instance, values in the range 1 to 500 ns are listed by Birks5 for aromatic hydrocarbons in solution. Although Parkers has observed that situations are conceivable in which fluorescence quenching might be used to increase selectivity, very little such use has occurred. The few examples of its use include the visual examination of thin-layer chromatograms, and selective spectrofluorimetric analyses for fluorescent hydrocarbons, heterocyclic compounds and other air pollutants under quenched conditions by Sawicki and his colleagues7 ; and Zander’s8 determination of perylene in the presence of benzo [alpyrene and tetracene, which were quenched by the addition of iodomethane.As the reported5 fluorescence lifetimes of these compounds are, in order, 4-9 to 6.9, 49, and 5-2 to 9.2 ns, the selective quenching of the long-lived benzo [alpyrene fluorescence, although not that of the relatively short-lived tetra- cene, is to be expected. The related use of heavy-atom solvents, such as iodomethane, to facilitate intersystem crossing from excited singlet to phosphorescent triplet statesg is a widely applied spectrophosphorimetric technique.1° A recent paper by Zanderll contains much pertinent data. Q SAC; Crown Copyright Reserved.730 LLOYD : PARTLY QUENCHED SYNCHRONOUSLY EXCITED FLUORESCENCE [Analyst, Vol.99 EXPERIMENTAL The spectra, which are uncorrected, are obtained with a Baird Atomic SFlOOE spectro- photofluorimeter (the excitation and emission monochromator drives of which can be actuated simultaneously at pre-set intervals in order to achieve synchronous excitation) fitted with a Hanovia 150 W xenon source lamp and an IP28 photomultiplier tube, and operated at a half-band width of 6 nm. Except when stated otherwise, the sample solutions are contained in a 5-mm cuvette and de-oxygenated with nitrogen pre-saturated with solvent. The cuvette is fitted with a PTFE lid (sealed in place with Araldite) through which pass two hypodermic needles.One of these needles, which reaches to the bottom of the cuvette, is used to introduce the nitrogen stream and to transfer solutions. The other needle, terminating immediately below the lid, serves as a nitrogen outlet. In order to minimise light-scattering effects, the inlet passes down the corner of the cuvette enclosed by the excitation and emission optical paths. Connections between the exterior of the instrument and the cuvette are made with PTFE tubing (0.5 mm id.). All of the solvents used are commercial spectrophotometric grades that exhibit no significant extraneous fluorescence emission or ultraviolet absorption. Iodomethane (Koch- Light Laboratories Ltd., puriss. grade) and bromoform (laboratory-reagent grade) are briefly shaken with solid sodium metabisulphite, filtered, distilled and passed through a 70-mm column of aluminium oxide, under nitrogen, immediately before use.No trace of free halogen (when tested with starch - iodide reagent) is present in the products. Experiments are conducted at ambient temperatures, the usual fluctuations in which are without significant effect on the spectra. The various amounts of quencher used in the Stern - Volmer experiments are transferred by microsyringe into the nitrogen inlet line, and mixed with the contents of the cuvette (0.5 ml) by ejection of a small amount into and out of the inlet line with a syringe attached to the outlet. Final concentrations of quencher are confirmed at the end of each experiment by ultraviolet spectroscopy. The plotted and tabulated results are corrected for absorption by the quencher at wavelengths of fluorescence excitation in the few instances when this is significant.Any effects on the reproduced spectra are indicated in the figure captions. At optimum excitation intervals sensitivities obtained with the synchronous excitation technique are identical with those attained in the conventional mode of operation with optimum fixed excitation wavelengths. Their esti- mated coefficient of variation is 4.2 per cent. at maximum spectrophotometer sensitivities. The estimated wavelength standard deviation is 1.5 nm. It is emphasised that because the spectra are uncorrected and, therefore, subject to distortions imposed by the spectral sensitivity distribution of the instrumentation used, the spectra are of significance in a comparative rather than strictly absolute sense and are liable to modification when measured with other instruments.This restriction does not affect the conclusions drawn. Fluorescence intensities are measured directly from the spectral base-line. DISCUSSION AND RESULTS The variation of the fluorescence of a mixture with concentration of quencher is deter- mined by the number and proportions of the components and by each individual component’s Stern - Volmer coefficient. If the observed fluorescence, F , at any particular wavelength is composed of emissions x,F + x,F + . . . x,F, where xi is the fractional contribution of the ith component, then for each component Hence, overall The variation of Fo(F)-l with Q tends to become linear as Q increases.In the two-compart- ment caseNovember, 19741 EMISSION SPECTRA IN THE CHARACTERISATION OF COMPLEX MIXTURES 731 In practice, the coefficient of [Q] is a good approximation to the gradient and becomes the Stern - Volmer coefficient of the remaining component when either xlo or x20 becomes zero, or the coefficient of either when K , = K,. However, the intercept term does not become unity in these circumstances. A better approximation is obtained when the intercept is equated with the difference between the approximate term in [Q] and the exact equation, with [Q] set equal to 1 moll-l. At such concentrations, even with relatively small coefficients characteristic of inefficient quenchers or short-lived fluorescences, the exact equation is essentially linear and values of the intercept calculated numerically vary negligibly above this concentration.With the substitution of 1 - x1 for x,, the equation becomes which fulfils the appropriate one-component criteria. Some results from the benzo[a]pyrene (component 1) - perylene (component 2) system are considered as a case in point. Mixtures of these compounds synchronously excited at intervals of 45 nm yield emissions centred at 409, 440 and 475 nm. The 409-nm emission is contributed solely by the benzo[a]pyrene and that at 475 nm solely by perylene, the 440-nm emission being due to both compounds. In Fig. I graphs for Fo(F)-l zleysus [Q] of data from such a mixture in de-aerated cyclohexane quenched with varied amounts of bromoform are shown. Quenching at 409 (x! = 1) and at 475 nm (x! = 0) varies linearly with concen- tration to give Stern - Volmer coefficients of 35.23 and 4.88 mol-l, the ratio of which (7.22) agrees within the limits of experimental error with the ratio (7.65) of the fluorescence lifetimes in cyclohexane of the two compounds.s Quenching at 440nm varies initially non-linearly with [Q], and is fitted by the line calculated from the results for 409 and 440 nm with xIo = 0.75, in agreement with the com- position of the mixture used.Beyond 0-2 mol 1-1 quenching becomes a linear function of concentration and the approximate equation is applicable. In general, this situation occurs at higher quencher concentrations as the lifetimes and proportions of long-lived components increase, and the calculated lines in Fig.1 show this effect for the present instance, The value of x! at which the deviation from the linearity is at a maximum is calculated to be 0.87 (from the point at which the differential of the intercept term in the linear equation with respect to x: is zero). The dependence on X! values of the [Q] coefficient is of obvious significance in quantitative analysis, provided that quenching follows diffusion-controlled kinetics and that sufficiently detailed information on the fluorescent components of the analyte is available. This aspect remains to be explored. The materials considered in the following are highly complex, and the spectra of their unseparated fluorescences are at present susceptible only to qualitative interpretation. The fluorescence of unused automobile engine oils is due to alkylated benzohomologues of thiophen4 and to benzene, naphthalene and phenanthrene derivatives,,, all of which emit largely in the ultraviolet region.With use, new emissions become apparent, particularly in the visible region, which are due to a considerable variety of polynuclear hydrocarbons con- taining up to thirty aromatic carbon atoms.* In Fig. 2 the uppermost spectrum, of a de- aerated solution in cyclohexane excited at an interval of 23nm, is an example from this type of material. When increasing amounts of iodomethane are added the over-all fluorescence is increasingly quenched (Fig. 2). The quenching is attributed to the heavy-atom solvent effect. Reduction of fluorescence emission caused by direct absorption of light by iodo- methane is not significant except in the most highly quenched spectrum, when emission intensities in the 350-nm region are, in consequence, reduced by 20 per cent.Quenching is most effective in the ultraviolet region and provides increased detail in the 350-nm region where, with an iodomethane concentration of 0.015 mol l-l, a new emission at 344 nm becomes apparent. The most prominent of the polynuclear hydrocarbon emissions, that at 405 nm, which, in this instance, is due largely to anthracene, is little affected. An initial rapid quenching occurs at 382 nm, to give a residual anthracene emission at higher732 LLOYD : PARTLY QUENCHED SYNCHRONOUSLY EXCITED FLUORESCENCE [AnaZyst, Vol. 99 levels of quenching, and relative intensities of the very weak anthanthrene and perylene emissions in the range 455 to 465 nm become inverted. The rapidly quenched 382-nm com- ponent is probably due to pyrene or an alkylated pyrene of which the fluorescence emissions are exceedingly long-lived." 0-2 0.4 0.6 0.8 1.0 1.2 [Q]/mol I-' Fig. 1. Stern - Volmer graphs and calcu- lated lines for benzo[a]pyrene (component 1) and perylene (component 2) mixtures in the presence of bromoform. The emissions were excited a t an interval of 45 nm. -4, Experi- mental points and least squares line, 409 nm (xlo = 1.0) ; R, lines calculated from the exact (full line) and approximate (broken line) equations with x10 = 0.9; C, calculated line and experimental points a t 440 nm for x10 = 0.75; D, calculated, x10 = 0.25; and E, experimental points and least squares line, 476 nm (x10 = 0, x20 = 1) 300 350 400 450 500 550 Wavelemthhm Fig.2. Used oil (0-2 mg ml-1) in cyclohexane synchronously excited at an interval of 23 nm in the presence of various amounts of iodomethane. Iodomethane concentrations: A, 0; B, 0.015; C, 0-046; D, 0.195; and E, 0.495 M. In E, absorption of excitation a t 350 nm is 20 per cent. Stern - Volmer graphs of emissions excited at intervals of 30 nm from a used oil sample in cyclohexane varyingly quenched with bromoform are shown in Fig. 3. Relative to the 23-nm intervals in Fig. 2, the increased interval tends to increase emission wavelengths by up to 7 nm, particularly in the visible region, and increases resolution at the extremities of the spectral range, to the detriment of the centre.The main features of the spectra are, however, unaltered. It follows from the foregoing that from single compounds or from mixtures of compounds with equal fluorescence lifetimes, linear Stern - Volmer graphs with intercepts of unity are to be expected. This condition is met only by the data for 335 nm, of which a least squares analysis yields a Stern - Volmer coefficient of 114 mol-l corresponding to a lifetime of 84 ns (based on a bimolecular quenching rate constant of 1-35 x loQ mol-l s-l, calculated from the results for benzo[a]pyrene and perylene on the assumption of lifetimes of 49 and 4.88 ns, respectively). This particular emission is assigned to dibenz~thiophens,~ for which fluorescence lifetimes are not available. All of the other graphs deviate from linearity to widely varying extents, but mostly in a manner expected for mixtures that undergo diffusion-controlled quenching.The 318-nm emission due to napht halenesll and benz~thiophens~ is exceptional. Quenching increases more rapidly than a linear dependence on quencher concentration requires, possibly because of interactions between the quencher and non-excited fluorescent molecules. The over-all variation in composite lifetimes represented by the Stern - Volmer graphs is 2.5 to 130 11s. In general, the long-lived emissions occur at shorter wavelengths, but the trend is irregular and hence enhances the value of quenching experiments in the characterisation of such materials.November, 19741 EMISSION SPECTRA IN THE CHARACTERISATION OF COMPLEX MIXTURES 733 Compared with the last example, a rather different distribution of polynuclear hydro- carbons is shown by the spectra of a soot extract in cyclohexane (Fig.4). At an excitation interval of 23 nm, the three unquenched composite emissions at 407, 430 and 456 nm are mainly due to benzo[a]pyrene (at 407 nm), benzo[k]fluoranthene (407 and 430 nm) and anthanthrene (430 and 456 nm). High concentrations of iodomethane (1.2 moll-l) transform the spectrum essentially into that of perylene (438 and 462 nm). At intermediate concen- trations the anthanthrene emission at 456 nm is reduced to an extent that permits the dual nature of the fluorescence in this region to be clearly seen. Similar considerations apply to the 430-nm region. The Stern - Volmer graphs (Fig.5 ) exhibit the expected curvature. Those graphs for emissions at 438 and 462 nm tend to become parallel, consistent with their origin , which is largely in one compound (perylene) . 14 12 10 r- I c- c 8 k 6 0 4 2 0 0.1 0.2 0.3 0.4 [Ql/mol I-' Fig. 3. Stern-Volmer graphs from a used oil (0.2 mg ml-l) in cyclohexane, quenched with bromoform and syn- chronously excited a t an interval of 30 nm. The lines are numbered according t o emission wavelengths 350 400 450 500 550 Wavelengthhm Fig. 4. Cyclohexane extract of a soot sample (10 p g ml-l) syn- chronously excited a t an interval of 23 nm in the presence of various amounts of iodomethane. Iodo- methane concentrations : A, 0; B, 0.045; C, 0.15; D. 0.60; and E, 1.20 M. The dotted spectrum is E plotted a t an increased sensitivity When many samples are presented for urgent examination, detailed Stern - Volmer analyses are unfortunately impracticable, but it is to be hoped that this situation will be remedied by the development of automated techniques.However, much useful information can be gained simply by comparison of changes that occur when samples are transferred from a non-quenching to a partially quenching solvent. From the point of view of repro- ducibility and manipulative convenience single solvents rather than mixtures are preferred ; but selective quenching effects due to varying quenching mechanisms are sometimes dis- played by solvent mixtures, so that their use in certain circumstances may be important. The effects of a selection of quenching and non-quenching solvents and mixtures on spectra, synchronously excited at intervals of 30 nm, from the previously used oil sample (Fig.2) are shown in Figs. 6 and 7. The relative emission intensities at 348 nm of these spectra and of various others, from aerated and de-aerated solutions, are shown in Table I. Spectra in polar and hydrocarbon solvents are generally similar. Of these, the best resolved, in acetonitrile, is shown in Fig. 6. The spectra of ethanol and cyclohexane solutions (not reproduced) are closely comparable, and differ significantly only in giving inferior resolution at 318 nm and a lower fluorescence signal. (Unfortunately, the rather low solubilities of hydrocarbons in acetonitrile limit the use of this otherwise suitable solvent.)734 LLOYD : PARTLY QUENCHED SYNCHRONOUSLY EXCITED FLUORESCENCE [Analyst, Vol. 99 [Q]/mol I-' Fig.5. Stern - Volmer graphs from the soot spectra of Fig. 4, and from others. The lines are numbered according to emission wavelengths (nm) In the chlorinated solvents emission intensities are reduced, as is to be expected from the heavy-atom solvent effect, according to the number of chlorine atoms present. Individual emissions vary in sensitivity to this effect, so that new features become apparent in the Thus, in dichloromethane, emission intensities in the region 330 to 370nm are 300 350 400 450 500 550 Wavelengthh m Fig. 6. Synchronously excited (30 nm) spectra of a used oil (0.2 mg rnl-l) in: A, acetonitrile; B, dichloromethane; C, chloroform ; and D, carbon tetrachloride. Spectrum E is the unused oil in carbon tetrachloride a t the same sensitivity as D.Sensitivities of A to D are varied arbi- trarily ; relative fluorescence yields are given in Table I 350 400 450 500 550 6 0 Wave I engt h/n m Fig. 7. Synchronously excited spectra (30 nm) of a used oil (0.2 mg ml-1) in cyclohexane containing : A, iodomethane, 0-17 M, aerated; B, as A but de-aerated; C, nitromethane, 0.017 M, aerated; and D, as C but de-aerated. The spectra are a t arbitrary sensitivities. Relative in- tensities are given in Table I. The nitro- methane spectra are subject to a 5 per cent. reduction in intensity in the 360-nm region because of absorption by quencherNovember, 19741 EMISSION SPECTRA IN THE CHARACTERISATION OF COMPLEX MIXTURES 735 modified relative to one another with the result that a new emission at 361 nm appears, and the weak 380-nm emission is quenched.In chloroform, extensive quenching of the region below 360 nm yields a spectrum dominated by a previously unsuspected emission at 373 nm, and by the now relatively strong, short-lived, polynuclear hydrocarbon emissions in the visible region. TABLE I VARIATION IN EMISSION INTENSITY (348 nm) OF A MINERAL OIL IN VARIOUS SOLVENTS AND MIXTURES, AERATED AND DE-AERATED Relative Relative Solvent and saturating gas intensity* Solvent and saturating gas intensity* ' * loo Carbon tetrachloride .. . . 47 .. .. 127 Methyl iodide (0.17 . . . . 44 in cyclohexane . . .. 111 Bromoform (0-2 mol I-') nitrogen . . { air . . air . . .. . . . . 63 in cyclohexane .. 83 Nitromethane (0.017 moll-1) .. 57 in cyclohexane Dichloromethane { i:pgen : : .. .. 19 .. .. 12 * Uncorrected for variations in refractive index. 0.2 0.2 22 19 8 8 40 29 Carbon tetrachloride quenches the whole of the ultraviolet emission and only much reduced polynuclear hydrocarbon emissions remain. These compounds accumulate in auto- mobile engine oils with use (6000 miles in the present instance), hence, in this solvent is displayed almost exclusively part of the contribution made by a mechanical environment to the oil. For comparison, the spectrum of the unused oil in carbon tetrachloride is included in Fig. 6. Emissions from anthracene are suppressed in the carbon tetrachloride spectra, the photochemical processes undergone by anthracene in carbon tetrachloride with con- siderable efficiency13 probably being responsible for this effect.In the presence of air, fluorescence intensities are reduced substantially (30 to 50 per cent.) by oxygen quenching, but changes in the appearance of spectra are usually small except for the complete disappearance of the 380-nm emission. However, a striking differential effect is observed when solutions quenched with nitromethane and iodomethane are compared. Fig. 7 shows the effect of oxygen quenching on solutions of used oil in cyclohexane quenched with either nitromethane or iodomethane. In the presence of the latter de-aeration slightly increases the 412-nm emission relative to the remainder. The spectrum of the aerated nitromethane solution is qualitatively comparable except for an enhancement at 350 nm and a reduction at 405 to 430 nm.On de-aeration the over-all emission is slightly increased, and new, resolved features appear at 405 nm (weakly) and 420 nm (strongly) that are not seen in the broad emission from iodomethane-containing solutions in this region. Evidently, the new emissions are relatively poorly quenched by nitromethane, and efficiently by iodomethane and by oxygen : a reflection of the sensitivity of quenching to the structure of both fluorescent and quencher molecules, particularly when the quenching processes are varied in mechanism as in the present instance. Long-lived emissions of polynuclear fluoranthenes may be responsible for this effect. Sawicki, Stanley and Elbert14 found that fluoranthenic hydrocarbon emissions are relatively weakly quenched in aerated nitromethane.From the few results a~ailable,~ diffusion-controlled oxygen quenching of this class of compounds will occur with a typical spread of lifetimes. The effects of partial quenching on the synchronously excited spectra of some other types of lubricating oil are shown in Figs. 8 and 9. In the naphthenic oil (Fig. €9, transfer from cyclohexane to chloroform results in a reduction in the major emission at 337 nrn of736 LLOYD : PARTLY QUENCHED SYNCHRONOUSLY EXCITED FLUORESCENCE [Analyst VOl. 99 about 90 per cent., which exposes new emissions at 350 and 360 nm, relatively intensifies that 383 nm and entirely quenches the 318-nm emission. In Fig. 9, the spectra of two oils, which, in cyclohexane, are doubtfully distinguishable from one another, are compared; in chloroform, however, they are clearly distinguished by new emissions revealed in the 425-nm region, prominently in one sample, and by the persistence of the 370-nm emission in the I I I I I 300 350 400 450 500 Wavelengthhm Fig.8. Synchronously excited spectra (30 nm) of a naphthenic oil Wavelengthhm (i mg ml-I) in: A, cyclohexane; and Fig. 9. Synchronously excited B, chloroform. B is a t ten times spectra (30 nm) of solutions greater sensitivity than A (0.2 mg ml-l) of gear oils. A and C are from one sample in cyclohexane and chloroform, respectively; B and D are from a sccond sample High-octane petrols contain substantial levels of polynuclear hydrocarbon~,~*4 which are revealed effectively by partial quenching in chloroform.In Fig. 10, the spectrum of a 98-octane fuel in cyclohexane is largely that of strong emissions in the region characteristic of mono- and dinuclear compounds. These emissions are quenched in chloroform, thus enabling comparisons to be made between the polynuclear hydrocarbon components of different petrols, as shown. Creosotes contain a wide range of fluorescent materials such as heterocyclic compounds and phenols, as well as aromatic hydrocarbons, which vary between different samples. In cyclohexane, emission is due mainly to the last of these groups. An example is shown in Fig. 11, in which emissions, synchronously excited at 30 nm, characteristic of pyrenes, anthracenes and benzo [alpyrene are prominent. In chloroform, these emissions and a broad complex envelope in the 340-nm region are suppressed relative to an emission at 374 nm, which results in highly distinctive spectra from trace amounts of this type of material.The basic fraction, which includes quinolines, isoquinolines, acridines and carbazoles, and their benzohomologues, which are all fluorescent in varying degrees, exhibits useful quenching effects. The transfer from cyclohexane to chloroform of a fraction containing creosote bases (Fig. 12) demonstrates the short-lived nature of the emission a t 332 nm and provides increased differentiation in the 400-nm region. In dilute aqueous acidic solution only three emissions are apparent, but more detail is obtained when potassium bromide is added. At a concentration of 0.1 mol l-l, the main emission at 370 nm is reduced approximately thirty-fold, relative intensities are considerably changed and the initial broad emission at 455 nm splits into two.These spectra are excited at intervals of 20 nm. A considerable variety of other components in this and other fractions become apparent at other excitation intervals.November, 19741 EMISSION SPECTRA IN THE CHARACTERISATION OF COMPLEX MIXTURES 737 300 350 400 450 500 Wavelengthhm Fig. 10. Synchronously excited spectra (30nm) of three 98-octane petrols, 5 p1 ml-l in chloroform (A, B and C) and cyclohexane (D). Spectrum D, the off-scale part of which is represented by the dotted line, is from the same sample as spectrum C. D is plotted a t a fifth of the sensitivity of the remainder Wavelength/nm Fig.11. Creosote (25 p g ml-l) ir cyclohexane (A) and chloroform (B). synchronously excited a t an interval of 30 nm. An approximate ten-fold increase in sensitivity was used in the case of B 350 400 450 500 Wave1 engt h/n m Fig. 12. Creosote bases: (a), in cyclohexane (dotted line) and chloroform (full line) ; and (b), in aqueous 0.1 N H,SO, solution with (full line) and without (dotted line) 0.1 M KBr. The quenched spectra were run at arbitrarily increased sensitivities738 LLOYD CONCLU s I ON There are many other examples of complex materials to which the described techniques can be applied; and many different quenching agents, such as those of Sawicki, Stanley and Johnson,15 which could be used to provide increased differentiation. Indeed, but for the limitations imposed by the range of lifetimes in any particular instance, and by the extent to which spectrophotometer sensitivities can accommodate the reduced emission intensities from quenched solutions, the number of possible permutations is considerable. As fluorescence techniques are often the only techniques that are applicable when trace amounts of material are available for comparison, quenching effects in conjunction with synchronous excitation clearly increase the number of distinguishing features potentially available and, it is expected, will increase the range of materials characterised on those occasions when only trace amounts remain. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Lloyd, J . B. F., Nature, Phys. Sci., 1971, 231, 64. -- , , J . Forens. Sci. S O ~ . , 1971, 11, 83. , Ibid., 1971, 11, 153. , Ibid., 1971, 11, 235. , I -- -- Birks, J. B., “Photophysics of Aromatic Molecules,’’ Wiley Interscience, New York and London, Parker, C. A., “Photoluminescence of Solutions,” Elsevier, Amsterdam, 1968, p. 460. Sawicki, E., T a h n t a , 1969, 16, 1231; and references cited therein. Zander, M., 2. analyt. Chem., 1967, 229, 352. Kasha, M., J . Chem. Phys., 1952, 20, 71. Zander, M., “Phosphorimetry,” Academic Press, New York and London, 1968. Zander, M., 2. analyt. Chem., 1973, 263, 19. Parker, C. A., op. cit., p . 458. Bowen, E. J., and Rohatgi, K. K., Discuss. Faraday Soc., 1953, 14, 146. Sawicki, E., Stanley, T. W., and Elbert, W. C., Talanta, 1964, 11, 1433. Sawicki, E., Stanley, T. W., and Johnson, H., Mikrochim. Acta, 1965, 1, 11. 1970, Chapter 4. Received May 21sf, 1974 Accepted June 24th, 1974

ISSN:0003-2654    

DOI:10.1039/AN9749900729

出版商:RSC    

年代:1974

数据来源: RSC