ISSN:0003-2654    

DOI:10.1039/AN97398FX033

出版商:RSC    

年代:1973

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97398BX035

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

i V THE ANALYST [September, 1973THE ANALYSTE D IT0 RIAL AD VI SO RY BOA R DChairman: H. J. Cluley (Wembley)*L. S. Bark (Salford)R. Belcher (Birmingham)L. J. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)E. A. M. F. Dahmen (The Netherlands)*J. B. Dawson (feeds)A. C. Docherty (BillinghamjD. Dyrssen (Sweden)*W. T. Elwell (Birmingham)*D. C. Garratt (London)*R. Goulden (Sittingbourne)J. Hoste (Belgium)D. N. Hume (U.S.A.)H. M. N. H. Irving (Leeds)A. G. Jones (Welwyn Garden City)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. I. Rees (LondonjE.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 SUBSCRIBERSSubscriptions for The Analyst, Analytical Abstracts and Proceedings should beThe Chemical Society, Publications Sales OfFice,Blackhorse Road, Letchworth, Herts.Rates for 1973(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 .. . . . . . . . . . . . . €45.00(other than Members of the Society)sent to:The 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) . . . . . . . . . . . . . . . . . . f35.00( 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 alonevi SUMMARIES OF PAPERS I N THIS ISSUE [September, 1973Summaries of Papers in this IssueMass and Charge Transfer Kinetics and Coulometric CurrentEfficienciesPart IX. An Examination of the Titanium(1V) - Titanium(II1)System and the Effects of Ultratrace Impurities in Sulphuric AcidTitanium (111) has found considerable use as a coulometric intermediate,and is the strongest reductant that can be generated with good efficiencyin aqueous media.Atsuch high concentrations, a surfactant impurity becomes adsorbed on theworking electrode and very seriously reduces the speed of the charge-transferprocess and therefore the current efficiency. A method of purification oftlie sulpliuric acid by electrosorption is shown to give a charge-transfer rateconstant in excess of 1 s-l, but charcoal column purification iswithout effect. Both 7 and 10 M sulpliuric acid media were examined, butbecause tlie mass-transfer rate is decreased by the higher viscosity of the10 M acid there is little difference in current efficiency.Addition of iron(II1)or iron(I1) hinders the adsorption of the impurity and the deactivation ofthe electrode, but offers no special benefit. Electrode kinetic parameters arereported for various treated and untreated media, and current efficienciesfor the generation of titanium (111) are computed. The behaviour in perchloricacid is re-interpreted.E. BISHOP and P. H. HITCHCOCKChemistry Department, University of Exeter, Stocker Road, Exeter, EX4 4QD.Analyst, 1973, 98, 625-634.The medium is restricted to strong sulphuric acid.Mass and Charge Transfer Kinetics and Coulometric CurrentEfficienciesat Platinum and Gold ElectrodesPart X.An Examination of the Tin(1V) - Tin(I1) - Tin(0) SystemsTin(I1) is more readily handled coulometrically than volumetrically, andthe electrode processes are critically evaluated. Zero-current tin (IV) - tin(I1)potentials cahnot be measured or derived from Lewartowicz plots, and charge-transfer kinetic parameters must therefore be referred to 0 V. Gold cathodesare deactivated towards reduction of hydrogen ion by specific adsorptionof bromide ion, and give well separated waves for the reduction of tin(1V)to tin(T1) and to tin metal, with limiting currents proportional to the tin(1V)concentration in 3-0 M bromide plus 0.4 M perchloric or hydrobromic acidmedium. Mass and charge transfer kinetic parameters are derived from thevoltamniogranis. Platinum cathodes become filmed and theories of thenature of the film are reviewed in the light of new work, which shows thatthe film is difficult to remove and was not completely removed in earlier work.'The generation of tin(I1) in bromide medium is superimposed on the filni-suppressed hydrogen-ion wave.Chloride media are unsuitable. Kineticparameters must be derived from anodic voltammograms of tin(II), whichare uncomplicated in the bromide medium. At low bromide concentrations,the platinum anode is filmed with a species such as [SnBr,OHI2-, which blockstlie reaction. The electrode mechanism is discussed. Charge-transfer para-meters vary with potential or current, but synthesised computer-plottedvoltammograms give a good fit with experimental curves.From the kineticparameters the current efficiency for tin(I1) generation is computed, andfor 0.2 M tin(1V) the efficiency is better than 99.98 per cent. over the currentdensity range 100 to 300 mA cm-2 but decreases rapidly outside this range.Sample concentrations must be chosen so as to maintain the intermediatecurrent within range throughout the determination, and platinum cathodesmust be filmed quickly, but not aged for too long.E. BISHOP and P. H. HITCHCOCKChemistry Department, University of Exeter, Stocker Road, Exeter, EX4 4QD.Analyst, 1973, 98, 635-646viii THE ANALYST [Scptcmbcr, 1973What'sfastautomatic and hoes60 proteinhydrolysate runs at weekends. ?and time wasting.protein hydrolysate run takes less thanWith Chromaspek, a singlefazest amino acid analyser in the business.Chromaspek is so sensitive that you can quantitate500 picomole peaks.The sample disc accommodates 60 separatesamples and will analyse each in successionautomatically, without attention, and record the results.So you can set up Chromaspekon Friday, let it workallweekend, and check the finished results on Mondaymorn i ng.And Chromaspek is so simple. It uses only2 buffer solutions and no near-neutral pH solutions.Sothere's no need for refrigeration, virtually no risk ofhow it performs.Why not send the coupon and ask us to fix a date?1 ------------Please contact me toarrangea demonstration of Chromaspek here in 1 my laboratory A1619 I Name I Position I Organisation I A d d r e s sIIII - I I II Telephone No. IANALYTICAL DIVISIONWestwood Industrial Estate, Ramsgate Road Margate Kent Telephone Margatc 2426

ISSN:0003-2654    

DOI:10.1039/AN97398FP097

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

September, 19731 THE ANALYST ixCLASSIFIED ADVERTISEMENTSThe rate f o r classified advertisements i s 35p a line (or s$weequivalent of a line) with an extra charge of l o p for theuse of a Rox Number. Semi-displayed classifiedadvertisements aye f;4 for szngle-column inch.Copy for classified advertisements required not later thaRthe 18th of the month preceding date of Publication whichis on the 16th 01 each month. Advertasements should beaddressed to J Arthur Cook 9 Lloyd Square, London,W C l X 9BA. ‘Tel.: 01-837 k31.5WANTEDLiterature on mctnllurgical analysis, 19th arid early 20th century,particularly iron and steel. W. Is. Clark?, 1 A1cestc.r Roxl, Stiidlcy,Warks. D80 7AN.Senior ControlAnalystA vacancy has arisen ir? the Analytical ControlLaboratory of Geigy Pharmaceuticals for theposition of Senior Control Analyst.The scope of the position embraces the dailycontrol of all Analytical matters emanating fromthe entire Control Section of the AnalyticalLaboratory. Reportability will be directly to theHead of the Analytical Department andresponsibility will be fQr the most effectiveprofessional use of four sections of theLaboratory, each separately controlled a tGraduate, or equivalent, level, assessing qualityof raw materials, bulk and finished goods againstapproved specifications and for specialinstrumentational analysis.Basic product scopeis a comprehensive range of excipients andactive ingredients, together with compoundedPharmaceutical dosage forms.The requirements are for a mature, well-qualifiedGraduate, probably of a chemical orpharmaceutical discipline, who has been exposedto a wide range of pharmaceutical analyticaltechniques, probable age range 25/35.Thecandidate will have several years‘ experience inthe Pharmaceutical Analytical area, with a desireto exploit potential for further career advancement,particularly in a more administrative direction.Attractive conditions of service, which includean excellent contributory pension scheme, areconsistent with the international reputation ofthe Company and assistance will be given withrelocation expenses where appropriate.Please apply in writing giving brief details of your age, qualifications and experience to:The Personnel Manager, GEIGY PHARMACEUTICALS, Hurdsfield Industrial Estate, Macclesfield S K l O 2LYPlease mentionI THE ANALYSTwhen replying to advertisementsBINDINGHave your back numbers of The Analystbound in the standard binding case.Send the parts and the appropriateindex(es) together with a remittancefor f2.40 to :HEFFER’S PRINTERS LIMITEDCAMBRIDGE, ENGLANSUMMARIES OF PAPERS I N THIS ISSUE [September, 1973The Determination of VoIatile Metal Chelates by Using aMicrowave- excited Emissive DetectorA microwave-excited emissive detector operated a t atmospheric pressurehas been used in conjunction with gas chromatography in order to achievethe separation and determination of various metal chelates of acetylacetoneand trifluoroacetylacetone.The operating parameters have been optimisedand the limits of detection and degrees of selectivity evaluated.The proposedmethod is both selective and very sensitive and can also confirm the presenceof a particular metal.R. M. DAGNALL,Biochemistry Division, Huntingdon Research Centre, Huntingdon, PE18 6ES.T. S. WEST and P. WHITEHEADChemistry Department, Imperial College of Science and Technology, London,SW7 2AY.Analyst, 1973, 98, 647-654.The Determination of Trace Amounts of Barium in CalciumCarbonate by Atomic-absorption SpectrophotometryBarium present a t the 1 to 20 p.p.m. level in calcium carbonate isseparated from the calcium matrix by co-precipitation on a lead sulphatecarrier. The lead sulphate - barium sulphate precipitate is dissolved in anammoniacal solution of ethylenediaminetetraacetic acid disodium salt andthe barium in the resulting solution is determined by atomic-absorptionspectrophotometry by using a nitrous oxide - acetylene flame and the atomicline a t 553.6 nm.The method ovcrcomcs interferences that occur in the atoiiiic-absorpti~)nspectropl-iotometric determination of barium in the presence of calcium.F.J. BAN0John & E. Sturge Limited, Lifford Chemical Works, Kings Norton, Birmingham,B30 3JW.Analyst, 1973, 98, 655-658.Spectrophotometric and Chelatometric Determination of Iron(II1)with 3- Hydroxypyridine-2- thiol3-Hydroxypyridine-2-thiol can be satisfactorily used in the spectroplioto-metric and chelatometric determination of iron(lT1). With the prescribedmethods the determination can be made over a wide pH range withoutinterference from many anions and cations. Structures are suggested forthe complexes formed.MOHAN KATYALSt.Stephen’s College, Delhi-7, India.Miss VEENA KUSHWAHA and R. P. SINGHDepartment of Chemistry, Delhi University, Delhi-7, India.Analyst, 1973, 98, 659-662.Assay of Micro-scale Amounts of Hydroperoxide and of Iodinein Aqueous Non-ionic Surfactant Solutions by aSpectrophotometric MethodAn iodimetric - ultraviolet spectrophotometric method has been developedfor the quantitative determination of trace concentrations of hydroperoxide(down to Itinvolves a simple extrapolation procedure based on the kinetics of iodinefading in these systems. This method was also applicable to macro-scaleamounts when the classical titrimetric method could not be used.Hydrogenperoxide and hydroperoxides of some polyethylene glycols, in the presence, of Cetomacrogol 1000, gave reproducible results that were in good agreementwith the results obtained by use of the titrimetric method.E. AZAZ, M. DONBROW and R. HAMBURGERPharmacy Department, School of Pharmacy, Hebrew Tlniversity of Jerusalem,Jerusalem, P.O.B. 12065, Israel.M) in the presence of etheric non-ionic surface-active agents.Analyst, 1973, 98, 663-672xii SUMMARIES OF PAPERS I N THIS ISSUEObservations on the Use of 2,4,6-Trinitrobenzenesulphonic Acidfor the Determination of Available Lysine in AnimalProtein ConcentratesThe determination of available lysine in proteins by their reactions with2,4,6-trinitrobenzenesulphonic acid has been examined in relation to thepreparation of the sample, the conditions of the reaction, the hydrolysis of thetrinitrophenylated protein and the effects of other amino-acids and relatedcompounds.From these observations a revised method is proposed for therapid routine screening of animal protein concentrates used in animal feeds.With the use of DL-lysine as a standard instead of etrinitrophenyllysine andwith a simplified hydrolysis procedure, the technique is suitable for examiningcomparatively large numbers of samples. Interference by other amino com-pounds is slight, except that by cadaverine, hydroxylysine and ornithine.Results are given for the levels of available lysine in a range of animalproteins, which compare closely with values obtained by the Carpenterprocedure in which l-fluoro-2,4-dinitrobenzene is used.R.J. HALL, N. TRINDER and D. I. GIVENSMinistry of Agriculture, Fisheries and Food ; Agricultural, Development and AdvisoryService, Government Buildings, Kenton Bar, Newcastle upon Tyne, NE1 2YA.Analyst, 1973, 98, 673-686.'[September, 1973Quantitative Determination of the Enantiomeric Purity ofSynthetic PyrethroidsPart I. The Chrysanthemic Acid MoietyClirysanthemic acid is liberated by hydrolysis of the pyrcthroid underbasic conditions and is made to react with (+)-cc-methylbenzylamine via theacid chloride. Complete separation of diastereoisomeric amides derived fromthe (+) and (-) forms of both cis- and tvans-chrysanthemic acids can beobtained by gas chromatography on a 50-m x 0-25-mm glass capillarycolumn coated with FFAP (free fatty acid phase).No racemisation or kineticresolution takes place during the reaction sequences. The method is suitablefor the analysis of bioresmethrin, bioallethrin, tetraniethrin and otherinsecticidal chrysanthemates.F. E. RICKETTThe Cooper Technical Bureau, Uerkhamsted, Hertfordshire.Analyst, 1973, 98, 687-691.Carbon Tetrachloride as a Possible Source of Interference DuringFumigant Residue AnalysisThe contamination of acetone with an impurity corresponding to carbontctrachloride on two different gas-chromatographic columns is described. 1 thas been shown that this contamination could be caused by carbon tetra-chloride in the laboratory atmosphere, which possibly arises from the useof aerosol propellent cans for spraying thin-layer chromatograms.P.B. BAKER, J. E. FARROW and R. A. HOODLESSDepartment of Trade and Industry, Laboratory of the Government Chemist, CornwallHouse, Stamford Street, London, SE1 9NQ.Analyst, 1973, 98, 692-693.The Use of a Laser for Cutting Bone SamplesPrior to Chemical AnalysisShort PapevJ. S. HISLOP and A. PARKER,4nalytical Sciences Division, Atomic Energy Research Establishment, Harwell,Ridcot, Berkshire,Analyst, 1973, 98, 694September, 19731 THE ANALYST xiiiTR AC E ELEMEN TAN ALY S I SP RO BLEMS?Activation Analysis could be the answer. Thistechnique offers the advantages of high sensitivityand specificity, good accuracy and precision, even atsub-p.p.m.concentrations. A bulk analysis isobtained, often non-destructively. The maindlsadvantage (the need for a nuclear reactor) can beovercome by making use of the Activation AnalysisService provided by the Universities Research Reactor,If you would I ike to discuss your particular problem,or would like to receive further details, please contact:Dr. G. R. Gilmore (Activation Analyst),Activation Analysis Service,Universities Research Reactor,Risley,Nr. Warrington,Lancs.Phone: Warr. 32680,33114BUREAU OF ANALYSEDSAMPLES LTD.announce the issue of the followingNEW SAMPLESChemical StandardsNo. 384 Hycomax I11 PermanentMagnet Alloy (34% Co,4% Ti, 1% Nb)No. 388 Zircon (66% ZrO,, 33% SO2)Spectroscopic StandardsNos.66 1-665 High Phosphorus Engin-eering Irons containingincrements of C, Si, Mn, Sand PFor further details please write to :-NEWHAM HALL, NEWBY,MIDDLESBROUGH, TEESSIDE, ENGLANDTS8 9EAor Telephone 0642 37216SPECIALIST ABSTRACTJOURNALSpublished bySCIENCE AND TECHNOLOGY AGENCYAtomic Absorption and FlameEmission Spectroscopy AbstractsVol. 5, 1973, bimonthly €30X-Ray Fluorescence SpectrometryAbstractsVol. 4, 1973, quarterly €28Thin-Layer Chromatography AbstractsVol. 3, 1973, bimonthly €28Gas Chromatography-MassSpectrometry AbstractsVol. 4, 1973, quarterly €37Nuclear Magnetic ResonanceSpectrometry AbstractsVol. 3, 1973, bimonthly €30Laser-Raman Spectroscopy AbstractsVol. 2, 1973, quarterly €30X-Ray Diffraction AbstractsVol.1-2,1973, quarterly €30Neutron Activation Analysis AbstractsVol. 2 3 , 1973, quarterly €30Electron Microscopy AbstractsVol. 1, 1973, quarterly €30Liquid Chromatography AbstractsVol. 1, 1973, quarterly €30Electron Spin Resonance SpectroscopyAbstractsVol. 1 , 1973, quarterly €30Sample copies on request from:SCIENCE AND TECHNOLOGY AGENCY,3 HARRINGTON ROAD,SOUTH KENSINGTON,LONDON, SW7 3ES01-584 808xiv THE ANALYST [September, 1973- ---- - - - -melting point determinedWhen accurate thermometric measurements are required, it is necessaryto calibrate thermometers or other thermal analysis instruments usingreference temperatures.Reference temperatures are usually the solid-liquid transition temperaturesof some organic substances of high purity.Thermometric standards are those substances which not only have meltingtemperatures perfectly known from the literature, but which guarantee becauseof their very high degree of purity, a consistent agreement with the meltingtemperature stated. Using the most advanced techniques of purification suchas zone melting, Carlo Etba has made available to scientists the followingproducts which can be regarded as primary thermometric standards:+Solid-liquid transition temperaturep-NITROTOLUENE 51.54 "CNAPHTHALENE 80.27PHENANTHRENE 99.90BENZOIC ACID 122.35DIPHENYLACETIC ACID 148.00ANlSlC ACID 182.98CHLOROANTHRAQUINONE 209.03CARBAZOLE 245.34ANTHRAQUINONE 284.59All products are available in 2 g. vials.CCARLO ERBA DlVlSlONE CHIMICA I VIA CARL0 IMBONATI 24 I 20159 MILAN

ISSN:0003-2654    

DOI:10.1039/AN97398BP103

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

SEPTEMBER, 1973 THE ANALYST Vol. 98, No. 11 70 Mass and Charge Transfer Kinetics and Coulometric Current Efficiencies Part IX.* An Examination of the Titanium(1V) - Titanium(II1) System and the Effects of Ultratrace Impurities in Sulphuric Acid BY E. BISHOP AND P. H. HITCHCOCK-/- (Chemistry Depavtment, University of Exeter, Stocker Road, Exeter, EX4 4QD) Titanium(II1) has found considerable use as a coulometric intermediate, and is the strongest reductant that can be generated with good efficiency in aqueous media. At such high concentrations, a surfactant impurity becomes adsorbed on the working electrode and very seriously reduces the speed of the charge-transfer process and therefore the current efficiency. A method of purification of the sulphuric acid by electrosorption is shown to give a charge-transfer rate constant in excess of 10-5 1 cm-2 s-l, but charcoal column purification is without effect.Both 7 and 10 M sulphuric acid media were examined, but because the mass-transfer rate is decreased by the higher viscosity of the 10 M acid there is little difference in current efficiency. Addition of iron(II1) or iron(I1) hinders the adsorption of the impurity and the deactivation of the electrode, but offers no special benefit. Electrode kinetic parameters are reported for various treated and untreated media, and current efficiencies for the generation of titanium(II1) are computed. The behaviour in perchloric acid is re-interpreted. The medium is restricted to strong sulphuric acid. TITANIUM(III) was first introduced as a coulometric aniperostatic intermediate by Arthur and Donahuel in 1952.It was one of the earliest intermediates to receive scientific rather than empirical e~aluation,~J and, although the literature is not e~tensive,l-~~ titanium(II1) shares with iron(I1) the greatest number of applications among reductant intermediates. On this account, and because it is the most powerful reductant that can be satisfactorily generated in aqueous media, a thorough investigation has been made both of the electrode kinetics and generation current efficiency of the titanium(1V) - titanium(II1) system and of the genera- tion medium. The system has been used in the determination of iron(II1) d i r e ~ t l y , ~ v ~ - ~ s ~ s ~ ~ and of other materials via iron (I1 I) ,8316917 9 21 ruthenium( IV) ,24 925 iridium (IV) , 25 molybdenum( VI) ,4 s5 s2' uranium(V1) ,11914926 phosphate via molybdenum(V1) ,13 chromium(V1) in standardisation,20 vanadium(V),8~11s15~21 titanium metal and its compounds,15~16*21 selenium( IV),4 tell~rium(IV),~s~ tellurium(VI),G 9-quinonedioximeg and organic dyestuff s.17~~8 Gold plated on platinum was originally used as the electrode material1 ; copper and copper amalgam,23 titanium metal,15 and mercury14J9,20 have been used, but platinum is the most popular material.3~6~10~11~15-17~21 Hydrochloric acid,l~59~~ citric acid,l4 and admixtures of hydrofluoric7 or phosphoric22 acid with sulphuric acid have been used as the medium, but sulphuric acid at high concentra- tion2~3~6,8~10,11~26 has met with most approval.Lingane and Kennedy2 found the charge-transfer process to be slow at the dropping- mercury electrode in perchloric acid of all concentrations and in dilute sulphuric and hydro- chloric acids, as well as in a 4.0 M solution of sodium sulphate in 0.2 M sulphuric acid, but fast in a 4.0 M solution of sodium hydrogen sulphate in 0.2 M sulphuric acid and in high concentrations of sulphuric and hydrochloric acids.The fast charge transfer was ascribed to complexation (ligand exchange with water molecules), and specifically to hydrogen sulphate ions in sulphuric acid media. The process is assisted also by the increase in conditional potential of the system in such media. Habashy28y29 confirmed these findings, but found that the titanium(1V) wave was not well separated from the hydrogen-ion wave at platinum. The conditional potential shifts faster than the background reduction wave with complexation, Present address: Ever Ready Co.(G.B.) Ltd., Central Research Laboratory, St. Ann's Road, * For Part VIII of this series, see p. 563; for Part X, see p. 635. London, N15 3TJ. @ SAC and the authors. 625626 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, Vol. 98 and so high sulphuric or hydrochloric acid concentrations are required. Even so, the titanium and hydrogen-ion waves merge to the extent of leaving the limiting current plateau. ill defined, so that high titanium(1V) concentrations are necessary in order to attain an acceptable current efficiency. Phosphoric acid fails to retain titanium( IV) in solution at concentrations above 0.01 M, and so is useless.If strong oxidants such as cerium(1V) or manganese(VI1) are to be determined, they will attack concentrated hydrochloric acid. The choice is therefore restricted to sulphuric acid media. EXPERIMENTAL Apparatus, reagents and general procedures have been described earlier.30s31 The only reasonably pure compound that is readily available is AnalaR potassium oxodioxalatotitanate(1V) ; other materials are liable to contain appreciable amounts of iron, silicon and vanadium among their impurities, and these three elements are particularly difficult to remove, so making the materials that contain them unsuitable for use in kinetic studies. A high-purity, but inert, titanium(1V) oxide was kindly supplied by Laporte Indus- tries Ltd., who quoted levels of iron(II1) of less than 10 p.p,m.and vanadium(V) of less than 5 p.p.m.; it is believed that the material was prepared by vapour-phase hydrolysis of purified tetra-n-butyl titanate(1V). Examination proved that the two compounds mentioned were of a satisfactory purity for use as starting materials. Titanium(1V) chloride-The titanium( IV) oxide was too inert to be dissolved in sulphuric acid, and so was converted into titanium(1V) chloride by Schreyer's method,32 which involves reaction with ccacca'a'or'-hexachloro-m-xylene (Fluka A.G.) . The product was distilled twice before further use. TitaniGm oxodi$wrchZorate-The titanium(1V) chloride was heated with 62 per cent. perchloric acid at a mole ratio of 1 : 2 at 30 to 40 "C and 15 mm of mercury so as to remove hydrogen chloride, and the product recrystallised from the minimum amount of 10 per cent.perchloric acid. The product is reported33 to be TiO(ClO,),.xH,O, where x = 6 to 7. Alterna- tively, hydrated titanium( IV) oxide was precipitated by hydrolysis of titanium(1V) chloride or potassium oxodioxalatotitanate(1V) by adding 50 per cent. ammonia solution to a saturated solution of the titanium salt. The hydrated titanium(1V) oxide was readily separated by centrifugation, and was washed with water until free from ammonia, then dissolved in 62 per cent. perchloric acid. This solution was used directly, but the product could be crystallised out by evaporation at 30 to 40 "C and 15 mm of mercury. Titarti~m(1V) su@hate-Freshly prepared hydrated titanium( IV) oxide was dissolved in concentrated sulphuric acid and the solution diluted to give a 0-5 M concentration of titan- ium(1V) in 10 M sulphuric acid.The solution was standardised by reduction with liquid zinc amalgam, separating the amalgam, running the reduced solution into an excess of iron(II1) solution and titrating the iron( 11) produced against standard potassium dichromate with sodium diphenylamine-4-sulphonate as indicator, or against potassium permanganate with a correction being made for the blank. Titanizcm(II1) s~l$kate-Titanium(III) sulphate was made by determinate reduction of standardised titanium(1V) sulphate with liquid zinc amalgam, and was stored over the liquid amalgam. PRE-TREATMENT OF THE WORKING ELECTRODE- Normally, treatment (c) (see Part VIIsl) was used: immerse the platinum electrode in fresh aqua regia at 50 to 60 "C for 2 minutes, anodise it in concentrated hydrochloric acid for 30 s at 100 mA cm-2, cathodise it in 0.1 M sulphuric acid for 5 minutes at 100 mA cm-2 and wash it thoroughly with water.PURIFICATION OF ELECTROLYTES- (a) Electrolytic method-A portion of 7 or 10 M sulphuric acid was prepared by dilution of concentrated AnalaR or Aristar sulphuric acid in an all-glass cell that carried a clamped cover bearing ground-glass joints lubricated with distilled water. Through the joints were inserted the electrodes, a nitrogen bubbler [the nitrogen was purified by passage through chromium(I1) chloride solution and water] and a siphon tube. The solution was stirred PREPARATION OF TITANIUM COMPOUNDS-September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIRNCIES. PART IX 627 magnetically, and the working cathode was a 100-cm2 piece of platinum black.Electrolysis was conducted at 0.3 V until the current, which was initially about 50 mA, decreased to about 0.1 mA. The time required was variable, but was often as long as 6 hours. The solution was then transferred into the clean voltammetric cell by closing the nitrogen outlet, thus causing the siphon to fill. The first few portions, breaking the syphon through a tap at its highest point, were used to rinse out the cell and then discarded. (b) Column adsorption-Sugar charcoal was prepared by the action of concentrated sulphuric acid on domestic granulated sugar. The washed product was ground in a mortar and packed into a glass column to a depth of 30 cm.The column was mounted over the voltammetric cell, and 7 or 10 M sulphuric acid allowed to percolate at about 5 ml min-l through the column into the cell. A second column filled with commercial granular activated charcoal was also used. RESULTS AND DISCUSSION Before the work had progressed very far, it became evident that the working platinum electrode was being deactivated by an impurity in the system, the presence and identity of which could not be detected other than by its effect in deactivating the electrodes. It was traced to the sulphuric acid, which is used at high concentrations, and was present, although at a much lower level, in Aristar as well as in AnalaR sulphuric acid. No mention of this was made in earlier work, and, while electrode pre-treatments were given attention, no report of purification of the electrolyte has appeared.It must be assumed, therefore, that all earlier work was carried out with untreated supporting electrolytes, and it is appro- priate to discuss the behaviour of the system in such conditions. UNTREATED ANALAR SULPHURIC ACID MEDIA- A treated electrode in 7.0 M sulphuric acid gave curve 1 in Fig. 1, with a suspicion of an adsorption pre-wave. A re-treated electrode gave curve 2 in Fig. 1 after addition of a mixture of titanium(1V) and titanium(II1). Repeated scanning without intervening pre- Electrode potential versus S.H.E./V Fig. 1. Voltammograms of titanium(1V) - titanium(II1) in untreated 7.0 M sulphuric acid: 1, 7-0 M sulphuric acid, freshly activated electrode; 2, 7.0 M sulphuric acid + 2.39 x lo-$ M titanium(II1) + 1.91 x M titanium(IV), freshly activated electrode; and 3, as for curve 2 but after completing seven cycles of cathodic and anodic scanning.Ramp speed 70 mV min-l628 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, VOl. 98 treatment of the electrode showed a drastic reduction in charge-transfer speed, and after six cycles gave the curves 3 in Fig. 1; further scans produced no further change. Removal of the electrode and re-treatment before scanning gave a curve identical with curve 2. The change is therefore in the electrode and not in the solution. The same de-activation occurred when a freshly activated electrode was immersed in the titanium - sulphuric acid solution on open circuit.Deactivation is not therefore due to the passage of current. Some activity could be restored by immersing the electrode for 25 minutes in 1.0 Rii sulphuric acid, but full activation required the electrochemical treatment. At least part of the deactivation is due to adsorption and not to recrystallisation of the platinum surface. Repetitive anodisa- tion and cathodisation [method (b), see Part VI131] at 100 mA c M 2 in 1.0 M sulphuric acid produced full activity as well as the stripping method given above; there was no difference between a cathodic and an anodic finish, because any oxide that remained on the platinum surface would immediately be reduced chemically on immersion in solutions that contain titanium(II1). The hysteresis between the forward and reverse voltage sweeps in curves 2 and 3 is also symptomatic of ad~orption.~~ Adsorption slows the titanium charge-transfer process, moves the titanium reduction wave so that it merges with the hydrogen-ion wave and seriously attenuates the current efficiency for the generation of titanium( 111).Adsorption deactivation probably accounts for the poor definition of the titanium waves recorded by Lingane and Kennedy3; their polarographic waves2 were not affected in this way because the electrode surf ace was continuously being renewed. PURIFIED SULPHURIC ACID MEDIA- With a freshly pre-treated electrode, very little hysteresis occurred in electrolytically puri- fied sulphuric acid, and the curves did not change appreciably when scans were repeated several times over a period of several hours without intervening re-activation of the electrode.The adsorbable impurities were therefore present in the sulphuric acid or the water30 that was used for dilution. The water was ruled out by purification of concentrated sulphuric acid by method (a) and then dilution with untreated water, when curves as good as those in purified 7-0 M sulphuric acid were obtained. -0.2 1 I I I I I 1 I I 0 5 0.4 0.3 0.2 0.1 Electrode potential versus S.H.E./V Fig. 2. Voltammograms of titanium(1V) - titanium(II1) in 7.0 M sulphuric acid purified by electrosorption: 1, treated 7.0 M sulphuric acid; 2, treated 7-0 M sulphuric acid + 1.0 x M titanium(1V); and 3, treated 7.0 M sulphuric acid + 1.08 x M titanium(1V) + 1.02 x M titanium(II1).Electrode activated before each cycle. Ramp speed 70 mV min-'September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES. PART I X 629 Column treatment failed to remove the surfactant impurity, and contaminated the solution with ionic impurities in one instance. The method was therefore abandoned. Fresh stock titanium(1V) and titanium(II1) solutions in electrolytically purified sulphuric acid were prepared. Fig. 2 shows scans of purified 7.0 M sulphuric acid (curve 1) with a very low background current over a large range of potentials, and of titanium(1V) alone (curve 2) and mixed with titanium(II1) (curve 3). Curves 2 and 3 show remarkably little hysteresis, which is a good indication that the platinum surface did not change appreciably over the potential range scanned.Repeated scans showed excellent reproducibility (better than 1 per cent.) for both titanium and background waves. Similar behaviour occurred in purified 1 0 ~ sulphuric acid, as shown in Fig. 3. The curves in Figs. 1 to 3 were analysed for the kinetic parameters listed in Table I by the diffusion-corrected Tafel plots of Lewartowic~~~ and also, for Fig. 1, by pattern theory.35 The difference between the treated and untreated supporting electrolyte is dramatic. 0 5 0.4 0.3 0 2 0.1 Electrode potential versus S.H.E./V Fig. 3. Voltammogram of titanium(1V) - titanium(II1) in 10.0 M sulphuric acid purified by electrosorption: 1, treated 10.0 M sulphuric acid; and 2, treated 10.0 M sulphuric acid + 2.1 x M titanium(1V) + 3.4 x 10-3 M titanium(II1).Electrode activated before each scan. Ramp speed 70 mV min-l When a titanium solution in purified sulphuric acid was left in the cell for 1 day, some gain in impurity occurred, as shown by increased hysteresis and slowing of the charge-transfer process. Repetition of the experiments in a completely sealed all-glass double cell showed detectable accession of impurities in 1 day. In view of other finding~,~'39~7 desorption of some substance from glass or platinum may be the source of the returning impurity, although the possibility that a trace amount of oil was not removed from the cylinder nitrogen in the scrubbers cannot be excluded. After several experiments involving contamination, the 1.0 M sulphuric acid used in the anodisation - cathodisation activation of the electrode was found to lose some effectiveness, in contrast to a fresh electrolyte.Thereafter, the electrolyte was discarded after each activation by either method. A solution of 10 M sulphuric acid was prepared by dilution of Aristar sulphuric acid taken from a freshly opened bottle. Background and titanium(1V) reduction waves are shownin Fig. 4. There was some hysteresis and repeated scans, without reactivating the electrode, showed a slowing down of the charge-transfer process, but full activity was restored by the normal electrode treatment methods. The screw-cap of the bottle was sealed with white wax and, although every care was taken in manipulation, it is possible that a trace amount of wax might have contaminated the acid. When purified by the electrosorption method, Aristar sulphuric acid gave the same behaviour as purified AnalaR sulphuric acid, and even in all-glass apparatus re-contamination was detected after about 20 hours.630 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, VOl.98 TABLE I KINETIC PARAMETERS FOR THE TITANIUM(IV) - TITANIUM(III) SYSTEM AT ACTIVATED Titanium concentration range 0.15 x to 1.3 x M PLATINUM ELECTRODES kmass ox/ kmass red/ Medium K / 1 cm-2 s-l a B E,’/V 1 cm-2 s-l 1 cm-z s-l 7-0 M H2S0,, pre- 2 1-4 x 0-58 to 0-69 0.57 to 0.65 0.215 1.9 X 1-9 X 10.0 M H,SO,, pre- 2 1.1 x 0.58 to 0.66 0.57 to 0.63 0-307 1-05 x loAs 0-99 x lop6 electrolysed electrol ysed treated (b) 10.1 x lo-* 0.615 7-0 M H,SO,, un- (u) 6-7 x lo-* 0.728 ::ii8} 0.215 * 1.9 x 10-6 7.0 M H,SO,pZus iron, (c) >, 0.56 0.57 0.215 1.9 x 1.9 x [Fe]/[Ti] = 1 ( d ) 5 x 10-7 0.6 1 0.46 0.215 * 1.9 x 10-6 (a) Aged electrode, negative-going scan.(b) Aged electrode, positive-going scan. (c) (d) Electrode immersed in solution for 30 minutes after activation. Electrode activated, used several times, then left in solution for 12 hours. * Limiting current region ill defined. It is notable that while the impurities exerted a gross effect on the shape of the titanium wave, very little change was observed in the hydrogen-ion reduction wave. UNTREATED SULPHURIC ACID MEDIA CONTAINING ADDED IRON- Voltammetric scans of a mixture of iron and titanium in untreated sulphuric acid showed that the iron mitigated the effect of impurities on the titanium wave. Fig. 5 shows the behaviour of an approximately equimolar solution of titanium and iron in the oxidation states 0.5 0.4 0.3 0 2 0.1 Electrode potential versus S.H.E./V Fig.4. Voltammograms of titanium(1V) in 10.0 M untreated Aristar sulphuric acid: 1, 10.0 M Aristar sulphuric acid, electrode freshly activated; 2, 10.0 M Aristar sulphuric acid + 2.3 x M titanium(IV), electrode used to record three cycles of background, then used for this scan without reactivation ; 3, recorded immediately after curve 2 without electrode activation; and 4, elec- trode activated, conditions as for curve 2. This figure is an actual copy of an original X - Y recorder chart, showing the noise typical of such plots Ramp speed 70 mV min-l.September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES.PART I X 63 I TiIv, FeIII and FeII. A freshly activated electrode used for three scans without reactivation gave titanium(1V) reduction waves that differed by only 5 per cent.; the iron(II1) wave was consistent within 1-5 per cent. The hysteresis was much less than in the absence of iron. After leaving the electrode in the solution on open circuit for 12 hours, curve 2 in Fig. 5 was recorded. Had there been no iron present, the curve would have been similar to curve 3 in Fig. 1. If the mole ratio of iron to titanium decreased much below 0-5, the deactivation of the electrode accelerated. At a constant mole ratio of iron to titanium, deactivation of the electrodes was retarded to the same extent whether the solution contained iron(I1) iron(II1) and titanium(IV), or iron(II), titanium(1V) and titanium(II1).The effect of adding iron was not, therefore, that of increasing the zero-current potential in the medium to a more positive potential at which adsorption did not occur. If an electrode that had been deactivated by exposure to an impure titanium - sulphuric acid solution was transferred to an equimolar titanium - iron solution, the observed rate of charge transfer was very markedly decreased. It would therefore seem that the function of the iron is to slow down the adsorption of impurities on the electrode, and not to speed up the reduction of titanium(1V) at a fouled electrode. Electrode potential versus S.H.E./V Fig. 6. Voltammograms of titanium(1V) and iron(II1) - iron(I1) in untreated 7.0 M sulphuric acid: 1, 7.0 M AnalaR sulphuric acid + 10.1 x M titaniuin(1V) + 4-5 X M iron(II1) + 3.4 x 10-3 M iron(II), activated electrode; and 2, electrode left in solution for 12 hours on open circuit, then this curve recorded without further activation PERCHLORIC ACID MEDIA- As perchlorate is the least cornplexing of anions, it was hoped to be able to determine the characteristics of the purely solvated oxotitanium(1V) and titanium(II1) ions in perchloric acid.Scans were performed in 1, 3, 7 and 10 M perchloric acid solutions of oxotitanium(1V) perchlorate, but no titanium reduction wave was found. The experiments were repeated in perchloric acid purified by the electrosorption process, and with electrodes pretreated by all meth0ds,~1 but still no wave was obtained. This result seemed to support Lingane and Kennedy’s contention2 that the titanium couple was “completely irreversible” at all con- centrations of perchloric acid.More exactly, it indicated that the conditional potential of the purely solvated ions and the charge-transfer rate parameters were such that hydrogen ion was reduced in preference to titanium(1V). However, treatment of oxotitanium(1V) perchlorate in perchloric acid with liquid zinc amalgam showed a reddish violet coloured layer in contact with the amalgam, which dis- appeared on shaking, and tests showed that the solution, which was initially free from632 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [ArtdySt, Vol. 98 chloride, now contained chloride. The titanium(II1) was reacting directly with perchlorate and reducing it to chloride.Titanium(II1) is therefore unstable in perchloric acid and so anodic scanning of such solutions would give misleading results that represent the sum of a homogeneous and a heterogeneous reaction. The homogeneous kinetics were later investi- gated,38 and prompted mechanistic di~coveries.~g THE ELECTRODE KINETIC PARAMETERS- Voltammetry in sulphuric acid purified by electrosorption showed that the limiting cathodic current was proportional to the titanium(1V) concentration and the anodic limiting current was proportional to the titanium( 111) concentration. Essin40 found the same relation- ship at a mercury-jet electrode, but Khomutov and Aboimov41 reported that the limiting cathodic current was not proportional to the titanium(1V) concentration, but they did not purify the electrolyte and it seems likely that electrode poisoning was responsible.Delahay and Tra~htenberg~~ studied the system at a mercury electrode and examined the effects of adsorption of cyclohexanol, n-pentanol and thymol on the charge-transfer rate in 1.0 M tartaric acid. They concluded that a monolayer of these substances on the electrode decreased the value of k from 5 x 1 cm-2 s-l at a clean electrode by several orders of magnitude, although there was little effect on the charge-transfer coefficients. This conclusion agrees with the results in Table I. The Lewartowicz plots for purified sulphuric acid give exchange currents that are very close to the limiting currents, which can therefore be regarded as being least possible values.The value of the charge-transfer rate constant, k , must also be regarded as being a least possible value. However, even if the true rate constant were considerably higher than 10-5 1 cm-2 s-1, very little change in the computed voltammograms would occur, because this value gives voltammograms that deviate only slightly from the infinitely fast mass transfer controlled curves. The tabulated values can be used with confidence in computing current efficiencies. Essin40 found that the sum of the charge-transfer coefficients, u + 8, was less than unity. In pure sulphuric acid, although hysteresis is slight, the sum is greater than unity. This may be because the generated species is not TiIIIL6, but a binuclear complex with a charge-transfer spectrum,39 which could be I-- l t 5 - 2 n L -1 or, as the hydrogen sulphate ion appearslto play a key r61e,2 83-33 In both formulations, an electron-tunnelling mechanism can be expected.THE GENERATION CURRENT EFFICIENCY FOR TITANIUM (111)- The background, reduction of hydrogen ion, rate parameters appear to be little affected by the condition of the electrode, and computer curve-fitting methods3O gave the values in Table 11. From these and the values in Table I, voltammograms were computed by VOLTAM- METRY 9,43 and gave an excellent fit with the experimental curves. There is therefore no potential dependence in k or a. Current efficiency curves were computed for 0.6 M titanium(1V)September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES. PART IX TABLE I1 RATE PARAMETERS FOR HYDROGEN EVOLUTION Medium [H3O+I /M k/l(a+aE) mol-(z+aH)Cm-2 s-1 a 7~ H,SO,, untreated .. .. 7 1.85 x 10-lo 0.5 7~ H,SO,, pre-electrolysed . . 7 1.85 x 0.5 1 0 ~ H,S04, pre-electrolysed . . 10 1.85 x 0.5 633 in 7.0 M sulphuric acid.3 Curves 1 and 2 in Fig. 6 show the situation for untreated and purified sulphuric acid, respectively. For a current efficiency loss of 1 p.p.m., the maximum usable current density predicted is 9.5 x A cm-2 for untreated and 5.92 x 10v2 A cm-2 for puri- fied sulphuric acid, and for a loss of 0.1 per cent. the values are 1.94 x and 0.109 A cM2, respectively. Lingane and Kennedy's value for a 0.1 per cent. loss in current efficiency was 3 x 10-3 A cm-2. However, they took no steps to purify the sulphuric acid, and experiment shows that different batches of sulphuric acid have different surfactant impurity levels.I I I I I I I I 0.02 , 0.06 0.10 0-14 0.18 Current density/A cm-* Fig. 6. Computed loss of current efficiency in the generation of titanium(II1) : 1, untreated 7.0 M sulphuric acid medium; 2, 7.0 M sulphuric acid cleaned by electrosorption; and 3, 10.0 M sulphuric acid cleaned by electrosorption. Parameter values as in Tables I and 11, with Sx = cm, [TiIvl~ = 0.6 M and [TP]B = 0 ; a for curve 1 = 0.56, otherwise 0.58 The generation efficiency in clean 10 M sulphuric acid is shown by curve 3 in Fig. 6. This is of interest, because the conditional potential in this medium (Table I) is more positive by 56 mV, yet the computed current efficiencies are about the same for 7 and 10 M sulphuric acid, because the mass-transfer rate constants are decreased by about 50 per cent.by the higher viscosity of the 10 M sulphuric acid and the effects nullify each other. It does appear that the addition of iron is of some use, and might be applied in the determination of vana- dium(V) by adding iron(III), or uranium(V1) by adding i r ~ n ( I I ) . ~ ~ CONCLUSIONS Voltammetry at platinum electrodes has shown that the purest sulphuric acids contain enough adsorbable impurities at the high concentrations used to reduce the charge-transfer rate of the titanium(1V) - titanium(II1) system by several orders of magnitude compared with the values in sulphuric acid purified by electrosorption. Such impurities seriously impair the current efficiency of the generation of titanium(II1). In perchloric acid, the immediately following chemical step of reduction of perchlorate to chloride makes the generation of titanium(II1) impossible in this medium.The non-unity value of a + /3 encourages the view that a binuclear complex of titanium(1V) and titanium(II1) is the electrode reaction product.634 BISHOP AND HITCHCOCK purity 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. We are deeply grateful to Imperial Chemical Industries Limited for a grant of research funds extending over 3 years. We also thank Laporte Industries Ltd. for the gift of high- titanium (IV) oxide. REFERENCES Arthur, P., and Donahue, J. F., Analyt.Chem., 1952, 24, 1612. Lingane, J . J., and Kennedy, J. H., Analytica Chim. Acta, 1956, 15, 294. __- , Ibid., 1956, 15, 465. Agasyan, L. B., Nikolaeva, E. R., and Agasyan, P. K., Zh. Analit. Khim., 1967, 22, 904. Binder, E., Goldstein, G., Lagrange, I?., and Schwing, J.-P., Bull. SOC. Chim. Fr., 1965, 2807. Chavdarova, R., and Sheltanov, Ch., C. R. Acad. Bulg. Sci., 1967, 20, 565. Clemency, C. V., and Hagner, A. F., Analyt. Chem., 1961, 33, 888. &ta, F., and Ji;, Y . C., Rev. ChirPz. Acad. Pop. Rom., 1962, 7, 123. Dobychin, S. I., and Kozulya, A. P., Zh. Analit. Khim., 1962, 17, 148. Hitchcock, P. H., Ph.D. Thesis, University of Exeter, 1969. Kennedy, J. H., and Lingane, J . J., Analytica Chim. Acta, 1958, 18, 240. KuCerovskS;, Z., PPibyl, M., and SiSka, M., Chem.Listy, 1965, 59, 604. Lagrange, P., and Schwing, J., Bull. SOC. Chim. Fr., 1965, 2811. Lingane, J . J., and Iwamoto, R. T., Analytica Chim. Acta, 1955, 13, 465. Malmstadt, H. V., and Roberts, C . B., Analyt. Chem., 1955, 27, 741. -- , Ibid., 1956, 28, 1884. Nikolaeva, E. R., Agasyan, P. K., Terenova, K. Kh., and Boikova, S. I., Vest. Mosk. GO~. Univ., Papier, J., Hkvue MLtall., Paris, 1954, 51, 723. Parsons, J . S., and Seaman, W., Analyt. Chem., 1955, 27, 210. Roberts, C. B., Diss. Abstr., 1956, 16, 1798. Sheitanov, Ch., Chavdarova, R., and Konstantinova, M., C. R. Acad. Bulg. Sci., 1966, 19, 1147. SlovAk, Z., and Pfibyl, M., 2. analyt. Chem., 1967, 228, 266. Stenina, N. I., and Agasyan, P. K., Zh. Analit. Khim., 1966, 21, 965. Stenina, N. I., Agasyan, P. K., and Berentsveig, G. A., Ibid., 1967, 22, 91. Takeuchi, T., Yoshimori, T., and Kato, T., Bunseki Kagaku, 1963, 12, 840. Yen, H.-Y., and Liu, Y. H., K’o Hsueh T’ung Pao, 1966, 17, 279. Habashy, G. M., Colln Czech. Chem. Commun., 1960, 25, 3166. -, 2. anorg. Chew., 1960, 306, 312. Bishop, E., and Hitchcock, P. H., Analyst, 1973, 98, 465. Schreyer, R. C., J . Amer. Chem. S O ~ . , 1958, 80, 3483. Krishnan, V., and Patel, C. C., Chem. & Ind., 1961, 321. James, S. D., Electrochim. Acta, 1967, 12, 939. Bishop, E., Analyst, 1972, 97, 761. Bockris, J. O’M., and Huq, A. K. M. S., Proc. 22. SOC., A , 1956, 237, 277. Warner, T. B., Schuldiner, S., and Piersma, B. J., J . Electrochem. Soc., 1967, 114, 1120. Bishop, E., and Evans, N., Talanta, 1970, 17, 1125. East, G. A., and Bishop, E., Proc. SOC. Analyt. Chem., 1972, 9, 186. Essin, 0. A., Acta Phys.-chim. URSS, 1940, 13, 429. Khomutov, N. E., and Aboimov, M. A., Trudy Mosk. Khim.-tekhnol. Inst., 1961, 32, 156. Delahay, P., and Trachtenberg, I., J - Anzer. Chem. Soc., 1958, 80, 2094. Bishop, E., Chemia Analit., 1972, 17, 511. Karp, S., and Meites, L., J . Amer. Chem. SOC., 1962, 84, 906. NOTE-References 30, 31, 36 and 43 are to Parts V, VII, 111 and I of this series, respectively. J , Ibid., 1956, 28, 1412. -- Khim., 1968, 23, 73. , , Ibid., 1973, 98, 553. -- Received February lst, 1973 Accepted ALIarck 27th, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800625

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, September, 1973, Vol. 98, Pp. 635-646 635 Mass and Charge Transfer Kinetics and Coulometric Current Efficiencies Part X.* An Examination of the Tin(1V) - Tin(I1) - Tin(0) Systems at Platinum and Gold Electrodes? BY E. BISHOP AND P. H. HITCHCOCK: (Chemistry Department, University of Exeter, Stocker Road, Excter, EX4 4QD) Tin(I1) is more readily handled coulometrically than volumetrically, and the electrode processes are critically evaluated. Zero-current tin(1V) - tin(I1) potentials cannot be measured or derived from Lewartowicz plots, and charge- transfer kinetic parameters must therefore be referred to 0 V. Gold cathodes are deactivated towards reduction of hydrogen ion by specific adsorption of bromide ion, and give well separated waves for the reduction of tin(1V) to tin(I1) and to tin metal, with limiting currents proportional to the tin(1V) concentration in 3.0 M bromide @us 0.4 M perchloric or hydrobromic acid medium.Mass and charge transfer kinetic parameters are derived from the voltammograms. Platinum cathodes become filmed and theories of the nature of the film are reviewed in the light of new work, which shows that the film is difficult to remove and was not completely removed in earlier work. The generation of tin(I1) in bromide medium is superimposed on the film- suppressed hydrogen-ion wave. Chloride media are unsuitable. Kinetic parameters must be derived from anodic voltammograms of tin(II), which are uncomplicated in the bromide medium. At low bromide concentrations, the platinum anode is filmed with a species such as [SnBr,OHI2-, which blocks the reaction.The electrode mechanism is discussed. Charge-transfer para- meters vary with potential or current, but synthesised computer-plotted voltammograms give a good fit with experimental curves. From the kinetic parameters the current efficiency for tin(I1) generation is computed, and for 0-2 M tin(1V) the efficiency is better than 99.98 per cent. over the current density range 100 to 300 mA cm-2 but decreases rapidly outside this range. Sample concentrations must be chosen so as to maintain the intermediate current within range throughout the determination, and platinum cathodes must be filmed quickly, but not aged for too long. TIN(II) is a well known volumetric reductant, but is infrequently used because it reacts with atmospheric oxygen, is prone to hydrolysis and its reaction kinetics lead to induced reactions.These objections do not arise if the tin(I1) is electrogenerated in an oxygen-free medium. Although it is the third most popular reductant intermediate, the electrode processes are not straightforward and the applications are not numerous. After its first introduction by Bard and Lingane,l generation of tin(I1) has been used in determinations of platinum(1V) ,2 g ~ l d ( I I f ) , ~ copper(II),4 iodine,5-7 br0mine,~7' ceri~m(IV),~,~ vanadi~rn(V),~ selenium(IV)8-10 and tel- lurium(IV).lO Little is known of the kinetics and mechanism of the reaction at solid electrodes. An empirical investigation of the current efficiency in bromide medial showed a current efficiency loss of 0.3 per cent., and in chloride media7 of 7 per cent.Early work2-4 indicated that platinum or gold electrodes in acidic halide media were most suitable. Platinum cathodes became filmed,l,ll whereby hydrogen-ion reduction is suppressed, but other reductions, e.g., of tin(1V) and antimony(III), continue with little change. No such film has been reported for gold electrodes. The current efficiency for the generation of tin(I1) appears to be better in bromide than in chloride media. The contention that this effect occurs because bromide complexes tin(1V) more strongly than chloride and thus makes the reduction potential more positive is neither logical nor supported by the extremely meagre data.12 Chloride-free media * For Part IX of this series, see p.625. t Presented a t the International Symposium on Analytical Chemistry, Birmingham, 1969. Present address: Ever Ready Co. (G.B.) Ltd., Central Research Laboratories, St. Ann's Road, London, N15 3TJ. @ SAC and the authors.636 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, Vol. 98 have not been investigated, and specific adsorption of bromide ion is probably the relevant factor. Finally, it can be demonstrated from his own resultsll that Bard did not always succeed in removing the film from his electrodes in experiments claimed to have been made with “clean electrodes.” More and better quantitative information on the fundamentals of the reaction is needed before it can be regarded as viable and useful. EXPERIMENTAL General apparatus, reagents and procedures have been described and water has been defined earlier.13 All potentials are quoted versus the standard hydrogen electrode (S.H.E.) unless otherwise specified.REAGENTS- Tin(l1) chloride, 0-05 M solution in 2.0 M hydrochloric acid-A sample of granulated tin was dissolved in hydrochloric acid, and the solution was diluted as required with de-oxy- genated water and hydrochloric acid. The solution was stored under scrubbed nitrogen in the reservoir bottle of an automatic zero burette (Baird and Tatlock Ltd., B 41/0500), modified so as to exclude air from the system. The tin(I1) content was determined by iodate titration in the presence of excess of iodide with the aid of fresh starch solution as indicator. The titration was performed under nitrogen in a beaker fitted with a machined Perspex cover. The approximate acid concentration was determined by titration with standard sodium carbonate solution, with the aid of screened methyl orange as indicator. Tin(11) bromide, 0-05 M solution in 2.0 M hydrobromic acid-With the substitution of hydrobromic for hydrochloric acid, preparation, storage and standardisation were the same as for the previous solution.Tin(1V) chloride, 0.05 M solution in 2.0 M hydrochloric acid-Tin(1V) chloride was prepared by the action of cylinder chlorine on heated tin, and the product was re-distilled and dissolved in 2.0 M hydrochloric acid so as to give the required solution. The tin(1V) content was determined by reduction to tin(I1) with zinc amalgam, and the tin(I1) titrated under nitrogen with standardised potassium permanganate solution. Tin(1V) byomide, 0.05 M solation in 2.0 M hydrobromic acid-A sample of granulated tin was refluxed with hydrobromic acid until the reaction ceased.A slight excess of bromine was added, and the excess removed by prolonged boiling. The solution was then diluted so as to give the required composition and standardised in the same way as the tin(1V) chloride solution. ELECTRODE ACTIVATION- Gold-wire electrodes-These electrodes were activated by two methods : (a) the electrode was etched in fresh aqua regia at 60 “C for 15 s and thoroughly washed; (b) the electrode was etched as in method (a), then anodised in 0.1 M sulphuric acid at 100 mA cm-2 for 5 s so as to produce a red oxide film on the electrode, which was then reduced to gold black by cathodisation for 10 s at 100mAcm-2 in the same medium.The two treatments gave essentially the same results. Platinum electrodes-These electrodes were activated by either method 1 or method 2, described below, or by method 1 followed by method 2. The latter treatment involving both methods was found to be essential for filmed electrodes. (1) The electrode was immersed in fresh aqua regia at 60 “C for 2 minutes, washed, anodised in concentrated hydrochloric acid for 30 s at 100 mA cm-2, washed, cathodised in 0-1 M sulphuric acid at 100 mA cm-2 for 10 minutes, and finally washed. (2) Alternate anodisation and cathodisation at 100 to 300 mA cm-2 was carried out in %OM sulphuric acid for 2 to 5 minutes each. The cycle was repeated at least once more and the electrode washed.In all instances the electrode was held potentiostatically at 0.4 V for 5 minutes in the supporting electrolyte so as to remove adsorbed hydrogen. RESULTS AND DISCUSSION ZERO-CURRENT AND CONDITIONAL POTENTIALS I N BROMIDE MEDIA- It proved impossible to measure the zero-current potential of either a gold or a platinum electrode treated by any of the activation processes in a mixture of tin(1V) and tin(I1) inSeptember, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES. PART x 637 any medium. Potentials drifted randomly at 3 mV min-l or above, and no two solutions or two electrodes gave potentials with satisfactory agreement. Evidently, the exchange current of the system is very low and it is possible that the electrodes become deactivated.This deactivation is not unexpected in view of similar findings in chloride media.l4,l5 Tafel and Lewartowicz plots are not helpful, which means that neither Ei nor E,, can be used as the reference potential in the derivation of charge-transfer parameters, and the conditional over-all rate constant, k , must therefore be referred to the zero point of the potential scale, that of the S.H.E., and will be annotated as kg (cathodic) and kz (anodic) in order to emphasise the distinction. Electrode potential versus S.H.E./V Fig. 1. Reduction of the supporting electrolyte a t a gold cathode: 1, 3.0 M sodium bromide, 0.38 M perchloric acid, fresh electrode, ramp speed -885 mV min-l; 2, same electrolyte, aged electrode, ramp speed -885 mV min-1; 3, same electrolyte, initially fresh electrode, manual plotting, duration 3.5 hours ; and 4, 0-38 M perchloric acid alone, fresh electrode, ramp speed -85 mV rnin-' CATHODIC VOLTAMMETRY AT GOLD ELECTRODES- It should be emphasised that earlier work was not carried out in entirely chloride-free media; solutions of tin chloride in bromide media were used.Media of 3 . 0 ~ bromide in 0.4 M hydrobromic acid or 0.4 M perchloric acid were used in this work and both media gave identical results. These supporting electrolytes gave the cathodic curves shown in Fig. 1. Curve 1 was obtained with a freshly activated electrode. After recording the curve, the electrode was returned to its starting potential and the scan repeated without intervening reactivation. Repetition of this process caused the scan to more to move negative potentials.If the electrode was then left on opencircuit in the solution, it recovered some of its activity and gave a scan at less negative potentials than the previous scan. The degree of recovery depended on the rest period, but the initial activity as in curve 1 was restored only on reactivation by method (a). Curve 2 in Fig. 1 was obtained with an electrode that had been activated, used to record five consecutive scans and left on open circuit for 2 hours. Curve 3 represents essentially the quasi-equilibrium situation : each point was obtained by maintaining a constant current until the potential drift decreased to below 1 mV in 5 minutes. This curve took 3.5 hours to produce, whereas the automatic scans required 8 minutes at a ramp speed of 85 mV min-I.Redzcction of tin(W)-A portion of the solution of tin(1V) bromide in hydrobromic acid was added to the supporting electrolyte and the acidity of the solution brought back to its original value by the addition of sodium hydroxide solution. The curve in Fig. 2 was recorded and shows the reduction waves of tin(1V) to tin(II), of tin(I1) to tin metal and the beginning638 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, Vol. 98 of the hydrogen wave. The electrode cannot be regarded as being gold over the whole potential range scanned, because after the second wave it would be a tin electrode, and the hydrogen-ion wave would be correspondingly moved to a more negative potential. Curves 3 and 1 in Fig. P can be regarded as the extreme limits for hydrogen evolution at a gold electrode in this medium, and can be used to define the maximum and minimum current efficiencies for the generation of tin(I1) if hydrogen evolution is the only competitive reaction.This conclusion also assumes that the presence of tin species in the contact layer does not alter the double-layer structure sufficiently to affect hydrogen evolution, which appears to be so. Quantitative measurement of the limiting current showed it to be proportional to the tin(1V) concentration over the range lW4 to M, which would not be so if an appreciable amount of hydrogen were produced, as predicted by curve 1 in Fig. 1. The electrode is therefore deactivated towards hydrogen evolution, but the degree of deactivation is not measurable on account of the further reduction to tin metal.Hydrogen evolution in the supporting electrolyte alone is not straightforward: curve 4 in Fig. 1 shows the hydrogen wave at a fresh gold surface in 0.4 M perchloric acid in the absence of bromide, in contrast to curve 1. The difference arises from specific adsorption of bromide ion. BreiteP found that in perchloric acid, a bromide concentration of M was sufficient to give a monolayer of bromide on a platinum electrode, and this monolayer altered the shape of cyclic voltammo- grams. The change in shape of the hydrogen region became more pronounced as the bromide concentration increased. Strong adsorption of bromide ions on gold has been dem0n~trated.l~ Anodic oxidations in bromide media were not attempted with gold electrodes because of the possible complications of attack of the electrodes.I I I I I 0.0 -0.2 -0.4 -0.6 Electrode potential versus S. H. E ./V Fig. 2. Reduction of tin species in bromide media a t a gold Freshly activated electrode, 3.0 M sodium bromide + M tin(1V) bromide, ramp cathode. 0.38 M perchloric acid + 0.25 x speed -85 mV min-l Kinetic $umzrneters-The reaction is obviously slow, and the reverse reaction can be neglected. Lewartowicz plots corrected for mass transfe1-1~ are shown in Fig. 3, and the change of slope at -0.195 V suggests that a two-step mechanism obtains. As EA is not measurable, the rate constants are referred to 0 V, and the parameters are shown in Table I. Advanced pattern theory analysis is limited in accuracy only by the experimental errors at such rates,l8 and shows that both kg and cc are potential dependent.CATHODIC VOLTAMMETRY AT PLATINUM ELECTRODES- The nature of a ‘ffilmed electrode”-Bard and Linganel found that tin(I1) was generated with greater efficiency at a used electrode than at a freshly activated electrode. They found that this effect was due to an invisible film on the electrode, which inhibited the reduction of hydrogen ions. Chemical strippingll of the film and spectroscopic analysis demonstrated the presence of tin, and it was concluded that the film was a hydrated tin oxide, the oxidation state of which depended on the conditions of formation. Bardll proposed that the initial reduction of hydrogen ions at a clean platinum electrode decreased the hydrogen-ion con- centration at the electrode surface to such an extent that hydrolysis of the tin compoundsSeptember, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES.PART x 639 1 I I I -0.15 - 0.20 -0.25 Electrode potential versus S.H.E./V Fig. 3. Lewartowicz plots for reduction of tin(1V) a t a gold cathode in 3.0 M sodium bromide + 0.38 M perchloric acid. Freshly activated electrode, ramp speed -855 mV min-l. [ S nII] / 1 0 -3 [SnIv] 10 --3 Curve moll-1 mol 1-1 1 0.25 0.25 2 0.25 0-50 3 0.50 0.50 4 0.50 0-75 occurred and the colloidal product stuck to the electrode. Conceivably, at the instant of starting the current, the hydrogen-ion concentration could transiently decrease to such a degree before the diffusion layer is set up by mass transport; otherwise, the current would have to be virtually the limiting current for hydrogen-ion transport, which is very high and was not approached. The tin(1V) oxide is less soluble than tin(I1) oxide, but the kinetics of hydrolysis are suchlg that tin(I1) oxide would be precipitated first, which accords with Bard’s observation that tin( 11) had a greater suppressive effect on hydrogen-ion reduction than tin(1V).However, the rate of hydrolysis hardly seems to permit the formation of oxides in the millisecond or so that is required to set up the diffusion layer for hydrogen ions, and the currents were too low to offer a sustained decrease in hydrogen-ion concentration. Some sort of film, however unlikely the conditions seem to be, is certainly formed, but it is possible that it might be, contain, or cover tin metal.Moreover, the film is very difficult to remove. Treatment by method 2 was expected to be adequate20 but was not, neither was method 1 alone, but the combination of method 1 followed by method 2 was found to be entirely reliable. Soaking in aqua regia and nitric acid was the method of pre-treating electrodes used by Bard, and does not seem to have given a reproducible surface: voltammograms for the reduction of hydrogen ion at a “clean electrode” in 3 M sodium bromide - 0.4 M perchloric acid differ in two of his diagrams that were supposedly recorded under identical conditions. In Figs. 10 and 11 of his paper,ll the foot of the hydrogen-ion wave appears at -0.1 V veysus640 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, VOl.98 S.C.E. and there is no evidence of a limiting current at the highest current density, 2 mA cm-2, reached. Yet in Fig. 2 of the same paper, the hydrogen wave starts at -0.2 V veysus S.C.E. and reaches a limiting current at 2 mA cm-2. The conditions used.should give a limiting current density of 4000 mA cm-2 at a clean platinum electrode. Repetition of Bard’s work gave agreement with his Figs. 10 and 11. Results similar to his Fig. 2 were obtained if a filmed platinum electrode was not completely cleaned before use. Bard also reported chrono- potentiometric transition times for hydrogen ion of 400 to 1200 s at a “clean” electrode, but calculated a value of 2 x 105s. Convection disturbs chronopotentiograms of such duration, but generally lengthens transition times instead of shortening them, and the short transition times again suggest inadequate cleaning.TABLE I MASS AND CHARGE TRANSFER KINETIC PARAMETERS FOR THE TIN(IV) - TIN(II) SYSTEM OVER THE CONCENTRATION RANGE 10-4 TO 3 x 10-3 M Working elec- Medium trode k;/l cm-2 s-1 k0,/1 cm-2 s-1 a ,!I kmaSs/l cm-2 s-1 3.0 M NaBr + 0.4 M HClO, 3.0 M NaBr + 0.4 M HBr 3.0 M NaBr + 0.4 M HC10, Pt (0.28 to i Au 6.0) x 3.0 M NaBr + 0-4 M HBr 1.0 M NaBr + 0.4 M HClO, 0.33 M NaBr + 0.4 M HC10, Pt 5.4 x 0.33 M NaBr + 0.4 M HClO, 3.0 M NaBr + 4.0 M HC10, Pt 8.13 x Pt Pt 1-1 x lO-g* 1.0 x 10-lo* + 2.7~NaC10, (u) (3.1 t o (b) (1.3 to (OX) 10-5 7.3) X lo-* to 0-37 - - 0.30 (red) to 0.35 5.7 x - - 0.16 1.4 x - - 0.16 1.1 x - - 0.22 2.3 x * For current densities < 0-1 mA cm-a.(a) Low current region. (b) High current region. Layers of tin are both strongly held and difficult to remove. Bowles and Cranshaw21 investigated the formation of monolayers of tin on platinum black at 0.1 V in an acidic solution of tin-119. The tin produced a “good Mossbauer spectrum” and it was concluded that the tin was plated on to the platinum surface, and not adsorbed in the electrical double layer, which would not have given a “good Mossbauer spectrum.” The Mossbauer results indicated that the tin existed as a monolayer that closely resembled tin metal, despite the use of an electrode potential that was anodic to that required for the bulk deposition of tin metal. Such adsorption for cadmium(I1) and thallium(1) on platinum has been reported by Frumkh22 He concluded that chemisorption led to the formation of a surface compound or surface alloy.Such a layer cannot be considered to be a pure metal or a true alloy, so its chemical and physical properties cannot be predicted with certainty, and it is difficult to devise unequivocal tests of the theory. It seems possible that in some instances Bard removed the hydrous tin oxide from the electrodes but a layer of tin remained that partially blocked the platinum surface for hydrogen-ion reduction. Generation of tin(I1) at platinum electrodes-Fig. 4 shows voltammograms of the sup- porting electrolyte (curve 1) and of tin(1V) in 3.0 M bromide - 0.4 M perchloric acid. Curve 2 shows the suppressed hydrogen-ion wave followed by reduction of tin(1V) to tin(I1). The transition from hydrogen-ion reduction to tin(1V) reduction is not distinct, but appears to occur at about -0-08 V; the third wave in curve 2 is not hydrogen-ion reduction, but reduction of tin(I1) and tin(1V) to tin metal.Curve 4 shows that a higher tin(1V) concentration sup- presses the hydrogen-ion wave to the extent that it cannot be detected without increasing the sensitivity of the current axis of the X - Y recorder. Curve 5 is a reverse sweep recorded immediately after curve 2 ; the maximum at -0.305 V represents the oxidation of the tin metal film. The reduction curves of tin(1V) to tin(I1) are superimposed on the suppressed hydrogen-ion reduction wave, and lack a well defined limiting current region, so that any deductions made from them can be of only a qualitative nature. The height of the tin(1V)September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES.PART x 641 wave increases with increasing tin( IV) concentration, but there is no strict proportionality. An electrode covered with an aged film gave a smaller hydrogen-ion reduction wave than a freshly filmed electrode. In addition, an aged film suppressed the tin(1V) reduction wave, although no effect was observed on the height of the wave that represented the formation of tin metal. A supporting electrolyte that consisted of a 4.5 M solution of ammonium chloride in 0.2 M hydrochloric acid was examined. Tin( IV) chloride suppressed the hydrogen-ion wave, but the suppressed wave appeared as a maximum, as reported by Bard. Suppression of the hydrogen-ion wave was less in chloride medium than in bromide medium.The repro- ducibility of the voltammograms was poor and the charge-transfer process was slower in chloride media. Oxidation of tin(11) at platinum electrodes-Clearly, the cathodic voltammograms at platinum are useless for the determination of kinetic parameters, and so anodic voltammo- grams were obtained in the same medium. Again, no difference was found between hydro- bromic and perchloric acids in the 3 . 0 ~ bromide medium. Fig. 5 shows anodic scans of the supporting electrolyte, with the bromine wave at 0.75 V, and the oxidation of tin(I1) with a single well defined limiting current region. The full pre-treatment process was used for activating the electrode, but the anodic scans were reproducible whether or not the elec- trode was pre-treated.Tafel and diffusion-corrected Lewartowicz plots are shown in Fig. 6, and the kinetic parameters derived are included in Table I. Pattern theory1* is highly precise in the context and gives parameter values in good agreement. Computer-plotted voltammo- grams23 prepared from the experimental values gave an exact fit with the experimental curves. I I 0.0 - 0.25 - 0-5 Electrode potential versus S.H.E./V Fig. 4. Reduction of tin(1V) a t a platinum cathode in 3.0 M sodium bromide + 0.4 M perchloric acid. Electrodes freshly activated by methods 1 and 2, ramp speed -85 mV min-l (+85 mV min-1 for curve 5). Curves: 1, background electrolyte alone; 2, 0.25 x M tin(1V) bromide; 3, 0.50 x M tin(1V) bromide; 4, 5.0 x M tin(1V) bromide; and 5, reverse scan of curve 2 Variations in hvdropen-ion and bromide-ion c I 1 0.8 0.6 0.Electrode potential versus S.H.E./V Fig. 5. Oxidation of tin(I1) a t a plati- num anode. Electrode activated by methods 1 and 2, ramp speed + 85 mV min-1. Curves: 1, 3-0 M sodium bromide + 0.4 M perchloric acid; and 2, as for curve 1 with the addition of 0.48 x M tin(I1) and 1.29 x M tin(1V) as bromides concentrations and ionic strength were examined. An elecirolyre that consisted of a 3.0 M solution of sodium bromidewin 4.0 M perchloric acid gave curves with the same shape as those in Fig. 5 ; the kinetic parameters are included in Table I. A decrease in bromide-ion concentration, while maintaining the ionic strength constant with sodium perchlorate, eventually produced a maximum (Fig.7)642 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [AnaZyst, Vol. 98 I 1 I 0.6 0.5 0.4 Electrode potential versus S.H.E./V Fig. 6. Tafel and Lewartowicz plots for oxidation of tin(I1) at a platinum anode. Electrodes activated by methods 1 and 2, 3.0 M sodium bromide + 0.4 M perchloric acid, ramp speed +86 mV min-l. Curves: 1, Lewartowicz plot, 0.48 x 10-3 M tin(I1) + 1.29 x lW3 M tiri(1V); 2, Tafel plot of same scan; 3, Lewartowicz plot, 0-96 x M tin(I1) + 1.32 x M tin(1V); and 4, Lewartowicz plot, 0.96 x M tin (11) + 2.03 x M tin (IV) in place of the limiting current plateau of Fig. 5 . The same effect occurred when no sodium perchlorate was added; the kinetic parameters are given in Table I. The reverse scan in Fig. 7 also shows a maximum, absent in curve 1 in Fig.5 , but coincident with the maximum in the forward scan in Fig. 7. It is probable that at low bromide concentrations a layer of tin(1V) species, such as [SnBr5OHl2- or the dihydroxo complex, is formed on the surface of the electrode, so blocking the electrode surface and causing the current to decrease after the maximum in the anodic scan, and that this species is reduced on the reverse, cathodic, scan and gives the small maximum shown in Fig. 7 . KINETICS AND MECHANISM OF THE TIN(IV) - TIN(II) PROCESS- Only the reduction of tin(1V) at a gold cathode and the oxidation of tin(I1) at a platinum anode in bromide media give sufficiently well defined waves for the determination of kinetic parameters. Variations in the tin(I1) and tin(1V) concentrations showed that the cathodicSeptember, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES.PART x 643 current is proportional to the tin(1V) concentration and independent of the tin(I1) concen- tration, and that the anodic current is proportional to the tin(I1) concentration and inde- pendent of the tin(1V) concentration. The reverse direction in either reaction is therefore without influence and the charge-transfer process is slow. The limiting currents were proportional to the stirring speed.13 All of these results accord with the over-all reaction SnIv + 2e + SnII taking place in a single two-electron step, or two consecutive one-electron steps. An alterna- tive mechanism would be an E.C.E. process in which the chemical step is the disproportionation of tin(II1).Vetter2* contends that two one-electron steps occur in reduction at mercury in chloride media, but there is no experimental evidence for this proposal and similar contentions concerning the thallium( 111) - thallium( I) 25 and quinone - hydroquinone26 systems have been ~hallenged.~’-~O Vetter used two criteria for distinguishing two consecutive one-electron steps from a single two-electron step : first, extrapolations of the mass transfer corrected Tafel plots to zero overpotential do not give the same exchange current, and second, the sum of a + /3 is not unity. For a single two-electron step, the exchange currents are identical and cc + ,L? is unity. However, potential-dependent adsorption2’ can so alter the electrode surface that these criteria are unreliable ; moreover, the potential difference between anodic and cathodic scans for such a slow system is great enough to produce other changes in the electrode surface.For tin, Vetter’s first criterion cannot be applied because the zero-current potential is not measurable, and the second criterion cannot be used in isolation because cc and ,8 are very susceptible to changes in adsorption on the electrode surface. The changes in slope of the Lewartowicz plots in Fig. 3 are in some agreement with Hurd’s predictions31 on successive steps, but are more likely to indicate a change in specific adsorption. The available evidence is insufficient to support a definite conclusion. 0.9 0-7 0.5 0.3 Electrode potential versus S. H . E./V Fig. 7. Oxidation of tin(I1) at a platinum anode in dilute bromide medium.Anode freshly activated by methods 1 and 2, 0.33 M sodium bromide + 0.4 M perchloric acid + 2.7 M sodium perchlorate. Curves: 1, iorward scan at + S S mV min-l; and 2, immediately following reverse scan at -85 mV min-1 Chemical studies of homogeneous reaction k i n e t i c ~ ~ ~ , ~ ~ show that the tin system accom- modates itself to the mechanism of the other reactant, and is a one-electron system for “one-electron reactants” such as cerium(IV), iron(III), chromium(V1) or manganese(VII), but a two-electron system for “two-electron reactants” such as hydrogen peroxide, iodine, bromine, thallium(II1) or mercury(I1). This property, however, is not of much help in the heterogeneous electrode process.644 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, Vol.98 TABLE I1 KINETIC PARAMETERS FOR BACKGROUND REACTIONS System k ~ / l @ -k ffH)mol-(2 aH)cm-2 s-1 a Freshly activated gold electrode, reduction of 0.4 M per- Gold electrode, manual plot, reduction of 3.0 M sodium chloric acid .. .. .. .. .. .. 1.93 x 10-1l 0.208 bromide + 0.4 M perchloric acid . . .. .. 2-24 x 0.13 Freshly activated platinum electrode, oxj dati on of 3.0 M sodium bromide + 0.4 M perchloric acid . . .. (mass transfer controlled, E: = 1.065 V) APPLICATION OF EXPERIMENTALLY DETERMINED KINETIC PARAMETERS IN COMPUTER CALCU- Parameters for the background reactions were derived by the curve-fitting process,13 pattern theory not having been fully developed for backgrounds at the time.34 The results, which were later confirmed by pattern theory, are shown in Table 11.The computer in use at the time did not have the capacity for an ecological matrix of k , cc and Ewe, and so LATIONS- 1 I 1 0.0 - 0.25 - 0.50 Electrode potential versus S.H.E./V Fig. 8. Computed voltammograms, reduction of tin(1V) a t gold cathode in 3.0 M sodium bromide + 0.4 M perchloric acid. [SnIv] = 2-5 x 10-4 M, [SnIIJ = 0. Curves A1 and A2, k H = 2.237 X 10-l1, aH = 0.133; B1 and B2, AH = 5.11 x 10-13, aH = 0.11; A1 and B1, Roc = 5.1 x lo-@, a = 0.435; A2 and B2, Roc = 34' x a = 0.348. A, Experimental pointsSeptember, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES. PART x 645 two voltammetric curves were plotted by using low and high current values for k and a for the tin(1V) - tin(I1) system, and the experimental points are shown superimposed on the computer plots in Fig.8. The effect of superimposition of hydrogen-ion reduction on the generation current efficiency for tin(I1) is shown in Fig. 9, again with two plots for low and high current values of the charge-transfer parameters being used.The results are in fair agreement with the experimental values of Bard and Lingane,l who claimed a current efficiency of 99.7 -+ 0-2 per cent. for current densities ranging from 10 to 84 mA cm-2. The computed efficiency loss is 120 to 500 p.p.m. in the same range. The computed current efficiency shows a sharp decrease at current densities below 5 mA cm-2, again confirming the earlier observa- ti0n.l It is likely that the background parameters used in the current efficiency computation underestimate the loss.but the commtation does show the maximum efficiencv that can be attained and the range of current ldensities that are available. J Fig. 9. Computed current efficiency curves for the generation of tin(I1) at a gold cathode. Parameters as given for curves A1 and A2 in Fig. 8; [SnIVJ = 0.2 M CONCLUSIONS Values of the mass and charge transfer kinetic parameters for the tin(1V) - tin(1I) and background systems have been determined within the limits of reproducibility of the voltam- mograms. Current efficiencies computed on this basis are in agreement with earlier work,l but show that in order to attain a current efficiency loss of about 100 p.p.m., the current density for reduction of tin must be maintained between 100 and 300 mA cm-2 for a 0.2 M solution of tin(1V) bromide in 3.0 M sodium bromide plus 0.4 M hydrobromic or perchloric acid.This conclusion means that the sample size and concentration must be such, and the generation current density must be so chosen, that the intermediate current density never falls below 100 mA cm-2 at the beginning of the determination or exceeds 300 mA cm-2 at the end. This constraint may necessitate very short increments at high currents in approaching the end-point. This procedure is now p0ssible.~5 The essential factor for platinum cathodes is to form the film as fast as possible, to maintain it throughout the determination, but not to allow it to age so much that the reduction of tin(1V) is affected. A special cleaning process is essential for the removal of the films.The behaviour of gold electrodes in bromide media is ascribed to specific adsorption of bromide ions, although this conclusion is not amenable to absolute proof. Platinum646 BISHOP AND HITCHCOCK electrodes in tin halide solutions do acquire a film, which is attributed by some workers1 to hydrous tin oxides and others22 to alloy-like compounds of platinum and tin metal, and in the present work the evidence favours a film of tin metal covered with a partially hydrolysed tin halide. The significant point is that the films do form and do block or partially block certain electrode reactions while permitting other electrode reactions to proceed with little or no inter- ference. This effect does increase the generation current efficiency of tin(II), but severely complicates any fundamental investigation of mechanisms.Similarly, the specific adsorption of bromide on gold electrodes hinders mechanistic investigation. We are deeply grateful to Imperial Chemical Industries Limited for the provision of research funds over a period of 3 years. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. REFERENCES Bard, A. J., and Lingane, J. J., Analytica Chim. Acta, 1959, 20, 463. Bard, A. J., Analyt. Chem., 1960, 32, 623. Bard, A. J., and Lingane, J. J., Analytica Chim. Acta, 1959, 20, 581. Lingane, J. J., Ibid., 1959, 21, 227. Sakurai, H., Kogyo Kagaku Zasshi, 1961, 64, 2121. Suzuki, S., Ibid., 1961, 64, 2112.Takahashi, T., and Sakurai, H., Talanta, 1962, 9, 74. Tso, T.-C., Li, H. L., and Sun, C. C., Hua Hsueh Hsueh Pao, 1964, 30, 301. Agasyan, L. B., Nikolaeva, E. R., Agasyan, P. K., and Lebedeva, 2. M., Izv. Vjjssh. Ucheb. Zaved. Bard, A. J., J . Electroanalyt. Chem., 1962, 3, 117. Sill&, L. G., and Martell, A. E., “Stability Constants of Metal - Ion Complexes.” Special Pub- lication No. 17, Chemical Society, London, 1964. Bishop, E., and Hitchcock, P. H., Analyst, 1973, 98, 553. Forbes, G. S., and Bartlett, E. P., J . Amer. Chem. SOC., 1914, 36, 2030; 1915, 37, 1201. Huey, C. S., and Tartar, H. V., Ibid., 1934, 56, 2585. Breiter, W. M., Electrochim. Acta, 1963, 8, 925. Bode, D. D., Andersen, T. N., and Eyring, H., J . Phys. Chem., 1967, 71, 792. Bishop, E., Analyst, 1972, 97, 761. GuBron, J., Bull. SOC. Chim. Fr., 1934, 1, 561. Bishop, E., and Hitchcock, P. H., Analyst, 1973, 98, 625. Bowles, B. J., and Cranshaw, T. E., Phys. Lett., 1967, 17, 258. Frumkin, A., in Yeager, E., Editor, “Transactions of the Symposium on Electrode Processes, Bishop, E., Chemia Analit., 1972, 17, 511. Vetter, K. J., “Electrochemical Kinetics,” New York, Academic Press, 1967, p. 481. Vetter, K. J., and Thiemke, G., Z . Elektrochem., 1960, 64, 805. Vetter, K. J., Ibid., 1952, 56, 797. James, S. D., Electrochim. Acta, 1967, 12, 939. Loshkarev, M. A., and Tomilov, B. I., Russ. J . Phys. Chern., 1960, 34, 836. Tomilov, B. I., and Loshkarev, M. A., Ibid., 1962, 36, 1027. Hurd, R. M., J . Electrochem. SOC., 1962, 109, 327. Higginson, W. C. E., Leigh, R. T., and Nightingale, R., J . Chem. Soc., 1962, 436. Welton, E. M. A., and Higginson, W. C. E., Ibid., 1965, 5890. Bishop, E., Analyst, 1972, 97, 772. Wright, D. T., and Bishop, E., Proc. SOC. Analyt. Chem., 1973, 10, in the press. , , , K’o Hsueh T’ung Pao, 1964, 163. --- Khim. Khim. Tekhnol., 1967, 10, 1316. 1959,” John Wiley & Sons Inc., New York, 1960, p. 1. 9 , Ibid., 1962, 36, 66. -- NOTE-References 13, 18, 20, 23 and 34 are to Parts VII, 111, IX, I and IV, respectively, of this series. Received January 24th, 1973 Accepted March 27th, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800635

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, September, 1973, Vol. 98, $9. 647-654 647 The Determination of Volatile Metal Chelates Using a Microwave-excited Emissive Detector BY R. M. DAGNALL, T. S. WEST AND P. WHITEHEAD (Biochemistry Division, Huntingdon Research Centre, Huntingdon, PE18 6ES) (Chemistry Department, Imperial College of Science and Technology, London, S W7 2A Y ) A microwave-excited emissive detector operated at atmospheric pressure has been used in conjunction with gas chromatography in order to achieve the separation and determination of various metal chelates of acetylacetone and trifluoroacetylacetone. The operating parameters have been optimised and the limits of detection and degrees of selectivity evaluated. The proposed method is both selective and very sensitive and can also confirm the presence of a particular metal.THE great advantages that gas chromatography possesses over many other separation techniques have raised considerable interest in its application to inorganic analysis. In particular, the gas chromatography of metal chelates has been extensively studied and has shown much greater potential than that of other metal derivatives. The principal chelates investigated have been those of acetylacetone and its fluorocarbon analogues, e g . , trifluoroacetylacetone and hexafluoroacetylacetone. All of these acetylacetone derivatives form stable chelates with a wide range of metals; Moshier and Sieversl have reviewed progress to 1965, although more recently the use of other ligands has been investi- gated. Fluoro-derivatives of dipivalylmet hane [ (CH ?) 3C .CO.CH,. CO. C (CH J 3] have been used in order to form fairly volatile chelates with alkali metals2 The chelates of monothio- p-diketones with a wide range of bivalent metals, including nickel, cobalt, zinc, palladium and platinum, are stable and volatile and have been chromatographed successfully.3-5 Amino-substituted /I-diketones also show potential as chelating agentsg In order to provide an acceptable method for the analysis of mixtures of metals, the gas- chromatographic determination of metal chelates requires good resolution of components, minimum interferences and, in particular, sensitivities at least as high as those of other methods in common use, e.g., atomic-absorption and emission spectroscopy. Hence the choice of detector is of major importance. The katharometer is non-selective and provides sensitivities in the range to 10-8 g.7~8 In addition, some decomposition of chelates occurs on the hot metal frlaments.l Flame- ionisation detectors have also been widely used but their response to metal chelates is less sensitive than that to organic compounds containing no metals.The presence of fluorine atoms and the metal atom itself in the chelate decreases the response of the detector and only .moderate sensitivities have been obtained.1° The electron-capture detector has been widely used for the determination of fluorinated chelates7,I1 and is highly sensitive.1°J2 The electron affinity of the chelates is a function of both the metal and any halogen atoms present ; however, the detector is much more sensitive to halogenated chelates.Under comparable conditions, the limit of detection for chromium( 111) acetylacetonate is 2.5 x 10-lo mol, for the trifluoroacetylacetonate 1.8 x mol and for the hexafluoroacetylacetonate 4.9 x mol.ll Often, the choice of chromatographic conditions is limited by the volatility and stability of the chelates used and hence complete resolution of mixtures is not always possible. Selective detection of individual metals would, therefore, be of considerable interest, but very little work has been reported on the application of selective detectors. Flame-photometric detection of both metal chelates and halides has been studied by Juvet and Durbin,l39l4 who viewed the emissions from an oxy-hydrogen flame by using a Beckrnann DU spectrophotometer. Although the sensitivity of the detector response was not high, good selectivity and a range @ SAC and the authors.648 DAGNALL, WEST AND WHITEHEAD : DETERMINATION OF VOLATILE [ A ?Za&!Lf, VOl.98 of linearity of lo4 were obtained. A similar system was used by Juvet and &do15 but with less sensitive results. It is only very recently that mass spectrometry has been applied in the determination of chromium and beryllium chelates a t the 10-l2,g level.16 Microwave-excited emissive detectors have proved highly sensitive and fairly selective in organic analysis,l7~l* but no application of microwave-excited emission to the determination of metal chelates has been reported. The potential usefulness of this type of emission for the determination of metals has been shown by the work of Bache and Lisk,lg who used emission a t the 253.7-nm atomic mercury line from a low-pressure helium plasma in order to measure the amounts of organomercury compounds present in fish, and by the work of Runnels and Gibson20 and Dagnall, Sharp and West.21 These two groups of workers used microwave plasmas in order to excite metal emissions from compounds that had been flash evaporated into the plasma from a platinum loop.This technique is highly sensitive, but is limited by the small volume (about 11.1) of solution that can be retained on the loop and to those compounds which can be volatilised under these conditions. Therefore, the response of the previously described22 microwave-excited emissive detector operated a t atmospheric pressure to some metal chelates was investigated.The chelating agents chosen were trifluoroacetylacetone and acetylacetone because they had been exten- sively used by other workers and comparison with other detectors would be possible. Pure metal chelates were bought or prepared, characterised, and their emission spectra in a micro- wave-excited plasma recorded. Solutions of the chelates in benzene were chromatographed and the major emission lines monitored. Samples of the eluates were collected and compared with the original compounds. EXPERIMENTAL PREPARATION OF METAL CHELATES- The acetylacetonates of aluminium(III), chromium(III), copper(I1) and iron(II1) were obtained commercially (Koch-Light Laboratories, Ltd., Colnbrook, Buckinghamshire) .Acetylacetone and trifluoroacetylacetone were also obtained from the same supplier and used to prepare the acetylacetonate* of scandium(II1) and the trifluoroacetylacetonates* of aluminium(III), chromium(III), copper(II), gallium(III), iron(III), scandium(II1) and vanadium( IV) . Trifluoroacetylacetone was found to deteriorate rapidly on exposure to air 'and was distilled immediately before use, the fraction boiling between 105 and 107 "C being collected. All the chelates, with the exception of aluminium(II1) and chromium(II1) trifluoro- acetylacetonates, were prepared by the method described by Berg and Tr~emper.~, In each instance, the pH of a suitable aqueous solution of the metal ion was adjusted to about 9.0 with 5 per cent. sodium acetate solution and an ethanolic solution of the chelating agent added.A precipitate formed on stirring, which was filtered, washed with water, air dried and either sublimed under vacuum or recrystallised from benzene or hexane. The preparation of Cr(tfa), was similar except that the pH was adjusted with dilute ammonia solution. Al(tfa), was prepared by the reaction of aluminium chloride in dry carbon tetrachloride with trifluoroacetylacetone. All the chelates used were characterised by means of their melting-points, and ultraviolet and infrared spectra. The melting-points were determined on a Kofler block, and agreement with values in the l i t e r a t u r e l ~ ~ ~ - ~ ~ was satisfactory. The infrared spectra of mulls of the chelates in Nujol were measured in the range 650 to 4000 cm-l by using a Perkin-Elmer 157 sodium chloride spectrophotometer. All the chelates of each ligand gave very similar spectra.The bands for the acetylacetonates were assigned by reference to the results of Nakamoto,28 whereas those for the trifluoroacetylacetonates were assigned from the results of Nakamoto, Morimoto and Marte11,29 and Morris, Moshier and S i e ~ e r s . ~ ~ The ultraviolet spectra of solutions of the complexes in ethanol, obtained by using a Perkin-Elmer 402 spectrophoto- meter, were relatively simple and contained, in most instances, one broad absorption band in the region 210 to 380 nm. The present results were very similar to those reported in the l i t e r a t ~ r e . ~ ~ ~ ~ l - ~ ~ * The abbreviation (acac) and (tfa) are used in the formulae of metal acetylacetonates and trifluoro- acetylacetonates, respectively.September, 19731 METAL CHELATES BY A MICROWAVE-EXCITED EMISSIVE DETECTOR 649 DETERMINATION OF THE EMISSION SPECTRA OF METAL CHELATES IN A MICROWAVE-EXCITED The emission spectrum of each metal chelate in the plasma was examined in order to determine the major emission lines that had the minimum of interferences from background, carbon and other metal emissions.An experimental system was set up so as to obtain an approximately constant, continuous level of vapour of the metal chelate in the argon entering the plasma, which enabled the entire spectrum to be scanned and more detailed studies to be made of the regions of interest. No attempt was made to obtain precise quantitative data on relative peak heights, as the concentration of the vapour of the chelates could not be maintained sufficiently constant.The microwave system was similar to that used in previous studies22 and consisted of a &wave Evenson-type microwave cavity, supplied from a microwave generator, and 1-mm i.d. silica tubing. When power was supplied to the cavity, a plasma could be maintained in the argon carrier gas flowing through the silica tube. A microwave power of 70 W was used throughout. The spectral emissions from the plasma were recorded by using a Beckmann DU monochromator and a d.c. read-out system. PLASMA- The apparatus used is shown in Fig. 1. PS = Power supply R = Read-out PM = Photomultiplier NI = Monochromator B = Sample boat HC = Heated column C = Cavity Q = Quartz tube S = Microwave power supply F = Furnace T = T-piece G = Pressure gauge P = Pressure controller Fig.1. Apparatus for the determination of the spectra of metal chelates It was found necessary to monitor each chelate as close as possible to the bottom of the plasma and to heat the connecting tubing. In the arrangement used, about 100 mg of the chelate under investigation were placed in a silica boat mounted in a brass T-piece just below the plasma. The silica tubing, which contained the plasma, passed into the middle of the T-piece just above the boat so that the vapour of the metal chelate could not easily come into contact with any hot metal surfaces. The carrier gas was pre-heated by passage through an empty steel column about 1 m in length and heated by means of heating tape to about 200 "C.A capillary restriction enabled the flow-rate of gas to be maintained at 3 1 h-1. The sample boat and silica tubing beneath the plasma were heated by means of a 12042 furnace, consisting of Nichrome wire round a 17 mm diameter silica tube, 90 mm long. The wire was covered with alumina cement. The furnace and the heating tape were controlled by using Variac variable transformers. For each set of spectral measurements, the chelate in the sample boat was introduced into the T-piece, the plasma initiated, the coil of tubing heated and finally the temperature650 DAGNALL, WEST AND WHITEHEAD : DETERMINATION OF VOLATILE [Analyst, Vol. 98 of the furnace raised slowly until carbon and metal emissions were observed. The variation of emission intensity with position in the plasma was measured as follows.The vertical height of the viewing slit of the monochromator was reduced to about 1 mm by a horizontal slit. Vapour of the sample was passed continuously into the plasma. The cavity, and hence the plasma, was moved vertically by means of a rack and pinion. With the plasma in the optimum position, the spectrum from 200 to 600 nm was scanned and when measurements for each chelate were completed, the silica tubing was replaced. Position of metal emissions in the plasma-The variation of the intensity of metal emissions with position in the plasma is shown in Fig. 2 for Al(acac), (396.2 nm) and Cu(acac), (327.4 nm). For comparison, the variation with position of the atomic carbon emission (247.9nm) for Al(acac), is included.In contrast to the carbon emission, metal emissions could be observed only near the beginning of the plasma. In addition, after the passage of relatively large amounts of a chelate into the plasma, a deposit with a metallic appearance also formed near the beginning of the plasma. It was concluded that the absence of metal emis- sions higher in the plasma was due to the metal atoms being removed from the gas phase on collision with the walls of the silica tubing. 0 20 40 Distance from bottom of plasma/mm Fig. 2. Variation of emission intensity with position in plasma. Graph A(+), 327.4 nm [Cu(acac),]; graph B ( x ), 396.2 nm [Al(acac),] ; and graph C( O), 247.9 nm [Al(acac)J The concentration of metal emissions in the first 10 mm of the plasma had two disadvan- tages.The extreme ends were the most unstable parts of a plasma and hence the variation in background emission was also greatest there. The metal deposit on the silica tubing gradually increased the background level of metal emission observed and after about milligram amounts of metal chelate had passed into the plasma the deposit significantly reduced the over-all intensity of light that reached the monochromator. Both of these effects were minimised by using small samples. Emission spectra of metal cdzelates-The major emissions observed for each met a1 species are shown in Table I. The recorded spectra also contained many less intense emissions that were considered to be less suitable for monitoring the passage of a gas-chromatographic eluate.However, they would be useful for providing a characteristic spectrum in order to confirm qualitatively the presence of a particular metal. In all instances when two compounds of a metal were studied, both of their spectra were found to be essentially identical. In order to confirm their identification, the emissions observed were compared with those obtained from electrodeless discharge lamps of each metal under the same experimental conditions and with those listed in the National Bureau of Standards Monograph No. 3236 obtained by using a 200-V d.c. arc. Owing to the relatively poor resolution of the mono- chromator used in the present study, many of the atomic lines listed could not be resolved.September, 19731 METAL CHELATES BY A MICROWAVE-EXCITED EMISSIVE DETECTOR 651 The major individual atomic lines are shown in parentheses after each composite emission peak observed.The spectra observed by using this experimental arrangement showed that suitable metal emissions could be obtained from a microwave-excited plasma for all the chelates used. Therefore, a gas-chromatographic system was set up and the major emissions were used to monitor the eluted chelates. TABLE I EMISSION CHARACTERISTICS OF METAL CHELATES Metal species Principal emissions/nm A1 396-2 Cr 357.9 425.4 520.5 (520-8, 520.6, 520-4) cu 324.7 327.4 Ga 287.4 294.4 (294.3, 294.4) 417.2 403.3 Metal species Principal emissions/nm Fe 344.1 357.0 373.5 s c 361.4 363-1 364.4 (364.5, 364.3) 357.5 (357-3, 357.6, 358.0) 424-7 vo 318.4 (318.3, 318.4, 318.5) 292.4 (292.4, 292-5) 438.1 249.0 (248.8, 249.0, 249.1) GAS~CHROMATOGRAPHY OF METAL CHELATES- The detector system was the same as that used for the above determination of spectra.The sample heater was replaced by a Pye Series 104 gas-chromatographic oven and column. Two columns were used- . . 0.6 m x 4.8 mm i.d. borosilicate glass packed with Universal B 0.6 m x 4.8 mm i.d. borosilicate glass packed with 0-5 per cent. Both columns were conditioned by heating them at 200 "C for 36 hours in a stream of argon. Poor peak shapes were obtained for Al(acac) 3r Cu(tfa), and VO(tfa), on column I and column I1 was therefore used in preference in these instances. In order to avoid decomposition or condensation of the metal chelates, the silica tubing containing the plasmas was connected directly to the column exit and the monochromator was positioned so as to view the plasma just above the roof of the oven.In order to reduce further the risk of condensation, the silica tubing between the column and the plasma was heated by means of a short coil of wire and a Variac variable transformer. The samples were injected directly on to a glass-wool pad in the top of the column, before the packing, and the top of the column was heated by means of the small electrical furnace described previously. The temperature of the top of the column could be maintained at about 30 "C above the oven temperature by applying a potential of 40 V across the furnace. Operating +arameters-The experimental parameters of the detector system were investi- gated by making 1-p1 injections of a 950 p.p.m.(m/V) solution of Ga(tfa), in benzene. Ga(tfa), was used because it was found to be conveniently eluted from column I at 125 "C after 3.1 minutes with a fairly good peak shape and the intense gallium atomic line at 294-4 nm lies in a region of very low background. Any variation of detector response at this wavelength could, therefore, be associated directly with changes in the metal emission. As expected from the studies with the continuous system, emissions characteristic of the metallic elements were obtained only from the first 10 mm of the plasma. In practice, in order to obtain a linear calibration graph that passes through the origin, it was found necessary to position the plasma so that emissions from the first 3 mm of the plasma could be monitored, and this position was used for all measurements of chelate emissions.The variation of the peak height at 294-4 nm with microwave power was measured by repeated injections of the Ga(tfa), solution. For each new power setting, the length of the plasma altered and the position of the plasma had to be re-optimised. The maximum peak height was obtained at a power of 70 W. The variation of noise with power was found to be very small and the maximum signal to noise ratio also occurred at 70 W. Column I .. Column I1 . . pre-coated with 10 per cent. Apiezon L. Apiezon L on glass micro-beads (0.2 mm diameter). . .652 DAGNALL, WEST AND WHITEHEAD : DETERMINATION OF VOLATILE [Analyst, Vol. 98 The variation of detector response with flow-rate of the carrier gas was also measured by repeated injections. The area response was found to be proportional to the reciprocal of the flow-rate, and the slowest convenient flow-rate (3.3 1 h-l) was used.In summary, the optimum experimental parameters were found to be the same as those for organic compounds.22 A microwave power of 70 W and a flow-rate of carrier gas of 3.3 1 h-l were used throughout the metal chelate determinations. Sensitivity and selectivity of emissions-The signal to noise ratios for the major emissions characteristic of the metallic elements observed previously were determined for solutions of each of the trifluoroacetylacetonates. In addition, the detector response at each wavelength to a 1000 p.p.m. solution of benzyl alcohol in benzene was measured in order to determine the selectivities of the metal emissions over carbon compounds.Benzyl alcohol was chosen because it has convenient retention times (1.0 to 3.5 minutes) on the columns used. The molar selectivity ratio was calculated in each instance as the ratio of the number of gram- atoms of carbon injected to the number of gram-atoms of the metal required to produce the same response. Benzene was used as the solvent for all the chelates. It was eluted after only 10 to 20 s and the plasma was not initiated until about 30 s after injection. The optimum experimental conditions used are shown in Table 11. All the trifluoro- acetylacetonates were chromatographed successfully, as were the acetylacetonates of alumin- ium, chromium and scandium. However, no peaks could be obtained for the copper and iron acetylacetonates. The most sensitive wavelength found for each metal, with the limit of detection and selectivity, are shown in Table 111.TABLE I1 OPTIMISED OPERATING PARAMETERS Chelate Column Al(acac), . . .. .. I1 Al(tfa), . . .. .. I Cr(acac), . . .. . . I Cr(tfa), . . .. .. I Cu(tfa), . . .. . . I1 Ga(tfa), , . .. . . I Fe(tfa), . . .. .. I Sc(acac), . . .. .. I Sc(tfa), . . .. .. I VO(tfa), . . .. .. I1 Temperaturel'C 175 100 190 130 140 125 135 135 160 160 Retention timelminutes 1.67 4.83 4-80 3.30 2.00 3.10 2.90 2.90 4-00 1.45 Interferences-Interference with the response of the detector to a metal chelate may be (1) spectral emissions from other metals or carbonaceous species at the wavelength being monitored; (2) suppression or enhancement of the detector response by another metal chelate that is eluted simultaneously; and (3) distortion of the plasma by a large excess of another compound.Type (3) is common to all compounds and methods of overcoming this problem, particu- larly in regard to the solvent, have been discussed previously.22 Types (1) and (2) will be considered separately. of three types- TABLE I11 LIMITS OF DETECTION AND SELECTIVITIES FOR SOME METAL CHELATES Chelate A1 (acac) , .. .. Al(tfa), . . .. .. Cr (acac), .. .. Cr(tfa), . . .. .. Cu(tfa), . . .. .. Ga(tfa), . . . . .. Fe(tfa), . . .. .. Sc(acac), .. .. Sc(tfa), . . .. .. VO(tfa), . . .. . . Wavelengthlnm Limit of detectionlg s-1 of metal 396.2 396.2 357.9 357.9 324.7 294.4 344.1 361-4 361.4 318.4 2.0 x 10-11 1.9 x 10-11 2-9 x 10-12 3.6 x 10-la 8.0 x 10-12 2.7 x 1.3 x 10-11 2.1 x 10-12 3.0 x 10-la 8.5 x 10-12 Selectivity 990 3930 2250 1170 1610 1620 1400 - - -September, 19731 METAL CHELATES BY A MICROWAVE-EXCITED EMISSIVE DETECTOR 653 The effect of emissions from benzyl alcohol, as shown by the selectivities in Table 111, was found to be, in general, very small.Possible spectral overlap of atomic metal lines was investigated by repeated injections of 1 pl of about 2 per cent. solutions of each chelate in benzene with the monochromator set in turn to each of wavelengtlis listed in Table 111. These concentrations are approximately the largest that could be injected without distortion of the plasma. Apart from the mutual interference of Cr(tfa), and Sc(tfa),, the only inter- ference observed was that of VO(tfa), on Cu(tfa),.Further investigations into spectral interferences were not carried out as the selectivity could easily be increased by reducing the slit width. If a particular interference presented a problem, alternative emissions (Table I) could often be used. The second type of interference was studied by preparing mixtures of Cr(tfa), and Ga(tfa) , in benzene and selecting gas-chromatographic conditions such that the peaks partially overlapped. Even in the presence of a 1000-fold excess of Ga(tfa),, the peak height of Cr(tfa), did not vary by more than 2 per cent. An example of the overlap of the peaks is shown in Fig. 3. Fig.. 3. Chromatorrrams of a mixturg of 985 p.p.m." of Ga(tfa), and 10 p.p.m.of Cr(tfa), in benzene. ( a ) , 357.8 nm (Cr); ( b ) , 294-4 nm (Ga); and ( G ) , 247-9 nm (C). R = re-ignition of plasma and I = injec- tion Characterisatiofi of the chromatographic ehates-The eluates were collected and charac- terised in order to confirm their identities and, in addition, to search for evidence of partial decomposition, which might be shown by the presence of an impure condensate. In order to collect a sufficient amount of each compound, 50p1 of about 1 per cent. solutions of each chelate in benzene were injected under the conditions used previously. The plasma was not initiated and the silica tubing not heated until the solvent was completely eluted. The tubing was cooled and the condensate collected. The melting-point and infrared and ultraviolet spectra of the condensate were measured in the same way as for the chelates.In all instances, the results were in close agreement with those obtained previously. CONCLUSIONS The microwave-excited emissive detector operated at atmospheric pressure responded to all the chelates used, both non-selectively by monitoring atomic carbon and selectively by using metal emissions. The detector response was linear and highly sensitive to the metals studied with limits of detection between 2 x 10-12 and 2 x 10-11 g s-1 (Table 111). The limits of detection of the acetylacetonates and trifluoroacetylacetonates of the same metals were approximately equal, indicating that the detector response may well prove to be independent of the chelating agent used.654 DAGNALL, WEST AND WHITEHEAD The sensitivity of this detector is considerably higher than those of other selective detec- tors used in the analysis of metal chelates, with the exception of mass spectrometry, and is of the same order as that of the electron-capture detector. Further, the dependence of the detector response on the metal, rather than on the chelate as a whole,‘ avoids the limitations of having a halogen atom or other electron-capturing species present in order to achieve high sensitivity. It is highly selective and, further, has the advantage that the pattern of atomic metal lines can be rapidly scanned in order to confirm the presence of a compound of a particular metal.The limitations of this detector for the analysis of mixtures of metals are similar to those for organic analysis.The principal problem is that of overloading, which imposes an upper limit on the working range of the detector. In addition, the formation of metallic deposits requires the tubing to be changed frequently when concentrated samples are used. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. REFERENCES Moshier, R. W., and Sievers, R. E., “Gas Chromatography of Metal Chelates,” Pergamon Press, Belcher, R., Dudeney, A. W. L., and Stephen, W. I., J . Inorg. Nucl. Chem., 1969, 31, 625. Stephen, W. I., Thompson, I. J., and Uden, P. C., Chem. Communs, 1969, 269. Belcher, R., Stephen, W. I., Thompson, I. J., and Uden, P. C., J .Inorg. Nucl. Chem., 1971,33, 1851. Stephen, W. I., Proc. SOC. Analyt. Chem., 1972, 9, 137. Sievers, R. E., Ponder, B. W., Morris, M. L., and Moshier, R. W., Inorg. Chem., 1963, 2, 693. Sievers, R. E., Chem. Engng News, 1963, 41, 41. Hill, R. D., and Gesser, H., J . Gas Chromat., 1963, 1, 11. Albert, D. K., Analyt. Chem., 1964, 36, 2034. Ross, W. D., Ibid., 1963, 35, 1596. Barratt, R. S., Proc. SOC. Analyt. Chem., 1972, 9, 86. Juvet, R. S., and Durbin, R. P., J . Gas Chromat., 1963, 1, 14. -- , Analyt. Chem., 1966, 38, 565. Juve’t, R. S., and Zado, F. M., Ibid., 1966, 38, 569. Wolf, W. R., Taylor, M. L., Hughes, B. M., Tiernan, T. O., and Sievers, R. E., Ibid., 1972, 44, 616. McCormack, A. J., Tong, S. C., and Cooke, W. D., Ibid., 1965, 37, 1470. Bache, C. A., and Lisk, D. J., Ibid., 1966, 38, 1757. -- , Ibid., 1971, 43, 951. Runiels, J . K., and Gibson, J. H., Ibid., 1967, 39, 1398, Dagnall, R. M., Sharp, B. L., and West, T. S., Nature, Phys. Sci., 1972, 235, 65. Dagnall, R. M., West, T. S., and Whitehead, P., Analytica Chim. Acta, 1972, 60, 25. Berg, E. W., and Truemper, J. T., J . Phys. Chem., 1960, 64, 487. Fay, R. C., and Piper, T. S., J . Amer. Chem. Soc., 1963, 85, 500. Henne, A. L., Newman, M. S., Quill, L. L., and Staniforth, R. A., Ibid., 1947, 69, 1819. Reid, J. C., and Calvin, M., Ibid., 1950, 72, 2948. Stanforth, R. A., Dissertation, Ohio State University, 1943. Nakamoto, K., “Infrared Spectra of Inorganic and Coordination Compounds,” Second Edition Nakamoto, K., Morimoto, Y., and Martell, A. E., J . Phys. Chem., 1962, 66, 346. Morris, M. L., Moshier. R. W., and Sievers, R. E., Inorg. Chem., 1963, 2, 411. Charles, R. G., J . Inorg. Nucl. Chem., 1958, 6, 42. Holm, R. H., and Cotton, F. A., J . Amer. Chem. SOC., 1958, 80, 5658. Bamun, D. W., J . Inorg. Nucl. Chem., 1961, 21, 221. -, Ibid., 1962, 22, 283. Hazeldine, R. N., Musgrave, W. K. R., Smith, F., and Turton, L. M., J . Chem. SOC., 1951, 609. Meggers, W. F., Corliss, C. H., and Scribner, B. F., “Tables of Spectral Line Intensities,” Natn. Received January 24th, 1973 Accepted A p r i l loth, 1973 Oxford, 1965. 8 J J , Chem. Communs, 1970, 1019. ---__. Wiley-Interscience, New York, 1970. BUY. Stand. Monog., No. 32, Part I, 1961.

ISSN:0003-2654    

DOI:10.1039/AN9739800647

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, September, 1973, Vol. 98, $$. 655-658 655 The Determination of Trace Amounts of Barium in Calcium Carbonate by Atomic-absorption Spectrophotometry" BY F. J. BANOf (John & E. Sturge Limited, Liffoord Chemical Works, Kings Norton, Birmingham, B30 3 J W ) Barium present at the 1 to 20 p.p.m. level in calcium carbonate is separated from the calcium matrix by co-precipitation on a lead sulphate carrier. The lead sulphate - barium sulphate precipitate is dissolved in an ammoniacal solution of ethylenedianiinetetraacetic acid disodium salt and the barium in the resulting solution is determined by atomic-absorption spectrophotometry by using a nitrous oxide - acetylene flame and the atomic line at 553.6 nm. The method overcomes interferences that occur in the atomic-absorption spectrophotometric determination of barium in the presence of calcium. THE determination of barium by atomic-absorption spectrophotometry in materials that contain large amounts of calcium has been reported by Bowman and Willis.1 Ingamells, Suhr, Tan and Anderson2 have reported a comprehensive study of the determination of strontium and barium in various rocks and minerals that contain calcium, and while they stressed the usefulness of atomic-absorption spectrophotometry for the determination of strontium, they were unable to obtain satisfactory results for barium by using this technique.Koirtyohann and Pickett3 reported the existence of a molecular spectral interference when barium was determined in the presence of calcium by atomic-absorption spectrophotometry. This interference was ascribed to absorption by the CaOH band.Capacho-Delgado and Sprague* claim that the interference can be overcome by using the nitrous oxide - acetylene flame. However, in these laboratories, it has been found impossible to overcome the inter- ference when trace amounts of barium are to be determined in calcium carbonate. As the interference could not be overcome by the use of the nitrous oxide - acetylene flame, a means of separating trace amounts of barium from a calcium matrix was investigated. Separation of the analyte barium from the calcium matrix, as well as overcoming the inter- ference effect, would also lead to an over-all increase in the sensitivity of the method. Feig16 has described a test for barium and other alkaline earth cations, which is based on the induced precipitation of lead sulphate when barium is added to a solution of lead sulphate in acetic acid.Feigl suggests that the precipitate formed in this test is a mixed crystal or addition compound of lead sulphate and barium sulphate. As lead sulphate is less soluble than calcium sulphate in aqueous media, it was thought that it might be possible to precipitate lead sulphate from a solution containing calcium and to scavenge the barium from the solution as the mixed lead sulphate - barium sulphate compound. In order to test this hypothesis, the following experiments were carried out. EXPERIMENTAL AND RESULTS REAGENTS- All reagents used were of analytical-reagent grade. Sulphuric acid, 2.5 M-Mix 136 ml of concentrated sulphuric acid (sp. gr.1-84> with Nitric acid, 5 M-Mix 320 ml of concentrated nitric acid (sp. gr. 1.42) with about 800 ml Ammonia solution, 5 M-Mix 280 ml of concentrated ammonia solution (sp. gr. 0.88) Lead nitrate. * Presented a t the Elwell Award Presentation Meeting of the Midlands Region of the SAC, held in t Present address : London and Scandinavian Metallurgical Co. Ltd., Fullerton Road, Rotherham, @ SAC and the author. about 800ml of water, allow the mixture to cool and make the volume up to 1 litre. of water and make the volume up to 1 litre. with about 800ml of water and make the volume up to 1 litre. Birmingham on January 16th, 1973. Yorkshire, S60 IDL. The author was adjudged the winner.656 BANO : THE DETERMINATION OF TRACE AMOUNTS OF BARIUM IN [Analyst, Vol.98 Ethylenediaminetetraacetic acid disodium salt. Calcium carbonate-Specpure (Johnson, Matthey & Co. Ltd.). Barium chloride stock solution-Dissolve 1.78 g of BaC1,.2H20 in about 100 ml of water and make the volume up to 1 litre. 1 ml of solution = 1000 pg of Ba. 1 ml of solution = 10 pg of Ba. Barium chloride working solution-Dilute 10 ml of barium chloride stock solution to 1 litre immediately before use. APPARATUS- Atomic-absorption measurements were made with an EEL 240 Mark I1 atomic-absorp- tion spectrophotometer (Evans Electroselenium Limited, Halstead, Essex) , fitted with a 50-mm slot nitrous oxide - acetylene burner and a barium hollow-cathode lamp. The instrumental conditions used were as follows- Nitrous oxide . . .. .. . . 5-51 min-l Acetylene .. . . .. . * 3.5 1 min-l Wavelength . . .. . . . . 553.6nm Slit width . . .. . . . . 0.14 mm (position 3) Lamp current . . .. .. .. 7.5mh Burner height , . .. .. . . Position 6 Scale expansion . . .. .. x10 RECOVERY OF BARIUM BY THE CO-PRECIPITATION PROCEDURE- A 100-pg amount of barium (in the form of 10 ml of barium chloride working solution) and 160mg of lead nitrate were dissolved in 100ml of 0.1 M nitric acid. The solution was heated to boiling and 5 ml of 2.5 M sulphuric acid were added. The solution was allowed to stand for 2 hours on a steam-bath, after which period a dense crystalline precipitate had settled out. The supernatant solution was decanted off and the precipitate was washed with two 10-ml portions of water. The supernatant liquid and washings were combined and evaporated to a volume of approximately 20 ml.To the precipitate were added 200 mg of ethylenedia- minetetraacetic acid disodium salt and 10 ml of 5 M ammonia solution. The mixture was gently warmed until all the solids had dissolved and the solution was transferred into a 25-ml calibrated flask and the volume made up to the calibration mark with water. To the supernatant solution were added 200 mg of ethylenediaminetetraacetic acid di- sodium salt, and the solution was transferred into a 25-ml calibrated flask and the volume made up to the calibration mark with water. The barium in each of the two solutions was determined by atomic-absorption spectro- photometry by using the instrumental conditions described above. The sodium derived from the ethylenediaminetetraacetic acid disodium salt functioned as an ionisation buffer.A large absorption signal was obtained for the solution that contained the dissolved lead sulphate, while no signal was obtained for the solution that contained the supernatant liquid. The absorption signal obtained from the solution that contained the dissolved precipitate was identical with that obtained from the solution to which 100 pg of barium (0.1 ml of barium chloride stock solution) had been added. It was therefore concluded that lead sulphate would function as a suitable carrier for barium. CONCENTRATION OF SULPHURIC ACID REQUIRED FOR THE PRECIPITATION OF LEAD SULPHATE- The concentration of sulphuric acid used for the precipitation of the lead sulphate carrier is critical, because too low a concentration of sulphuric acid would lead to incomplete precipitation of lead sulphate while too high a concentration would cause precipitation of calcium sulphate.Five 5 & 0.01-g samples of Specpure calcium carbonate were weighed into 150-ml beakers, 25 ml of 5 M nitric acid were added to each beaker and the solutions were warmed so as to effect complete dissolution of the calcium carbonate. Nitric acid was used because hydrochloric acid was found to inhibit the precipitation of lead sulphate. To each beaker were added 100 pg of barium (10 ml of barium chloride working solution) and 160 mg of lead nitrate, and the solutions were heated to incipient boiling. To the separate solutions wereSeptember, 19731 CALCIUM CARBONATE BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 657 added 1, 2, 3, 4 and 5 ml of 2.5 M sulphuric acid, with stirring, and the solutions were allowed to stand on a steam-bath for 2 hours.After this period of time, no precipitate was observed to have settled out of the solution to which 1 ml of sulphuric acid had been added, while the solutions to which 3, 4 and 5 ml of sulphuric acid had been added were found to have precipitated substantial amounts of needle-shaped crystals, which are characteristic of calcium sulphate. The recovery of barium from the solution to which 2 ml of sulphuric acid had been added was checked in the above manner and found to be satisfactory. CALIBRATION GRAPH- Weigh 160 mg of lead nitrate and 200 mg of ethylenediaminetetraacetic acid disodium salt into each of six 50-ml beakers.Dissolve the solids in approximately 10 ml of 5 M am- monia solution. Add to the separate beakers 0,2,4,6,8 and 10 ml of barium chloride working solution (containing 10 pg ml-l of barium), transfer the solutions into 25-ml calibrated flasks and make the volumes up to the calibration mark with water. Spray the solutions into the flame of the atomic-absorption spectrophotometer, obtain the absorbance readings and plot them against barium concentration. PROCEDURE- Weigh 5 & 0.01 g of sample and dissolve it in 25 ml of 5 M nitric acid in a 150-ml beaker. Dilute the solution to 100 ml with water and add 160 mg of lead nitrate, then heat the mixture to incipient boiling and add 2 ml of 2.5 M sulphuric acid by means of a pipette.Allow the solution to stand on a steam-bath for 2 hours, then decant off the supernatant liquid, wash the precipitate with two 10-ml portions of water, taking care not to lose a significant amount of precipitate, and add 200 mg of ethylenediaminetetraacetic acid disodium salt. Add 10 ml of 5 M ammonia solution and warm the mixture gently so as to effect complete dissolution of the solids. Transfer the solution into a 25-ml calibrated flask and make tlie volume up to the calibration mark with water. Spray the solution into the flame of the atomic-absorption spectrophotometer, obtain the absorbance reading, and read off the amount of barium in micrograms from the calibration graph. PRECISION- The precision of the method at the 5, 10 and 20 p.p.m. levels of barium was established by determining the barium contents of solutions of Specpure calcium carbonate to which 25, 50 and 100 pg of barium, respectively, had been added.The determination at each level was repeated six times. The Specpure calcium carbonate was known to have a nominal barium content of 1 p.p.m. The results are shown in Table I. TABLE I PRECISION OF THE METHOD METHOD Barium level, Mean barium content, Standard deviation, Coefficient of variation, p.p.m. p.p.m. p.p.m, per cent. 5 6.9 0.86 12.5 10 10.9 0.65 6-0 20 21.1 0.42 2.0 ACCURACY- The accuracy of the method was established by comparing the mean value of barium found in samples of calcium carbonate to which barium had been added with those in synthetic standard solutions containing known amounts of barium. It should be noted that the calcium carbonate used for determining the accuracy of the method was known to contain approxi- mately 1 p.p.m.of barium. The results are shown in Table 11. TABLE I1 ACCURACY OF THE METHOD Barium found less the 5 pg Barium added to of barium in the sample Recovery, 5-g samplelpg Barium found1p.g (equivalent to 1 p.p.m.)/pg per cent. 25 34.5 29.5 118 50 54.5 49.5 99 100 105-5 100.5 101658 BAN0 CONCLUSION The method described should provide a rapid and accurate method for the determination of barium in calcium carbonate. The method should be applicable -to the determination Qf barium in limestone and other materials that contain calcium, such as the rocks and minerals discussed by Ingamells, Suhr, Tan and Anderson.2 The author thanks the Board of Directors of John & E. Sturge Limited for permission to publish this paper. REFERENCES 1. Bowman, J. A., and Willis, J. B., Analyt. Chem., 1967, 39, 1210. 2. Ingamells, C. O., Suhr, N. H., Tan, F. C., and Anderson, D. H., Analytica Chim. Acta, 1971, 53, 345. 3. Koirtyohann, S. R., and Pickett, E. E., Analyt. Chem., 1966, 38, 585. 4. Capacho-Delgado, L., and Sprague, S., Atom. Absorfition Newsl., 1965, 4, 363. 5. Feigl, F., “Spot Tests in Inorganic Analysis,” Fifth Edition, Elsevier Publishing Company, Received March 2nd, 1973 Accepted Mads 29th, 1973 Amsterdam, 1958, p. 218.

ISSN:0003-2654    

DOI:10.1039/AN9739800655

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, September, 1973, Vol. 98,@. 659-662 659 Spectrophotometric and Chelatometric Determination of Iron(II1) with 3 =Hydroxypyridine--2=thiol BY MOHAN KATYAL, (St. Stephen's College, Delhi-7, India) MISS VEENA KUSHWAHA AND R. P. SINGH (Department of Chemistry, Delhi University, Delhi-7, India) 3-Hydroxypyridine-2-thiol can be satisfactorily used in the spectrophoto- metric and chelatometric determination of iron(II1). With the prescribed methods the determination can be made over a wide pH range without interference from many anions and cations. Structures are suggested for the complexes formed. IN continuation of studies on hydroxypyridines as analytical reagent~,l-~ it has been found that the ligand 3-hydroxypyridine-2-thiol (I) forms coloured complexes with iron(II1). The reaction is selective and sensitive.The colour of the complex formed is discharged by the addition of EDTA. In the Dresent DaDer. use has been made of this comDlex and its subseauent decomposition by EDTA in the siectrophotometric and chelatom&-ic determinati6n of iron (111). n I EXPERIMENTAL REAGENTS- Standard iron(ll1) solutions-Standard solutions of iron( 111) were prepared by dissolving precipitated iron( 111) hydroxide in the appropriate acid (perchloric, hydrochloric or sulphuric acid). The solutions obtained were then standardised gravimetrically for iron content and titrimetrically for acid content. Bufer solutions-Buff er solutions were prepared by mixing 0.1 N potassium hydrogen phthalate solution with 0.1 N hydrochloric acid or 0.1 N sodium hydroxide solution as required.3-Hydroxypyridine-2-t~ioZ yeagent solution, 0.01 M-The solution of 3-hydroxypyridine-2- thiol (Fluka, Switzerland) was prepared by dissolving the appropriate amount of the reagent in methanol. The reagent solution can be kept for several weeks in an amber-glass bottle or in a refrigerator without appreciable decomposition occurring. All other solutions were prepared by dissolving analytical-reagent grade chemicals in double-distilled water. APPARATUS- Metrohm E350 pH meter was used for pH measurements. Absorbance readings were taken with a Unicam SP600 spectrophotometer, and a SPECTROPHOTOMETRIC DETERMINATION When a few drops of a methanolic solution of 3-hydroxypyridine-2-thiol are added to an iron(II1) solution, a green complex is instantly formed, the green colour of the complex changing with increase in pH first to blue and finally to red.Iron(I1) appears to form similar species with this reagent but these have not been studied in detail. The system is reversible and can be represented as follows: OH- OH- H+ H+ Green complex + Blue complex + Red complex @ SAC and the authors.660 [Analyst, Vol. 98 3-Hydroxypyridine-2-thiol shows maximum absorption a t wavelengths 275 and 360 nm with corresponding molar absorptivities of 9.1 x lo3 and 1.1 x lo*. The spectra of the solutions obtained by mixing iron(II1) solution with appropriate amounts of reagent solution were recorded under various pH conditions. The green complex formed is stable in the range of acidity 0.2 to 0.01 N perchloric acid medium with maximum absorption at 675 nm, the blue complex is stable a t pH between 3-5 and 11.0 with maximum absorption at 600 nm and the red complex is stable at pH above 11.0 with maximum absorption at 520 nm.The blue complex is the most stable of these three complexes with respect to time and permits the use of a wide pH range for the determination of iron(II1). A 50 per cent. methanolic medium was used throughout this study. The system adheres to Beer's law for concentrations of iron up to 3.36 pg ml-l. The sensitivity of the reaction is 0.012 pg of iron for 0.001 absorbance unit in a 1-cm cell. The molar extinction coefficient and apparent stability constant of the complex in a 50 per cent. methanolic medium at pH 5-0 and at 20 "C are 4.7 x lo3 and 3.82 x los, respectively. EFFECT OF pH- The effect of pH on the 3-hydroxypyridine-2-thiol reagent and the blue complex that it forms with iron(II1) was studied.The reagent itself does not absorb at the wavelength of maximum absorption of its iron complex. The absorption of the blue complex remains unaltered between pH 4.2 and 5-6, and solutions were therefore maintained at pH 5.0 in the subsequent work. EFFECT OF REAGENT CONCENTRATION- MI, in- creasing amounts of reagent solution were added while maintaining a 50 per cent. concen- tration of methanol and the pH at 5.0. It was found that a twenty-fold molar excess of 3-hydroxypyridine-2-thiol was sufficient for full colour development ; the excess of reagent did not interfere. COMPOSITION OF THE COMPLEXES- Job's curves were plotted for the complexes: green (at pH 2.0 only), blue (at pH 3.5 to 11.0) and red (at pH above 11-0), at different wavelengths ranging from 450 to 700 nm.It is concluded that in each instance iron(II1) and 3-hydroxypyridine-2-thiol combine in a 1 : 2 molar ratio. Taking into account earlier work,ls3 the following tentative structures are proposed for the species: green complex (11), where X is Clop, C1- or SO,2-; blue complex (111); and red complex (IV): KATYAL et al. : SPECTROPHOTOMETRIC AND CHELATOhlBTRIC To a fixed volume of iron(II1) solution [final concentration of iron(II1) 5 x r- 1- RECOMMENDED PROCEDURE- To a solution containing up to 30 pg of iron, add 5.0 ml of 0.01 M solution of 3-hydroxy- pyridine-2-thiol in methanol and a volume of buffer solution such that the pH of the final solution is about 5.0 when the volume is made up to a total of 10.0 rnl (or 25 ml or more if necessary) in a calibrated flask.Measure the absorbance of the complex at 600 nm against a reagent blank and from the standard graph calculate the amount of iron in the unknown solution. INTERFERENCES- In the determination of 1 pg ml-1 of iron by this method, 200 pg ml-l of each of the species CH3COO-, NO2-, NO3-, C1-, Br-, I-, B033-, citrate, tartrate, thiourea, zinc(II), cadmiumfII), barium(II), strontium(II), aluminium(III), lead(II), arsenic(II1) and man- ganese(II), and 100 pg ml-1 of each of F-, C,O,2- and PO,3- did not interfere. Also, theSeptember, 19731 DETERMINATION OF IRON(II1) WITH 3-HYDROXYPYRIDINE-2-THIOL 661 following ions can be tolerated at the levels (pgml-l) given in parentheses: silver (25), copper (15), bismuth (50), thorium (25), uranyl (50), molybdate (lo), thiocyanate (50), cyanide (120) and sulphide (250).Silver, bismuth, thorium and barium gave precipitates that were removed by centrifugation before taking the reading. Platinum metals caused serious interference. The determination can be accomplished at low pH (about 2-0) with more selectivity by making use of the green complex. In this instance the readings should be taken within the 30-minute period during which the absorbance remains steady. In all of the above tests the ions examined were present before the reagent was added. CHELATOMETRIC DETERMINATION As previously indicated, the colour of the iron(II1) - 3-hydroxypyridine-2-thiol complex varies from green to red, depending on the level of acidity.Thus, if a titration with EDTA solution is carried out at a pH below 3-5, the colour of the solution changes from green to colourless (or yellowish with large amounts of iron). If the pH of the iron solution to be analysed is above 3.5 and below 11.0, thus giving a blue complex with the reagent, the end-point is shown by the disappearance of the blue colour. The optimum conditions for titrations with EDTA solution have been established. EFFECT OF pH- Titrations of iron(II1) against EDTA solution were carried out a t different pH values. The metal-ion and indicator (3-hydroxypyridine-2-thiol) concentrations were kept constant and the pH was adjusted by adding potassium hydrogen phthalate buffer solution.The results are shown in Table I. TABLE I TITRATIONS OF IRON(III) AT DIFFERENT pH VALUES Volume of iron(II1) solution taken, 2-0 ml of 0.01 M concentration ( ~ 1 . 1 1 7 mg of the metal) ; volume of buffer solution added, 10.0 ml; amount of 0-5 per cent. methanolic indicator solution added, 5 drops ; concentration of EDTA solution used, 0.01 M ; total volume made up to approximately 20 ml Volume of EDTA pfI solution used/ml 0-8 2.10 1.0 2.08 1.2 2.02 1.5 2.00 1.8 2.00 2.0 2-00 Volume of EDTA PH solution used/rnl 2.4 2-00 2.8 2.00 3.2 2.00 3.6 2.00 4.0 2.00 4.4 2.00 Volume of EDTA PH solution used/ml 4.5 2.00 4.6 2.01 4-8 2.02 5.0 2.04 5.2 2.08 5.5 2-12 It is evident that satisfactory results can be obtained in the pH range 1.5 to 4.5.REQUIREMENTS FOR CONCENTRATION OF IRON AND INDICATOR- Titrations with different amounts of iron (1 to 15 mg in 20 ml of solution) were carried out at various pH values from 1-5 to 4.5. It was found that when iron is present in amounts greater than 3 mg, the end-point is given by the appearance of the pronounced yellow colour of the iron(II1) - EDTA complex. Iron solutions (containing 1.117 mg of the metal) a t different pH values (1.5 to 4.5) were titrated by using different volumes of 0.5 per cent. solution of the indicator in methanol. It was found that the presence of 2 to 5 drops of the indicator produced a sharp change of colour. However, the use of slightly larger amounts of the indicator did not affect the results. EFFECT OF TEMPERATURE- Titrations were performed in the temperature range 15 to 70 "C; accurate results were obtained betwen 20 and 45 "C.At lower temperatures the reaction is slow, while at higher temperatures a false end-point is reached, presumably because of the dissociation of the iron( 111) - 3-hydroxypyridine-2-thiol complex.662 KATYAL, KUSHWAHA AND SINGH RECOMMENDED PROCEDURE- Adjust the pH of a solution containing 1 to 15 mg of iron(II1) to between 1-5 and 4.5 by adding 10.0 ml of potassium hydrogen phthalate buffer, add 5 to 10 drops of 0.5 per cent. methanolic indicator solution and dilute the solution to about 20 ml. Titrate the solution slowly with standard 0.01 M EDTA solution, occasionally shaking it until the green colour (pH below 3.5) or blue colour (pH between 3.5 and 4.5) completely disappears.EFFECT OF DIVERSE IONS- 1-5 to 2.0. The results are incorporated in Table 11. The effect of diverse ions was examined in the determination of iron(II1) in the pH range TABLE I1 EFFECT OF DIVERSE IONS Amount of iron(III), 1.117 mg ( ~ 2 . 0 ml of 0.01 M EDTA solution) Amount of foreign ion EDTA solution Volume of 0.01 M Foreign ion* added/mg usedlml Masking agent CH3COO- . . .. .. .. .. 700 2.00 - NO,-, C1-, Br-, I-, SO,2-, citrate, tartrate.. 100 2.00 - NO2- .. .. . . .. . . .. 100 1.98 - .. . . .. .. 100 2.01 - 1 3 0 ~ 3 - . . .. .. .. .. 60 1.98 - so32- . . .. . . . . .. . . 50 1.97 - F- . . .. .. .. .. .. 50 1.98 - ~ 0 ~ 3 - . . .. . . CNS- .. .. . . .. .. .. 30 2.02 - . . .. . . 25 2.00 - CN- .. .. .. .. . . . . .. 10 1.97 - c,o*2- .. .. .. .. .. .. 50 2.00 - S2- .. . . Silver(1) . . .. .. .. .. .. 50 2.00 c1- .. 10 1.98 - Barium(I1) . . .. .. * . .. .. . . 10 1-97 - Calcium(I1) . . * . Bismuth(II1) . . .. . . .. .. 10 2.00 c1- Lead(I1) . . .. .. .. .. . . 10 2.02 CH3COO- Mercury(I1) . . .. .. .. .. 6 2.02 CN- Zinc(I1) . . .. .. .. .. .. 5 2.01 CN- Cadmium(I1) . . .. .. .. .. 5 2.01 CN- Aluminium( 111) .. .. .. .. 5 2.01 CH3COO- Manganese(I1) . . .. .. .. .. 4 2.01 Tartrate Cobalt(I1) . , .. .. .. .. 2 2.02 CN- Copper( 11) .. .. .. .. 1.3 2-01 CN- Molybdenum(V1j ‘ . . .. .. . . 1-2 2.03 CNS- * These ions were present before the 3-hydroxypyridine-2-thiol reagent was added. DISCUSSION The use of 3-hydroxypyridine-2-thiol as a reagent for iron(II1) compares favourably in respect of selectivity and sensitivity with that of its precursors, viz., 2,3-dihydro~ypyridinel-~ and 2-hydroxy-6-methylpyridine-3-carboxylic acid.4 The only disadvantage in using the green complex formed at low pH (below 3.5) for the spectrophotometric determination of iron(II1) is that the absorbance readings require to be taken within half an hour, after which time the absorbance does not remain steady. The complexometric titrations can be carried out over a much wider pH range (between 1.5 and 4.5). Even though the intensity of the colour of the green complex may decrease slightly with time, the end-point is not affected when the solution is titrated against EDTA solution. The University Grants Commission (India) is gratefully acknowledged for financial assis- tance in the award of the scheme “Water Pollution.” REFERENCES 1 . 2. 3. 4. Katyal, M., Goel, D. P., and Singh, R. P., Talanta, 1968, 15, 711. Goel, D. P., and Singh, R. P., Analyst, 1971, 96, 123. Kushwaha, V., Singh, R. P., and Katyal, M., Mikrochim. Acta, 1972, 807. -,-,- , Talanta, 1973, 20, 431. Received December 8th. 1972 Accepted March 16th, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800659

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Analyst, September, 1973, Vol. 98, $9. 663-672 663 Assay of Micro-scale Amounts of Hydroperoxide and of Iodine in Aqueous Non-ionic Surfactant Solutions by a Spectrophotometric Method BY E. AZAZ, M. DONBROW AND R. HAMBURGER (Pharmacy Department, School of Pharmacy, Hebrew University of Jerusalem, Jerusalem, P.O.B. 12065, Israel) An iodimetric - ultraviolet spectrophotometric method has been developed for the quantitative determination of trace concentrations of hydroperoxide (down to It involves a simple extrapolation procedure based on the kinetics of iodine fading in these systems. This method was also applicable to macro-scale amounts when the classical titrimetric method could not be used. Hydrogen peroxide and hydroperoxides of some polyethylene glycols, in the presence of Cetomacrogol 1000, gave reproducible results that were in good agreement with the results obtained by use of the titrimetric method.M) in the presence of etheric non-ionic surface-active agents. THE most universally used method for the determination of hydroperoxide is that developed by Lea.lS2 It is based on the titration of iodine liberated from potassium iodide by oxidising species. This method was found not to be applicable in the presence of non-ionic surfactants as up to 40 per cent. of the iodine was unavailable for titrati0n.~-6 Another disadvantage of Lea’s method is that it is sensitive only to concentrations above 5 x M ; lower concen- trations give errors of up to 50 per cent.7 Nevertheless, the necessity of carrying out such determinations was encountered in studies of autoxidisable substances, such as benzaldehyde,* linoleic acidg and vitamin A,l0 which were rendered soluble with the aid of non-ionic surface- active agents.The development of micro-scale methods for the quantitative determination of hydroperoxides in the presence of non-ionic surf ace-active agents would enable the initial stages of autoxidation to be investigated. Micro-assays of iodine by ultraviolet spectroscopy in solvents other than aqueous non- ionic surfactants have been de~eloped.~9ll According to Heaton and Uri,7 calibration graphs of absorbance against concentration for molecular iodine added to potassium iodide solutions were identical with those obtained for iodine liberated from potassium iodide by hydro- peroxides. They also found that up to concentrations of 5 x 10-4 M Beer’s law was obeyed.The object of the present work was to find conditions under which the ultraviolet absorbance of iodine would be applicable to the determination of hydroperoxides present in non-ionic surfactants. The ultraviolet absorbance of iodine at 360 nm in non-ionic surfactants was used by several workers for the determination of the critical micellar concentration.12-14 However, there seem to be discrepancies as to the position of the iodine absorption peaks. Earlier papers5?l5 reported the peak to be at 370 nm. Ross arid Olivier14 state that the critical micellar concentration has to be determined within 1 hour from the preparation of the solution as otherwise the ultraviolet spectrum changes. Spectral changes were also observed by other workers, occurring with change in concentration ratio, 1 2 9 1 3 9 1 6 storage timelsJ7 and electrolyte composition.13 A maximum absorbance in the region of 360 to 370 nm for I,- was also reported for non-aqueous solvents.18-20 After standardisation of the concentration of surfactant, the temperature, the storage time and the electrolyte composition, this absorption region (360 to 370 nm) provided reproducible assays of iodine liberated by peroxides present in the surfactant system, as can be seen from the results given below.A variety of other methods for the determination of micro-scale amounts of hydroperoxide, none of which was checked for systems containing surfactants, have been described. However, the iodimetric method is the most generally applicable.21 The method described in the present paper is used in routine testing in our laboratory for monitoring oxidisable drugs, for which it has proved to be indispensable and results obtained will be published elsewhere.@ SAC and the authors.664 MATERIALS AND REAGENTS- Low Mills, Leeds. AZAZ et al. : ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [Analyst, Vol. 98 METHODS Cetomacrogol 1000 B.P.C. (Texofor A IP)-Supplied by Glovers Chemicals Ltd. , Wortley Iodine. Potassium iodide (iodate-free)-Analytical-reagent grade material was used. Sodium thiosulphate solution-BDH Chemicals Ltd. Concentrated Volumetric Solution Ammonium molybdate. Nitrogen gas, 99-999 per cent. Sodium dihydrogen orthophosphate. Disodium hydrogen orthophosphate.PoZyethylene glycol 400, 1500, 4000 and 6000. Hydrogen peroxide solution-E. Merck's 30 per cent. analytical-reagent grade material was used. Buflered solution of 9otassium iodide in 0.1 M phosphate (pH 6), 5 per cent.-This solution was prepared by mixing 87.7 ml of 0-2 M sodium dihydrogen orthophosphate solution with 12.3 ml of 0.2 M disodium hydrogen phosphate solution,22 adding 10 g of potassium iodide and diluting to 200 ml. Iodine, sodium thiosulphate and hydrogen peroxide solutions were prepared and standard- ised according to Kolthoff and Sande1LZ3 Cetomacrogol 1000 solutions were freshly prepared from approximately 20 per cent. stock solutions and standardised by making refractive index measurements on an Abbk 60 refractometer (Bellingham & Stanley Ltd., London) (see Fig.1). For each batch a separate calibration graph was drawn by using accurately weighed and diluted samples. The differences were not great (n: for 20 per cent. solutions ranged be- tween 1-3580 and 1.3602). was used. 1.330 -o 15 0 5 Concentration, per cent. Fig. 1. Refractive index of Cetomacrogol a t 25 "C uemws con- centration (batch A) APPARATUS- A Hilger Ultrascan H999 recording spectrophotometer was used for qualitative work, and a Hilger Uvispek H700 spectrophotometer with temperature-regulating water-jacket for quantitative, time-dependent readings. Readings were made in the concentration range of iodine from 1 x PROCEDURES- 1. Spectrophotometric determination of iodine in the presence of Cetomacrogo LIodine solutions in concentrations of about 0.1 M were diluted with 5 per cent.potassium iodide solution to about 1 x M. Final dilutions to concentrations of 1 x lou4 to to 1 x M by using 0.5 to 4-cm stoppered cells. to 1 xSeptember, 19731 AND IODINE IN AQUEOUS NON-IONIC SURFACTANTS 665 1 x M were made by adding appropriate amounts of these iodine solutions to calibrated flasks containing appropriate amounts of concentrated, freshly prepared Cetomacrogol solution (5 to 10 per cent.) in buffered (at pH 6) 5 per cent. potassium iodide solution. The contents of the flasks were diluted to volume with more buffered 5 per cent. potassium iodide solution to make a final sample of buffered solution (pH 6) containing 1 per cent. of Cetomacrogol and 5 per cent. of potassium iodide with a concentration of 1 x All of the dilution steps were performed in flasks protected from the light and under purified nitrogen and the iodide was free from iodate.The time of bringing the iodine solution into contact with Cetomacrogol solution was taken as zero time. Absorbances were measured a t a wave- length of 370 nm a t 5 minute intervals and at controlled temperatures for the first half hour. Results were extrapolated to zero time as in Fig. 5 (a) and the concentration of iodine was calculated by using 20 000 as the molar extinction coefficient. This determination and the following determination were performed in triplicate. The results were identical throughout. 2. Spectrophotometric determination of hydrogen pevoxide in the presence of Cetomacrogol- Standardised 0.05 M hydrogen peroxide solutions were diluted with buffered Cetomacrogol and potassium iodide solutions as described above for iodine, except that before final dilution to volume 0-05 ml of 3 per cent.ammonium molybdate solution was added to catalyse the liberation of iodine. The time of mixing hydrogen peroxide solution with Cetomacrogol and potassium iodide solution was taken as zero storage time of liberated iodine. For both iodine and hydrogen peroxide determinations, the reference cell was filled with a freshly prepared solution containing buffered 1 per cent. Cetomacrogol containing 5 per cent. of potassium iodide (iodate-free). 3. Spectrophotometric determination of hydroperoxides of etheric substances in the Presence of Cetomacrogol-Aged samples of polyethylene glycol 400 and of 10 per cent.solutions of polyethylene glycol 1500, 4000 and 600Q were diluted as described above for iodine and hydrogen peroxide and the hydroperoxides determined by the above spectrophotometric method. Spectrophotometric method f o r the determinatioa of unknown amounts of hydroperoxides in micellar solutions-The micellar solution containing the hydroperoxides was added to freshly prepared solutions of Cetomacrogol buffered with potassium iodide solution (pH 6) in amounts such that the final concentration was 1 per cent. of Cetomacrogol in 0-1 M phosphate buffer at pH 6 containing 5 per cent. of potassium iodide in flasks protected from the light and swept with purified nitrogen. The time of addition of the hydroperoxide to the potassium iodide solution was taken as zero time and the absorbances were read in stoppered, well filled cells every 5 minutes for half an hour at 370 nm, being extrapolated to zero time as in Fig.5 (a). The reference solution contained all components except the hydroperoxides. When the ab- sorbance was unsuitable, preliminary dilutions were performed to being the final concentrations of iodine within the range lo-* to Each hydroperoxide group is equivalent to 1 mol of iodine liberated and the concentration is calculated by using 20 000 as the molecular extinction coefficient. 5. Titrimetric determination of hydroperoxides in etheric materials-The hydroperoxide content of aged samples of polyethylene glycol 400 and 10 per cent. aqueous solutions of polyethylene glycol 1500, 4000 or 6000 were determined as follows.The sample was placed in a 250-ml stoppered Erlenmeyer flask, 20 ml of a mixture of acetic acid - chloroform (3 + 2) were added, the solution was purged with nitrogen and 2 ml of a freshly prepared saturated aqueous solution of iodate-free potassium iodide (nitrogen-saturated) were added as quickly as possible. The stoppered flask was then shaken for about 10 minutes. Water and starch solution were added and the sample was titrated against standard 0.01 M sodium thiosulphate solution. All determinations were carried out in triplicate. RESULTS AND DISCUSSION Fig. 2 (a) shows the ultraviolet spectrum of iodine, recorded at 25 "C, 5 and 20 minutes after the iodine was added to a buffered 1 per cent. Cetomacrogol solution in the presence of excess (5 per cent.) of potassium iodide.(See Factors that affect the rate of iodine fading in the presence of Cetomacrogol.) The spectrum shows peaks a t 294 and at 370 nm, the intensi- ties of which decrease with time. to 1 x 10-6 M of iodine. 4. M.666 AZAZ et d.: ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [Analyst, VOl. 98 , 7 0.8 - fa) fb) 0.6 - - 0-4 - I 1 1 1 1 294 300 350 370 400 294 300 350 370 400 Wavelength] nm Fig. 2. Absorption spectra of iodine in a buffered 1 per cent. Cetomacrogol solution contain- ing 5 per cent. of potassium iodide at 25 "C, measured 5 minutes (solid line) and 20 minutes (broken line) after: (a), direct addition of iodine to the system; and (b), liberation of iodine by hydrogen peroxide added to the same system Table I summarises the data reported in the literature on aqueous and micellar iodine The differences may be accounted for in part by variations in the equilibrium spectra.situations encountered. For iodine in pure water, equilibrium in the reaction I, + I- + I, lies predominantly to the left and the low-intensity bands of hydrated molecular iodine at 270 and 460 nm are observed,lg whereas in the presence of potassium iodide, equilibrium is to the right and the high-intensity bands in the 290 and 350-nm regions are due to I,. The bands are sensitive to I- concentration,ll the reaction being incomplete even at high I- ion concentration^^^ (Table I, A and B). The I, bands are not detected at the concentrations used for measurement of the 1, absorption.TABLE I ULTRAVIOLET ABSORPTION DATA FOR IODINE SPECIES IN AQUEOUS AND I N MICELLAR SYSTEMS Medium Iodine species Amax./nm cmax. Water . . .. .. . . I, 270 121 Water with excess of KI . . I, 460 704, 746 See B - See A - 13- 1,- 350 26 040 352 352 25 120 353 26 400 285 - 287-5 40 000 289 41 300 25 417 to 26 968* - Aqueous Cetomacrogol (and I, micellar 390 other non-ionic surfactants) I, + I, micellar % I,- 388 * s $ 3 § , + * § *, 1, 9 ** $ 8 § * + 360 to 390 360 293 to 294 * + 370 D + I 9 Aqueous Cetomacrogol in I, micellar 4 I,- 362 to 363 + presence of excess of KI - 370 20 6oot 293 to 294 294 36 6OOt * Absorption intensities rise with addition of potassium iodide. t Data from present work. E values by extrapolation method. 3 Amax and intensity depend on surfactant concentration.Amax. and intensity depend on storage time. Reference 16, 19, 24 - - - 16, 19, 24 11 t 24 19 24 t 16 17 12, 13, 16, t 14 4, 15, t 16, t 16 t 16 tSeptember, 19731 AND IODINE IN AQUEOUS NON-IONIC SURFACTANTS 667 In the surfactant systems (Table I, C and D), peaks are observed between 360 and 390 nm, which are very sensitive to surfactant concentration in the absence of iodide, and at 293 to 294 nm.12-16 These are probably overlapping peaks of equilibrium mixtures of hydrated and micellar iodine and triiodide, together with a molecular complex of iodine and surfactant (Amax. 390 nm), the last peak being distinctive in the vicinity of the critical micellar con- centration.16 On addition of iodide to the iodine - surfactant system the intensities rise significantly.The 390-nm peak is replaced by the peak in the 360-nm region and the spectrum is much less sensitive to surfactant concentration,l6 presumably because of the formation of I;. The degree of conversion of I, into 1, is less than in non-micellar systems because, with the surfact- ant, species other than 1, may be formed. This may account for the lower extinction co- efficient obtained in the presence of Cetomacrogol, as well as for the discrepancies in the litera- ture (reporting peaks from 360 to 370 nm), which may be due to differences in composition, such as pH and salt concentrations, which alter the equilibrium. 0.50 I I I I I I I I 0 2 4 6 Ti me / hours Fig. 3. Dependence of iodine absorbance at 370 nm on time a t 25 "C. Iodine added to buffered 1 per cent.Cetomacrogol solution containing 5 per cent. of potassium iodide A systematic study of the absorption of iodine at 370 nm, covering 8 hours in controlled conditions (1 per cent. of Cetomacrogol and 5 per cent. of potassium iodide in phosphate buffer at pH 6 and at a temperature of 25 "C) is shown in Fig. 3. After a period of about 1 hour the intensity drops drastically and might reach such a low value as to be below the optimum range of the instrument. Moreover, the intensity does not reach a constant value within 24 hours. Therefore, no single time selected during this period would be applicable to quantitative analyses ; calculations reported in the literature are therefore incorrect. On the other hand, a graph of the logarithm of absorbance at either of the observed peaks (370 and 294 nm) against storage time gives a straight line within the first half hour after addition of iodine to the system (Fig.4). Extrapolation of the line to zero storage time gives an intercept that proves to be the absorption of the amount of iodine actually added to the system. At controlled pH and temperature, the slopes of such lines differ with the Cetomacrogol batch used and with the initial iodine concentration [Fig. 5 ( a ) ] . However, the values extrapolated to zero storage time depend solely on the amount of iodine added. A log - log graph of these extrapolated absorption values against initial I, concentration is a straight line and gives an intercept representing log E , where E is the molar extinction coefficient of iodine in Cetoma- crogol [Fig.5 (b)]. It is evident from Fig. 5 (a) and (b) that Beer's law is observed within the iodine concentration range from 1 x to 1 x 10-6 M . Similar behaviour was observed at668 the 244-nm peak. 36 000 at 294 nm. stant at 1:1.18 for the first 30 minutes of storage time. subsequent studies so as to minimise possible interferences. AZAZ et al.: ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [Analyst, Vol. 98 The molar extinction coefficient was found to be 20 000 at 370 nm and As can be seen from Fig. 4, the ratio of the intensities at 370 nm and 294 nm remains con- The 370-nm peak was chosen for 0.2 I I I I 0 10 20 Time/ minutes Fig. 4. Semi-logarithmic graph of absorbance of iodine against time: A, a t 294 nm; and B, a t 370 nm (2.7 x M iodine in buffered 1 per cent.Cetomacrogol - 5 per cent. potassium iodide solution a t 25 "C) The validity of the above extinction coefficient for iodine liberated from potassium iodide by hydroperoxides present in the system is evident from Figs. 2 ( b ) and 6, and from Table 11. The spectrum of the iodine liberated from potassium iodide by hydrogen peroxide is identical with that of molecular iodine [Fig. 2 (a) and (b)]. Further, amounts of molecular iodine identical with those liberated by hydrogen peroxide from potassium iodide give coincident curves when the logarithm of absorbance is plotted against the storage time in minutes (Fig. 6). Table I1 shows that hydrogen peroxide concentrations determined by classical titrimetry are in good agreement with those obtained spectrophotometrically by extrapolating the readings at 370 nm to zero storage time at 25 "C.It is important to emphasise that the titrations were carried out on macro-scale amounts of peroxide in non-micellar stock solutions, hence the microdetermination provides an alterna- tive method of determining macro-scale amounts of peroxide in micellar solutions. Although Cetomacrogol was chosen for systematic investigation, similar linear relation- ships were obtained with other non-ionic surfactants such as Tweens. FACTORS THAT AFFECT THE RATE OF IODINE FADING IN THE PRESENCE OF CETOMACROGOL- Efect of temPerature-As the graph of the logarithm of absorbance of iodine against stor- age time is linear [Figs. 4 and 5 (a)], the slope can be used to evaluate an apparent first-order rate constant, which is convenient for evaluating the influence of factors on the rate of the reaction.Fig. 7 (a) shows that the rate constant of fading, for a given amount of iodine in the presence of Cetomacrogol and potassium iodide, rises with temperature. However, the intercept on the log-concentration axis is independent of temperature so that there is no need for temperature control in the analytical extrapolation procedure proposed, provided that solution temperatures do not change during the period of measurement, graphs obtained being linear only under such conditions. Rate constants given in Fig. 7 (a) are nevertheless specific for a given Cetomacrogol batch and for a given initial iodine concentration. Within these limitations, the apparent rate constant changes with temperature in accordance with the Arrhenius equation [Fig.7 ( b ) ] , the energy of activation for this particular system being 15.7 kcal mol-1. Batch efects-With controlled temperature, pH and initial iodine concentration, the slope of the graph of the logarithm of absorbance against time changes with the batch of Cetomacrogol used (Fig. 8). The extinction coefficient obtained by extrapolation to zero storage time was, however, independent of the batch used [Fig. 5 (a) and (b)].September, 19731 AND IODINE IN AQUEOUS NON-IONIC SURFACTANTS lo Oo0 t I t a, C m e z a n 0 10 20 30 40 6 4 u 2 669 Time/ minutes -Llog concentration Fig. 5. (u) Semi-logarithmic graph of absorbance of iodine a t 370 nm per l-cm path length against time.Intercepts show initial absorbances. Initial iodine concentrations as determined by titrimetry : 1, 1.05 x 2, 5-40 x 3, 4.50 x 4, 4.00 x 5, 2.00 x 6, 5.00 x and 7, 2.00 x M (all in buffered 1 per cent. Ceto- macrogol - 5 per cent. potassium iodide solution at 25 "C). Cetomacrogol batches: a, batch A ; 0, batch B; and A, batch C. (b) Extrapolated initial iodine absorbances vemxs initial conceri- trations corresponding to (a). The intercept of this graph (log absorbance = log E + log concentration, for a 1-cm path length) gives E = 20 000 Efect of initial iodine concentration-Fig. 5 (a) and (b) shows that a t controlled pH, temperature and with a given Cetomacrogol batch, the slope changes with initial iodine concentration without affecting the extinction coefficient obtained by extrapolation to zero time.TABLE I1 COMPARISON OF HYDROGEN PEROXIDE CONTENT AS DETERMINED IN THE PRESENCE AND ABSENCE OF CETOMACROGOL BY TWO METHODS Ultraviolet Experimental Theoretical absorption a t zero concentration from concentration (from Sample Cetomacrogol storage time ultraviolet standardised stock number batch used (l-cm cell) * absorptionliu x lo5* solution)/M x lO5t A A A A D D D D 0.501 0.794 0.318 1.047 0.603 0.50 1 1.189 0.479 2-51 3-97 1-59 5.24 3.02 2.51 5.95 2-40 2.43 3.90 1.67 5-25 3.00 2.44 6.00 2.40 * Proposed method in the presence of surfactant. t Absence of surfactant (Kolthoff and Sande1P3).670 AZAZ et d.: ASSAY OF MICRO-SCALE AMOUNTS OF HYDROPEROXIDE [AW4lySt, VOl. 98 Efect of CetomacrogoZ concentration-A minimal concentration of 0.5 per cent.of Ceto- macrogol was needed for a linear relationship of the type shown in Figs. 4 and 5 (a) to occur. Although throughout this work 1 per cent. of Cetomacrogol was used,-a range of 0.5 to 5 per cent. was also found to be satisfactory (Fig. 9). Efect ofpH-The pattern of iodine fading changes significantly from that described above when the pH extends outside the range 5.8 to 7.0. Only slight changes in the slopes of the graphs described in Fig. 5 ( a ) were observed within this range, which were without effect on the extrapolated zero storage time reading. Because of possible variations in the acidity or alkalinity of surfactant samples, all systems were buffered to pH 6.0 by using phosphate buffer.1.01 1 I I I I I 0 10 20 30 Time /minutes Fig. 6. Comparison of absorption a t 370 nm of iodine added to system (0) with that of iodine liberated from potassium iodide solution by equivalent amounts of hydrogen peroxide (a). Initial iodine concentration in buffered 1 per cent. Cetomacrogol - 5 per cent. potassium iodide solution a t 25 OC: A, 2.5 x M ; and B, 3.2 x 10-5 M APPLICABILITY OF THE METHOD- Clearly, changes in the slopes observed will occur as a result of differences in the batch of surf actant and the initial iodine concentration, whereas temperature, pH and surfactant concentration effects can be fully controlled. However, under the conditions described, slope variations have been shown not to affect the accuracy of the procedure, which invariably gives the correct initial iodine concentration on extrapolation of the absorbances measured over a half-hour period back to zero time.Table I11 summarises results obtained in the quantitative determination of hydro- peroxides present in aged samples of polyethylene glycols25 by two methods, titration in an acetic acid - chloroform medium and the spectrophotometric method in a micellar medium {see Methods). The applicability to hydroperoxides present in all the aged samples of polyethylene glycol 400 to 6000, together with the results obtained on standard peroxide solutions, establishes the technique as a general method for the analysis of macro-scale or micro-scale amounts of these hydroperoxides.September, 19731 AND IODINE I N AQUEOUS NON-IONIC SURFACTANTS 1.0.67 1 - I I I I 1 0 5 10 15 20 25 30 3-2 3-3 3.4 3-5 Time/minutes 1 xi03 T Fig. 7. (a) Rate of fading of iodine a t A, 18 "C; B, 22.5 "C; and C, 30 "C. First- order rate constants ( k ) obtained from slopes of semi-logarithmic graphs for Ceto- macrogol (batch A) and for an initial iodine concentration of 3.98 x M : 1.76 h-1 a t 18 "C; 2.63 h-1 at 22.5 "C; and 5.12 h-1 at 30 "C. (b) Arrhenius plot of results from (a). k = first-order rate constant (apparent) for fading of iodine; and T = ab- solute temperature. Energy of activation calculated from slope = 15.7 kcal mol-1 CONCLUSIONS The kinetics of iodine fading in the presence of an etheric non-ionic surfactant have been followed spectrophotometrically and factors influencing the kinetics have been examined.A procedure is given for the determination of micro-scale or macro-scale amounts of hydroperoxides in the presence of etheric non-ionic micelles based on the liberation of iodine ahd i& spectrophotometric determination o.2 ~ by the extrapolation method described. 0 10 20 3 Time/ minutes Fig. 9. Semi-logarithmic graph of iodine absorbance at 370 nm against time at three different Cetomacrogol concentrations : A, 5 per cent.; 0, 1 per cent.; and 0, 0.5 per cent. Cetomacrogol (batch A) buffered in 5 per cent. potassium iodide solution at 25 "C672 AZAZ, DONBROW AND HAMBERGER TABLE I11 COMPARISON OF HYDROPEROXIDE CONTENT OF AGED SAMPLES OF POLYETHYLENE GLYCOLS AS DETERMINED IN THE PRESENCE AND ABSENCE OF CETOMACROGOL BY TWO METHODS Hydroperoxide content/M x lo3 r -I A Determined by titration in Determined spectrophotometrically Material acetic acid - chloroform medium” in micellar medium PEG 400 2.30 2.30 PEG 400 4-45 4.50 PEG l500t 5.35 5.20 PEG 4000t 5.15 5.15 PEG 6000t 5-95 6-15 * In the absence of Cetomacrogol (Procedure 5).t 10 per cent. solutions. This method has been substantiated by comparison with titrimetric procedures on macro- scale amounts by use of both hydrogen peroxide and hydroperoxides contained in aged samples of polyethylene glycols. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. REFERENCES Lea, C. H., Proc. R. Soc., 1931, 108B, 175. -, J . Soc. Chem. Ind., Trans., 1946, 65, 286. Henderson, G., and Newton, J. M., Pharm. Acta Helv., 1966, 41, 228. , Ibid., 1969, 44, 129. Allawala, N. A., and Riegelman, S., J . Amer. Pharm. Ass., Sci. Edn, 1953, 42, 396. Hugo, W. B., and Newton, J. M., J . Pharm. Pharmac., 1963, 15, 731. Heaton, F. W., and Uri, N., J . Sci. Fd Agric., 1958, 9, 781. Mitchell, A. G., and Wan, L. S. C., J . Pharm. Sci., 1965, 54, 699. Rhodes, C. T., Can. J . Pharm. Sci., 1967, 2, 16. Kern, C. J., and Antoshkiw, T., Ind. Engng Chem., 1950, 42, 709. Custer, J. J., and Natelson, S., Analyt. Chem., 1949, 21, 1005. Carless, J. E., Challis, R. A., and Mulley, B. A., J . Colloid Sci., 1964, 19, 201. Elworthy, P. H., J . Pharm. Pharmac., 1960, 12, 293. Ross, S., and Olivier, J. P., J . Phys. Chem., 1959, 63, 1671. Osol, A., and Pines, C. C., J . Amer. Pharm. Ass., Sci. Edn, 1952, 41, 634. Woodward, R. J., Ph.D. Thesis, University of London, 1962. Ross, S., and Baldwin, V. H., jun., J . Colloid Interface Sci., 1966, 21, 284. Benesi, H. A., and Hildebrand, J. H., J . Amer. Chew. Sac., 1950, 72, 2273. Katzin, L. I., J . Chem. Phys., 1953, 21, 490. Klaeboe, P., Acta Chem. Scand., 1964, 18, 27. Johnson, R. M., and Siddiqui, J. W., “The Determination of Organic Peroxides,” Pergamon Press, Colowick, S. P., and Kaplan, N. O., Editors, “Methods in Enzymology,” Volume I, Academic Kolthoff, J. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Awtrey, A. D., and Connick, R. E., J . Amer. Chew. Soc., 1951, 73, 1842. Lloyd, W. G., J . Polym. Sci., Part A , 1963, 1, 3551. J -- Oxford, 1970, pp. 43 and 113. Press Inc., New York, 1955, p. 143. The Macmillan Company, New York, 1962. Received January 19th, 1973 Accepted April 3rd, 1973

ISSN:0003-2654    

DOI:10.1039/AN9739800663

出版商:RSC    

年代:1973

数据来源: RSC