摘要:

THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETYEDITORIAL ADVISORY BOARD"Chairman: H. J. Cluley (Wembley)'L. S. Bark (Salford)R. Belcher (Birmingham)L. J. Bellarny, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)L. R. P. Butler (South Africa)E. A. M. F. Dahmen (The Netherlands)A. C. Docherty (Billingham)D. Dyrssen (Sweden)'W. T. Elwell (Birmingham)J. Hoste (Belgium)'J. A. Hunter (Edinburgh)H. M. N. H. Irving (Leeds)M. T. Kelley (U.S.A.)W. Kemula (Poland)' G . F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)H. W. Nurnberg (W. Germany)'J. M. Ottaway (Glasgow)'G. E. Penketh (Wilton)'T. B. Pierce (Harwell)E. Pungor (Hungary)D. I. Rees (London)'R. Sawyer (London)P. H. Scholes (Sheffield)'W.H. C. Shaw (Greenford)S. Siggia (U.S.A.)A. A. Smales, O.B.E. (Harwell)A. Walsh (Australia)T. S. West (Aberdeen)A. L. Wilson (Medmenham)P. Zuman (U.S.A.)'A. Townshend (Birmingham)'Members of the Board serving on The Analyst Publications CommitteeREGIONAL ADVISORY EDITORSDr. J. Aggett, Department of Chemistry, University of Auckland, Private Bag, Auckland, NEWDr. G. Ghersini, Laboratori CISE, Casella Postale 3986, 20100 Milano, ITALY.Professor L. Gierst. Universitb Libre de Bruxelles, Facult6 des Sciences, Avenue F.-D. Roosevelt 50,Professor R . Herrrnann. Abteilung fur Med. Physik., 63 Giessen, Schlangenzahl 29, W. GERMANY,Professor W. A. E. McBryde. Dean of Faculty of Science, University of Waterloo, Waterloo, Ontario,Dr. W.Wayne Meinke. KMS Fusion Inc., 3941 Research Park Drive, P.O. Box 1567, Ann Arbor,Dr. I. RubeBka, Geological Survey of Czechoslovakia, Kostelni 26, Praha 7, CZECHOSLOVAKIA.Dr. J. Rbii6a. Chemistry Department A, Technical University of Denmark, 2800 Lyngby, DENMARK.Professor K. Saito. Department of Chemistry, Tohoku University, Sendai, JAPAN.Dr. A. Strasheim, National Physical Research Laboratory, P.O. Box 395, Pretoria, SOUTH AFRICA.ZEALAND.Bruxelles, BELGIUM.CANADA.Mich. 481 06, U.S.A.Published by The Chemical SocietyEditorial: The Director of Publications, The Chemical Society, Burlington House,London, WIV OBN. Telephone 01 -734 9864. Telex No. 268001.Advertisements: Advertisement Department, The Chemical Society, Burlington House, Piccadilly,London W1 V OBN. Telephone 01 -734 9864Subscriptions (nonmembers): The Chemical Society, Distribution Centre, Blackhorse Road,Letchworth, Herts., SG6 1 HNVolume 103 No 1223@ The Chemical Society 1978February 197

ISSN:0003-2654    

DOI:10.1039/AN97803FX005

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

ANALAO 1 03 (1 223) 1 13-1 92 (1 978)I SSN 0003-2654February 1978THE ANALYSTTHE ANALYTICAL JOURNAL OF THE CHEMICAL SOCIETY1131221281 34140149156162176179183185186189CONTENTSModified Oscillating Mirror Rapid Scanning Spectrometer as a Detector forSimultaneous Multi-element Determination-Oliver Rose, Jr., William R.Heineman, Joseph A. Caruso and Fred L. FrickeDetermination of Lead in Plastic Containers for Pharmaceutical Products byAtomic-absorption Spectrophotometry Using a Carbon Rod Atomiser-Pamela Girgis-Takla and loannis ChroneosInfrared Determination o f Quartz, Kaolin, Corundum, Silicon Carbide andOrthoclase i n Respirable Dust from Grinding Wheels-Jose L. NietoDetermination of Nickel, Cobalt and Copper by Direct Photometric TitrationBiacetyl Bis(4-phenyl-3-thiosemicarbazone) as a Reagent for the Spectrophoto-Potentiometric Determination of Copper in Palm Oil w i t h a Copper(l1) lon-Determination of Ethinyloestradiol in Single Tablets and Its Separation fromOther Steroids by High-performance Liquid Chromatography-Kay R.Bagonand E. W. HammondPart XXIV.with Cyanide-M. A. Leonard and R. Murphymetric Determination o f Copper-A. G. Asuero and J. M. Can0selective Electrode-Y. S. Fung and K. W. FungDifferential Electrolytic Potentiometry w i t h Periodic Polarisation.An Appraisal o f the Amplitude-biassed Mode-E. Bishop and P. Cofr6SHORT PAPERSDifferential-pulse Polarography of Selenium(1V) i n the Presence o f Metal lons-G. P. Bound and S. ForbesDetermination of Small Amounts of Bismuth i n Antimony(ll1) Oxide by UsingTitrimetric Determination of Halogens i n Halo- and Dihalo-p-diketones-Relationship Between Refractive Index and Specific Gravity of Aqueous GlycerolS-Benzylthiuronium Chloride as an Acidimetric Standard i n Non-aqueousAnodic Stripping Voltammetry-P.PetBk and V. KoubovfiC. Muhammad Ashraf, Riaz Ahmed, Amjad Aqeel and M. Saleem KhalidSolutions-Dov BaskerTitrimetry-Miss Neelam Khanna and K. S. BoparaiBook ReviewsSummaries of Papers in this Issue-Pages iv, v, viii, i xPrinted by Heffers Printers Ltd, Cambridge, EnglandEntered as Second Class at New York, USA, Post OfficJOURNALS * BOOKSMONOGRAPHSOrders for all publicationsformerly published by the Societyfor Analytical Chemistry shouldbe sent direct or through abookseller toTHE CHEMICAL SOCIETY,Distribution Centre,Blackhorse Road, LetchworthHerts., SG6 1 HN~ ISelected Annual Reviews ~of the Analytical Sciences ,Volume 4 ICONTENTS I'Advances in Voltammetric Techniques,' byB.Fleet and R. D. Jee'High-frequency Electrodeless PlasmaSpectrometry,' by B. L. SharpPp. vi + 73 f9.50ISBN 0 85990 204 8CS Members' price f3.00Orders should be sent direct, with remittance, or ~through your usual bookseller to- I1 THE CHEMICAL SOCIETYI Distribution Centre,Blackhorse Road, Letchworth,Herts. SG6 1 HN, England'~ enclosing the appropriate remittance.CS Members must write direct to the above address 'INOTICE TO SUBSCRIBERS(other than Members of the Society)Subscriptions for The Analyst, Analytical Abstracts and Proceedings shouldbe sent to:The Chemical Society, Distribution Centre,Blackhorse Road, Letchworth, Herts., SG6 1 HNRates for 1978The Analyst, Analytical Abstracts and Proceedings (including indexes):(a) The Analyst, Analytical Abstracts and Proceedings .. . . .. . . f99.00(b) The Analyst, Analytical Abstracts printed on one side of the paper, andProceedings . . * . . . . . . . .. . . . . . . f105.00The Analyst and Analytical Abstracts without Proceedings (including indexes) :The Analyst, and Analytical Abstracts printed on one side of the paper(c) The Analyst, and Analytical Abstracts . . .. .. .. .. . . f87,OO(d) . . f93.00(Subscriptions are NOT accepted for The Analyst and/or for Proceedings alone)Analytical Abstracts only (two volumes per year, including indexes):(e) Analytical Abstracts . . .. .. .. .. .. .. . . f67.00(f) Analytical Abstracts printed on one side of the paper . . .. .. . . f73.0

ISSN:0003-2654    

DOI:10.1039/AN97803BX007

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

iv SUhlhlAKIES OF PAPERS IS THIS I S S U E February, 1978Summaries of Papers in this IssueModified Oscillating Mirror Rapid Scanning Spectrometer as aDetector for Simultaneous Multi-element DeterminationAn oscillating mirror rapid scanning spectrometer has been modified toimprove its light throughput and resolution. Results of the application ofthis spectrometer t o simultaneous microwave-induced atomic-emissionspectrometry and simultaneous carbon furnace atomic-absorption spectro-metry are given.Keywords : Oscillating mirror rapid scanning spectrometer ; microwaue-induced atomic-emission spectrometry ; carbon furnace atomic-absorptionspectrometry ; multi-element determinationOLIVER ROSE, Jr., WILLIAM R. HEINEMAN and JOSEPH A. CARUSODepartment of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221, USA.and FRED L.FRICKECincinnati District Food and Drug Administration, 1141 Central Parkway, Cincinnati,Ohio 45202, USA.Analyst, 1978, 103, 113-121.Determination of Lead in Plastic Containers for PharmaceuticalProducts by Atomic-absorption SpectrophotometryUsing a Carbon Rod Atomiser-\ method is described for the determination of lead in a variety of plasticpharmaceutical containers by direct solid-sample analysis using a carbon rodatomiser. The recovery of added standard is measured in order to establishsuitable assay conditions for each type of plastic. I n the absence of recoverytests, low results for lead in the presence of plastics can result from mechanicalloss during ashing or because the lead has been rendered more difficult toatomise.The amounts of lead found ranged from about 0.03 to nearly1 p.p.m., with a mean relative standard deviation of 1596, and were consider-ably less than the British Pharmacopoeia limit of 50 p.p.m. Vt'here compara-tive tests were possible, results are shown t o be in good agreement with thoseobtained by measuring lead in a chelate - organic solvent extract of wet-ashedplastic.Keywords : L.ead determination ; plastic pharmaceutical contninevs ; atomic-absorption spectrophotometry ; carbon rod atomisationPAMELA GIRGIS-TAKLA and IOANNIS CHRONEOS'L5'elsh School of Pharmacy, University of Wales Institute of Science and Technology,King Edward VII Avenue, Cardiff, CF1 3SU.Analyst, 1978, 103, 122-127.Infrared Determination of Quartz, Kaolin, Corundum, SiliconCarbide and Orthoclase in Respirable Dust from Grinding WheelsA method is proposed for the solid-state, quantitative, infrared determinationof the common components found in respirable dust from grinding wheels.The entire spectral region, 1300-300 cm-', is used and the masses of com-ponents are computed by the least-squares method.The practical detectionlimit is 20 pg and the working range extends up t o 600 pg, which is sufficientfor this type of sample. Known synthetic mixtures yielded mean recoveriesof 92-108%, depending on the component under consideration, the relativestandard deviations being of the order of 5-18%.Keywords : Multi-component quavztitatiue determination; infrared spectra-scopy; respivable grinding wheel dustJOSE L.NIETOServicio Social de Higiene y Seguridad del Trabajo, Instituto Territorial de Madrid,Departamento de Higiene Industrial, Torrelaguna, 73 Madrid-27, Spain.Analyst, 1978, 103, 128-133February, 1978 SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Nickel, Cobalt and Copper by DirectPhotometric Titration with CyanideXickel(II), cobalt(I1) and copper(I1) at the 0.1 M level are titrated photo-metrically with cyanide in ammoniacal solution. The nickel - cyanidereaction shows reproducible 1 : 4 stoicheiometry and can be used for titration.The reaction with cobalt shows firm 1 : 5 stoicheiometry but is complicated byformation of oxygen-containing species. The reaction with copper showsuncertain stoicheiometry of about 1 : 4.In the nickel titration zinc does notinterfere but cobalt and copper add on, both showing 1: 4 metal - cyanidestoicheiometry .Keywords : Nickel determination ; cobalt determinatiovz ; copper determination ;pkotometric determination ; cyanideM. A. LEONARDDepartment of Analytical Chemistry, Queen's University of Belfast, Belfast,BT9 5AG.and R. MURPHYDepartment of Analytical Chemistry, Queen's University of Belfast, Belfast,BT9 5AG, and Messrs. Gallaher Ltd., 138 York Street, Belfast, BT15 1 JE.Altalyst, 1978, 103, 134-139.Biacetyl Bis(4-phenyl-3-thiosemicarbazone) as a Reagent for theSpectrophotometric Determination of CopperThe synthesis, characteristics and analytical applications of biacetyl bis-(4-phenyl-3-thiosemicarbazone) (BBPT) are described.The reaction betweencopper(I1) and BBPT has been studied by spectrophotometry. The reddishorange 1 : 1 copper - BBPT complex ( C = 12.7 x lo3 1 mol-l cm-l at 485 nmand 8.2 x lo3 1 mol-1 cm-I at 530 nm) is formed at p H 1.8-11.9, in a solutioncontaining 60% VjV of dimethylformamide. The effect of interferences wasstudied. A rapid and simple method for the spectrophotometric deter-mination of copper in a white metal, in blende and in a waste water has beendevised. The method is compared with others proposed for the spectrophoto-metric determination of copper with thiosemicarbazone reagents.Keywords : Biacetyl bis(4-plzenyl-3-thiosemicarbazone) reagent; copper detev-mination; spectrophotometryVA.G. ASUERO and J. M. CAN0Department of Analytical Chemistry, Faculty of Sciences, The University, Sevilk-4,Spain.Analyst, 1978, 103, 140-148.Potentiometric Determination of Copper in Palm Oil with aCopper( 11) Ion-selective ElectrodeThe applicability of a copper(I1) ion-selective electrode for the determination ofthe copper content of crude and hydrogenated palm oils was investigated. Dryashing was used for destroying the organic matrix and a porcelain crucible wasused as container. The recovery of copper in dry ashing and the leaching ofcopper from the internal glazing of a porcelain crucible were investigated. Acomplexing antioxidant buffer was used in order t o minimise the interferingeffect of iron(II1) on t h e determination of copper(I1). A direct potentio-metric method can be applied t o determine the copper content of palm oilat a concentration higher than 10 pg kg-I in the presence of 30 mg kg-1 of iron,Keywords : Copper determination; palm oil; direct potentionzetry ; coppev(II)ion-selective electrodeY . S . FUNG and K. W. FUNGDepartment of Chemistry, University of Hong Kong, Hong Kong.Analyst, 1978, 103, 149-155

ISSN:0003-2654    

DOI:10.1039/AN97803FP009

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

...V l l l SUMMARIES OF PAPERS I N THIS ISSUEDetermination of Ethinyloestradiol in Single Tablets andIts Separation from Other Steroids by High-performanceLiquid ChromatographyFebribary, 3978A reversed-phase chromatographic system is used to separate 17a-ethinyl-oestradio1-1,3,5(10)-triene-3,17 8-diol (ethinyloestradiol) from structurallyrelated steroidal compounds. A direct and highly sensitive method of assayis given for uncoated tablets containing ethinyloestradiol, as used for contra-ception. The method is modified for sugar-coated tablets used for relief ofmenopause symptoms and is sufficiently sensitive to detect ethinyloestradiolif it is present with other steroids as a contaminant.Keywords 1 Ethinyloestradiol determination ; tablets; high-performance liquidchromatographyKAY R.BAGON and E. W. HAMMONDDepartment of Industry, Laboratory of the Government Chemist, Cornwall House,Stamford Street, London, SE1 9NQ.Analyst, 1978, 103, 156-161.Differential Electrolytic Potentiometry with Periodic PolarisationPart XXIV. An Appraisal of the Amplitude-biassed ModeA simple device for amplitude biassing a periodic waveform without distortionhas been developed to overcome the difficulty with earlier instrumentationof obtaining more than a 2% bias. Amplitude-biassed differential electrolyticpotentiometry has been compared with zero-current potentiometry, classicald.c., symmetrical periodic and time-biassed periodic differential electrolyticpotentiometry in oxidation - reduction titrimetry for electrode reactions ofvarious rates, and in acid - base titrimetry.The behaviour of individualelectrodes with reference to zero-current and standard electrodes has beenexamined. The behaviour observed is in accord with theoretical prediction,and is analogous to the application of a d.c. offset to symmetrical periodicpolarisation. The advantages of rapid potential equilibration and extendedelectrode life between activations accrue, b u t with appreciable bias the end-point periodic signal deviates from the equivalence point.Keywords ; Differential electvolytic potentionzetry ; peviodic polarisation ;amplitude-biassed modeE. BISHOP and P. COFROChemistry Department, University of E;xeter, Stocker Road, Exeter, EX4 4QD.Analyst, 1978, 103, 162-175.Differential-pulse Polarography of Selenium( IV) in the Presenceof Metal IonsShovt PapevKeywords : Selenium(I 17) determznation; diffeel.ential-pulse polarogvaphy ;metal-ion intevferentsG.P. BOUND and S. FORBESThe Macaulay Institute for Soil Research, Craigiebucltler, .4berdeen, hR9 2Q J.Analyst, 1978, 103, 176-179February, 1978 SC'MMARIES OF PAPERS IK THIS ISSUE ixDetermination of Small Amounts of Bismuth in Antimony( 111)Oxide by Using Anodic Stripping VoltammetryShort PapevKeywords : Bismuth deternzimtion; antimony (111) oxide ; anodic strippingvoltamntetvyP. PETAK and V. KOUBOVAOre Research Institute, 147 13 Prague, Czechoslovakia.Analyst, 1978, 103, 179-182.Titrimetric Determination of Halogens inHalo- and Dihalo-P-diketonesShort PaperKeywords : Halogen determination ; halo- and dihalo-P-diketolzes ; titvimetryC.MUHAMMAD ASHRAFChemistry Department, Makerere University, P.O. Box 7062, Kampala, Uganda.RIA2 AHMED, AMJAD AQEEL and M. SALEEM KHALIDInstitute of Chemistry, University of the Punjab, Lahore, Pakistan.Analyst, 1978, 103, 183--185.Relationship Between Refractive Index and Specific Gravity ofAqueous Glycerol SolutionsShovt PaperKeyziovds : Refvactzve index ; specafic gvaz'ity ; glycevolDOV BASKERDivision of Food Technology, Agricultural Research Organization, P.O. Box 6,Bet Dagan, Israel.Analyst, 1978, 103, 185-186.S-Benzylthiuronium Chloride as an Acidimetric Standard inNon-aqueous TitrimetryShort PaperKeywords: S-Benzyltlaiuvonium chloride; non-aqueous titrimetry; acidimetvicstandardMiss NEELAM KHANNA and K.S . BOPARAISchool of Studies in Chemistry, Vikram University, Cjjain 456 010, India.Analyst, 1978, 103, 186-188x THE ANALYST February, 1.978INSTITUTE OFGEOLOGICAL SCIENCESANALYSED SAMPLESFollowing an inter-laboratory survey of ore analysis by theInstitute of Geological Sciences, a sene6 of standard ores andconcentrates is now available. The samples have been analysedby up to thirty-one laboratories in fourteen countries and afterstatistical elimination of questionable results, mean concentra-tions and 05% confidence limits have been calculated. Samplesvary in weight from forty to one hundred grams and are asfollcws:Approx. Priceweight, i; Ster-Sample Description Main analyscs grams lingIGS 20 Kickel concen- Ni ;.I%%, Cu ,5,79%, 63trate Co0.33%, Zn 0.1470/0,Fe R8.0iyo21 Sickel ore Ni 1.97 %, Cu O.iDRq;, 2022 Kickel ore Xi 1 .255%, Cu 0.1060/,, 4523 Nickel ore Ni 1.68’:;, Fe 2fi.8304, 5024 Cobalt ore Co 0.323% Cu 4.9!,$, 1025 W’olframite W :39.500/o, Sn 0.423/,, 5526 Tin-tungsten Sn 33,36O/,, W 1:L6:W/,, 452 i Molybdenum- >lo 0.2Y:/o, W 0.036%, 65(norite) C00.0690/6, Fe23.400:,(serpentinite) Co 0.0510,0, Fe 22.73:h(laterite) Cu 0.013u/b, CoO.0819;Fe 2.94:/,, Ni 0 .0 4 ~ ,,,Fe 19.43ygore Fe 12.20glo, Cu 2.110,0tungsten ore Fe l . i 6 7 ;28 Lead-zinc Pb 84.0104, Zn4.21:/,, 100concentrate Fe 11.14q0, As 4.24:;,Ag 1740 ppm,Sb 0.14%, Cu O.lSS?;,Cd 0.0211 ”/,29 Pyrolusite NnO, Y:3.:3l:/,, Ba 0.211a& 4030 Chromite Cr??,.Y2~<,Fe11.21%, 5 5Ti 0.14:;31 Ilmenite Ti32.iXo~,~e31).3601, 43Fe++ l U , l l o , o ,Fe+++ 1l.SYq;32 Rutile33 Columbite34 Tantalite35 Zircon36 Monazite3 i Uranium ore38 Baryte39 FluoriteOrders for samples, which will be despatched to overseasdestinations b y air mail.should include the appropriate pay-ments (cheques etc. should be made payable to the NaturalEnvironment Research Council Deposit Account) and should beaddressed to the Institute of Geological Scirnces, Geochemistryand Petrology Division, 64-78 Gray’s Inn Road, LondonWClX 8NG, United Kingdom and marked STAXDARDSAMPLES. Requests for further information should be madeto the same address.“ A N A L O I D ”COMPRESSED CHEMICALREAGENTSoffer a saving in the use of lab-oratory chemicals.A range of over50 chemicals includes Oxidizingand Reducing Agents, Reagents forColorimetric Analysis and Indicatorsfor Complexometric Titrations.For full particulars send for ListNO. 458 to:-RIDSDALE & CO. LTD.Newham Hall, Newby,Middlesbrough,Cleveland TS8 9EAortelephoneMiddlesbrough317216CLASSIFIED ADVERTISEMENTST h e Rate for Classified Advertisements is i 2 . p pe7 sin&column centimetre (minimum E4.60)Box Numhers are chargd an extra sop.Deadline for classifi;iipd copy is 20th of the month pmccdingmonth of issue.All space orders, copy instructions and enquiries should beaddressed to T h e Advertisement Department,The Chemical Sociefy, Burlington House, Piccadiiiy,London WrV oBNTelephone 01.734 9864 Telex 268001ADVERTISERSPLEASE NOTEXI1 advertisements forTHE ANALYSTshould from now on be addressed to our ownAdvertisement Department,The Chemical Society,Burlington House,Piccadilly, London W1V OBNTel: 01-734 9864Please send all space orders, copy, enquiries etc.to this address

ISSN:0003-2654    

DOI:10.1039/AN97803BP013

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

FEBRUARY 1978 The Analyst Vol. 103 No. 1223 Modified Oscillating Mirror Rapid Scanning Spectrometer as a Detector for Simultaneous Multi-element Determination Oliver Rose, Jr.," William R. Heineman and Joseph A. Caruso and Fred L. Fricke DepartmeBt of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221, USA Cincinnati District Food and Drug Administration, 1141 Central Parkway, Cincinnati, Ohio 45202, USA An oscillating mirror rapid scanning spectrometer has been modified to improve its light throughput and resolution. Results of the application of this spectrometer to simultaneous microwave-induced atomic-emission spectrometry and simultaneous carbon furnace atomic-absorption spectro- metry are given. Keywords Oscillating mirror rapid scanning spectrometer ; microwaue- induced atomic-emission spectrometry; carbon furnace atomic-absorption spectrometry ; multi-element determination An oscillating mirror rapid scanning spectrometer1 can be classified as a sequential linear scan spectrometer.2 The oscillating galvanometer provides a scanning wavelength window and each spectral element is detected in rapid sequence by a single photomultiplier tube. In a preliminary evaluation3 of this type of spectrometer as applied to simultaneous multi- element emission spectrometry, it was found that its photomultiplier detection system and wide optical range would make it suitable as a multi-element detector.However, modifica- tions were needed in order to improve the optical throughput and resolution. This paper describes the modified rapid scanning spectrometer (RSS) and its application to simultaneous atomic-emission spectrometry using the microwave-induced plasma (MIP) as the source and to simultaneous carbon furnace atomic-absorption spectrometry.Experimental Apparatus Detection system Basically, the detection system, consisting of the Harrick Rapid Scan Spectrometer (Model RSS-B), the RSS signal processing module, the Nicolet NMR-80 minicomputer with the FT-74 software package, was similar to that described previ~usly.~ In order to avoid a lengthy re-alignment of the optics from one application of the RSS to the next, the instrument being also used extensively in other projects as a dual-beam spectrophotometer with a xenon arc lamp source, two modifications of the optical layout of the Model RSS-B described by Denton et aZ.* were made.Firstly, a telescope entrance lens (5 cm focal length, 2.5 cm diameter), aligned with the entrance slit (S,) and the first spherical mirror (SM,), was. added with the first plane mirror removed, and secondly, a 1P28 photo- multiplier tube was inserted immediately after the exit slit. Fig. 1 is a diagram of the modified optical layout and indicates that only the optical T was found to be necessary for atomic spectrometry. A photomultiplier power supply (MPI Model MP-1031) and a laboratory-constructed signal amplifier were used. By using the 1 P28 photomultiplier tube and the separate amplifier, the signal to noise ratio was improved. (The amplifier circuit supplied by the manufacturer contained a pre-amplifier section and a log amplifier section.Only the output of the pre-amplifier section was displayed in the preliminary ~ t u d y . ~ ) This 113 * Present address: Shepherd ChemicaI Company, 4900 Beech Street, Cincinnati, Ohio 45212, USA.114 ROSE et al. OSCILLATING MIRROR RAPID SCANNING SPECTROMETER Analyst, ‘Vd. 103 improvement of the signal to noise ratio permitted the slit widths to be reduced, resulting in better resolution. Slit widths of 0.06 and 0.18 mm for entrance and exit slits, respec- tively, were used in the present emission studies. S1 SM. Red 52 Fig. 1. Diagram of modified optical layout of Harrick RSS-B rapid scan spectrometer as used for trace metal emission work: S is the emission source; S,. the entrance slit; SM,, the first spherical mirror; GM, the galvanometer mirror; SM,, the second spherical mirror; GR, the grating ; M,, the flat mirror ; S,, the exit slit; and PMT, the photomultiplier tube.In order to evaluate the performance of the RSS with optical components other than those supplied by the manufacturer, other sources including a copper hollow-cathode lamp (Tekmar) , a zinc, cadmium, lead and copper multi-element hollow-cathode lamp (Tekmar) and a low-pressure mercury vapour lamp were utilised for well defined spectral lines. The hollow-cathode lamps were powdered by a d.c. supply (Model GPS-2, Barnes Engineering Co.) and both lamps were operated at 8mA. Before spectra were recorded following a change of an optical component, the optical alignment was optimised. To improve the resolution, the grating supplied by the manufacturer, which was ruled a t 300 lines mm-1 and blazed a t 300 nm, was replaced with one ruled at 1200 lines mm-1 and blazed at 250 nm.In order to enhance the sensitivity by increasing the light throughput, and to improve the resolution by illuminating more lines of the grating, the galvanometer mirror (4.8 mm diameter) was replaced by a larger mirror (8.9 mm diameter) after a study of various galvanometer mirror sizes with respect to repetition rate. To fill the larger mirror with radiation, a larger first spherical mirror (SM,) was installed. An ultraviolet sapphire hemicylinder, 20 x 25 mm with a 10-mm radius, fitted in well with the use of the 8.9 mm diameter galvanometer mirror. In the preliminary study,3 the minicomputer determined the repetition rate of the RSS.However, it was found that identical spectra could be stored in different 1K memory loca- tions and coincide more precisely (permitting exact subtraction of background) if the galvano- meter mirror oscillated continuously. A square-wave generator (Model 126C, Exact Electronics Inc.) was then used to trigger each scan of the galvanometer mirror and each sweep of the minicomputer. Thus, while the RSS collected spectral information continu- ously, the minicomputer stored data only upon command. A comparison of the two gratings is made in Table I. Carbon cu$ vaporisation assembly - MIP system The carbon cup vaporisation assembly and the procedure for introducing analyte into theFebruary, 1978 AS A DETECTOR FOR SIMULTANEOUS MULTI-ELEMENT DETERMINATION 1 15 MIP has been described previ~usly.~ The set power level of the microwave generator was 50 W direct and 0.5 W reflected, while the optimum flow-rate of argon was 600 ml min-1. Desolvation a$@aratus - M I P system This system was similar to that described by Margoshes and Veillon6 and modified by Skogerboe and Coleman,' except the microwave cavity was an Evenson &-wave cavity.8 Small droplets of water splashed into the side-arm below the condenser that led to the quartz tube and extinguished the plasma.A piece of PTFE tape was suspended in front of the side-am opening in order to minimise this interruption of the signal. The microwave power was 90 W direct and 0.5 W reflected. The flow-rate of argon was 800 ml min-1 and the solution aspiration rate was 2.2 ml min-l.Carbon furnace atomic-absor$tion system Radiation from the zinc, cadmium, lead and copper multi-element hollow-cathode lamp operated at 8 mA was passed through the carbon rod atomiser (Model 63, Varian Techtron) and then detected by the RSS. The atomisation temperature was chosen to be the optimum for the determination of copper with a corresponding flow-rate of argon of 4 1 min-l. Reagents Aqueous standards pipetted into the carbon cup or carbon rod were diluted from stock solutions, which were either Fisher atomic-absorption standards or solutions prepared by dissolving pure metals or laboratory-reagent grade salts in an acid and diluting the solution with de-ionised water. Solutions aspirated into the desolvation apparatus were made 0.1 N in hydrochloric acid.All other reagents were of ACS reagent grade. Data Acquisition As in the preliminary study,3 the difference between the signal observed for the sample and that observed for the solvent was used as a direct measure of the emission (absorption for the carbon furnace atomic-absorption system) of the elements. For a determination using the carbon cup vaporisation assembly - MIP and the carbon furnace atomic-absorption systems, a 130-nm wavelength window was observed by the RSS with a repetition rate of 73 Hz. For both systems, the minicomputer accumulated data only when the transient signals were detected by the RSS. For the emission system, 200 spectra were accumulated, while for the carbon furnace atomic-absorption system, 140 spectra were accumulated.For a determination involving the use of the desolvation apparatus - MIP system, an 80-nm wavelength window was observed by the RSS with a repetition rate of 20 Hz and 500 spectra were accumulated after aspiration of the sample solution. Results and Discussion Evaluation of Modified RSS Grating The use of a more finely ruled grating in the RSS results in a compromise between increased resolution and reduced maximum wavelength window size as the angular displacement (excursion) of the galvanometer mirror needs to be greater to display the same wavelength window size than that with less spectral dispersion. Table I lists the maximum wavelength TABLE I COMPARISON OF PERFORMANCE OF THE RSS WITH MOUNTING OF DIFFERENT GRATINGS 300 lines mm-l grating 1 200 lines mm-1 grating Maximum wavelength window sizelnm .. . . 500* 170t Line half-width of Cu 324.8-nm linelnm: . . . . 1.9 0.4 * Limited by spectral response of 1P28 photomultiplier. t Limited by size of second spherical mirror (see Fig. 1). $ Wavelength window size 17 nm.116 ROSE f?t d . : OSCILLATING MIRROR RAPID SCANNING SPECTROMETER Analyst, VOl. 103 window size and line half-width observed when the 300 and the 1200 lines mm-l gratings were mounted with all other components unchanged. The improvement in resolution was substantial as the line half-width (full width at half maximum height) of the copper 324.8-nm line decreased from 1.9 to 0.4 nm. Fig. 2 shows the same 17-nm window of the radiation from the copper hollow-cathode lamp, which includes the 324.8- and 327.4-nm lines recorded after optical components were changed. A comparison of spectra (a) and (b) indicates the enchancement of resolution with a more finely ruled grating mounted in the RSS.However, the maximum possible wavelength window decreased from 500 to 170 nm. With the 300 lines mm-1 grating mounted in the RSS, the maximum wavelength window size was determined by the spectral response of the photomultiplier tube. For example, with the photomultiplier tube supplied by the manufacturer (RCA C31025Q), the maximum wavelength window is 730 nm, while with the 1P28, the maximum width is 500 nm. The spherical mirror, SM,, limits the maximum wavelength window size with the 1200 lines mm-l grating mounted in the RSS. Any greater angular displacement (excursion) of the galvanometer mirror does not result in an increase in wavelength window because the path of the light beam produced by the oscillating mirror is greater than the width of SM,.Galvanometer nziwor size The use of the 4.8 mm diameter galvanometer mirror supplied by the manufacturer provided a repetition rate fast enough to permit a sufficient accumulation of transient signals.3 In order to select the size of a larger mirror that would increase the light through- put and resolution, but still provide a reasonable repetition rate, the performance of several galvanometer mirrors was studied by mounting these on the same optical scanner (General Scanning, Model 0612). For example, the use of a 12.7 mm square galvanometer mirror (area 161.3 mm2), which increased the light throughput markedly, decreased the repetition rate to 16 Hz (at a scanning rate of 41 Hz) with the wavelength window of 50 nm.(Any subsequent reference to wavelength window size implies that the 1200 lines mm-1 grating was mounted in the RSS.) Any faster repetition rate caused significant distortion of the spectrum displayed. This decrease in repetition rate is due, in part, to the unfavourable ratio of mirror inertia to scanner armature inertia.9 The time required for the mirror to come to rest after each scan is much greater than that required for a smaller mirror and there is much more inertia to overcome at the beginning of each scan. The maximum repetition rate would decrease still further if the wavelength window size were increased because the rate of angular displacement would be larger.This increased rate provides greater momentum to reverse as the mirror re-sets after the scan is completed and more time is required to bring to rest the resulting increased surface distortions of the mirror. However, in applications that need only a slower repetition rate (less than 20 Hz), but high light throughput, the use of a larger galvanometer mirror can be an important modification TABLE I1 COMPARISON OF PERFORMANCE OF THE RSS WITH MOUNTING OF DIFFERENT GALVANOMETER MIRRORS 4.8 mm diameter mirror (area 18.1 mm2) 8.9 mm diameter mirror (area 62.2 mma) Maximum repetition rate/Hz* . . 73 ratelHz . . .. .. .. 101 Scan rate a t maximum repetition Relative intensity of Cu Line half-width of Cu Linearity of scan 324.8-nm line . .. . ,. 1 324.8-nm linet . . .. .. 0.4 (correlation coefficient of wavelength vevsm position plot$) . . .. 0.999 4 20 60.5 3.1 0.3 0.999 4 * Wavelength window size 130 nm. t Wavelength window size 17 nm. $ Thirteen emission lines of radiation from a low-pressure mercury vapour lamp.February, 1978 AS A DETECTOR FOR SIMULTANEOUS MULTI-ELEMENT DETERMINATION 117 to the RSS-B spectrometer. It was not possible to utilise a larger optical scanner without totally rebuilding the RSS-B. By use of a different monochromator, much larger optical scanners and corresponding mirrors (e-g., an 875-mm2 mirror such as are found on General Scanning 300 Series modelsg) could be employed. An 8.9 mm diameter galvanometer mirror was chosen as a compromise between increased light throughput and repetition rate for the RSS-B spectrometer.Table 11 gives a compari- son of the performance of the RSS when the 4.8 mm diameter and 8.9 mm diameter mirrors are mounted in the RSS with all other optical components unchanged. The light throughput was increased by a factor of 3.1 even though the actual observation time (scan period multiplied by the repetition rate) was decreased and the resolution of the RSS increased. A comparison of spectra (b) and (c) in Fig. 2 indicates the improvement of the spectral output of the RSS with the use of the larger galvanometer mirror. Spectrum (c) has been attenuated by a factor of two for display purposes. The maximum repetition rate was reduced to 20Hz with a wavelength window of 130 nm, compared with 73 Hz for the same wavelength window width with the smaller mirror mounted in the RSS.With the Model 0612 optical scanner, the larger mirror is best utilised in applications in which the observed signals are continuous so that a more favourable signal to noise ratio is obtained. It should be indicated at this point that even with the larger mirror mounted, the grating was severely under-illuminated, which appears not to be a drawback in other applicationsl94 but needs to be improved for future atomic-emission work, probably by building a new instrument. 327.4 nm a Wavelengthhm Fig. 2. Identical wave- length windows (width 17 nrn) of the emission of a copper hollow-cathode lamp. Slit widths: en- trance 0.06 mm, exit 0.18 mm. Display attenuation : (a), 16 000 counts cm-l; (b), 16 000 counts cm-l; ( c ) , 32 000 counts cm-l.Optical components mounted in the RSS: (a), 4.8 mm diameter gal- vanometer mirror, 300 lines mm-l grating; (b), 4.8 mm diameter galvan- ometer mirror, 1 200 lines mm-l grating; (t), 8.9 mm diameter galvanometer mirror, 1200 lines mm-1 grating. (a) Cd Znand Cd \/ lines unresolved 228.8 nm 214.4 nm 213.9 nm Fig. 3. Identical wave- length windows (width 35nm) of the emission of a zinc, cadmium, lead and copper hollow-cathode lamp. Slit widths: en- trance 0.06 mm, exit 0.18 mm. Optical components mounted in the RSS: (a), 4.8 mm diameter gal- vanometer mirror, 300 lines mm-1 grating; (b), 8.9 mm diameter galvan- ometer mirror, 1 200 Iines mm-1 grating.118 ROSE et aZ. : OSCILLATING MIRROR RAPID SCANNING SPECTROMETER Analyst, VoZ.103 In the preliminary study of the RSS,3 it was noted that zinc and cadmium could not be determined simultaneously as the zinc 213.9-nm and cadmium 214.4-nm lines were un- resolved. By mounting the 8.9 mm diameter galvanometer mirror and the 1 200 lines mm-1 grating in the RSS, these two lines can be resolved. Fig. 3 gives a comparison of the same 35-nm window of radiation from a zinc, cadmium, lead and copper multi-element hollow- cathode lamp recorded with different optical components mounted in the RSS. An increase in the wavelength window size decreases the resolution because fewer data points per nano- meter are accumulated. The maximum wavelength window size that can be used for multi- element determinations while still resolving the zinc 213.9- and cadmium 214.4-nm lines according to the Rayleigh criterion of resolution was 80 nm.Atomic-emission and -absorption Spectrometry After modifying the RSS spectrometer as described above, it was again evaluated as a detector in simultaneous multi-element determinations. Both the transient signals produced by the carbon cup vaporisation assembly - MIP system and the carbon furnace atomic- absorption system, and the continuous signals produced by the desolvation apparatus - MIP system, were observed. In order to obtain an enhanced signal to noise ratio, the 4.8 mm diameter galvanometer mirror was mounted in the RSS for the detection of the transient signals, although the 8.9mm diameter mirror was mounted for the detection of continuous signals.These limits are not as low as it is possible to achieve with the MIP, as Skogerboe and Coleman indicate It is likely that these high limits result from the low duty cycle (see below) as well as from the rather limited optical throughput available with the RSS-B. With the carbon cup vaporisation assembly - MIP system, bismuth, cadmium, manganese, lead, magnesium and copper were determined simultaneously; Fig. 4 shows a typical background-corrected emission spectrum of this determination. Of note is the resolution of the manganese triplet that appeared as one peak in an identical wavelength window recorded in the preliminary study.3 As a result of the intensity and randomness of the argon background, its subtraction was not precise. Zinc was determined separately owing to the spectral overlap of the zinc 213.9- and cadmium 214.4-nm lines.Typical analytical graphs for this system are shown in Figs. 5 and 6. Two sets of units are given in the abscissae as the system is mass sensitive. In this study, the sample volume used was 5 pl. The linear range is approximately one order of magnitude. This range is more compressed than is desirable; however, that is to be expected when the limited optical throughput of the system, which in part leads to higher detection limits than might otherwise be possible, is considered. Indeed, scanners capable of supporting mirrors greater than ten times the largest area that we could mount in the RSS-B are readily available with sufficient frequency response (see above). In order to utilise such scanners, however, a different instrument is required as size restrictions in the RSS-B prohibit further modification.The onset of self-absorption gives rise to a slope of about 0.5 at higher concentrations. This is not unexpected, particularly with the carbon cup experiments where the particle density in the transient pulse is likely to be high enough to give rise to this phenomenon. Table I11 lists the limits of detection for the three systems. TABLE I11 LIMITS OF DETECTION* (p.p.m.) Element Zn Bi Cd Mn Pb Mg c u Wavelength/nm 213.9 223.1 228.8 257.6 283.3 285.2 324.8 Carbon cup vaporisation assembly - MIRf 0.1 0.2 0.03 0.04 0.2 0.06 0.05 Desolvation Carbon furnace apparatus - MIP atomic absorptiont 0.07 - 0.7 - 0.02 0.000 3 0.04 - - 0.007 0.02 - 0.09 0.005 * Defined as that concentration which produces a net signal twice the standard deviation of the t Sample volume 5 p1.background signal.February, 1978 AS A DETECTOR FOR SIMULTANEOUS MULTI-ELEMENT DETERMINATION 119 Unquestionably, the MIP is capable of providing lower detection limits and extended linear ranges.10 Nevertheless, the limits of detection for the carbon cup vaporisation system and the desolvation system are in the sub-parts per million range and are mutually comparable. These detection limits are surprising in view of the low duty cycle for the analytical line (line sampling time). For example, consider a typical carbon cup experiment with a 130-nm window, a repetition rate of 73 Hz and a scanning rate of 101 Hz. Also, assume that the 220 I 300 W avel en gth/nm Fig.4. Typical background-corrected emission spectrum obtained for a simultaneous determination of bismuth, cadmium, manganese, lead, magnesium and copper using the carbon cup vaporisation assembly - MIP system. maximum residence time for cadmium in the plasma is 1 s and that 0.5 nm represents the cadmium “slice” of the 130-nm total. Based on the repetition and scanning rates, 73 spectra are taken per second with the total time per 130-nm spectrum given as 0.0099 s ( l / l O l ) . Thus, in the residence time of 1 s, a total of 0.73 s is spent in sampling the 130-nm window giving 2.8 ms “on time” for the cadmium line. Thus, while the detection levels for direct-reading plasma emission spectrographs are less, the observation time per analytical line is at least 1 000 times greater.> C a f a + m c, .= 100 + .- .- 5 10 u 1 ( Cd 228.8 nrn I I I 1 .o 10 Concentration, p.p.rn. 0.5 5.0 50.0 I 1 I Arnountlng Fig. 5. Log - log plots of analytical graphs for cadmium, copper and bis- muth obtained using the carbon cup vaporisation assembly - MIP system. i Mn257.6nm 1 0.1 1 .o 10.0 0.5 5.0 50.0 Concentration, p.p.m. I 1 I Arnount/ng Fig. 6. Log-log plots of analytical graphs for manganese and lead obtained using the carbon cup vaporisation assem- bly - MIP system.120 ROSE et al. : OSCILLATING MIRROR RAPID SCANNING SPECTROMETER AnaZyst, VoZ. 103 As the larger mirror was mounted for determination using the desolvation apparatus- MIP system, zinc and cadmium were determined simultaneously in an 80-nm window that also includes bismuth, manganese and magnesium.The emission of lead at low concentra- tion levels was obscured by background emission. Copper was determined in a separate window. The linear range is also approximately one order of magnitude for this system. Typical analytical graphs for this system are shown in Figs. 7 and 8. g 100 C Q) 4-l c Q) + .- .- = 10 a I I 0.1 1 .o 10.0 Concentration, p.p.m. .El00 c aJ C Q, m 4-l .- .- .+ - 0, 10 a 0.1 1 .o 10.0 Concentration, p.p.m. Fig. 7. Log - log plots of analytical Fig. 8. Log - log plots of analytical graphs for magnesium, manganese and graphs for cadmium and zinc obtained copper obtained using the desolvation using the desolvation assembly - MIP assembly - MIP system. system. Cadmium, lead and copper were determined simultaneously by using the carbon furnace atomic-absorption system.No data were obtained for zinc owing to the presence of zinc in the de-ionised water and the problems associated with it for furnace atomic-absorption determination.11 However, the determination of zinc is sensitive, being in the parts per lo9 range. As is to be expected, considering the sensitivity of the carbon furnace atomic- absorption method, the limits of detection for this system are superior to those for the two emission systems. However, the linear range is narrow, especially for cadmium, making an optimum dilution difficult when working with complex samples. Matrix Effects In order to study possible matrix effects for the carbon cup experiment, 500 ng of sodium (5 p1 of a 100 p.p.m.sodium solution from a stock solution of sodium nitrate) were vaporised into the plasma. As the sodium passed through the plasma the intense orange emission was noted, but the plasma appeared to be stable. The direct power increased only slightly and the reflected power did not change. After several vaporisations, the quartz tube above the plasma was coated with a white residue. Table IV gives the limits of detection of cadmium, manganese, magnesium, copper and zinc in 100 p.p.m. solutions of a sodium salt. There was little difference between these values in comparisonwith the limits of detection in aqueous solutions (Table 111). As can be seen, the concentration of the matrix element is usually more than 1 000 times that of the analyte and in one instance about 5 000 times.Five micro- litres of a 1 000 p.p.m. solution of sodium were injected into the cup and 5 000 ng of sodium were vaporised into the plasma, causing the plasma to contract appreciably. Also, there was a considerable memory effect as the sodium signal persisted for several minutes. How- ever, casual observations indicate that it may be possible to tune the M I P with a continuous flow of sodium salt solution comparable with the amount that passes through the plasma upon vaporisation of 5 OOO ng of sodium in a carbon cup. Tuning the MIP to the matrix is very important in sustaining a plasma. Skogerboe and Coleman aspirated solutions containing 1000 p.p.m. of sodium into a desolvation - MIP system without extinguishing the plasma.' However, when a 1 000 p.p.m.sodium solution was aspirated in the desolvation apparatus described above, it was not possible to sustain a plasma, although it must be mentioned that the tuner had beenFebruary, 1978 AS A DETECTOR FOR SIMULTANEOUS MULTI-ELEMENT DETERMINATION 121 TABLE IV LIMITS OF DETECTION* (p.p.m.) IN THE PRESENCE OF A MATRIX ELEMENT (100 p.p.m. OF SODIUM) Carbon cup vaporisation Desolvation Element Wavelength/nm assembly - MIP apparatus - MIP 213.9 0.1 0.4 228.8 0.04 0.06 257.6 0.02 0.2 285.2 0.1 0.3 324.8 0.1 0.2 :it Mn Mg c u t * Defined as that concentration which results in a net signal twice the standard t Zinc determined separately in the carbon cup experiment; copper determined deviation of the background. separately in the desolvation experiment.rebuilt in our laboratories and was not operating completely satisfactorily. A plasma was sustained with a 100 p.p.m. solution of sodium aspirating into the desolvation apparatus after the aspiration rate had been decreased to 0.9 ml min-1 and the argon flow-rate increased to 1.9 1 min-l. With the aspiration of the 100 p.p.m. sodium solution into the desolvation apparatus the entrance lens became coated with salt, because the stream of gas leaving the quartz tube is directed at the lens. Limits of detection for zinc, cadmium, manganese, magnesium and copper in a 100 p.p.m. solution are also given in Table IV. The higher detection limits in solution in 0.1 N hydrochloric acid in the presence of the 100 p.p.m. solution of sodium as compared with those in just the 0.1 N hydrochloric acid (Table 111) are attributed to the reduction in transport efficiency in the presence of sodium,' to changes in the plasma energy characteristics as a result of the large excess of sodium' and, to alarge extent, the decrease in the aspiration rate.Conclusions The carbon cup vaporisation assembly - MIP system has the potential for simultaneous determination when sample volume is critical. The desolvation apparatus - MIP system offers the advantage of observation of a continuous signal. For example, in this study, the larger mirror could only be mounted in the RSS for the detection of a continuous signal. The carbon furnace atomic-absorption system gives superior sensitivity of detection, but is limited by the availability of multi-element sources of radiation.This study indicates that rapid scanning spectrometry may provide a useful detector for simultaneous atomic spectrometry. However, future work must increase the light through- put still further in order to lower the limits of detection, thereby increasing the linear range and fulfilling the capabilities of the MIP.lO Also, more fully illuminated optics will enhance the resolution still further. The general concept of an oscillating mirror rapid scanning spectrometer with photomultiplier detection may well provide an answer to the growing need for an inexpensive detector for simultaneous multi-element determinations. O.R., W.R.H. and J.A.C. gratefully acknowledge partial financial support provided by the National Institute of Occupational Safety and Health and by the National Science Foundation with Grants ROH-00415 and CHE 74-02641, respectively. References Each of the three systems has distinct advantages. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Strojek, J , W., Gruver, G. A., and Kuwana, T., Analyt. Chem., 1969, 41, 481. Winefordner, J . D., Fitzgerald, J. J., and Omenetto, N., Appl. Spectrosc., 1975,29, 369. Rose, O., Jr., Mincey, D. W., Yacynych, A. M., Heineman, W. R., and Caruso, J. A., Analyst, 1976, Denton, M. S., DeAngelis, T. P., Yacynych, A. M., Heineman, W. R., and Gilbert, T. W., Analyt. Fricke, F. L., Rose, O., Jr., and Caruso, J. A., AnaZyt. Chem., 1975, 47, 2018. Margoshes, M., and Veillon, C., Spectrochim. A d a , 1968, 23B, 503. Skogerboe, R. K., and Coleman, G. N., AppZ. Spectrosc., 1976, 30, 504. Fehsenfeld, F. C., Evenson, K. M., and Broida, H. P., Rev. Scient. Instrum., 1965, 36, 294. Technical Data available on Optical Scanners, General Scanning Inc., Watertown, Mass., 02172, USA. Skogerboe, R. K., and Coleman, G. N., Analyt. Chem., 1976, 48, 611A. Evenson, M. A., and Anderson, C. T., Clin. Chem., 1975, 21, 537. 101, 753. Chem., 1976, 48, 20. Received June loth, 1977 Accepted August 25th, 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300113

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

122 Analyst, Febrztary, 1978, Vol. 103, PP. 122-127 Determination of Lead in Plastic Containers for Pharmaceutical Products by Atomic-absorption Spectrophotometry Using a Carbon Rod Atomiser Pamela Girgis-Takla and loannis Chroneos Welsh School of Pharmacy, University of Wales Institute of Science and TechLnology, King Edward V I I Avenue, Cardifl, CF1 3NU A method is described for the determination of lead in a variety of plastic pharmaceutical containers by direct solid-sample analysis using a carbon rod atomiser. The recovery of added standard is measured in order to establish suitable assay conditions for each type of plastic. In the absence of recovery tests, low results for lead in the presence of plastics can result from mechanical loss during ashing or because the lead has been rendered more difficult to atomise.The amounts of lead found ranged from about 0.03 to nearly 1 p.p.m., with a mean relative standard deviation of 15%, and were consider- ably less than the British Pharmacopoeia limit of 50 p.p.m. Where compara- tive tests were possible, results are shown to be in good agreement with those obtained by measuring lead in a chelate - organic solvent extract of wet-ashed plastic. Keywords : Lead determination ; plastic pharmaceutical containers ; atomic- absorption spectrophotometry ; carbon rod atomisation The extent to which pharmaceutical products are liable to be contaminated with lead from plastic containers is difficult to define, and comparatively little information is available concerning the content of lead in different plastic materials.In view of its physiologically harmful nature, however, the need to limit the occurrence of lead as an impurity is recognised and the British Pharmacopoeia1 specifies that the heavy metal content of certain plastic containers for pharmaceutical use should not exceed the equivalent of 50 p.p.m. of lead. The determination of lead in plastics by use of flame atomic absorption generally requires a preliminary ashing procedure, owing to the relative insolubility of plastics in most solvents. Two publications2s3 by a Sub-committee of the Analytical Methods Committee recommend wet oxidation with 50% hydrogen peroxide in the presence of hot concentrated sulphuric acid, although the recovery of lead from plastics following the wet-oxidation procedure has not been reported.The direct aspiration of aqueous solutions after ashing does not usually provide sufficient ~ensitivity,~ and the lead is therefore complexed with a suitable organic reagent495 in order to effect its extraction into a small volume of organic solvent that can be aspirated into the flame. In some instances, procedures can be made less time consuming by dissolving the plastic directly in an organic solvent. have used dimethylacetamide as a solvent for poly(viny1 chloride) that is suitable for direct aspiration into the flame. However, Olivier’ has compared the different methods of sample preparation and reported that detection limits for lead, when 2y0 solutions of plastics are aspirated into a flame, are low (10 p.p.m.), while at higher concentrations the viscosity of the solution prevents efficient aspiration.As a con- sequence of this, it appeared that the direct analysis of solid samples of plastic using an electrothermal atomiser would be the most convenient procedure. Robinson et aZ.8 have reported briefly on the determination of lead in polyethylene using a “hollow-T” atomiser. The purpose of this work has been to investigate this electrothermal atomisation method further, using a carbon rod atomiser, and to apply the method to determining the levels of lead normally found in a variety of plastic containers. Musha et Experimental Lead nitrate stock standard solzttion. This is a 1.00 mg ml-l solution of lead as lead nitrate Reagents and Apparatus in approximately 1 N nitric acid.GIRGIS-TAKLA AND CHRONEOS 123 A Varian Techtron CRA-90 carbon rod atomiser was used in conjunction with a Varian Techtron, Model 1000, atomic-absorption spectrophotometer.The CRA-90 is essentially similar to the earlier Model 63 carbon rod atomiser, which has been described briefly by Culver et al.,9 but it has been modified in order to permit temperature read-out. It was kept cool with water flowing at a rate of 2 1 min-l, and set up using standard size cup atomisers and with nitrogen as the inert gas at a pressure of 10 lb Peak signals were recorded with a Rikadenki, Model B-34, or a Varian, A-25, recorder. A single-element lead hollow- cathode lamp (Varian Techtron) was used as a line source in order to determine lead at 217.0 or 283.3 nm. A lamp current of 5 mA and a spectral band width setting of 1.0 nm were used.Systems were examined sequentially for background absorption using a hydrogen lamp (Varian Techtron) at the wavelength and band width settings used for the determination, the lamp current being adjusted so that the gain required to produce a lOOyo transmittance reading was the same as for the line source. They were also examined using the lead lamp at the non-lead resonance lines near 220 and 280 nm. Solid samples for the carbon rod were weighed by using a Stanton Micro-Balance (Model MC5). Liquid samples were injected into the cup of the carbon rod using a 5-pl Autopette injection syringe (Excalibur Laboratories Ltd.) fitted with disposable polypropylene tips. These tips were cleaned individually by drawing a solution of dilute nitric acid (20%) into them several times, followed by distilled water.The precautions necessary for trace analysis were taken at all times in order to prevent contamination. Procedure Solid-sample preparation Wash the container thoroughly with distilled water, allow it to dry in air at room tempera- ture, protected from aerial contamination, and then cut it into small pieces of approximately the required mass using stainless-steel scissors, handling the plastic with stainless-steel forceps. Weigh a cut sample accurately to the nearest microgram and rinse it well by swirl- ing in distilled water in a small beaker. Without drying, transfer the weighed sample to the atomiser cup in order to determine the lead. Selection of carbon rod atomiser control unit settings Adjust the CRA-90 control unit settings initially to the following: dry, 130 "C for 90 s; ash, 800 "C for 60 s; and atomise, 1600 "C for a hold time of 2 s and a ramp rate of 400 "C s-1.Select a sample of the type of plastic material being examined that shows not more than a negligible or small signal for lead at the x 1 sensitivity setting of the instrument and ensure, by using the hydrogen lamp, that this signal is not the result of background absorption. Inject dilute standard lead solution containing 0.25 pg ml-l of lead into the atomiser cup and deter- mine the lead absorbance at 217.0 nm. The sensitivity to lead under these conditions is about 1.4 x 10-ll g to give 1% absorption. In order to measure the recovery of the standard from the plastic, inject a further volume of the dilute standard lead solution into the cup and dry it at 130 "C for 90 s.Transfer about 1 mg, accurately weighed, of the plastic to the atomiser cup and operate the complete CRA-90 cycle (dry - ash - atomise) for the lead standard plus the plastic. Determine the recovery of the lead standard, correcting if necessary for any signal normally given by the plastic under the conditions applied. Atomise aqueous lead standards alone between solid sample additions to the cup to ensure that the normal lead response is being maintained. If the lead absorbance is found to be low in the presence of plastic, reduce the temperature of ashing and ash at the reduced temperature until the ash peak returns to zero on the recorder chart.Then complete the ashing by increasing the temperature to 800 "C and maintain this temperature for at least 30 s, or as long as is necessary to ensure that, if a second ash peak is produced, it returns to zero and remains there for at least 15-20 s before commencement of the atomisation stage. Continue in this way, reducing the initial ash temperature suitably, until further reduction does not result in an increase in the lead absorbance; extend the duration of ashing as necessary. Inadequate ashing can itself cause an apparent increase in the recovery of lead because of interference from molecular absorption, and for this reason it is always advisable to carry out a final ashing at 800 "C. If the recovery of lead is still not complete, as was found with124 GIRGIS-TAKLA AND CHRONEOS : DETERMINATION OF LEAD IN PLASTIC Analyst, VoZ.103 poly (vinyl chloride) and polypropylene, select the most suitable ashing conditions and in- crease the atomisation temperature to 1800 "C in order to overcome the deficit. Stan- dards and standard recovery tests in the presence of plastic should still give the same res- ponse at the higher atomisation temperature. Having established suitable conditions using approximately l-mg samples of plastic, determine the recovery of added standard from larger amounts of the plastic so as to establish the maximum amount of plastic that can be ashed effectively without loss of lead. Finally, ensure that the background absorption is zero when measured using the hydrogen lamp and with the larger amount of plastic in the atomiser cup.The CRA-90 control unit settings and the maximum plastic sample masses that were found to give complete recovery of lead, without background interference, from different types of plastics are shown in Table I. It is doubtful whether the recommended settings would reproduce the required conditions exactly from one instrument to another, and it would be advisable, before commencing an assay, always to establish the correct conditions in the manner described. TABLE I ASSAY CONDITIONS FOR DIFFERENT PLASTICS IN THE CRA-90 CARBON ROD ATOMISER All samples were dried a t 130 "C for 90 s before ashing. Control unit settings r 1 Atomisation temperature with hold time 2 s and Plastic mass/mg Ash ramp rate 400 "C s-l/"C Maximum sample Polyethylene .. .. . . 3.0 650 "C for 30 s 1600 Polypropylene . . .. . . 1.3 600 "C for 120 s 1800 Poly(viny1 chloride) . . . . 2.0 380 "C for 60 s 1800 Polystyrene . . .. . . 4.0 800 "C for 60 s 1600 800 "C for 30 s 800 "C for 60 s 620 "C for 60 s 800 "C for 60 s Solid-sample analysis Transfer a weighed sample of plastic to the atomiser cup and determine the lead (usually at 217.0 nm) with the sensitivity (scale expansion) suitably adjusted such that the meter reading lies within the range 0.2-0.8 absorbance unit. Inject 5-p1 volumes of a dilute standard lead solution, containing approximately the same amount of lead as the plastic sample, into the atomiser cup before and after each sample measurement in order to ensure that the stan- dard response, obtained under the same conditions as the sample response, still shows con- formity with the calibration graph.Determine the mean response for at least five sample measurements. For calculation of the lead content, refer to a calibration graph that has been prepared at the same sensitivity setting as that used for the sample measurements by using four or five standard aqueous dilutions in order to cover the range of measurement. In this work, calibration graphs were obtained for the concentration range 0.1-0.7 pg ml-1 of lead at 217.0 nm with a scale expansion of x 1 and at 283.3 nm with a scale expansion of approximately x3, and for the concentration range 0.005-0.030 pg ml-l of lead at 217.0 nm using maximum scale expansion (approximately x 17). Results and Discussion The amount of sample that can be placed in the atomiser cup for solid-sample analysis is necessarily limited by the density of the material and the size of the cup, but with the plastics examined the main limiting factor was the need to decompose and oxidise organic material efficiently without loss of lead before atomisation.The CRA-90 atomiser has a maximum ashing temperature of 1700 "C, which is normally maintained for up to 60 s, although the duration of ashing can be extended by switching off the atomiser and re-operating the cycle. Trial runs with standard amounts of lead (3.75 ng) showed that no loss of metal occurred from the CRA-90 with ashing temperatures of up to 1000 "C, but that at higher settings lower absorbance readings were obtained during atomisation, and small signals indicatingFebruary, 1978 CONTAINERS BY AAS USING A CARBON ROD ATOMISER 125 loss of lead were observed during the ashing stage.An ashing temperature of 800 "C for 60 s was therefore selected as being suitable, as it sufficed to ash completely up to about 4 mg of plastic for all of the samples examined. For atomisation, any temperature setting in the range 1500-1 900 "C was satisfactory, and 1600 "C was selected so as to prolong the life of the atomiser cup. Findlay et aZ.lO have observed that the loss rate in charring samples for lead determinations can differ between different matrices. In this work, when standard recovery tests for lead were carried out as described in the procedure, losses of lead were initially found to reach 70% in the presence of polypropylene and were also high in the presence of poly(viny1 chloride).The poor re- coveries of lead were possibly caused by mechanical loss during ashing, or else by the low atomisation efficiency resulting when polypropylene and poly(viny1 chloride) are incompletely ashed. As is shown in Table I, it was found necessary with both of these plastics to commence ashing at temperatures below 800 "C, and to operate the ashing stage three times. Poly- ethylene and polystyrene, however, could be ashed a t 800 "C for 60 s with good recoveries of lead, although with amounts of sample greater than 2 mg it was preferable to ash poly- ethylene at a lower temperature for a brief initial period. In all instances, it was found to be advisable to carry out the final ashing at 800 "C in order to avoid the appearance of a background signal in the atomisation stage.Low recoveries of lead also resulted from the lead becoming more difficult to atomise after being present with the plastic during ashing. Thus, with polypropylene and poly(viny1 chloride), it was found necessary to increase the atomisation temperature from 1600 "C to 1800 "C in order to obtain complete recovery of the lead. TABLE I1 LEAD FOUND IN VARIOUS PLASTICS BY DIRECT SOLID-SAMPLE ANALYSIS IN THE CRA-90 Plastic Standard Lead found, p.p.m. deviation I-A-, (9 degrees Scale Wavelength/ Mean Range of freedom), expansion nm p.p.m. Brown polyethylene bottles x l (sample 1) . . . . . . .. x 3 Brown polyethylene bottles x l (sample 2) . . . . .. .. x 3 Poly(viny1 chloride) tubing .. .. x17 White polyethylene bottles . . .. x17 White polypropylene bottles .. x17 Amber polystyrene vials . . .. x17 217.0 283.3 217.0 283.3 217.0 217.0 217.0 217.0 0.74 0.92 0.46 0.51 0.032 0.089 0.082 0.040 0.65-0.84 0.73-1.05 0.41-0.51 0.45-0.54 0.0 1 7-0.05 1 0.067-0.10 7 0.067-0.092 0.026-0.052 0.07 0.11 0.03 0.04 0.012 0.015 0.008 0.009 The results obtained by the direct solid-sample analysis procedure with samples taken from different types of plastic container are shown in Table 11. No correction was necessary for background absorption in any instance. The assay results were confirmed approxi- mately by the standard additions method. The atomisation peaks were sharp in all instances, and were the same for both solid-sample and solution atomisation.The plastic samples containing less than 0.1 p.p.m. of lead were determined at 217.0 nm with maximum scale expansion (approximately x 17). Careful sample preparation and handling were essential when the very high sensitivity setting was used. Thus, it was found that when samples were rinsed after being weighed, instead of being placed directly into the atomiser cup, the improvement in precision was more than two-fold. As both types of brown polyethylene bottle, which had been obtained from the same supplier, had relatively high contents of lead, it was found possible to analyse them at 217.0 nm and at the less sensitive 283.3-nm line, because at the latter wavelength ashing peaks are negligible and there is less likelihood of background interference.Results could not be compared at the two wavelengths using the same sensitivity setting because of the limitation on sample size, and for comparable amounts of sample it was necessary to work with a scale expansion of about x3 at 283.3 nm while the x 1 setting was used at 217.0 nm. The agreement between results obtained at the two wavelengths was good with one sample, but less so with the second sample, which showed a higher standard deviation of results at both wavelengths. The level of noise was negligible at both settings.126 GIRGIS-TAKLA AND CHRONEOS: DETERMINATION OF LEAD IN PLASTIC Analyst, VoZ. 103 In order to establish the accuracy of results obtained by direct solid-sample analysis, the lead content of the two brown polyethylene samples was also determined by a different method, which involved wet ashing weighed samples of the plastic with sulphuric acid and hydrogen peroxide solution,2 and extracting lead from the acidic solution into 5 ml of a solution of diethylammonium diethyldithiocarbamate in xylene, as described by Ro~chnik.~ Reaction blanks were carried out at the same time under the same conditions, but omitting the plastic.The lead atomic absorption was measured a t 217.0 nm by use of the CRA-90 atomiser, and the content calculated by reference to a calibration graph that had been pre- pared by subjecting solutions containing 0-3.0 pg of lead to the extraction procedure. In many of the brown polyethylene bottles the pigmentation was not even, and in order to obtain corroborative results it was found necessary to select samples in which the brown pigment was, as far as possible, uniformly distributed.The direct solid-sample analysis procedure was then carried out, using portions taken at random from different parts of the bottles in order to establish the uniformity of lead distribution, so that these results could be compared with those obtained by the wet-oxidation procedure. A small sample of uniform appearance was then selected from each of the larger samples and suitable amounts were analysed alternately by the direct solid-sample procedure and by a standard additions method. The latter was carried out by injecting dilute standard lead solution containing 0.2 pg ml-l of lead into the atomiser cup and evaporating it to dryness before adding the plastic sample, and was carried out by using two different mass ranges of plastic. The lead content of the plastic was calculated by deducting the amount of lead added in the standard solution from the total amount of lead found.The results obtained are shown in Table 111. Results obtained TABLE I11 LEAD FOUND IN PLASTIC SAMPLES BY VARIOUS METHODS Lead content, p.p.m. Sample Method of analysis samplelmg Mean Range Amounts of r-*-, Brown Direct solid sample polyethylene using random (sample 3) bottles samples 0.7 15-2.278 Direct solid sample using selected sample 0.982-1.901 Standard addition to solid sample using selected sample 0.657-1.346 Wet oxidation and chelate - xylene extraction 1.385 0- 2.679 6 g Brown Direct solid sample polyethylene using random (sample 4) bottles samples 1.085-2.825 Direct solid sample using selected sample 1.230-1.945 Standard addition to solid sample using selected sample 0.688-1.591 Wet oxidation and chelate - xylene extraction 1.277 0- 2.558 6 g 1.18 1.16 1.04 1.18 0.82 0.84 0.81 0.91 0.90-1.47 0.8 7-1.32 0.85-1.28 0.93-1.41 0.54-1.15 0.60-1.11 0.6 7-0.98 0.80-1.02 Degrees of freedom 16 9 9 4 37 9 9 4 Standard deviation, p.p.m.0.18 0.16 0.13 0.19 0.17 0.19 0.12 0.08 by the standard additions method were 90 and 96%, respectively, of those obtained by the direct solid-sample analysis procedure using small selected samples. Results obtained by direct solid-sample analysis using random samples were 100 and SO%, respectively, of those obtained by the wet-oxidation method. The wet-oxidation method could not be applied to plastics with a low lead content (less than 0.1 p.p.m.) because blank readings were too high.127 February, 1978 CONTAINERS BY AAS USING A CARBON ROD ATOMISER References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. “British Pharmacopoeia 1973,” HM Stationery Office, London, 1973, p . A129. Analytical Methods Committee, Analyst, 1967, 92, 403. Analytical Methods Committee, Analyst, 1976, 101, 62. Roschnik, R. K., A n d y s t , 1973, 98, 596. Analytical Methods Committee, Analyst, 1976, 100, 899. Musha, S., Munemori, M., and Nakanishi, Y., Japan Analyst, 1964, 13, 330; Analyt. Abstr., 1966, 13, Olivier, M., Atom. Absorption Newsl., 1971, 10, 12. Robinson, J . W., Wolcott, D. K., and Rhodes, L., Analytica Chim. A d a , 1975, 78, 285. Culver, 8. R., Lech, J. F., and Pradhan, N. K., Fd Technol., Chicago, 1975, 29, 16. Findlay, W. J., Zdrojewski, A., and Quickert, N., Spectrosc. Lett., 1974, 7, 355. 2489. Received February l l t h , 1977 Amended September Bth, 1977 Accepted October 6th, 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300122

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

128 Analyst, February, 1978, Vol. 103, p p . 128-133 Infrared Determination of Quartz, Kaolin, Corundum, Silicon Carbide and Orthoclase in Respirable Dust from Grinding Wheels Jose L. Nieto Servicio Social de Higiene y Seguridad del Trabajo, Instituto Territorial de Madrid, Departamento de Higiene Industrial, Torrelaguna, 73 Madrid-27, Spain A method is proposed for the solid-state, quantitative, infrared determination of the common components found in respirable dust from grinding wheels. The entire spectral region, 1300-300 cm-l, is used and the masses of com- ponents are computed by the least-squares method.. The practical detection limit is 20 pg and the working range extends up to 600 pg, which is sufficient for this type of sample. Known synthetic mixtures yielded mean recoveries of 92-108y0, depending on the component under consideration, the relative standard deviations being of the order of 5-1 8 % .Keywords ; Multi-component quantitative determination ; infrared spectro- scopy; respirable grinding wheel dust The infrared spectroscopic technique has been little used for the quantitative multi-com- ponent analysis of solid ~ubstances.~-~ Among the reasons4s5 for this are scattering, in- homogeneity of the components in the potassium bromide pellet, pick-up of moisture and changes in the crystalline structure with pressure. Another reason is the change in the band contours of the components as a result of different particle size distribution, which requires controlled grinding procedures for the samples if it is to be avoided .With respirable dust samples this last problem does hot arise. However, the most restrictive limitation for its use is that it is necessary to know which components can be present in the dust. Granite dust is a typical example, and some works has been carried out with this type of dust, using non-overlapping bands in order to obtain the concentrations of each component. In Spain, the grinding wheels are composed of binders (kaolin, feldspar and organic substances) and abrasives (quartz, carborundum and corundum). The determination of quartz by use of infrared spectroscopy in a sample of that type of dust is not possible if the classical base-line technique at 795 cm-1 is used. The presence of carborundum and corundum seriously affects both the absorbance of the base line and the 795 cm-l peak, so that no measurements can be taken.Nevertheless, if the over-all contour of the spectrum is used, it is possible to determine the concentrations of the components simultaneously. The aim of the present work was the simultaneous determination of five components (quartz, kaolin, corundum, carborundum and orthoclase) in samples of respirable dust arising from grinding wheels by collection of the dust on filters. Dust from grinding wheels is of importance in. industrial hygiene. Experimental Calibration Mixtures The finer fraction was then separated by means of sedimentation in ethanol. The particle size distribution was measured by direct counting, using scanning electron microscopy, and is shown in Fig. 1. This distribution was confirmed in an X-ray dispersion particle size analyser.Stock mixtures (0.5%) were prepared for the five components mentioned above by dilution in potassium bromide, followed by mechanical mixing and then manual grinding in an agate mortar. The components were mechanically ground for 4 h in a tungsten carbide ball mill. The mixtures were dried in a furnace at 200 "C.NIETO 129 100 80 4 F 8 ai 60 .- v) a, D 3 C .- + 40 s n a a, 4 4 rn L .- 20 0 0.8 1 2 3 4 6 8 1 0 20 30 I l l 1 I Equivalent spherical d iameter/pm Fig. 1. Particle size distribution of the components used as standards. A total of 65 duplicate calibration mixtures of the same mass (300 mg), containing known The Pellets were made in the usual form, amounts of the components, were prepared by taking aliquots of the stock mixtures.mass of each component was in the range 0-500 pg. weighed and scanned in the spectrophotometer without attenuation in the reference beam. Apparatus, Materials and Reagents Double-beam infrared spectrophotometer, Perkin-Elmer, Model 325. Computer facilities, C I I Model IRIS-50. M u B e furnace, Heraeus Model KR-170. Laboratory press, 30-ton. Evacuable die, 13 mm. Porcelain crucibles. Agate mortar and pestle. Mixer, Lab Line Instruments. Ball mill, Fritsch Model Pulverisette-6. Membrme jilters, PVC, 37 mm diameter, 5.0 pm pore size, Mine Safety Appliances Ty$e Potassium bromide, Merck Uvasol grade. Quartz, Merck reagent grade. Aluminium silicate, Riedel de Haen reagent grade. Aluminium oxide, Carlo Erba reagent grade. Carborundum, Prolabo reagent grade.Potash feldspar (orthoclase), mineral specimen, Natural Science Museum, Spaia. FWS-B. Method The amounts of each component can be obtained from the spectrum contour. These amounts give rise to a computed spectrum that fits the experimental one as near as possible in a least-squares The entire 1300-300 cm-1 region of the infrared spectrum is used.130 NIETO : INFRARED DETERMINATION OF COMPONENTS Analyst, VoZ. 103 sense. Particle size effects in several bands are minimised when the entire spectral region is used. The mathematical treatment necessary for the spectrophotometric analysis of a multi- component mixture is well known.'-ll A Fortran program, which performs the calculations, has been written. Transmittances were measured at 10-cm-l intervals, resulting in a total of 100 points per spectrum.The transmittance at 1500 cm-l, where an absorbance minimum exists for every component, was taken as the reference transmittance for computing absorbances. The alternative of using an internal standard that supplies a reference transmittance was discarded on the basis of not including another component in the mixture. Pellet opacity due to moisture causes light scattering and therefore causes changes in the base-line straight- ness. This factor is not significant, provided that special care is taken in drying the potassium bromide and the components. Care should also be taken to avoid eventual displacement of the over-all spectrum as a result of incorrect wavelength setting. Fig. 2 shows the infrared spectrum for the components isolated. The overlapping bands at 795 cm-1 can be clearly observed.This overlap is the reason for the inapplicability of the classical base line and maximum absorption method12 for the determination of quartz in a multi-component type of sample. Zarborundum Corundum Frequencylcm-' Fig. 2. Infrared spectra of the individual components of grinding wheels. Dust samZples Respirable dust samples can be collected on PVC filters by using a particle size selector (cyclone) and an aspirating device (personal pump sampler). Low mass (less than 1 mg) and fine particle size (less than 10 pm) are the most important features of these dust samples. After sampling, the filter is weighed on an analytical balance and the mass of the dust obtained by difference.Then, it is transferred to a porcelain crucible and ashed at 380 "C for several hours in a muffle furnace in order to eliminate both the filter itself and the organic matter. At that temperature no appreciable change is observed in either the kaolin infrared spectrum or in that of the other components; the filter ash shows no infrared background absorption.February, 1978 OF GRINDING WHEELS IN RESPIRABLE DUST 131 Losses are minimised if the crucible is partially covered and if the muffle temperature is raised gradually in order to avoid instantaneous ignition of the filter. Next, 300 mg of potassium bromide are added to the crucible and mechanically mixed with the ash; the whole is then transferred to the agate mortar and a pellet is made in the usual manner.Finally, the pellet is scanned in the infrared instrument. Results The matrix of the absorptivity of each component at each frequency was compiled, but is not included here because of its size. The relative standard deviation of each absorptivity was also computed, resulting in values of the order of 5% for the frequencies at which absorp- tion bands exist. An accuracy test was prepared by use of ten synthetic mixtures of known composition. The results obtained in the simultaneous determination of the components of the mixture are shown in Table I. The recoveries are expressed as percentages of the nominal value; also included is the mean recovery for the ten mixtures and for each component. The best results were obtained for corundum, with a mean recovery of 98% and a small standard deviation of &5%.The worst was for orthoclase, with a mean recovery of 92% and a standard deviation of &18%. For the components the nominal masses of which in the mixtures were zero, a small computed mass was obtained, less than 0.02 mg in every instance. The typical good agreement between observed and computed spectra for the test mixtures is shown in Fig. 3. Correlation coefficients obtained for the mixtures ranged from 0.985 to 0.997. TABLE I RECOVERIES OBTAINED IN THE ANALYSIS OF KNOWN SYNTHETIC MIXTURES Quartz Kaolin Corundum Carborundum Orthoclase & & & t-A-, ,-*-> Mixture Present/ Recovery, Present/ Recovery, Present/ Recovery, Present/ Recovery, Present/ Recovery, number pg % Pg % Pg % CLQ % CLg % 1 2 3 4 5 6 7 8 9 10 - 500 96 330 103 - 173 76 167 - 167 150 120 150 150 108 150 150 99 150 150 89 150 - - - - - - - - - - - - - - 170 101 82 107 167 99 98 103 91 81 101 104 105 - 150 99 101 176 93 - - - - - - - - - - 170 59 L - 332 109 173 91 - - 167 95 101 100 - - 100 104 100 121 100 103 100 103 100 108 100 114 100 83 - - - - Mean* 99& 14 9 6 h l l 9 8 t b lOS&ll 92&18 * Mean recovery & standard deviation for the ten mixtures.Lower accuracy is to be expected with dust samples than with synthetic mixtures, owing to the presence of some unexpected compounds in low concentration, whose effect will be to give a worse fit. Particle size distribution differences between the standards and the respirable dust samples can affect the results, but only slightly, because both distributions are rather similar and, although some bands are affected by the particle size, all of them are included in the calculation. As an example, Fig.4 shows the agreement between observed and com- puted spectra for a respirable dust sample collected from a place in which grinding wheels are used. The computed masses for this sample are shown in Table 11. Accuracy and reproducibility could not be tested with filter samples because of the lack of suitable standard filters with guaranteed masses of the five components deposited on them. In certain regions of the spectrum the fit is not as good as above.132 NIETO: INFRARED DETERMINATION OF COMPONENTS Analyst, VoZ. 103 0.6 OI c * 0.4 51 a -0 0.2 0.0 300 500 700 900 1100 1300 Frequency/cm-’ Fig.3. mixture. calculated spectrum. Calculated and observed spectra of a typical unknown synthetic The full line shows the observed spectrum and the broken line the 0.4 I I - a C 4 0.2 - 2 - 51 0.0 - 300 500 700 900 1100 1300 F req u e nc y Icm -’ Fig. 4. Calculated and observed spectra of an actual dust sample. The full line shows the observed spectrum and the broken line the calculated spectrum. Discussion It is necessary to note that all mineral compounds other than the five mentioned, if present in appreciable amounts, will interfere in this type of analysis, when the over-all infrared spectrum is being used. This is not a very restrictive condition when grinding wheel dust is to be analysed. It is not important if some of the five components are missing from the dust, because the calculated masses for them will be nearly zero; the masses for the remainder of the components will be obtained simultaneously.TABLE 11 COMPUTED MASSES FOR THE RESPIRABLE DUST SAMPLE, THE SPECTRUM OF WHICH IS SHOWN IN FIG. 4 Quartz .. .. .. .. < 0.02 Component Masslmg Kaolin . . .. .. .. < 0.02 Corundum .. * . .. 0.20 Carborundum . , .. .. 0.10 Orthoclase .. .. .. (0.02February, 1978 OF GRINDING WHEELS IN RESPIRABLE DUST 133 If a good fit is obtained, it is not a complete guarantee that exact values have been obtained for the masses of the components,l3 Nevertheless, when the overlapping among the intense bands of the components is not high, good results can be obtained. In the present work quartz, kaolin and orthoclase have an intense band at 1100 cm-l, while corundum and carborundum do not.It is significant that the last two give rise to minor dispersions in the results (see Table I). Another factor that also influences the results is the minimisation of the squares of the absolute differences in absorbance between observed and computed spectra, rather than relative differences. This gives rise to higher statistical weight for the intense bands. A minimisation of the relative differences at each frequency could give better results, but in any case, from the point of view of industrial hygiene samples, this is not the main pr0b1em.l~ I thank Miss A. Moreno for her invaluable help in the experimental work, and Mrs. C. Serrano for her co-operation with programming. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. References Hunt, J. M., and Turner, D. S., Analyt. Chem., 1953, 25, 1169. Lyon, R. J. P., Tuddenham, W. M., and Thompson, C. S., Econ. Geol., 1959, 54, 1047. Step, P. A., Kovach, J . J., and Karr, C., Analyt. Chem., 1968, 40, 358. Baker, A. W., J . Phys. Chem., 1957, 61, 450. Duyckaerts, G., Analyst, 1959, 84, 201. Cares, W. J., Goldin, A. S., Lynch J . J., and Burgess, W. A., Am. Ind. Hyg. Ass. J., 1973, 34, 298. Barnett, P., and Bartoli, A., Analyt. Chem., 1960, 32, 1153. Sternberg, J. C., Stillo, H. S., and Schwenderman, R. H., AnaZyt. Chem., 1960, 34, 84. Zscheile, F. P., Murray, H. C., Baker, G. A., and Peddicord, R. G., Analyt. Chem., 1962, 34, 1776. Sustek, J., Analyt. Chem., 1974, 46, 1676. Blackburn, J. A., Analyt. Chem., 1965, 37, 1000. Larsen, D. J., Doenhoff, L. J., and Crable, J. V., Am. Ind. Hyg. Ass. J., 1972, 33, 367. Vandeginste, B. G. M., and de Galan, L., Analyt. Chem., 1975, 47, 2124. Anderson, P. L., Am. Ind. Hyg. Ass. J., 1975, 36, 767. Received July 12th, 1977 Accepted September 14th, 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300128

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

134 An,alyst, February, 1978, Vol. 103, $p. 134-139 Determination of Nickel, Cobalt and Copper by Direct Photometric Titration with Cyanide M. A. Leonard and R. Murphy Department of Analytical Chemistry, Queen’s University (f Belfast, Belfast, BT9 5AG Defiartment of Analytical Chemistry, Queen’s University of Belfast, Belfast, BT9 5AG, and Messrs. Gallaher Ltd., 138 York Street, Belfast, BT15 1JE Nickel(II), cobalt(I1) and copper(I1) at the 0.1 M level are titrated photo- metrically with cyanide in ammoniacal solution. The nickel - cyanide reaction shows reproducible 1:4 stoicheiometry and can be used for titration. The reaction with cobalt shows firm 1:5 stoicheiometry but is complicated by formation of oxygen-containing species. The reaction with copper shows uncertain stoicheiometry of about 1:4.In the nickel titration zinc does not interfere but cobalt and copper add on, both showing 1 :4 metal - cyanide stoicheiometry . Keywords : Nickel determination ; cobalt determination ; copper determination ; photometric determination ; cyanide Nickel(I1) and cobalt(II), in the presence of excess of cyanide, stoicheiometrically form the complex ions [Ni(CN) J2- and [Co(CN),(H20)I3-. Before the introduction of EDTA, nickel and cobalt were determined by reaction with a known excess amount of cyanide in mildly ammoniacal solution followed by back-titration of the remaining cyanide with standard silver nitrate solution. Visual turbidimetric or potentiometric equivalence point indication was used. When ammoniacal solutions of nickel, copper and cobalt are titrated with cyanide, marked colour changes occur and it was decided to see if such reactions could form the basis of useful photometric titrations of these metals.The use of cyanide as titrant would introduce selec- tivity towards “B type” and “right-hand side transition” metal ions, an advantage in com- parison with the broad-spectrum titrant EDTA. Use of dilute ammonia solution as solvent would introduce further selectivity. Experimental Reagents Potassium cyanide solution, 0.4 M. Prepared from fresh solid reagent and standardised by potentiometric titration against standard silver nitrate using silver and mercury(1) sulphate electrodes. Nickel sulphate, copper(II) sulfihate and cobalt(I1) nitrate solutions, 0.1 M. The solutions were standardised against standard EDTA solution.Instrumentation Solution absorption spectra were produced on a Pye-Unicam SP8000 spectrophotometer. Titrations were performed on an EEL photometric titrator using Ilford standard spectrum filters, a 40-ml cell and a 10-ml microburette. Results Titration of Nickel Upon addition of cyanide a smooth transition to the yellow tetracyano complex occurs as shown by the regular change in absorption spectrum (Fig. 1). Fig. 2 shows nickel - cyanide titrations at 575 nm and using various over-all concentrations of ammonia. A C, of at least 1 M is required in order to avoid precipitation of insoluble nickel complexes. The hexaamminenickel( 11) complex [Ni(NH3),J2+ is coloured blue with A,,,. 582 nm.LEONARD AND MURPHY 135 0.8 (u m 0.6 e en 2 0.4 0.2 I 400 500 600 700 80 0 0 W avel engthh m Fig.1. addition of cyanide. 40 mm. and E, 1:4. Change in the absorption spectrum of [Ni(NH3),]a+ upon the CNi = 0.0402 M ; CNns = 1.0 M ; cell path length = Molar ratios of Ni to CN: curve A, 1: 0; B, 1: 1; C, 1 : Z ; D, 1: 3; A small, but distinct and reproducible, break occurs at the 1 : 2 nickel to cyanide molar ratio and the final end-point break seems always to occur about 0.7% low, based on 1 : 4 stoicheio- metry. 0 4 8 12 16 Amount of 0.3944 M CN-/ml Fig. 2. Photometric titration of nickel with cyanide in ammoniacal solution. Filter 606 ( hmax.T = 575 nm). Taken, 14.0 ml of 0.100 4 M NiSO, solution plus n ml of concentrated NH, solution. Starting volume, 26 ml. Curve A, n = 1 ml; B, 12 = 2 ml; C, n = 4 ml; and D, n = 8 ml. Theoreti- cal end-point for Ni to CN ratio 1 : 4 = 14.25 ml.It is unexpected that the initial absorbance decreases with increase in ammonia concentra- For tion but the final end-point is independent of this variable. log p4 = 31.3 and for Ni2+ + 4 CN- + [Ni(CN),I2- Ni2+ + 6 NH, e [Ni(NH,),12+136 LEONARD AND MURPHY: DETERMINATION OF NICKEL, COBALT AND log p6 = 8.49,1 where ,8 is the over-all formation constant. readiness with which cyanide can displace ammonia in nickel complexes. Analyst, VoZ. 103 This difference illustrates the Titration of Cobalt The reaction of hexaamminecobalt (11) with oxygen : 2 [Co(NH3)6I2+ + 0 2 [(NH3)5Co(02)Co(NH3)~14+ + 2NH3 causes complications. occurs with the cyanocobaltate(I1) complex in the presence of oxygen: The colour change is froin pink to dark brown.A similar reaction 2[Co(CN)5(H@)J3- + 0 2 + [(C~)5C0(02)C0(CN),16- the colour change being from yellow-green to brown. Because of these reactions all cobalt titrations were carried out under nitrogen. The absorption spectra of [CO(NH,),]~+ and [Co(CN),H20I3- are shown in Fig. 3. The high absorbance of [Co(CN),(H20)I3- at 800 nm is rapidly and completely destroyed by oxygen. 400 500 600 700 800 Wave1 engthhm Fig. 3. Absorption spectra of [Co(CN),(H,0)l3- and [Co(NH,),12+. For the cyanide complex Cc0 = 0.008 M ; Cm3 = 1.4 M ; and Cm = 0.048 M. For the ammine, Cco = 0.008 M ; and CNH3 = 3.7 M. Cell path length = 20 mm. Both solutions kept under nitrogen. Fig. 4 shows the titration of hexaamminecobalt(I1) with cyanide at 435 nm; a set of curves increasing in absorbance is evident, as would be expected from the absorption spectra. An over-all ammonia concentration greater than 2 M is required in order to avoid precipitation.For a range of ammonia concentrations distinct breaks occur at 1 : 5 stoicheiometry; the reaction is [CO(NH,)6]2+ + 5CN- + [Co(CN)5(H20)I3- + 6NH3 or Precipitation is presumably due to the formation of uncharged species such as [Co(NH,) 4(CN)2]0 or mixed hydroxy complexes. Titration of cobalt at 550nm, as would be anticipated from the spectra, gives a slowly decreasing absorbance but this fall is terminated by an upward step at 1 : 5 stoicheiometry. At wavelengths above 640 nm (filter 608) a titration curve similar to that obtained at 435 nm is found.Again an acceleration of absorbance increase is evident as 1 : 5 stoicheiometry is approached. [Co2(CN),016- Such precipitates readily dissolve in excess of cyanide. Titration of Copper colourless transition are shown in Fig. 5. The absorbance decrease at regular with cyanide addition. molar ratio, caused presumably by the insolubility of [Cu(NH,),(CN),]O. The absorption spectra of tetraamminecopper( 11) -cyanide complexes illustrating the blue to 605 nm is not Precipitation tends to occur at the 1 : 2 copper to cyanideFebruary, 1978 COPPER BY DIRECT PHOTOMETRIC TITRATION WITH CYANIDE 0.8 0.6 a S + 2 0.4 2 0.2 t- 137 0 1 .o 2.0 3.0 Amount of 0.394 4 M CN-/mi Fig. 4. Photometric titration of cobalt with cyanide in ammoniacal solution under nitrogen.Filter 601 ( h m a x . ~ = 435 nm). Taken, 2.0 ml of 0.100 2 M Co(NO,), solution plus n ml of concentra- ted NH, solution. Starting volume, 30 ml. Curve A, n = 4 ml; B, n = 6 ml; and C, n = 10 ml. Theo- retical end-point for Co to CN ratio 1 : 5 = 2.54 ml. Copper - cyanide titration curves are shown in Fig. 6. The end-points, as determined by extrapolation of the linear portions, vary somewhat with over-all ammonia concentration but the general stoicheiometry is 1 : 4, ie., the reaction is basically [CU(NH,),]~+ + 4CN- + e + [CU(CN),]~- + 4NH3 For Cu+ + 4CN- + Cu(CN)i- log f14 = 30.3 and for Cu2+ + 4NH3 + Cu(NH3)2,+ log fl, = 12.6. These values show the ability of cyanide to displace ammonia but it will be seen from Fig. 6 that end-point curvature increases with increasing ammonia concentration , thus illustrating the decrease in the conditional formation constants of copper - cyanide species with increase in ammonia concentration.1 .o A 0.8 a 0.6 e B a 0.4 0.2 0 Tendency to / ' I precipitate \ \\ 500 600 700 WavelengWnm 800 Fig. 5. Changes in the absorption spectrum of [Cu(NH,),I2+ upon the addi- tion of cyanide. CcU = 0.004 M; C,, = 0.30 M ; cell path length = 40 mm. Molar ratios of Cu to CN: curve A, 1 : O ; B, 1: 1; C, 1:2; D, 1 : 3 ; and E, 1:4.138 LEONARD AND MURPHY: DETERMINATION OF NICKEL, COBALT AND Analyst, VoZ. 203 1 .o 5 0.8 d 0.6 0.4 6.2 0, % a 0 0.4 0.8 1.2 1.6 2.0 2.4 Amount of 0.394 4 M KCN/ml Fig. 6. Photometric titration of copper with cyanide in ammoniacal solution. Filter 607 (Amax.* = 600 nm).Taken, 2.0 ml of 0.100 1 M CuSO, plus n ml of concentrated NH, solution. Starting volume, 35 ml. Curve A, n = 1 ml; B, n = 2 ml; C, n = 4ml; D, n = 8 ml; and E, n = 16 ml. Theoretical end-point for Cu to CN ratio 1 : 4 = 2.03 ml. Titration of Mixtures The only really feasible determination in this system is that of nickel, and we therefore The interfering ions chosen examined the effect of interferences only on the nickel titration. were the soluble ammine formers zinc, cobalt and copper. Titration of Nickel and Zinc in Fig. 2, although with a less obvious break a t 1 : 2 Ni: CN stoicheiometry. log p4 = 16.7 and for log f14 = 9.1. This result is reasonable given the low log p4 difference and high ammonia concentration existing in the titration. Zinc in a 1 : 1 molar ratio with nickel causes no interference.The titration curve appears as For Zn2+ + 4CN- + Zn(CN)2,- Zn2+ + 4NH3 + Zn(NH3)2,+ Titration of Nickel and Cobalt adds on to the nickel titre exactly. ratios of 1 : 2 and 1 : 4. Under the conditions shown in Fig. 7(a), cobalt appears to react with 1 :4 stoicheiometry and Small but distinct breaks occur at nickel to cyanide molar The nickel is titrated first. Titration of Nickel and Copper Fig. 7(b) shows that the end-point is equivalent to the sum of nickel plus copper. The titration curve is an interesting shape and shows that nickel is titrated before copper; there is some indication of completion of nickel complexation. log fi4 [Ni(CN)i-]-log Is, [Ni(NH,)i+] = 22.8 log 194 [Cu(CN)S,-]-log f l 4 [Cu(NH,);+] = 17.7 These relationships explain why nickel is titrated first.Discussion The titration system described could be useful for the determination of nickel in the presence of ions that do not complex strongly with cyanide or that precipitate in dilute ammonia solution. Also, in a different sphere of interest, photometric titrations of this type are reallyFebwary, 1978 COPPER BY DIRECT PHOTOMETRIC TITRATION WITH CYANIDE 139 continuous Yoe Q) m 2 0.6 -2 2 n Q 0.4 0.2 c 0 2 4 6 8 1 0 1 2 0 2 4 6 8 1 0 1 2 Amount of 0.394 4 M KCN/ml Amount of 0.394 4 M KCN/ml Fig. 7. (a), Photometric titration of a mixture of nickel and cobalt with cyanide in ammoniacal solution under nitrogen. Filter 606. Curve A: taken, 8.0 ml of 0.1004 M NiSO, solution, 1.0 ml of 0.1002 M Co(NO,), solution and 4 ml of concentrated NH, solution. Starting volume, 26 ml. Curve B, titration of 1.0 ml of 0.100 2 M cobalt only under the same conditions. Theoretical end-point for Ni to CN ratio 1 : 4 and Co to CN ratio 1:4 = 9.17 ml; for Ni to CN ratio 1:4 and Co to CN ratio 1: 5 = 9.42 ml. ( b ) , Photometric titration of a mixture of nickel and copper with cyanide in ammoniacal solution. Filter 606. Curve A, 1.0 ml of 0.1001 M CuSO, solution and 4 ml of concentrated NH, solution. Curve B, 8.0 ml of 0.1004 M NiSO, solution and 4ml of concentrated NH, solution. Curve C, 1.0 ml of 0.100 1 M CuSO, solution, 8.0 ml of 0.1004 M NiSO, solution and 4 ml of concentrated NH, solution. All starting volumes 26 ml. Theoretical end-point for nickel only = 8.15 ml : for Ni to CN ratio 1 : 4 plus Cu to CN ratio 1 : 4 = 9.16 ml. Jones molar ratio plots and can be valuable aids to the understanding of complexation reactions. Reference 1. Ringborn, A., “Complexation in Analytical Chemistry,” Interscience, New York and London, 1963. Received August 1st 1977 Accepted August 22nd, 1977

ISSN:0003-2654    

DOI:10.1039/AN9780300134

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

140 Analyst, February, 1978, Vol. 103, $@. 140-148 Biacetyl Bis(4-phenyl-3-thiosemicarbazone) as a Reagent for the Spectrophotometric Determination of Copper A. G. Asuero and J. M. Cano Department of Analytical Chemistry, Faculty of Sciences, The University, Seville-4, Spain The synthesis, characteristics and analytical applications of biacetyl bis- (4-phenyl-3-thiosemicarbazone) (BBPT) are described. The reaction between copper(I1) and BBPT has been studied by spectrophotometry. The reddish orange 1 : 1 copper - BBPT complex (E = 12.7 x lo3 1 mol-1 cm-1 a t 485 nm and 8.2 x lo3 1 mol-1 cm-l at 530 nm) is formed at pH 1.8-11.9, in a solution containing 60% V/V of dimethylformamide. The effect of interferences was studied. A rapid and simple method for the spectrophotometric determination of copper in a white metal, in blende and in a waste water has beendevised.The method is compared with others proposed for the spectrophotometric determination of copper with thiosemicarbazone reagents. Keywords : Bzacetyl bis(4-phenyl-S-thiosemicarbaaone) reagent ; copper deter- mination ; sfiectrophotometry Thiosemicarbazones have been widely used for the spectrophotometric determination of metal ions and several papers have dealt with the use of dithiosemicarbazones as analytical reagents. The most studied have been derived from glyoxa1,l bipyridylglyoxal,2~3 cyclo- hexane-l,2-dione4-6 and ~yclohexane-l,3-dione.~ The introduction of the phenyl radical at the end of the thiosemicarbazide molecule is an excellent example of how group action in organic compounds can be modified to provide increased sensitivity ; the molar absorptivities of metal - thiosemicarbazone com- plexes are favourably influenced by this substitution, which also causes a shift of absorption peaks towards longer wavelengths.Some phenylthiosemicarbazones have been studied as reagents*-11 ; nevertheless, compounds with the bis(pheny1thiosemicarbazone) grouping have received very little attention. A search of the literature revealed that Niederschulte and Ballschmiter12 and Ball~chmiterl~ have studied some bis(pheny1thiosemicarbazones) derived from biacetyl and glyoxal, varying the phenyl substituent. However, the principal applica- tion reported by these workers for this type of reagent has been the separation of complex chelate mixtures by thin-layer chromatography on aluminium oxide or by liquid chromato- The use of biacetyl bis(4-phenyl-3-thiosemicarbazone) (BBPT) for the selective deter- mination of copper has been investigated, and this paper, which forms part of an investigation into the use of diphenylthiosemicarbazones as analytical reagents, describes the development of a simple procedure that has high sensitivity and selectivity.The determination of small amounts of copper in different materials is described. graphy. BBPT Experimental Synthesis of BBPT Phenylthiosemicarbazide (3.63 g) was dissolved in 100 ml of ethanol - water (1 + 1) andASUERO AND CAN0 141 the solution boiled under reflux. One millilitre of biacetyl dissolved in 20 ml of ethanol and two drops of glacial acetic acid were added and the mixture was refluxed for 1 h.The solution was allowed to cool to room temperature and the product was separated by filtration. The yellow powder obtained was washed with boiling ethanol and dried in a vacuum desiccator (yield 30%). The product had a melting-point above 300 "C and elemental analysis gave the following results: C 56.2, H 5.3, N 21.9 and S 16.8%; C,,H,,N,S, requires C 56.25, H 5.20, N 21.87 and S 16.67%. Apparatus A Unicam SP800 spectrophotometer was used for recording spectra in the ultraviolet and visible regions of the spectrum and a Coleman 55 (digital) instrument was used for measure- ments at fixed wavelengths. Quartz cells (1-cm path length) were used throughout the work. A Philips PW 9408 pH meter, with a glass and calomel electrode pair, was used for pH measurements.Throughout this paper pH is used to denote pH-meter reading and not the actual concentration of hydrogen ions in solution. A Metrohm E.1009 photometric titrator with a 4.0-cm glass cell was used in the deter- mination of the pK of the reagent. Reagents All solutions were prepared with analytical-reagent grade chemicals using distilled water. Biacetyl bis(4-~henyZ-3-thiosemicarbazone) stock solution. A 0.5% m/V solution was pre- pared in dimethylformamide. This solution was stored in an amber-glass bottle in a refrigerator. It was diluted to 0.033% for use in the spectrophotometric procedure. The stock solution was stable for several weeks when kept under these conditions. This solution was prepared by dissolving copper(I1) sulphate pentahydrate in water and was standardised by using a complexometric method with 1 -(2-pyridylazo)-2-naphthol as a metallochromic indi~at0r.l~ Sodium acetate trihydrate (105 g) and glacial acetic acid (100 ml) were diluted to 1 1 with distilled water.CoPper(I1) solz&on, 1 .OOO 3 mg mZ-1. B u f e r solution, PH 4.3. Dowex 50-X8 resin, sodium form, and Dowex 1-X8 resin, chloride form. Procedure Determination of copper in acetate medium To the copper solution (10-150 pg of copper) in a 25-ml calibrated flask, add 15 nil of 0.033% m/V BBPT solution in dimethylformamide, 1 ml of pH 4.3 acetate buffer and dilute to volume with water. Measure the absorbance a t 485 or 530 nm against distilled water. A reagent blank has negligible absorbance at these wavelengths.Determination of copper in moderately acidic medium To the copper solution (10-15Opg of copper) in a 25-ml calibrated flask, add 15 ml of 0.033% m/V BBPT solution in dimethylformamide. Adjust the pH to 2.0 & 0.05 with dilute hydrochloric acid and dilute to volume with water. Measure the absorbance at 485 or 530 nm against a reagent blank prepared simultaneously with the sample. Determination of copper in waste water, white metal and zinc blert.de Add 1 ml of 0.1 M EDTA to solutions of the first two materials and 2.5 ml of 0.1 M EDTA to solutions of zinc blende before adding the BBPT reagent solution, in order to prevent interferences by foreign ions. Calibration graphs were prepared by using standard solutions of copper(II), treated in the same way as in the recommended procedures. Results and Discussion Biacetyl BBPT Reagents BBPT has salubilitres in chloroform, methanol, ethanol, nitrobenzene and dimethyl-142 ASUERO AND CAN0 : BIACETYL BIS(4-PHENYL-3-THIOSEMICARBAZONE) Analyst, VOZ.103 formamide of 1.1, 0.2, 0.1, 1.2 and 31.3 g l-l, respectively. The solubility in water is less than 0.001 g 1-l. M) stored in darkness at low tempera- tures were stable for at least 1 week. A water - dimethylformamide medium (7 + 3 V/V) was used in the study of the reagent in order to prevent the precipitation of BBPT. In this medium a BBPT solution of 1.6 x M concentration shows maximum absorption at 343 nm, with a molar absorptivity of 4.24 x lo4 1 mol-l cm-l. Slow hydrolysis of the reagent to phenylthiosemicarbazide and biacetyl occurred in dilute solutions (1.6 x The rate of hydrolysis of the ligand increased at pH values below 4 and above 10, but decreased when the dimethylformamide to water ratio was increased. A bathochromic shift was produced at first in alkaline medium before the reagent underwent hydrolysis.A simultaneous potentiometric - photometric method15 was used in the determination of the ionisation constant when the average pK value was found to be 10.95. This behaviour may be caused by deprotonation of thiol groups. BBPT appears to be a tetradentate ligand with a convenient steric arrangement of its donor groups and contains a conjugated system of v-electrons connected with the donor system. A medium containing 60% of dimethylformamide and 40% of water was chosen for further experimental work.The main advantage of such a medium is that the chelates as well as the reagent were soluble in it at the concentration used in the photometric procedure. The characteristics of the most important RBPT complexes are shown in Table I. Dilute solutions in dimethylformamide (3.2 x M) at pH values of 6.0 and 10. The chelates of BBPT are uncharged. TABLE I CHARACTERISTICS OF BBPT COMPLEXES IN SOLUTION Metal ion Optimum pH Zn(I1) . . . . 5.5-9.5 Cd(I1) . . . . 6.2-10.7 Hg(I1) . . . . 4.2-9.7 Cu(I1) . . . . 1.8-11.9 Pb(I1) . . . . 6.5-10.5 Bi(II1) . . . . 5.9-7.2 Fe(II1) . . . . 4.8-7.5 Fe(I1) . . . . 4.5-7.1 Co(I1) . . . . 6.5-8.5 Pd(I1) . . . . 1.5-10.0 Ni(I1) . . . . 2.0-10.9 Amax./nm 440 430 420 485 530 440 450 400 400 600 400 428 582 405-420 Molar absorptivity/ lo4 1 mol-l cm-1 21.5 22.4 14.5 12.7 8.2 16.5 25.0 25.5 31.5 23.5 3.4 30.4 25.3 2.8 Colour of complex Yellow Yellow Yellow Reddish orange Yellow Orange Yellow Yellow Green Brown Green Study of Copper - BBPT System Addition of a solution of BBPT to a solution of copper(I1) ions produced a red - purple complex when the copper was in large excess; the colour changed progressively through red - purple to reddish orange as diphenylthiosemicarbazone was added in excess (Fig.1). The absorption spectra of solutions having constant contents of the reagent but increasing contents of the metal ion show an isosbestic point at 505 nm, but as the copper content of the solution exceeds the content of BBPT, the absorptivity curves no longer pass through this point.This shift indicates that a new species is being formed in the solution and also shows that the complex formation takes place stepwise. Stoicheio~netry of the complexes Job’s curves were plotted at different pH values and wavelengths (Fig. 2). At pH 10, 6.7 and 4.1 the curves intersected at 0.5 and 0.67 molar fraction of copper. This effect indicated the existence of two complexes in which the ratios of copper to BBPT were 1 : 1 and 2: 1, respectively. In acidic media, at pH 4.1 and 2.0, the ratio of metal to ligandFebrzcary, 1978 AS A SPECTROPHOTOMETRIC REAGENT FOR COPPER 143 differed from 1 : 1. This effect was attributed to competition of hydrogen ions for the reagent, which prevented the deprotonation of thiol groups. kkvelength/nm Fig.1. Absorption spectra of copper - BBPT complexes formed from various metal to ligand ratios (pH = 6.7; acetate buffer). Copper to ligand ratios: A, 1O:l (33 pg ml-l of copper); B, 1 : l (3.3 p g ml-l of copper) ; and C, 1 : 10 (3.3 pg ml-l of copper). Oxidation state of copper and structure complex previously mentioned with copper(I1). From experimental evidence it was concluded that the reagent forms the reddish orange The presence of ascorbic acid in the - Ratio [Cul /I[Cul + [BBPT] 1 Ratio [Cul /([Cul + [BBPT] 1 Fig. 2. Composition of copper - BBPT complexes by the con- (a) pH = 10: absorbance at A, 405 nm; (b) pH = 6.7 : absorbance a t A, 395 nm; (c) pH = 4.1 : absorbance at (6) pH = 2.0: absorbance a t tinuous variation method. B, 485 nm; and C, 545 nm.B, 485 nm; C, 560 nm; and D, 600 nm. A, 410 nm; B, 485 nm; and C, 540 nm. A, 485 nm; B, 540 nm; and C, 680 nm.144 ASUERO AND CAN0 BIACETYL BIS(4-PHENYL-3-THIOSEMICARBAZONE) Analyst, 'vd. 103 solution before adding the BBPT reagent did not alter the absorption spectra of the complex formed at high ligand to copper ratios. Hydrogen peroxide altered the absorption peak situated at 395 nm, because oxidising agents destroy the reagent (Fig. 3). The reddish orange complex was not retained on either a cationic or an anionic ion-exchange resin, indicating that it was uncharged. Wave I ength /n m Fig. 3. Influence of hydrogen peroxide (1 ml of 30% m/ V ) with time on the absorption spectra of copper - BBPT complex (2 pg ml-1 of copper; 15 ml of BBPT, 0.033% solution) : A, immediately; B, after 10 min; C, after 1 h ; D, blank; and E, blank after 1 h.pH = 6.7, acetate buffer. The exact configuration of a complex of this type, cc-diketone bis(thiosemicarbazone) - copper(II), has been established.l8~l7 It is apparent that BBPT acts as a tetradentate ligand; the co-ordination occurs by bonding from the first two nitrogen atoms of each phenyl- thiosemicarbazide residue and from the two sulphur atoms of the thiol groups so that three five-membered chelate rings are produced. This fact explains the greater stability of the reddish orange complex. When the ratio of copper to ligand was 10: 1, a new yellow complex appeared in the solution in the presence of ascorbic acid. The study of the copper - BBPT system when amounts of copper exceed the amounts of the reagent will be the subject of a future report.This forrnula has been previously suggested by Bahr.18 Injuence of $H When the pH was varied by addition of sodium hydroxide or hydrochloric acid, the absorb- ance remained constant over the pH range 1.8-11.9 (Fig. 4). The copper - BBPT chelate differed from cuproine analogues,lg in that the former was formed in moderately acidic as well as basic media. ~ - x - K ~ - x - x - - x - x - x ~ ~ - x - x - ~ cf 0.2 4 8 12 Ph Fig. 4. Influence of pH on the forma- tion of reddish orange copper complex. Absorbance measured at ( 0) 395 nm, ( x ) 486 nm and (a) 630 nm.Febrztary, 1978 AS A SPECTROPHOTOMETRIC REAGENT FOR COPPER 145 Temperature and colozcr development time When the temperature varied between 15 and 60 "C the absorbances remained constant within the limits of experimental error.The complex was stable for at least 24 h over the pH range 3.6-10 and it remained stable for at least 2 h at pH 1.9 (Table 11). The complex was formed immediately on addition of the reagent. TABLE I1 STABILITY OF COPPER - BBPT COMPLEX Absorbance of solution containing 2 pg ml-l of copper at pH 1.9 measured against distilled water. Timelmin f A 10 20 30 60 90 1 2 7 Absorbance a t 485 nm .. . . 0.436 0.442 0.447 0.454 0.452 0.452 Absorbance of blank . . .. . . 0.035 0.043 0.046 0.055 0.055 0.054 Absorbance difference , . * . . . 0.401 0.399 0.401 0.399 0.397 0.398 Absorbance at 530 nm 0.287 0.288 0.294 0.300 0.298 0.295 Absorbance of blank . . .. .. 0.029 0.029 0.033 0.040 0.042 0.040 Absorbance difference . , .. . . 0.258 0.259 0.261 0.260 0.266 0.265 Solvent extraction The complex can be extracted into chloroform, toluene, ethyl acetate, tributyl phosphate and benzene. The other metal - thiosemicarbazone complexes were also extracted by these solvents. Spectrophotometric Determination of Copper with BBPT Based on the experimental work two methods are proposed for the determination of trace amounts of copper. Beer's law is obeyed between 0.2 and 4 pg ml-l at 485 nm and 0.5 and 6 pg ml-l at 530 nm. The optimum concentration range, evaluated by Ringbom's method, is 1-3 pg ml-l of copper at 485 nm and 1 . 2 4 pg ml-l at 530 nm. The reddish orange complex gave values for the molar absorptivity of E = 12.710 x los 1 mol-l cm-l at 485 nm and 8.260 x lo3 1 mol-1 cm-1 at 530 nm.The sensitivities of the method according to Sandell are 0.005 and 0.0077 pg cm-2 at 485 and 530 nm, respectively. The relative error of the method is kO.19 and &0.27% at 485 and 530nm, respectively, when the pH is 6.2. At pH 2.0 0.05, the method gave a relative error (P = 0.05) of 0.27 and 0.38%. The photometric characteristics are similar in both instances. TABLE I11 EFFECT OF FOREIGN IONS ON THE DETERMINATION OF 50 pg OF COPPER AT pH 6.2 Amount tolerated/ Foreign ion pg ml-l SCN-, Br-, I-, BaOla-, PO,8-, citrate, tartrate, S,0,2- Cr(III), Mo(VI), W(VI), Tl(I), As(V), Se(IV), Mn(II), Al(III), C,042- 120 Pb(II), Zn(II), Cd(II), Hg(II), Pt(IV), Rh(III), Sn(I1) 2 000 200 Ce(IV), Th(IV), UOa(II), La(III), alkali and alkaline earth metals, Cr04a-, F- 20 V(V) Au (I1 I) 10 6 Ag(1).Bi(III), Pd(I1) 0.4 Ag(I), Pb(II), Hg(II), Au(II1) ; (SzOsa-, 2 ml of 0.26 N solution) Sb(V) ; (tartrate, 60 mg) Fe(II1) ; (F-, 180 pg d-l) 200 200 20146 ASUERO AND CAN0 : BIACETYL BIS(4-PHENYL-3-THIOSEMICARBAZONE) Analyst, VOZ. 103 TABLE IV EFFECT OF FOREIGN IONS ON THE DETERMINATION O F 50 pg OF COPPER AT pH 2.0 Absorbance measured at 530 nm. Amount tolerated/ Amount tolerated/ Ion added pg ml-1 Ion added pg ml-l Pd(II), Bi(II1) 1 C1-, Br-, I-, Se(IV), Sb(V) 5 SCN-, CH,COO-, Pt(IV), Sn(I1) 20 NO,-, phthalate, Hg(II), Fe(I1) 30 Bp0,2-, tartrate, 0 s (VIII) 80 citrate Cr( 111) 1 500 EDTA Rh( I1 I) 180 Pop- Mn(I1) 2 500 c10,- Pb(I1) 3 500 C,O,a- Zn(II), Cd(II), F- Al(III), Be(I1) 8 000 s20,2- As(V) and As(II1) 10 000 Mo(V1) > 700 Ti(IV), W(VI), Zr(1V) > 300 10 000 8 000 4 000 3 000 2 800 2 200 2 000 For the determination of 50 pg of copper by these methods, foreign ions can be tolerated at the levels given in Tables 111, IV and V.Silver and gold are reduced by BBPT to the elemental state and interfered at very low concentrations. The good results obtained by masking them with sodium thiosulphate were due to the fact that the amount of dimethyl- formamide contained in the samples prevented the decomposition of the metal - thiosulphate complexes. TABLE V ELIMINATION OF INTERFERENCES BY ADDITION OF MASKING AGENTS pH of solution 2.0. Amount tolerated/pg ml-I Without With masking Foreign ion masking agent agent Fe(II1) . . - 300 35 Ni(I1) .. - 250 Sb(V) .. 5 800 .. 5 2 000 2 000 1 000 ZF!) .. 3500 V(V) .. 20 50" - 200f. 1 30 1 000: Hg(I1) .. Au(II1) . . - >400f Fe(II1) . . - 800 Co(I1) . . - Bi(II1) . . 1 2 000 Sn(I1) . . 20 - * Masking agent EDTA; 2.5 ml of 0.1 M solution F- EDTA; 2.5 ml of 0.1 M solution EDTA; 2.5 ml of 0.1 M solution EDTA; 2.5 ml of 0.1 M solution EDTA; 2.5 ml of 0.1 M solution Tartrate; 0.25 g EDTA; 1 ml of 0.1 M solution EDTA; 2.5 ml of 0.1 M solution EDTA; 2.5 ml of 0.1 M solution S20a2-; 4 ml of 0.1 N solution S,0,2-; 4 ml of 0.1 N solution S2OS2-; 4 ml of 0.1 N solution * Absorbance measured directly after addition of reagent. t pH adjusted with dilute nitric acid. $ At pH 3.0. Applications The method has been applied to the determination of copper in industrial effluents from a sulphuric acid plant (pyrites process), a white metal and a zinc mineral.The waste water solution was filtered through an asbestos mat contained in a Gooch crucible. The average composition of the waste water analysed (seven samples) at pH 1.4 &- 0.1 was chloride 300, zinc 27.8, iron 53.1, chromium 0.05, copper 15.6, manganese 0.5, lead 0.05, arsenic 11.0 and calcium carbonate 534 pg ml-l. The copper content found by spectrophotometric determination was 15.6 0.1 pg ml-l (mean result of six determinations).February, 1978 AS A SPECTROPHOTOMETRIC REAGENT FOR COPPER 147 Zinc blende was treated first with a mixture of concentrated nitric acid and bromine (2 + 1 V / V ) and then with dilute nitric acid (1 + 1) and boiled to ensure complete dissolution of the are and to remove the oxides of nitrogen as well as the excess of bromine.The silicic acid that was precipitated was dehydrated and re- moved. White metal 8e had the following certificate composition: copper 4.57, lead 3.13, antimony 9.5, zinc 0.04, cadmium 0.14 and tin 82.6Zy0. The copper content found was 4.57 & 0.02%. A zinc blende I11 ore had the following certificate composition: zinc 52.99, lead 3.89, copper 0.12, tin 0.0018, mercury 0.039, arsenic 0.13, iron 6.59, manganese 1.45, cadmium 0.17, germanium 0.004, silver 0.005 and sulphur 30.04%. The copper content found was 0.124 These results are in exact agreement with those quoted in the certificates of analysis. The method is superior to the existing colorimetric procedure involving the use of sodium diethyldithiocarbamate,20 because it is less time consuming and involves fewer chemical manipulations.White metal was dissolved in aqua regia. Triplicate results were obtained in both instances. o.oo4~0. Conclusion Many methods are available for the determination of trace amounts of copper with thiosemicarbazone reagents (Table VI), but in our experience none of them is completely TABLE VI CHARACTERISTICS OF COPPER - THIOSEMICARBAZONE COMPLEXES Compound Picolinaldehyde thiosemicarbazone Picolinaldehyde phenylthiosemicarbazone Biacetylmonoxime thiosemicarbazone Biacetylmonoxime phenylthiosemicarbazone Thiophenaldehyde thiosemicarbazone Bipyridylglyoxal dithiosemicarbazone Cyclohexane- 1,2-dione dithiosemicarbazone Biacetyl diphenylthiosemicarbazone Optimum PH 8.9-1 0.7 8.5-10 8.5-9.5 8.5-9.7 4.5-8.2 7-9 4-10 1.8-1 1.9 Xmax./nm 410 400 345 360 372 390 467 485 630 Molar absorptivity/ 1 mol-1 cm-1 Reference 6 300 21 23 500 22 10 600 23 12 700 8 39 000 24 9 550 25 5 700 6 12 700 - 8 260 satisfactory. The great ability which the atoms of sulphur have for co-ordinating metal cations imposes a serious limitation on the use of these reagents, as it makes the establish- ment of selective methods of analysis difficult.This paper describes a study of the optimum conditions for a selective and sensitive spectrophotometric method for the determination of copper. We are grateful to Professor F. Pino for his interest and encouragement. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. References Budesinsky, B. W., and Svec, J., Analytica Chim.Acta, 1971, 55, 115. Bahamonde, J. L., Bendito, D. P., and Pino, F., Talanta, 1973, 20, 694. Bahamonde, J . L., Bendito, D. P., and Pino, F., Analyst, 1974, 99, 355. Muiioz, J. A., Cano, J. M., and Pino, F., Infcidn. Quim. Analit. Pura ApZ. Ind., 1972, 26, 226. Muiioz, J. A., Cano, J. M., and Pino, F., An. Quim., 1976, 72, 392. Muiioz, J. A., Cano, J. M., and Pino, F., Quim. Analit., 1974, 28, 90. Berzas, J. J., Muiioz, J. A., and Roman, M., Talanta, 1976, 23, 257. Cano, J. M., Jimenez, J. C., and Pino, F., Alzalytica Chim. Acta, 1975, 75, 336. Cano, J. M., and Pino, F., Analyt. Lett., 1974, 7, 159. Ariza, J. C., Cano, J. M., and Pino, F., Talanta, 1976, 27, 460. Ariza, J. G., and Cano, J. M., Analyt. Lett., 1976, 9, 677. Niederschulte, U., and Ballschmiter, K., 2. Analyt. Chem., 1972, 261, 191. Ballschmiter, K., 2. Analyt. Chem., 1973, 263, 203. Bermejo, F., and Prieto, A., “Applicaciones Analiticas del AEDT Y AnAlogos,” Departamento de Muiioz, J. A., and Pino, F., Infcidn. Quim. Analit. Pura. Apl. Ind., 1973, 27, 67. Quimica Analitica, Santiago de Compostela, 1960, p. 339.148 16. ASUERO AND CAN0 Taylor, M. R., Gabe, E. J., Glusker, J. P., Minikin, J. A., and Patterson, A. L., J . Am. Chem. SOL, Warren, L., Horner, S., and Hatfield, W., J . Am. Chem. SOC., 1972, 94, 6392. Bahr, G., 2. Anorg. A&. Chem., 1952, 268, 351. Diehl, H., and Smith, G. F., “The Copper Reagents : Cuproine, Neocuproine and Bathocuproine,” IUPAC “Spectrophotometric Data for Colorimetric Analysis,” Butterworths, London, 1963, p. 154. Cano, J. M., and Pino, F., Talanta, 1972, 19, 1959. Ariza, J. G., PhD Thesis, University of Seville, 1976. Valcarcel, M., and Bendito, D. P., Infcidn. Quim. Analit. Pura A@. Ind., 1970, 24, 49. Muiioz, J. A., Cano, J. M., and Pino, F., An. Quim., 1973, 69, 251. Bahamonde, J. L., PhD Thesis, University of Seville, 1973. 1966, 88, 1845. G. Frederick Smith Chemical Co., Columbus, Ohio, 1958. Received June 3rd, 1977 Accepted August 12th, 1977 17. 18. 19. 20. 21. 22. 23. 24. 25.

ISSN:0003-2654    

DOI:10.1039/AN9780300140

出版商:RSC    

年代:1978

数据来源: RSC

摘要:

Analyst, February, 1978, Vol. 103, pp. 149-155 149 Potentiometric Determination of Copper in Palm Oil with a Copper( 11) Ion-selective Electrode Y. S. Fung and K. W. Fung Department of Chemistry, University of Hong Kong, Hong Kong The applicability of a copper(I1) ion-selective electrode for the determination of the copper content of crude and hydrogenated palm oils was investigated. Dry ashing was used for destroying the organic matrix and a porcelain crucible was used as container. The recovery of copper in dry ashing and the leaching of copper from the internal glazing of a porcelain crucible were investigated. A complexing antioxidant buffer was used in order to minimise the interfering effect of iron(II1) on the determination of copper(I1). A direct potentio- metric method can be applied to determine the copper content of palm oil at a concentration higher than 10 pg kg-l in the presence of 30 mg kg-1 of iron.Keywords : Copper determination; palm oil; direct potentionzetry ; copper(II) ion-selective electrode The deteriorating effect on the quality of oils and fats caused by the presence of trace amounts of some heavy metals is well kn~wn.l-~ They act as pro-oxidants to catalyse the oxidation of oils and fats by the oxygen in air. The end-products of the oxidation are ketones, alde- hydes and other compounds that contribute the off-flavour to oils and fats3-' Copper is a strong pro-oxidant and its deteriorating effect on soybean oil even at very low levels (approximately 30pgkgl) has been reported.2 The effect of copper on the stability and bleachability of palm oil has been investigateds; the rate of oxidation during storage is strongly accelerated by the presence of trace amounts of copper.The determination of copper in palm oil is part of the routine quality control procedure. The low levels of copper present in good quality palm oils are beyond the range of conventional flame atomic-absorption spectrophotometry or visible spectrophotometry with detection limits of about 0.1 mg 1-1 in s o l u t i ~ n . ~ J ~ Large samples must therefore be used for the quantitative determination of copper by these methods and preparing the sample for chemical analysis requires a considerable time. Although the levels are above the detection limit obtainable with carbon furnace atomic-absorption spectrophotometry, and this method can be applied without prior pre-treatment,llJ2 it suffers from drawbacks such as poor repro- ducibility, non-atomic absorption, higher running costs and the need for higher operator skill and expensive instrumentation. The recently developed solid-state copper(I1) ion-selective electrode is selective and has a limit of detection of about M.13 Thus, direct potentiometric measurement is a promising method for the quantitative determination of copper in palm oil.In this paper we describe a method of analysis in detail. The levels of copper and iron in palm oil were found to vary between 10 pg k g l and 5 mg k g l and 0.7 and 30 mg k g l , re~pectively.~ The interfering effect of iron(II1) on the determination of copper(I1) with a copper( 11) ion-selective electrode is well known.15-17 Hence the methods must be capable of measuring the concentration of copper(I1) at the 10 pg k g l level even in the presence of 30 mg k g l of iron(II1).A complexing antioxidant buffer (CAB) has been used by Smith and Manahan16 to minimise the interfering effect of iron(II1) on the determination of copper(I1) with a copper(I1) ion-selective electrode. However, its capacity to reduce the interference of iron(II1) is not known and the solution contains a high concentration of acetate so that it is not suitable for direct potentiometric measurement. A modified buffer solution is described. Dry ashing was employed for the destruction of the organic matrix in the palm oil in this study because wet ashing is not recommended in view of safety hazards and contamination of the solution.lE-21 Acid extraction is not desirable in a direct potentiometric method because of possible contamination and its effect on the control of the pH of the resulting150 FUNG AND FUNG: POTENTIOMETRIC DETERMINATION OF COPPER Analyst, VoZ.103 solution. Controversial results for the amount of metal recovered have been reported by different gro~ps~2-2~ using porcelain crucibles for dry ashing prior to the determination of metals. We studied the recovery of copper in palm oil as well as the leaching of copper from the crucible. The applicability of the proposed method for the determination of copper in palm oil was checked by performing atomic-absorption spectrophotometric determinations in parallel.The standard additions procedure was used for checking the matrix effect of the resulting solution on the copper determination with a copper( 11) ion-selective electrode. Experimental Apparatus An Orion, Model 94-29A, copper(I1) ion-selective electrode and a Model 90-02 double- junction reference electrode were used with an Orion, Model 701A, digital meter for the potential measurement. A Thomas, Model 80, digital printer was used for recording the readings at 15-s or 1-min intervals, depending on the response time of the electrode. A Radiometer 26 pH meter with its associated glass microelectrode (type G2222C) and saturated calomel electrode were used for the pH measurements. A Varian Techtron, Model 1200, atomic-absorption spectrophotometer with an air - ace- tylene atomiser or a Model 63 carbon rod atomiser were used for atomic-absorption spectro- photometric analysis. A 5-p1 Excalibur Autopipette was used to introduce the sample solution into the carbon tube furnace and the signal was recorded with an Esterline Angus, Model 575, recorder.An electric furnace was used for ashing the palm oil in a porcelain crucible that had been immersed in a mixture of nitric and sulphuric acids overnight before use. Reagents The nitric and hydrochloric acids were triply distilled in glass from analytical-reagent grade reagents. All other chemicals used in this study were analytical-reagent grade and were used without further purification. All solutions were made up with water that had been doubly distilled in glass. The stock solutions of copper(I1) and iron(II1) were prepared by dissolving freshly cleaned copper foil (99.9% pure) and iron wire in dilute nitric acid.The pH of the stock solution of iron(II1) was adjusted to below 2 in order to keep the iron(II1) in solution. Fresh standard solutions of copper(I1) were prepared daily. The standard bis( l-phenylbutane-1,3-dione) - copper(I1) complex was obtained from the National Bureau of Standards, USA. The buffer solution (CAB) used for complexing iron(II1) and adjusting the pH of the final solution was made up with the following composition: 1 x 10-1 M sodium perchlorate, 1 x 10-1 M potassium hydroxide, 1 x M sodium acetate, 1 x 10-1 M sodium fluoride and 1 x 10-1 M formaldehyde, Distilled diethyl ether and light petroleum (boiling range 60-80 "C) were used in the column- chromatographic separation.M acetic acid, 1 x Procedure The palm oil samples were melted in a water-bath a t 60 "C and 10 ml of the sample (approxi- mately 9 g) were transferred into a weighed 25-ml porcelain crucible, which was re-weighed to obtain the mass of sample taken. The covered crucible was heated in a furnace for 2 h at 350 "C and then for 2-3 h at 480 "C to burn off all the carbon. The crucible was then removed from the furnace and 10 ml of the glass-distilled nitric acid were added in order to dissolve the ash. The acid was boiled off on a water-bath and 4.5 ml of perchlorate solution (containing 1 x 10-2 M perchloric acid and 1 x M sodium perchlorate) were then added to dissolve the residue. When solution was complete 0.5 ml of CAB was added to adjust the pH of the solution to 4-5 and to complex the iron(II1).The resulting solution was analysed by potentiometric and atomic-absorption spectrophotometric (flame or electrothermal atomisation) methods. Potentiometric measurements were made in the crucible at 25 "C in darkness in order to prevent errors caused by the photovoltaic effect. Before each measure- ment the electrodes were rinsed thoroughly with doubly glass distilled water and then dried.Febrztary, 1978 IN PALM OIL WITH A COPPER(II) ION-SELECTIVE ELECTRODE 151 The timing of the experiment was started as soon as the electrodes were immersed in the solution and readings were taken at 15-s or 1-min intervals until the reading was constant. The blank oil was prepared by repeated extraction of palm oil with constant-boiling hydro- chloric acid and O.Olyo EDTA solution.The oil was then washed with doubly distilled water several times and the water was removed by heating under vacuum. The oil was dissolved in a mixture of light petroleum (boiling range 60-80 "C) and diethyl ether in the ratio of 1 : 9 and passed through a silica-gel column. The adsorbed layers in the column were then eluted with diethyl ether - light petroleum (1 + 9). The solvent was removed by vacuum distillation. The recovery of copper was studied by adding known amounts of NBS copper standard to the melted palm oil with a known copper content in order to make up the concentration of copper to approximately 100 mg k g l and determining the copper content as described above.A column-chromatographic method was also employed for preparing a blank oil. Results and Discussion Porcelain crucibles are not recommended for the quantitative determination of trace amounts of copper in organic m a t t e F ~ ~ ~ because of the leaching of copper from the internal glazing. Porcelain crucibles are inexpensive containers for dry ashing and have often been used for this purpose in different laboratories and so it was useful to find out whether or not the problem of leaching was serious in the determination of copper in palm oil. Blank crude palm oil was analysed for copper, using the same method of pre-treatment as for the sample, by carbon furnace atomic-absorption spectrophotometry and the results are given in Table I. The contamination of copper from the reagents was investigated by following the same procedure as in the sample pre-treatment in the absence of palm oil and using a platinum crucible.The copper content in the resulting solution was found to be 3.5 pg l-l, which was about the same as that in the blank palm oil. Thus, leaching out of copper from the internal glazing of a porcelain crucible is not a problem in the quantitative determination of copper in palm oil. The copper content in the blank hydrogenated palm oil prepared by repeated acid extraction of a hydrogenated oil is slightly higher, which may be attributed to the incomplete extraction of copper owing to the higher viscosity and melting-point of the hydro- genated palm oil. TABLE I CONCENTRATION OF COPPER IN BLANK PALM OIL Determinations carried out on a Model 1200 atomic-absorption spectrophotometer and Model 63 carbon rod atomiser operating under the following conditions: drying, 50 s a t 100 " C ; ashing, 20 s a t 800 "C; atomisation, 3 s a t 2000 "C; wavelength, 324.7 nm; spectral band width, 0.2 nm; and lamp current, 3 mA.Sample Mean/ Standard deviation/ Pre-treatment Number of samples tested pg kg-l PQ kg-l Crude palm oil Acid extraction 8 Column chromatography 1 Hydrogenated palm oil Acid extraction 7 5.5 1.6 3.3 8.1 4.8 - The recovery of trace metals from the residue obtained by dry ashing without the addition of an ashing aid is a controversial problem. Loss due to volatilisation and retention on the crucible has been reported.24 Volatilisation of copper is not a problem if the ashing temperature is kept below 550 0C23 but retention of copper on the wall of the crucible has been reported.24 In order to check whether or not the loss of copper in the proposed method was significant, standard samples containing a known amount of copper were pre-treated and analysed by both potentiometry and flame atomic-absorption spectrophotometry.Table I1 shows that close to 100~o recovery was obtained. Retention of copper is not significant in the proposed sample preparation procedure. A complexing antioxidant buffer (CAB), which consists of sodium acetate, acetic acid, sodium fluoride and formaldehyde, was used by Smith and Manahan16 for adjusting the pH, complexing copper(I1) and interfering iron(III), regulating ionic strength and providing a moderately reducing medium in the determination of trace amounts of copper in water by potentiometry.A high concentration of Such a solution was modified to suit our purpose.152 FUNG AND FUNG: POTENTIOMETRIC DETERMINATION OF COPPER AnaZyst, Vol. 103 TABLE I1 RECOVERY OF NBS COPPER STANDARD IN PALM OIL Recovery, yo r 1 Sample Number of samples tested Amount of copper added/pg POT* AASt Crude palm oil . . .. 2 3.2 103 93 4 83.8 103 103 3 90.6 103 102 4 258.8 103 102 Hydrogenated palm oil . . 1 2 2 16.51 109 103 68.6 93 95 86.3 100 99 * POT = determination by direct potentiometry. t AAS = determination by flame atomic-absorption spectrophotometry. acetate ion depresses the response of the electrode by forming a complex with copper(I1) and is not desirable for direct potentiometric measurement of concentrations just above the limit of detection. A low buffer capacity is adequate for adjusting the pH of the final solution under a controlled sample pre-treatment process and the matrix effect is less significant after dry ashing.Thus, 1 x 1 0 - 3 ~ sodium acetate and 1 x 1 0 - 3 ~ acetic acid were used for adjusting the pH of the final solution to 4-5. At such a low concentration, the acetate ion does not affect the response of the electrode, as shown in Table 111. The selectivity ratio was estimated by the mixed solution method.25 Sodium fluoride (1 x lop2 M) was used for complexing iron(II1) to prevent the loss of copper(I1) by co-precipitation16 and 1 x 10-2 M formaldehyde was used to provide a moderately reducing medium for stabilising the response of the electrode in very dilute copper(I1) solutions.Sodium perchlorate (1 x 1 0 - 2 ~ ) was used for adjusting the ionic strength of the resulting solution and 1 x 1 0 - 2 ~ potassium hydroxide was used for neutralising the acidic perchlorate solution that was used for dis- solving the residue. None of these ions in these concentration ranges in CAB affects the response of the copper(I1) ion-selective electrode as shown in Table 111. In the presence of the modified CAB, the selectivity ratio of copper(I1) to iron(II1) is approximately 8 x (as shown in Fig. l), while the selectivity ratio required for the determination of 10 pg k g l of copper(I1) in the presence of 30 mg k g l of iron(II1) is approximately 3 x Thus, the modified CAB is capable of handling the worst combination of copper(I1) and iron(II1) in Dalm oil.Nitric acid was used in order to ensure complete oxidation of the organic matrix ind the copper to copper(II), as well as to dissolve crucible. TABLE I11 SELECTIVITY RATIOS OF THE COMPONENTS IN THE Interfering species Acetate? . . .. .. Iron(III)$ . . .. .. Formaldehyde? . . .. Fluoride? . . .. .. Perchlorate . . .. the retained copper on the wall of the COMPLEXING ANTIOXIDANT BUFFER Selectivity ratio* 4 x 10-4 8 x 10-4 < 10-6 < 10-6 < 10-6 * The selectivity ratio (kg:In) is defined as the ratio [Cu(II)]/[In], where In = interfering species. t In 0.1 M sodium perchlorate solution. $ In CAB solution. Concentration of copper(I1) is 50 p g 1-l.The calibration graph, which was constructed with known amounts of copper(I1) in the solution with the same matrix as the sample solution, is shown in Fig. 2. The slope of the straight line is about 30 mV per decade increase in concentration and the detection limit (when the slope is reduced to 30%) is about 10 pg 1-I. The deviation from Nernstian slope at low concentrations may be attributed to trace amounts of copper in the reagents, as discussed previously. The reproducibility of the electrode at various concentrations is about h0.2 mV. The response time (which is taken as the time for the potential of the electrode to reachFebrGuary, 1978 IN PALM OIL WITH A COPPER(II) ION-SELECTIVE ELECTRODE 120 110 > E 2 t loo,; w 90 80 153 - 1 a - - I I I Fig. 1.Interference of iron(II1) on determination of 50 pg 1-1 of copper in the complexing antioxidant buffer. h0.2 mV of the equilibrium value) of the electrode depends on the history of the electrode and the direction of the change of concentration. It will take only 10 min to reach the equilibrium potential if the concentration of the solution changes from 10 to 20 pg 1-l. How- ever, it will take l h to reach the equilibrium potential if the concentration of the solution changes from 0.3 mg 1-1 to 10 pg 1-l. The response time is usually fast (less than 10 min) when the concentration of copper(I1) is higher than 0.1 mg 1-l. Only the presence of mercury- (11), silver(I), sulphur(I1) and strong oxidising agents will interfere in the determinati0nl59~5 and it is unlikely that such species would be present in the sample solution.200 > < 150 +. E. w I I I I 0.001 0.01 0.1 1 .o 10 1 Log [copper (I I) concentration/mg I-'] Fig. 2. Calibration graph for copper(I1) in the complex- ing antioxidant buffer. One of the problems in direct potentiometric measurements is the matching of the com- position of the matrix of the standard solution to that of the sample solution. The presence of some foreign ions can affect the activity of copper(I1). Trace amounts of phosphate (a few milligrams per kilogram) were found in some sample solutions; phosphorus is known to be present in natural impurities in palm oil as phosphatides and phosphoric acid is added to the oil during the refining process for the extraction of metals.6,' In order to investigate the matrix effect on the proposed method, the standard additions technique was also employed to determine the concentration of copper(I1) in the sample solution.The results obtained by the two techniques are similar, as shown in Table IV, with no statistically significant difference. The applicability of the method was checked with the recommended flame atomic-absorp- tion spectrophotometric method. The sample solution was analysed in parallel by the two Thus, matrix effects are not significant in this instance.Sample AAS* 1 .. . . 0.47 2 .. .. 0.59 3 .. . . 0.24 4 .. ... 0.58 5 .. .. 0.31 6 .. . . 0.41 7 ,. . . 0.22 8 .. . . 0.40 9 .. . . 0.35 10 .. . . 0.31 11 .. . . 0.35 12 .. . . 0.50 POTt 0.49 0.56 0.20 0.56 0.36 0.43 0.21 0.48 0.32 0.34 0.30 0.53 154 FUNG AND FUNG: POTENTIOMETRIC DETERMINATION OF COPPER Analyst, VoZ. 103 TABLE IV COMPARISON OF METHODS FOR DETERMINATION OF COPPER IN PALM OIL S atistical test for the differences between matched pairs of copper concentrations in crude palm oils determ- ined by atomic-absorption spectrophotometry and direct potentiometry.Concentration of copper/mg kg-l Difference Difference - -- AAS-POT/ POT-STD/ STD? mg kg-1 mg kg-l 0.49 0.53 0.22 0.52 0.37 0.41 0.20 0.44 -0.02 0.03 0.04 0.02 -0.05 -0.02 0.01 -0.08 0.03 -0.03 0.05 -0.03 0.00 0.03 -0.02 0.04 -0.01 0.02 0.01 0.04 Methods Number of Mean compared samples value/mg kg-l AAS-POT .. 12 -0.042 POT-STD .. 8 0.014 Standard deviation /mg kg-1 tcalc.$ ttheo.11 0.040 -0.36 &2.201 0.023 1.72 k2.365 * AAS = determination by atomic-absorption spectrophotometry.t POT = determination by direct potentiometry with calibration graph. 9 The statistical test for matched pairs, tcalc., is calculated26 by STD = determination by direct potentiometry with standard additions technique. where po = 0 for Ho : px = 0 Ha : px # 0 s is sample variance P is sample mean n is number of samples. 11 All of the theoretical t-values, ttheo., are found a t 5';/0 significance level from Ref. 27. methods and the results obtained were similar, as shown in Table IV, with no statistically significant difference. The reproducibility of the method was studied by repeated analysis of a hydrogenated palm oil sample. The relative percentage error is larger than that caused by the reproducibility of the electrode potential (0.2 mV; approximately 2%).This difference may be attributed to the inhomogeneity of the copper content in the palm oil sample and to errors introduced during the procedure. However, the reproduci- bility of the direct potentiometric method is better than that of the flame atomic-absorption spectrophotometric method at such low levels of copper. The results are given in Table V. TABLE V DETERMINATION OF COPPER I N HYDROGENATED PALM OIL BY DIRECT POTENTIOMETRY AND FLAME ATOMIC-ABSORPTION SPECTROPHOTOMETRY No. of determinations = 7. Mean Standard 95% copper content/ deviation/ confidence limits/ Relative standard Method mg kg-l mg kg-l mg kg-I deviation, yo POT .. .. 0.147 0.01, *0.01, 10 AAS .. .. 0.13, 0.02, f0.01, 20February, 1978 IN PALM OIL WITH A COPPER(II) ION-SELECTIVE ELECTRODE Conclusions Dry ashing in a porcelain crucible and a direct potentiometric method with a copper(I1) ion-selective electrode can be used for determining trace amounts of copper in palm oil.The use of expensive platinum crucibles for dry ashing is not necessary. The detection limit is about 10 pg kg-l, which is lower than that of the recommended flame atomic-absorption method. The method does not suffer from matrix effects and the reproducibility is reason- able. The apparatus assembly is simple, the running cost is low and the capital cost is less for the proposed method. The method can also be applied to the determination of copper in other oils and fats. 155 We thank Mr. K. L. Er, Lam Soon Oil and Soap Manufacturing Sdn.Bhd., Malaysia, and Dr. H. T. Toh, Chemistry Department, University of Malaya, Malaysia, for supplying the palm oil samples. This research was supported by the Higher Degrees and Research Grants Committee of the University of Hong Kong. 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. References Marcuse, R., “Metal Catalysed Lipid Oxidation,” Swedish Institute of Food Preservation Research Evans, C. D., Schwab, A. W., Mosu, H. A., Hawley. J. E., and Melvin, E. H., J . Am. Oil Chem. Hampson, G. C., and Hudson, B. J. F., “Physical and Chemical Properties of the Constituents of Eskin, N. A. M., J . Am. Oil Chem. Soc., 1976, 53, 746. Scott, G., “Atmospheric Oxidation and Antioxidants,” Elsevier, Amsterdam, 1965, pp.357-368. Lundberg, W. 0.. in Lundberg, W. O., Editor, “Autoxidation and Antioxidants,” Volume 11, Inter- Williams, P. N., in Devine, J., and Williams, P. N., Editors, “The Chemistry and Technology of “Malaysian Palm Oil Technical Bulletin,” No. 1, Malaysian Palm Oil Producer’s Association, Kuala Analytical Methods Committee, Analyst, 1963, 88, 253. Analytical Methods Committee, Analyst, 1971, 96, 741. Kundu, M. K., and Prevot, A., Analyt. Chem., 1974, 46, 1591. Olejko, J. T., J . A m . Oil Chem. SOG., 1976, 53, 480. Blaedel, W. J., and Dinwiddie, D. E., Analyt. Chem., 1974, 46, 873. Jacobsberg, B., and Jacqmain, D., Olkagineux, 1973, 28, 25. Ross, J . W., Jr., in Durst, R. A., Editor, “Solid State and Liquid Membrane Ion-Selective Electrodes,” NBS Special Publication, No. 314, National Bureau of Standards, Washington, D.C., 1969, p. 57. Smith, M. J., and Manahan, S. E., Analyt. Chem., 1973, 45, 836. Fung, Y. S., and Fung, K. W., Analyt. Chem., 1977, 49, 497. Analytical Methods Committee, Analyst, 1960, 85, 643. Taubinger, R. P., and Wilson, J. R., Analyst, 1965, 90, 429. Analytical Methods Committee, Analyst, 1973, 98, 458. Feldman, C., Analyt. Chem., 1974, 46, 1606. Grys, S., Mikrochim. Acta, 1976, 7, 147. Gorsuch, T. T., Analyst, 1959, 84, 135. Analytical Methods Committee, Analyst, 1976, 101, 62. Moody, G. J., and Thomas, J. D. R., “Selective Ion Sensitive Electrodes,” Merrow, Watford, 1971. Harnett, D. L., “Introduction to Statistical Methods,” Second Edition, Addison-Wesley Publishing “Handbook of Chemistry and Physics,” Forty-seventh Edition, The Chemical Rubber Co., Cleveland, Received J u l y 26th, 1977 Accepted September 5th. 1977 (SIK), Goteborg, Sweden, 1968. Soc., 1951, 28, 68. Edible Oils and Fats,” Pergamon Press, London, 1961, pp. 13-16. science, New York, 1962, pp. 452-460. Edible Oils and Fats,” Pergamon Press, Oxford, 1961, pp. 40-42. Lumpur, Malaysia, 1973. Co., Menlo Park, Calif., 1975, pp. 295-298. Ohio, 1966-1967, p. 156.

ISSN:0003-2654    

DOI:10.1039/AN9780300149

出版商:RSC    

年代:1978

数据来源: RSC