ISSN:0003-2654    

DOI:10.1039/AN97196FX045

出版商:RSC    

年代:1971

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97196BX047

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

iv THE ANALYSTTHE ANALYSTEDITORIAL ADVISORY BOARDChairman: H. J. Cluley (Wernbley)*T. Allen (Bradford)*L. S. Bark (Salford)R. Belcher (Birmingham)L. 1. Bellamy, C.B.E. (Waltham Abbey)L. S. Birks (U.S.A.)E. Bishop (Exeter)*R. C. Chirnside (Wembley)*A. C. Docherty (Billingham)D. Dyrssen (Sweden)*W. T. Elwell (Birmingham)*D. C. Garratt (London)*R. Goulden (Sittingbourne)J. Hoste (Belgium)D. N. Hume (U.S.A.)H, M. N . H. Irving (Leeds)A. G. Jones (Welwyn Garden City)*J. A. Hunter (Edinburgh)M. T. Kelley (U.S.A.)W. Kemula (Poland)*G. F. Kirkbright (London)G. W. C. Milner (Harwell)G. H. Morrison (U.S.A.)*G. Nickless (Bristol)*J. M. Ottaway (Glasgow)*G. E. Penketh (Billingham)S. A. Price (Tadworth)D. 1. Rees (London)E. B, Sandell (U.S.A.)A.A. Smafes, O.B.E. (Harwell)H. E. Stagg (Munchester)E. Stahl (Germany)A. Walsh (Australia)T. S. West (London)P. Zuman (U.S.A.)*Members of the Board serving on the Executive Committee.[December, 1971NOTICE TO SUBSCRIBERS(other than Members of the Society)Subscriptions for The Analyst, Analytical Abstracts and Proceedings should besent through a subscription agent or direct to:The Chemical Society, Publications Sales Omce,Blackhorse Road, Letchworth, Herts.Rates for 1972(a) The Analyst, Analyticul Abstracts, and Proceedings, with indexes . . ..index), and Proceedings .. .. .. .. .. ..index), and Proceedings ,. .. . . .. 6 . ..(b)(c)The Anulyst, Analytical Abstracts printed on one side of the paper (withoutThe Analyst, Analytical Abstracts printed on one side of the paper (withThe Analyst and Analytical Abstracts without Proceedings-The Analyst,and Analytical Abstracts printed on onesideofthe paper (withoutThe Analyst, and Analytical Abstracts printed on one side of the paper (with(d) The Analyst and Analytical Abstracts, with indexes .... ..index) . . . . . . ,. .. .. .. .. ..index) . . .. .. .. .. .. .. .. * .(e)(f)€33.50f 34.50f40.50€3 I -00f 32.00€38.00$80.40$82.80$97.20$74.40$76.80$9 1.20(Subscriptions are NOT accepted for The Anulyst and/or for Proceedings alone)Members should send their subscriptions to the Hon. Treasurevi SUMMARIES OF PAPERS I N THIS ISSUE [December, 1971Summaries of Papers in this IssueA Comprehensive Scheme for the Analysis of a Wide Range ofSteels by Atomic-absorption SpectrophotometryThe determination of manganese, nickel, chromium, molybdenum, copper,vanadium, cobalt, titanium, tin, aluminium and lead in steel by a directatomic-absorption method, which involves a single dissolution based onperchloric acid, is described.The scheme effects considerable savings in timecompared with traditional methods, and is of comparable accuracy.The serious depressive interferences caused by iron on the response ofsome elements are overcome in the nitrous oxide - acetylene flame, and minoreffects are corrected for by inclusion of iron in calibration standards. Noother inter-elemental interferences were encountered in perchloric acid basedsolutions in the presence of iron.Silicon and tungsten, which are not retained in solution in this scheme,are determined after a separate dissolution of the sample.Results obtained with British Chemical Standard steels are tabulated,and accuracies discussed.D.R. THOMERSON and W. J. PRICEPye Unicam Ltd., York Street, Cambridge, CB1 2PX.Analyst, 1071, 96, 825-834Determination of Sulphur in Steel by a Modified CombustionProcedure and Coulometric TitrationThe quantitative determination of sulphur in steel by the combustionmethod is inefficient as the sulphur content of samples indicated by thismethod is lower than that given by other methods because sulphur trioxideoccurring in the combustion furnace tends to condense in the piping.Anattempt has been made to establish how this sulphur trioxide condensationoccurs, and a unique, automatic circulation type of adsorption method hasbeen devised to ensure complete collection of gaseous sulphur occurring inthe combustion. For this purpose, a practical sulphur analyser with highsensitivity and accuracy has been developed in which an automatic coulo-metric titrator and a high-frequency induction furnace are combined. Standardsamples of various types of iron and steel analysed with this apparatus showedthat the results agreed well with their standard values.R. KAJIYAMANippon Yakjn Kogyo Co. Ltd., Kawasaki, Japan.and K. HOSHINOKokusai Electric Co. Ltd., Hamura, Tokyo, Japan.Analyst, 1971, 96, 835-842.The Determination of Uranium by Atomic- absorptionSpectrophotometryThe determination of uranium by atomic-absorption spectrophotometryis complicated by significant interference effects and demands critical controlof fuel composition and burner adjustment.For the determination ofuranium in uranium-based nickel - uranium catalysts a solvent-extractionmethod was used to avoid the matrix effects.MARGARET J. MARTINThe Gas Council, London Research Station, Michael Road, Fulham, London, S.W.6.Analyst, 1971, 96, 843-846...Vlll SUMMARIES OF PAPERS I N THIS ISSUESelective Atomic-absorption Determination of Inorganic Mercuryand Methylmercury in Undigested Biological SamplesA simple method for the determination of total mercury in biologicalsamples contaminated with inorganic mercury and methylmercury is des-cribed.The method is based on the rapid conversion of organomercurialsfirst into inorganic mercury and then into atomic mercury suitable for aspira-tion arough the gas cell of a mercury vapour concentration meter, by acombined tin(I1) chloride - cadmium chloride reagent. It was found thatif 100 mg of tin(I1) chloride alone were added instead of the tin(I1) chloride -cadmium chloride reagent, only the release of inorganic mercury influencedthe peak deflection of the potentiometer, thus permitting the selective deter-mination of inorganic mercury in the presence of methylmercury. I t waspossible first to release inorganic mercury then, after re-acidification of thereaction mixture, methylmercury, by adding the tin(I1) chloride - cadmiumchloride reagent and sodium hydroxide.When total mercury and inorganicmercury were determined separately, the difference between results gavethe methylmercury content of the sample.L. MAGOSMedical Research Council Laboratories, Toxicology Unit, Woodmansterne Road,Carshalton, Surrey.Analyst, 1971, 96, 847-853.[December, 1971A Critical Study of Brilliant Green as a SpectrophotometricReagent : The Determination of Perchlorate Particularlyin Potassium ChlorateA modified procedure for the extraction of perchlorate with Brilliantgreen into benzene is described, which gives a reproducible and nearly quanti-tative recovery of perchlorate (apparent E~~~~~ = 94 000) and low blanks( A = 0.02 to 0.04). Studies on the use of untreated glassware, silanisedglassware and polypropylene-ware have indicated that adsorption effects arenot very important.Nevertheless, the standard and sample solutions shouldbe treated in the same vessels and the absorbance measurements made inquartz cells.Brilliant green perchlorate can be extracted from aqueous solutions inthe pH range 3 to 6.5, but the solutions should be buffered at a constantpH value.The procedure has been applied successfully to the determination ofperchlorate in samples of potassium chlorate.A. G. FOGG, C. BURGESS and D. THORBURN BURNSDepartment of Chemistry, University of Technology, Loughborough, Leicestershire.Analyst, 1971, 96, 854-857.Rapid Polarographic Microdetermination of DissolvedOxygen in Water with Flavin EnzymeThe rapid polarographic microdetermination of dissolved oxygen inwater with flavin enzyme, P-D-glucose oxidase [EC 1.1.3.4; /I-D-glucose :oxygen oxidoreductase] or D-amino-acid oxidase [EC 1.4.3.3 ; D-amino-acid :oxygen oxidoreductase] is described in detail. The method is based on themeasurement of the ratio of oxygen consumed by the enzyme in the presenceof a limited amount of substrate as an internal standard to the total dissolvedoxygen, which can be measured by the addition of an excess amount ofsubstrate or sodium hydrosulphite. It was found that this method requiresonly 1 ml of test water a t a temperature between 5 and 40 "C, and givessatisfactory results over a wide range of oxygen concentrations up to 100 percent. saturation, and the result can be obtained in only 5 minutes. Thevalues found are almost identical with those generally accepted.JUN OKUDA, TSUNEAKI INOUE and ICHITOMO MIWAFaculty of Pharmaceutical Science, Meijo University, Showa-ku, Nagoya, Japan.Analyst, 1971, 96, 858-864

ISSN:0003-2654    

DOI:10.1039/AN97196FP181

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

xii SUMMARIES OF PAPERS I N THIS ISSUEAn Improved Oxalate Method for the Determination of Active Oxygen[December, 1971in Manganese DioxideOf the various methods available for the determination of active oxygenin manganese dioxide, the oxalate method is preferred because of its simplicityand reproducibility. However, it tends to give higher results than the iron(I1)and Bunsen methods, the discrepancies being greater with synthetic materialsthan with ores. In confirmation of the work of other authors it has beenshown that there is a Mn2+ catalysed loss of oxalate that causes the highresults. It is concluded that the loss is caused by oxidation by oxygendissolved in the reagent solutions. With the addition of Cu2+ or Fe3+, theoxalate method gives results that are comparable with those of the iron(I1)method.The stabilisation of oxalate with Fe3+ has been used to speed upthe time taken to dissolve samples by boiling them under reflux.D. S. FREEMAN and W. G. CHAPMANThe Ever Ready Company (G.B.) Limited, Central Laboratories, London, N15 3TJ.Analyst, 1971, 96, 865-869.Electrophoresis of Dyed Polysaccharides on “PhoroSlides”Polysaccharides that have been previously treated with reactive dyestuffscan be separated within 1 to 3 minutes by electrophoresis on PhoroSlides atpotential gradients of 40 to 56 V cm-1. In addition to providing the requiredself-indicating system, dyeing increases the electrophoretic mobility ofcharged polysaccharides. The difficulty experienced in dyeing polysaccharidesthat have small proportions of primary hydroxyl groups (e.g., plant gums,pectin, alginic acid and sulphated polysaccharides) can be overcome byincreasing the amount of sodium carbonate added in the dyeing process.D.M. W. ANDERSON, A. HENDRIE, J. R. A. MILLAR and A. C. MUNROChemistry Department, The University, Edinburgh, EH9 3 J J.Analyst, 1971, 96, 870-874.The Detection of Ten Components of a Multi-vitaminPreparation by Chromatographic MethodsA method in which chromatography is used for the analysis of multi-vitamin preparations is described. Fat-soluble vitamins are extracted afterdissolving the preparation in aqueous ammonia solution and are detected bythin-layer or gas chromatography. Water-soluble materials are then separatedin three thin-layer chromatographic systems and active constituents detectedwith inorganic phosphors and selective spray reagents.R.T. NUTTALL and B. BUSHChemistry Department, Thames Polytechnic, London, S.E.18.Analyst, 1971, 96, 875-878.An Improved Titration Medium for Sulphate-ion IndicatorsThe indicators sulphonazo 111, dimethylsulphonazo I11 and carboxy-arsenazo have been compared for the determination of sulphur by flaskcombustion or by furnace-tube combustion. Carboxyarsenazo is preferred,the optimum conditions for its use being a 2 + 1 acetone - water solutionbuffered with pyridine and perchloric acid.E. E. ARCHER, D. C. WHITE and R. MACKISONBP Chemicals International Ltd., Great Burgh, Yew Tree Bottom Road, Epsom,Surrey.Analyst, 1971, 96, 879-880xiv SUMMARIES OF PAPERS I N THIS ISSUEThermometric Determination of Aromatic AldehydesThe thermometric determination of some aromatic aldehydes is des-cribed. A known and excess amount of 2,4-dinitrophenylhydrazine is addedto the aldehyde to be determined and, after stirring the mixture for 10 t o 15minutes, the excess of 2,4-dinitrophenylhydrazine is titrated with a standardaldehyde, namely 2-methoxybenzaldehyde.The solvent used for both titrantand titrand is a solution of 4 per cent. v/v of water and 4 per cent. v/v of sulph-uric acid in isobutyl alcohol. The time of titration is about 1 minute and theaccuracy is within &l per cent.[December, 1971L. S. BARK and P. BATEDepartment of Chemistry and Applied Chemistry, The University of Salford, Salford,M5 4WT, Lancs.Analyst, 1971, 96, 881-884.A Stable Low-current Source for Electrode PolarisationA constant-current source for the range nanoamperes to microamperesbased upon the use of cheap operational amplifiers is described.It has provedsatisfactory for use in differential electrolytic potentiometry and otherbipotentiometric titrations.Mrs. L. G. HARTSHORN and E. BISHOPChemistry Department, University of Exeter, Stocker Road, Exeter, Devon.Analyst, 1971, 96, 885-886.Application of Gas - Liquid Chromatography to the Analysisof Essential Oils. Part I. Determination of CedrolReport prepared by the Essential Oils Sub-committee.ANALYTICAL METHODS COMMITTEE9/10 Savile Row, London, W1X 1AF.Analyst, 1971, 96, 887-894xviii THE ANALYST [December, 197 1Keprints of Review PapersREPRINTS of the following Review Papers published in The Analyst since January, 1963, areavailable from The Society for Analytical Chemistry, Book Department, 9/10 Savile Row, London,W1X 1AF (not through Trade Agents).Orders MUST be accompanied by a remittance for thecorrect amount made out to “Society for Analytical Chemistry.”“Classification of Methods for Determining Particle Size,” by the Particle Size Analysis“Methods of Separation of Long-chain Unsaturated Fatty Acids,” by A. T. James (August,“Beer’s Law and its Use in Analysis,” by G. F. Lothian (September, 1963).“A Review of the Methods Available for the Detection and Determination of Small Amounts“Circular Dichroism,” by R.D. Gillard (November, 1963).“Information Retrieval in the Analytical Laboratory,” by D. R. Curry (November, 1963).“Thermogravimetric Analysis,” by A. W. Coats and J. P. Redfern (December, 1963). Price“Some Analytical Problems Involved in Determining the Structure of Proteins and Peptides,”“The Faraday Effect, Magnetic Rotatory Dispersion and Magnetic Circular Dichroism,” by“Electrophoresis in Stabilizing Media,” by D. Gross (July, 1965).“Recent Developments in the Measurement of Nucleic Acids in Biological Materials,” by“Radioisotope X-ray Spectrometry,” by J. R. Rhodes (November, 1966).“The Determination of Iron(I1) Oxide in Silicate and Refractory Materials,” by H. N. S.“Activation Analysis,” by R. F. Coleman and T. B. Pierce (January, 1967).“Techniques in Gas Chromatography. Choice of Solid Supports,” by F.J. Palframan“Heterocyclic Azo Dyestuffs in Analytical Chemistry,” by R. G. Anderson and G. Nickless“Determination of Residues of Organophosphorus Pesticides in Food,” by D. C. Abbott and“Radioactive Tracer Methods in Inorganic Trace Analysis : Recent Advances,” by J. W.“Gamma-activation Analysis,” by C. A. Baker (October, 1967).“Precipitation from Homogeneous Solution,” by P. F. S. Cartwright, E. J. Newman andD. W. Wilson (November, 1967).“Industrial Gas Analysis,” by (the late) H. N. Wilson and G. M. S. Duff (December, 1967).Price 35p.“The Application of Atomic-absorption Spectrophotometry to the Analysis of Iron andSteel,” by P. H. Scholes (April, 1968).“Inorganic Ion Exchange in Organic and Aqueous - Organic Solvents,” by G.J. Moody andJ. D. R. Thomas (September, 1968).“Radiometric Methods for the Determination of Fluorine,” by J. K. Foreman (June, 1969).Price 25p.“Techniques in Gas Chromatography. Developments in the van Deemter RateTheory of Column Performance,” by E. A. Walker and J. F. Palframan (August, 1969).Price 25p.Choice of Detectors,” by T. A. Gough andE. A. Walker (January, 1970).Sub-committee of the Analytical Methods Committee (March, 1963).1963). Price 25p.Price 25p.Price 25p.of Cyanide,” by L. S. Bark and H. G. Higson (October, 1963). Price 25p.Price 15p.Price 15p.25p.by Derek G. Smyth and D. F. Elliott (February, 1964).J . G. Dawber (December, 1964).Price 25p.Price 25p.Price 25p.H. N. Munro and A. Fleck (February, 1966). Price 25p.Price 25p.Schafer (December, 1966). Price 25p.Price 25p.Part I.and E. A. Walker (February, 1967).(April, 1967). Price 25p.H. Egan (August, 1967).McMillan (September, 1967). Price 25p.Price 25p.Price 25p.Price 25p.Price 25p.Price 25p.Price 35p.Part 11.“Techniques in Gas Chromatography.Price 25p.“Laser Raman Spectroscopy,” by P. J. Hendra and C. J. Vear (April, 1970).“Ion-selective Membrane Electrodes,” by Ern0 Pungor and KlAra T6th (July, 1970). Price“X-ray Fluorescence Analysis,” by K. G. Cam-Brion and K. W. Payne (December, 1970).“Mass Spectrometry for the Analysis of Organic Compounds,” by A. E. Williams and H. E.Part 111.Price 35p.35p.Price 25p.Stagg (January, 1971). Price 35p

ISSN:0003-2654    

DOI:10.1039/AN97196BP189

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

DECEMBER, 1971 THE ANALYST Vol. 96, No. I149 A Coniprehensive Scheme for the Analysis of a Wide Range of Steels by Atomic-absorption Spectrophotometry* BY D. R. THOMERSON AND W. J. PRICE (Pye Unicam Ltd., York Street, Cambridge, CB1 2 P X ) The determination of manganese, nickel, chromium, molybdenum, copper, vanadium, cobalt, titanium, tin, aluminium and lead in steel by a direct a ‘omic-absorption method, which involves a single dissolution based on perchloric acid, is described. The scheme effects considerable savings in time compared with traditional methods, and is of comparable accuracy. The serious depressive interferences caused by iron on the response of some elements are overcome in the nitrous oxide - acetylene flame, and minor effects are corrected for by inclusion of iron in calibration standards.No other inter-elemental interferences were encountered in perchloric acid based solutions in the presence of iron. Silicon and tungsten, which are not retained in solution in this scheme, are determined after a separate dissolution of the sample. Results obtained with British Chemical Standard steels are tabulated, and accuracies discussed. THE problems arising in the determination of chromium and molybdenum have hitherto prevented the development of a general scheme for the analysis of irons and steels by atomic absorption. A recently published methodl in which these problems appear to have been successfully overcome has provided the clue to such a scheme, in which all the elements commonly encountered in steels can be determined by simple direct procedures.The scheme covers the determination of twelve elements, all of which, with the exception of tungsten, are determined on a single sample solution. A l-g sample of steel is dissolved in hydrochloric and nitric acids and the solution is finally evaporated to fumes with perchloric acid. The sample solution is then compared directly with calibration standards prepared by adding aqueous solutiom of the elements to l-g portions of pure iron and dissolving the latter in a manner identical to that used for the samples. In this way manganese, nickel, chromium, molybdenum, copper, vanadium, cobalt, titanium, tin, aluminium and lead were determined in a wide range of steels. An alternative procedure for the determination of tungsten and for molybdenum in the presence of tungsten is necessary.This procedure is based on the use of a mixture of perchloric, phosphoric and sulphuric acids, which retains the tungsten in solution. APPARATUS- All measurements were made with a Unicam SPSOA, Series 2, atomic-absorption spectro- photometer incorporating an SP91 lamp turret accessory and an SP94 nitrous oxide accessory. Air was supplied through an SP93 air compressor and the nitrous oxide and acetylene from cylinders. The instrument was fitted with an inert nebuliser and Unicam hollow-cathode lamps. The absorbance peaks were displayed on an SP22 chart recorder. The only non-standard modification was the fitting of 0-125-mm metal shims to the nitrous oxide burner to increase the jaw width from 0.46 to 0.59 mm.This was necessary to prevent blockage by solutions of steel of concentrations higher than about 0.5 per cent. The fitting of the shims to the nitrous oxide - acetylene burner naturally reduces to some extent the previous operating safety levels. Therefore, it is important to maintain the gas flow-rates at about the recom- mended values. On no account should shims that are thicker than 0.125 mm be used; the maximum burner jaw width is therefore 0-59 mm. No modification to the air - acetylene burner was necessary. * Presented at the Third SAC Conference, 1971, Durham. 0 SAC and the authors. 825826 THOMEKSON AND PRICE : A COMPREHENSIVE SCHEME FOR THE ANALYSIS OF [Analyst, Vol. 96 MANGANESE- Belcher and Kinson2 reported a slight enhancement of manganese absorption in 10 per cent. nitric acid solution and a slight depression in 10 per cent.solutions of hydrochloric, sulphuric and orthophosphoric acids, and in mixtures of phosphoric and sulphuric acids. They also noted depression by 1 per cent. of iron, whereas we found that the presence of 1 per cent. of iron slightly enhanced the signal. They also noted interference by 20 per cent. of chromium in a sulphuric acid - orthophosphoric acid mixture, which was overcome by using a smaller section of the flame, by placing an iris at the front of the hollow-cathode lamp. This interference did not occur with an air - acetylene flame in our perchloric acid medium as we obtained excellent results on a steel containing 4 per cent. of chromium (B.C.S. 341). However, Belcher and Kinson’s results do show that very good accuracy and reproducibility can be achieved with atomic absorption. We found no interference from either aluminium or molybdenum and, among other workers, only Atsuya3 found it necessary to add these elements to the calibration solutions.Ramirez-Munoz and Roth4 reported that better results were obtained with standards prepared from standard steels than with synthetic standards. Hubbard and Monks6 used a mixture of hydrochloric and nitric acids as the solvent and finally evaporated the solution to fumes with perchloric acid. They found that perchloric acid interfered to a lesser extent than any other acid (which confirms our findings) but reported small interferences by large amounts of cobalt, tungsten, chromium, nickel and molybdenum in the air - acetylene flame.All these interferences, together with the small depressive effect these workers reported for iron, can be completely overcome by the addition of 50 per cent. aqueous ethanol solution, which also serves to increase considerably the sensitivity . Of the many other ~ o r k e r s , ~ - l ~ most used a hydrochloric acid - nitric acid solvent for samples and aqueous calibration solutions containing only additions of iron. Results obtained were usually excellent for standard steels, which confirms our view that few difficulties are experienced with manganese in steel. As some pure irons contain a relatively high proportion of manganese this treatment, coupled with the good sensitivity of manganese, can result in a large blank value, e.g., B.C.S.260/2 containing 0.013 per cent. of manganese gave an absorbance reading of 0.06 for a calibration range of 1 to 20 mg 1-l. This signal, which results entirely from the manganese absorption and not from any interference effects in the flame, must be subtracted from each of the calibration absorbance values. The problem is reduced by using pure iron containing less manganese when available, e.g., B.C.S. 260/3 containing 0.002 per cent. of manganese gives an absorbance signal of less than 0.01, which in most instances can be ignored without adversely affecting the results. To assess the reproducibility of the manganese determination and also the long-term stability of the solutions, the four manganese determinations below were repeated with the same solutions on each of ten successive days.A statistical analysis of the results obtained yielded the following figures. Manganese, per cent. Coefficient of variation, B.C.S. No. (certificate value) Standard deviation per cent. 321 0.13 0.0013 1.04 312 0.20 0.0037 1.77 341 0.43 0.0076 1.83 23512 0.89 0-0090 0.96 NICKEL- The determination of nickel in steel with the air - acetylene flame has usually been found to be easy and interference free. Kinson and Belcherls proposed the use of phosphoric acid - sulphuric acid to retain tungsten in solution, as did Knight and Pyzyna.16 The latter observed that a slight depression of the nickel absorption was caused by most acids and it became necessary to incorporate iron in their standards. Other wm-kers8-15JgJ0 concerned with the determination of nickel in steel used a hydro- chloric acid - nitric acid solvent, which acts more rapidly and has the advantage of dissolving all of the copper whereas a phosphoric acid - sulphuric acid mixture may not.Nickel absorption wavelengths are discussed and some workers1°-13 found it preferable to use theDecember, 19711 A WIDE RANGE O F STEELS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 827 more sensitive line at 232.0 nm for levels of nickel below about 1 per cent., but they noted that the calibration graph for the 341.5 nm line is far more linear than that for the 232.0 nm line. We found that nickel was similar to manganese in giving rise to a small blank value, but whereas the manganese blank resulted entirely from manganese impurities in the iron, with the nickel it was caused only by physical interferences in the flame.However, this effect was completely overcome in the calibration without adversely affecting the analytical results. CHROMIUM AND MOLYBDENUM- In a recent paper1 the present authors described how difficulties encountered in the determination of chromium and molybdenum in steel are overcome. Depression of absorption response is minimised by using nitrous oxide - acetylene instead of air - acetylene flames. Variations caused by the difference in oxidation state of molybdenum and chromium between the samples and standards are overcome by evaporating both samples and standards to fumes with perchloric acid. COPPER- Wallace21,22 obtained high results for copper by using the air - acetylene flame and dis- solving the sample in hydrochloric acid; with phosphoric acid - sulphuric acid solvent the values were satisfactory.Kinson and BelcherZ3 found that although iron suppressed the copper signal no other interferences were present. Phosphoric acid - sulphuric acid solvent, coupled with the addition of iron to the calibration solution, gave good analytical results. Since then, as with manganese and nickel, most ~ 0 r k e r ~ ~ ~ ~ - - 1 ~ s ~ ~ J 6 , ~ * have preferred the hydrochloric acid-nitric acid solvent, combined with the use of either standard steels or pure iron added to aqueous copper solution for the calibration solutions. Exceptionally, Atsuya3 extracted copper with isobutyl methyl ketone as the copper diethyldithiocarbamate complex. Good analytical results were obtained with no detectable interferences.We found that the presence of 1 per cent. of iron depressed the copper absorp- tion in perchloric acid by about 10 per cent., but satisfactory results were obtained when iron was incorporated in the calibration solutions. VANADIUM- Capacho-Delgado and Manning25 successfully determined vanadium in steels by a direct method in which a nitrous oxide - acetylene flame and aqueous vanadium calibration solutions were used. They used the phosphoric acid - sulphuric acid mixture for dissolution of the samples and noted that sulphuric acid depressed the vanadium signal, while phosphoric acid enhanced it. They assumed that these two effects cancelled each other out as their published results were in good agreement with certificate values.In perchloric acid solution, we found that the presence of 1 per cent. of iron enhanced the vanadium signal by about 20 per cent., thus necessitating the addition of iron to the calibration solutions. However, we found no evidence that any other inter-elemental inter- ferences affected the response. As with some of the above elements a small blank value (about 0.03 absorbance unit on a range of 10 to 100 mg 1-l) was evident. After aspirating at a non-absorbing wavelength it was shown that this blank resulted entirely from physical interferences by iron in the flame operating on both samples and standards and had no adverse effects on the precision of the determination when disregarded. COBALT- McPherson, Price and Scaife,26 who used a 2-g sample dissolved in hydrochloric acid - nitric acid and diluted the solution to 50m1, reported a sensitivity of 0.001 per cent.of cobalt in steel with an air - acetylene flame. They also used phosphoric acid - sulphuric acid to dissolve a wide range of high and low-alloy steels when higher levels of cobalt were present. In most subsequent publications on cobalt in ~ t e e 1 ~ , ~ ~ , ~ ~ , ~ ~ , ~ ~ , ~ ~ , ~ ~ no interference problems were reported. We found that in the perchloric acid solvent, the cobalt response was enhanced by about 26 per cent. in the presence of 1 per cent. of iron. A moderate blank resulted from the cobalt calibration, which was caused almost completely by physical interferences in the flame.828 THOMERSON AND PRICE: A COMPREHENSIVE SCHEME FOR THE ANALYSIS OF [Analyst, vol.96 Some pure irons, however, contain appreciable amounts of cobalt, so it is necessary to choose calibration material with care. If cobalt-free iron is not readily available it will be necessary to determine how much of the signal is caused by the physical interference (matrix effect) of iron and how much by the cobalt absorption. To differentiate between these effects the blank solution should be aspirated and measured at a non-absorbing wavelength emitted by the cobalt lamp, with identical scale expansion, which will show how much of the signal results from matrix interference, the difference between this signal and the blank signal at 240nm being caused by the absorption of the cobalt in the iron blank.For most determinations, however, by using a pure iron similar in composition to B.C.S. 260/2 (0.009 per cent. of cobalt), satisfactory results are obtained by ignoring the blank, which in this instance is caused almost completely by matrix interference. TITANIUM- Atomic absorption was successfully applied to the determination of titanium in steel only after the advent of the nitrous oxide - acetylene flame. Bowman and Willis28 dissolved samples in hydrochloric acid - nitric acid and evaporated the solutions to fumes with sulphuric acid. They noticed that the interference effect of iron was a function not only of the sulphuric acid concentration, but also of the flame conditions. Calibration solutions had to be made up so that they contained iron, nickel, chromium and cobalt in concentrations similar to those of the samples.Headridge and H ~ b b a r d ~ ~ dissolved the sample in hydrofluoric and nitric acids, finally making up the solution so that it contained 50 per cent. of ethanol (a similar approach was used5 for manganese). This treatment enhanced the titanium response and overcame all inter-elemental interferences except that of iron, which still had to be added to the calibration solutions. Low results were reported by Mostyn and Cunningham30 when using either hydrochloric or sulphuric acid as a solvent. But when using hydrochloric acid - nitric acid mixture and making a single addition of potassium chloride, all the interferences, including that by iron, were overcome. It was therefore not necessary to match the iron content of samples and standards.With perchloric acid solvent we experienced no interference problems for titanium from any of the more usual elements in steel; even iron had no effect. We therefore confirm that unmatched aqueous standards can be used satisfactorily. However, our titanium values given here were obtained with iron present in the standards, as mixed standards were used throughout. TIN- No previous papers dealing with the determination of tin in steel by atomic-absorption spectrophotometry could be found, probably because of the very poor sensitivity of tin, which makes its direct determination at low levels difficult. With a 2 per cent. sample solution in perchloric acid we were able to determine down to about 0.01 per cent. of tin by using a nitrous oxide - acetylene flame and a wavelength of 224.0 nm.A small enhancement from 1 per cent, of iron was apparent, but no other interferences arose. The noise level of the signal, however, was still rather high, even when operating with maximum damping facilities. ALUMINIUM- The sensitivity of aluminium in the air - acetylene flame is so poor that it was not seriously considered for useful application in steel analysis, especially as most steels contain only very small amounts of aluminium. Before the introduction of the nitrous oxide - acetylene flame, Nikolaev31 distilled alu- minium from solid samples of refractory metals into a graphite cell that was heated to a high temperature, which readily permitted the detection of to per cent. of aluminium. By using the nitrous oxide - acetylene flame, Amos and Thomas32 reported good results for aluminium in the range 0.1 to 6 per cent.after adding iron to the calibration solutions. Pricell described the use of hydrochloric acid - nitric acid for the direct determination of acid-soluble aluminium, which necessitated fusing the residue and adding the resulting solution to the main sample solution for the determination of the total aluminium. A similar method was also used by other workers.24~33~34 Clarke and Cookel2 improved the sensitivityDecember, 19711 A WIDE RANGE OF STEELS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 829 of the determination by using an organic solvent to give a 2-fold enhancement of the signal. A 10 or 20-fold concentration of the aluminium was also demonstrated after extracting the aluminium as cupferrate into isobutyl methyl ketone, which solution could be aspirated directly.In perchloric acid medium, we found that the presence of 1 per cent. of iron gave rise to an enhancement of about 15 per cent. of the aluminium absorption, giving adequate sensitivity for the steels analysed. LEAD- As early as 1961, Elwell and G i d l e ~ ~ ~ used the hydrochloric acid - nitric acid dissolution technique to obtain a 2 per cent. sample solution, which was compared with a range of aqueous lead calibration solutions and corrected by the response of one iron-based standard. The absorption line 283.3 nm was used and good analytical values were reported for the range of 0-05 to 0.5 per cent. of lead. Dagnall, West and Young36 removed the iron matrix with isopentyl acetate and then extracted lead, as the iodide, into isobutyl methyl ketone.This technique yielded good results down to 0.001 per cent. of lead in a wide range of steels. Pricell and Clarke and Cooke12 found that a 2 per cent. sample solution in hydrochloric acid - nitric acid, aspirated into the air - propane flame at a scale expansion of five times on an SP90 atomic-absorption spectro- photometer, enabled lead in the range 0.0025 to 0.015 per cent. to be successfully determined in steel and cast iron. In our perchloric acid solution, the presence of 1 per cent. of iron enhanced the lead response by about 35 per cent., which, coupled with the use of the more sensitive 217.0 nm wavelength, enables lead levels down to 0.001 per cent.to be determined directly on a 2 per cent. sample solution. TUNGSTEN- By using a mixture of sulphuric, phosphoric and perchloric acids to effect sample dis- solution and with standard steels for calibration purposes, Knight and PyzynalG reported good agreement with other methods of analysis at about the 1 to 6 per cent. of tungsten level. We found that the effect of adding 1 per cent. of iron to a solution of sodium tungstate (50 p.p.m. of tungsten) was to suppress the tungsten signal by about 50 per cent., whereas in the presence of perchloric acid - phosphoric acid - sulphuric acid (mixture A, Method 2) the signal was enhanced, resulting in an over-all slight suppression in our final solutions. With this direct method it should be possible to determine any level of tungsten down to about 0.1 PROPOSED METHOD REAGENTS- 50ml of hydrochloric acid (sp.gr.1.18) and dilute to 1 litre. 5 N nitric acid and dilute to 1 litre. Stock manganese solution, 1000 mg I-l-Dissolve 1.0000 g of pure manganese Stock nickel solution, 1000 mg I-l-Dissolve 1-0000 g of pure nickel metal in per cent. metal in 40ml of Stock chromium solution, 1000 mg I-l-Dissolve 1.0000 g of pure chromium metal in 30 ml Stock molybdenum solution, 1000 mg kl-Dissolve 1.829 g of analytical-reagent grade Stock copper solution, 1000 mg Z-l-Dissolve 14000 g of pure copper metal in 50 ml of Stock vanadium solution, 1000 mg I-l-Dissolve 2.296 g of analytical-reagent grade am- Stock cobalt solution, 1000 mg I-1-Dissolve 1.0000 g of pure cobalt metal in 50 ml of 6 N Stock titanium solution, 1000 mg I-l-Dissolve 7.394 g of analytical-reagent grade potas- Stock tin solution, 1000 mg 2-l-Dissolve 1.0000 g of pure tin metal in 50 nil of hydro- Add 150 ml of hydrochloric of hydrochloric acid (sp.gr.1.18) and dilute to 1 litre. ammonium molybdate in water and dilute to 1 litre. 5 N nitric acid and dilute to 1 litre. monium vanadate in 20 ml of 100-volume hydrogen peroxide and dilute to 1 litre. nitric acid and dilute to 1 litre. sium titanium oxalate in water and dilute to 1 litre. chloric acid (sp.gr. 1.18) plus 5 ml of nitric acid (sp.gr. 1.42). acid (sp.gr. 1.18) and dilute to 1 litre.830 THOMERSON AND PRICE : A COMPREHENSIVE SCHEME FOR THE ANALYSIS OF [Analyst, vol. 96 Stock aluminium solution, 1000 mg I-1-Dissolve 1.0000 g of pure aluminium metal in 25 ml of hydrochloric acid (sp.gr.1-18) plus a few drops of nitric acid (sp.gr. 1.42) and dilute to 1 litre. Stock lead solution, 1000 mg 1-1-Dissolve 1.0000 g of pure lead metal in 10 ml of 2 N nitric acid and dilute to 1 litre. Stock tungsten solution, 1000 mg 1-1-Dissolve 1.420 g of ammonium tungstate in water and dilute to 1 litre. Dilute solutions, when needed, are prepared by diluting the above concentrated solutions, and should be prepared daily. Stock iron solution, 5 per cent.-Dissolve 5 g of high purity iron (B.C.S. 260/3) in 40 ml of hydrochloric acid (spgr. 1.18) plus 5 ml of nitric acid (sp.gr. 1.42). When the reaction is complete, add 20ml of perchloric acid (sp.gr. 1.54) and evaporate until fumes of perchloric acid just appear.Cool and dilute to 100ml with water. Perchloric acid, sp.gr. 1.54. Hydrochloric acid, sp.gr. 1.18. Nitric acid, sp.gr. 1-42. Perchloric acid - phosphoric acid - sulphuric acid (mixture A)-To 300 ml of water, add 100 ml of perchloric acid (spgr. 1.54), 100 ml of phosphoric acid (sp.gr. 1.75) and 100 ml of sulphuric acid (sp.gr. 1.84). METHOD 1. DETERMINATION OF MANGANESE, NICKEL, CHROMIUM, MOLYBDENUM, COPPER, Preparation of sample solutions-Weigh 1.0000 g of sample into a 250-ml beaker and dissolve it in 10ml of hydrochloric acid (spgr. 1.18) plus 5ml of nitric acid (sp.gr. 1.42). After the initial reaction has subsided, add 10 ml of perchloric acid (sp.gr. 1.54) and evaporate the solution until it is fully oxidised and fumes of perchloric acid appear.(This is achieved when the solution turns red and perchloric acid is seen to reflux on the sides of the beaker. For samples not containing chromium, the solution will not turn red but is merely allowed to fume for 5 minutes as specified below.) Allow to fume for about 5 minutes, cool and dissolve the soluble salts in about 50 ml of water. Filter the solution through a Whatman No. 541 filter-paper, wash well with water, adding the washings to the filtrate, and dilute to 100ml. Preparation of calibration solutions-To each of seven 250-ml beakers transfer 1.0 g of pure iron and suitable volumes of stock solutions (Table I). Use the same dissolution procedure as specified above for the samples. It is imperative that the chromium and molybdenum stock solutions be added before fuming takes place. However, if the chromium and molybdenum are to be omitted from VANADIUM, COBALT, TITANIUM, TIN, ALUMINIUM AND LEAD- TABLE I CALIBRATION SOLUTIONS These are based on a 1.0-g sample (except when otherwise stated) and a final volume of 100ml is used throughout Manganese- Dilute stock solution (200 mg l-l)/rnl Concentration/mg 1-1 .. .. Manganese, per cent. .. .. Concentration/mg 1-1 . . .. Nickel, per cent. . . .. .. Concentration/mg 1-1 . . . . Chromium, per cent. .. . . Concentration/mg 1-1 . . .. NickeJ- Stock solution (200 mg l-l)/ml . . Chromium ( u p to 0-5 per cent.)- Stock solution (500 mg 1-f)/ml . . Chromium (up to 4.0 per cent.)- Stock solution (1000 mg l-l)/ml . . Chromium, per cent. (0-25-g sample) 0 1.0 0 1 0 0.01 0 1.0 0 2 0 0.02 0 1.0 0 5 0 0.05 0 1.0 0 10 0 0.4 2.0 2 0.02 2.5 5 0.05 2.0 0.1 2.0 0.8 10 20 5.0 5 0.05 5.0 0.10 4-0 0.2 4.0 1.6 10 20 40 10.0 10 0.10 10.0 20 0.20 6.0 0.3 6.0 2.4 30 60 15.0 15 0.15 15.0 30 0.30 8.0 0.4 8-0 3.2 40 80 20.0 20 0.20 20.0 40 ._ 0-40 10.0 50 0-5 10.0 4.0 100December, 19711 A WIDE RANGE OF STEELS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 831 TABLE I-continued Chromium (up to 12.0 per cent.) *- Stock solution (1000 mg l-l)/ml .. Chromium, per cent. (0.25-g sample) Dilute stock solution (100 mg l-l)/ml Concentration/mg 1-l . . .. Molybdenum (up to 0.2 per cent.)- Concentration/mg 1-1 . . .. Molybdenum, per cent. . . .. Concentration/mg 1-' . . .. Molybdenum, per cent. . . .. Concentration/mg 1-1 . . .. Molybdenum, per cent.. . .. Concentration/mg 1-l . . .. Copper, per cent. . . .. .. Concentration/mg 1-1 . . .. Vanadium, per cent. .. .. Concentration/mg 1-' . . .. Cobalt, per cent. . . .. .. Concentration/mg 1-' . . .. Titanium, per cent. .. .. Concentration/mg 1-1 . . .. Tin, per cent. .. .. .. Concentration/mg 1-1 . . .. Aluminium, per cent. . . .. Concentration/mg 1-1 . . .. Lead, per cent. . . .. .. Concentration/mg 1-l . . . . Tungsten, per cent. . . .. .. Molybdenum (up to 0.5 per cent.)- Stock solution (500 mg 1-I)In-d . . Molybdenum (up to 1.0 per cent.)- Stock solution (1000 mg l-l)/ml . . Copper- Dilute stock solution (100 mg l-l)/ml Vanadium- Stock solution (1000 mg l-l)/rnl . . Cobalt- Stock solution (500 mg l-l)/ml . . Titanium- Stock solution (1000 mg l-l)/ml . . Tin- Stock solution (500 mg l-I)/rnl .. Aluminium- Stock solution (500 mg l-l)/ml . . Lead- Dilute stock solution (100 mg l-l)/ml Tungstent- Stock solution (1000 mg l-l)/ml . . 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2.5 25 1 2-0 2 0.02 1.0 5 0.05 1.0 0.1 1.0 1 0.01 1.0 0.1 1.0 5 0.05 1.0 0.10 1.0 5 0.05 1.0 5 0.05 1.0 1 0.01 5.0 0.5 10 10 10 50 5.0 50 2 5-0 5 0.05 2-0 0.1 2.0 0-2 2.0 2 0-02 2.0 0.2 2-0 0.10 2.5 0.25 2.0 0.10 2.0 0.10 2-0 2 0.02 10 20 20 10 25 10 10 10.0 100 1 7.5 75 3 10.0 10 0.10 4.0 0.2 4.0 0.4 5.0 5 0-05 4.0 0.4 4.0 0.20 6.0 0.50 4.0 0.20 4.0 0.20 5.0 5 0.05 20 40 40 20 50 20 20 20.0 200 2 10.0 100 4 15.0 15 0.15 6.0 0.3 6.0 0.6 10.0 10 30 60 0.10 6.0 0.6 6-0 0.30 60 30 10.0 1.0 6.0 0.30 6-0 0.30 100 30 30 10.0 10 0.10 30.0 300 3 20 200 8 20.0 20 0.20 8.0 0-4 8.0 0.8 15.0 15 40 80 0.15 8.0 0.8 8.0 0.40 80 40 15.0 1.6 8.0 0.40 8.0 0.40 150 40 40 10.0 15 0.15 40.0 400 4 30 300 12 10.0 50 0.5 10.0 1.0 20.0 20 100 0.20 10.0 1.0 10.0 50 100 0.50 20.0 2.0 10.0 50 200 0.50 10.0 50 0.50 20.0 20 0.20 * For this range the burner must be in the fully rotated position.t 50 ml of acid mixture (A) must be added to each of the tungsten calibration solutions (Method 2). the scheme, the stock solutions can be added after the iron has been dissolved for the calibra- tion ranges. Alternatively, the pure iron stock solution can be added to the aqueous standards before dilution (i.e., 20 ml of pure iron stock solution in a final volume of 100 rnl is equivalent to a 1 per cent. sample solution). METHOD 2.DETERMINATION OF TUNGSTEN AND OF MOLYBDENUM WHEN THE CONCENTRATION OF TUNGSTEN IS HIGHER THAN 0.5 PER CENT.- Preparation of sample solutions-Weigh 1.0000 g of sample into a 250-ml beaker, add 50 ml of the acid mixture A and heat gently. When the sample has dissolved, oxidise the solution by adding nitric acid (sp.gr. 1.42) dropwise, and evaporate the solution until the first fumes of perchloric acid appear. Cool, dilute to 50 ml, filter the solution through a Whatman No. 541 filter-paper and dilute to 100ml.832 THOMERSON AND PRICE: A COMPREHENSIVE SCHEME FOR THE ANALYSIS OF [Analyst, Vol. 96 Preparation of calibration solutions-To each of seven 250-ml beakers, add 1.0 g of pure iron and suitable volumes of stock solutions. Use the dissolution procedure specified above for the samples.EXTENSION OF CALIBRATION RANGES- When it is required to determine higher concentrations of alloying elements than those provided for by the recommended calibration ranges, the sample solution should be diluted by an appropriate factor such that the final solution still contains 1 per cent. of iron. All observations of freedom from inter-elemental interferences were made on solutions containing 1 per cent. of iron. I t may therefore be taken as a general principle that, provided both samples and standards contain 1 per cent. of iron, it is possible to use any degree of dilution. In preparing samples known to require range extension the appropriate amount of 5 per cent. stock iron solution can be added before finally making up to volume. A sample already made up to volume can simply be diluted as required with 1 per cent.iron solution derived by appropriate dilution of the 5 per cent. stock solution. ANALYSIS- Recommended instrumental conditions for common constituent elements are summarised in Table 11. The blank (“zero” standard) and calibration solutions should be aspirated followed by the sample solutions. For highest accuracy the blank solution should be run between all other standards and samples and a high standard repeated after every five or ten samples. The calibration graph of absorbance versus concentration is plotted for each element and the concentrations of the elements in the sample solutions are read off. TABLE I1 INSTRUMENTAL CONDITIONS Mn Ni Cr Mo Cu V Co Ti Sn A1 Pb W Wavelength/nm 279.5 232.0 357.9 313-3 324.8 318.4 240.7 364.3 224.0 309.3 217.0 255.1 Slit width/mm 0.03 0.05 0.05 0.05 0.05 0.05 0.03 0.05 0.10 0-05 0.05 0.05 Burner A A N N A N A N N N A N Observation heightlcm 0.8 0-8 0.5 0.5 0.8 0.7 0.8 0.7 0.7 0.7 0.8 0.8 5.0 - - - Air/l min-l 5.0 5.0 - - 6.0 - 5.0 - Acetylene/l min-l* 1.4 0.8 4.2 4.7t 1.0 4.5 1.2 4.5 4.4 4.2 1.2 4-5 Nitrous oxide/l min-l - - 5.0 5.0 - 5.0 - 5.0 5.0 5.0 - 5.0 A = 10 cm air - acetylene burner.N = 5 cm nitrous oxide - acetylene burner with 0.59-mm jaw width (see Apparatus). * These values should be used as a guide only and the flow-rate must be adjusted to give the best sensitivity for each element. t This acetylene flow-rate is critical and must be carefully adjusted to give a flame with a maximum red “feather” height without luminescence.For the chromium range 25 to 300 mg I-’ the burner must be in the fully rotated position to reduce the sensitivity (for the SP90 atomic-absorption spectrophotometer this is about 50” from the optical axis). DISCUSSION The use of perchloric acid as the final medium was adhered to because it has the advantage that it causes the least interference of all common acids used. Originally proposed for the determination of chromium and molybdenum,l its use has proved ideal for the determination of the remaining elements attempted here. The only addition necessary to the aqueous calibration solutions throughout the scheme is that of iron. Therefore, calibration solutions of mixed elements can be used whenever possible with only one simple addition of the stock iron solution or prepared from I-g portions of pure iron (Proposed method).Titanium, however, can be satisfactorily determined by direct comparison with aqueous standards, but can equally acceptably be incorporated into mixed element standards containing iron. Most of the results given in Tables I11 and IV were obtained by using only one calibration range for each element. If, therefore, a better choice of calibration range were made for specific samples greater precision should be possible in most instances.December, 19711 A WIDE RANGE OF STEELS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 833 TABLE I11 RESULTS OF STEEL ANALYSIS BY USING METHOD 1 Element Manganese Nickel Chromium Molybdenum Copper Vanadium Cobalt Titanium Tin Aluminium (acid soluble) Lead Steel type Mild Permanent magnet alloy Ferritic stainless Stainless Mild Ferritic stainless Stainless Permanent magnet alloy Mild Low alloy Ferritic stainless Austenitic stainless Stainless Mild Mild Low alloy Austenitic stainless Mild Ferritic stainless Permanent magnet alloy Permanent magnet alloy Stainless Stainless Mild High speed Mild Stainless Mild High speed Mild Stainless Permanent magnet alloy Permanent magnet alloy High speed Mild Mild Mild Permanent magnet alloy Permanent magnet alloy Mild Mild B.C.S.No. 321 312 341 23512 32 1 341 23512 233 321 25711 339 334 341 32 1 324 2521 1 336 329 341 312 233 261 23512 329 326 23512 329 321 23512 233 312 24111 324 329 321 233 312 326 329 22011 22011 Value found, per cent.0.126, 0-126, 0.124 0.208, 0.208, 0.208 0-425, 0.425, 0.425 0.90, 0.90, 0.88 0.102, 0.102, 0.098 0.58, 0.59, 0.59 9.6, 9.4, 9-4 11.4, 11-4, 11.4 0.12, 0.11, 0.11 2-98, 2-88, 2.94 12.4, 12.4, 12.5 25.4, 25.7, 25.4 24.3, 23.9, 23.7 0.07, 0.07, 0-07 0-17, 0-17, 0.17 1.08, 1.10, 1.08 2-36, 2.40, 2.36 0.070, 0.070, 0.068 0-12, 0.12, 0-12 3.12, 3-12, 3-12 5.13, 5.11, 5-08 0.02, 0.03, 0.03 0-04, 0.05, 0.04 0.10, 0.08, 0.09 2-05, 1.98, 1.98 -0.02, 0.02, 0.02 0.06, 0.06, 0.06 0.05, 0.05, 0.05 0.12, 0.12, 0.12 0.13, 0-12, 0.12 0.31, 0.30, 0-31 0.83, 0.82, 0.83 1-21, 1-19, 1-17 0.02, 0.03, 0.03 0-13, 0.13, 0.12 0.06, 0.06, 0.06 0.13, 0.13, 0.14 6.96, 6.87, 6.87 7.78, 7.84, 7-92 0.016, 0.016, 0.016 0.050, 0.050, 0-048 TABLE IV Certificate value, per cent. 0.13 0.20 0.43 0.89 0.099 0.56 9.38 11.22 0.11 2.97 12.4 25.6 24.0 0.068 0.17 1.11 2.43 0.072 0.10 3.10 5.09 0.03 0.04 0.083 2-09 0.023 0.056 0.07 0.13 0-13 0.32 0.79 1.19 0.025 0.13 0.053 0.12 6.98 7.87 0.014 0.050 RESULTS OF STEEL ANALYSES BY USING METHOD 2 Certificate Element Steel type B.C.S.No. per cent. per cent. Value found, value, Tungsten Mild High speed High speed Molybdenum High speed High speed 323 0-25, 0.25, 0.25 0.25 22011 6.88, 6.88, 6.88 6-86 24111 19.4, 19.8, 19.4 19.61 24111 0.56, 0.56, 0-56* 0-52 22011 5.30, 5.30, 5-30? 5.20 Certificate range, per cent. 0.12-0.13 0.19-0.22 0.41-0.44 0.88-0.90 0.096-0.105 9-34-9.42 11.14-11*28 0.1 0-0.1 1 2.95-3.01 12.3-12-5 25.5-25- 7 23.9-24.1 0*064-0-072 0.55-0.58 0.16-0.18 1 -09- 1.1 3 2.39-2.46 0.07 1-0.075 2.99-3-15 - 5.04-5.16 - 0.03-0.04 0*079-0.085 2.08-2.13 0.02 1-0.025 0.056-0.057 0.068-0.072 0.12-0.13 0.127-0.139 0.3 1-0.33 0.78-0.80 1 - 16-1-23 0.024-0*028 0.12-0- 14 0.049-0.057 6.89-7-08 7-7 7-7-93 0.01 2-0.0 17 0.042-0.052 - Certificate range, per cent.0.24-0.26 19.561 9.72 6.7 8-7-00 0.51-0.52 5.15-5.2 7 * Results obtained in presence of 19-6 per cent. of tungsten. t Results obtained in presence of 6.9 per cent. of tungsten.834 THOMERSON AND PRICE Based on the fact that interference and suppressions are minimised, and between samples and standards are equalised, in the presence of 1 per cent. of iron, the method is readily ex- tended to higher concentration ranges simply by the dilution of samples and addition of iron to give a concentration of 1 per cent. An increase in reading accuracy can also be achieved by using the difference method that has previously been described.1937 The only restriction on the use of perchloric acid results from its failure to retain silicon and tungsten in solution.The tungsten can be determined separately (Method 2) without any problem arising. The silicon recovery, after filtering the sample solution, is quantitative, so we recommend the use of either the traditional simple gravimetric method or a separate atomic-absorption spectrophotometric method of analysis as exemplified by Price and Roos38 and by M~Auliffe.~Y A great advantage most atomic-absorption methods have over traditional techniques is the small sample volume or weight required. This scheme for the analysis of steels is no exception and it is possible to carry out the complete analysis on only 1 g of sample (for the determination of tungsten and silicon separate sample weighings are, of course, needed).Consequently much time can be saved compared with most laboratory methods in common use. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. REFERENCES Thomerson, D. R., and Price, W. J., Analyst, 1971, 96, 321. Belcher, C. B., and Kinson, K., Analytica Chim. Acta, 1964, 30, 483. Atsuya, I., Sci. Rep. Res. Insts Tdhoku Univ., A , 1967, 19, 59. Ramirez-Munoz, J., and Roth, M. E., Beckman Flame Notes, 1969, 2, 3. Hubbard, D. P., and Monks, H. H., Analytica Chim. Acta, 1969, 47, 197. McPherson, G. L., “Proceedings of the 17th BISRA Chemists Conference, Scarborough, 1964,” Suzuki, M., and Takeuchi, T., J .Chem. SOC. Japan, Ind. Chem. Sect., 1964, 67, 1207. Beyer, M., Atomic Absorption Newsletter, 1965, 4, 212. Sprague, S., and Slavin, W., Dev. Appl. Spectrosc., 1965, 4, 433. Clarke, W. E., British Cast Iron Research Association Report No. 873, 1967, p. 243. Price, W. J., “XIIIth Colloquium Spectroscopicum Internationale,” Ottawa, 1967. Clarke, W. E., and Cooke, P. A., British Cast Iron Research Association Report No. 891, 1967, Price, W. J., and Cooke, P. A., Spectrovision, 1967, 18, 2. Heinz, K., and Ohls, K., Arch. EisenhtiittWes., 1968, 39, 925. Jimenez Seco, J. L., and Gomez Coedo, A., Revta Metalurgia, 1968, 4, 621. Knight, D. M., and Pyzyna, M. K., Atomic Absorption Newsletter, 1968, 8, 129. Feldman, F. J., Blasi, J. A., and Smith, S. B., Analyt. Chem., 1969, 41, 1095. Kinson, K., and Belcher, C. B., Analytica Chim. Acta, 1964, 30, 64. Ramirez-Munoz, J., and Roth, M. E., Beckman Flame Notes, 1969, 4, 102. Carper, J. L., Atomic Absorption Newsletter, 1970, 9, 2. Wallace, F. J., Foseco Dev., 1961, 7, 54. -, Hilger J., 1963, 7, 65. Kinson, K., and Belcher, C. B., Analytica Chim,.. Acta, 1964, 31, 180. Konig, P., Heinz, K., and Thieman, E., Arch. EisenhiittWes., 1969, 40, 53. Capacho-Delgado, L., and Manning, D. C., Atomic Absorption Newsletter, 1966, 5, 1. McPherson, G. L., Price, J. W., and Scaife, P. H., Nature, Lond., 1963, 199, 371. Lockyer, R. L., Hilger Watts Research Report, BR.26, 1965. Bowman, J. A., and Willis, J. B., Analyt. Chem., 1967, 39, 1210. Headridge, J. B., and Hubbard, D. P., Analytica Chim. Acta, 1967, 37, 161. Mostyn, R. A., and Cunningham, A. F., Atomic Absorption Newsletter, 1967, 6, 86. Nikolaev, G. I., J . Analyt. Chem. U.S.S.R., 1965, 20, 412. Amos, M. D., and Thomas, P. E., Analytica Chim. Acta, 1965, 32, 139. Endo, Y., Ohata, H., and Nakahara, Y., Japan Analyst, 1967, 16, 364. Konig, P., Schmitz, K. H., and Thieman, E., 2. analyt. Chem.. 1969, 244, 232. Elwell, W. T., and Gidley, J. A. F., Analytica Chim. Acta, 1961, 24, 71. Dagnall, R. AT., West, T. S., and Young, P., Analyt. Ckem., 1966, 38, 358. Thomerson, D. R., Spectrovision, 1971, 25, 12. Price, W. J., and Roos, J. T. H., Analyst, 2968, 93, 709. McAuliffe, J. J., Atomic Absorption Newsletter, 1967, 6, 69. 1964, p. 12. p. 526. Received February 3rd, 197 1 Accepted July 30th, 197 1

ISSN:0003-2654    

DOI:10.1039/AN9719600825

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, December, 1971, Vol. 96, $9. 835-842 835 Determination of Sulphur in Steel by a Modified Combustion Procedure and Coulometric Titration BY R. KAJIYAMA (Nzppon Yakin Kogyo Co. Ltd., Kawasaki, Japan) AND K. HOSHINO (Kokusai Electric Co. Ltd., Harnura, Tokyo, Japan) The quantitative determination of sulphur in steel by the combustion method is inefficient as the sulphur content of samples indicated by this method is lower than that given by other methods because sulphur trioxide occurring in the combustion furnace tends to condense in the piping. An attempt has been made to establish how this sulphur trioxide condensation occurs, and a unique, automatic circulation type of adsorption method has been devised to ensure complete collection of gaseous sulphur occurring in the combustion.For this purpose, a practical sulphur analyser with high sensitivity and accuracy has been developed in which an automatic coulo- metric titrator and a high-frequency induction furnace are combined. Standard samples of various types of iron and steel analysed with this apparatus showed that the results agreed well with their standard values. AMONG the existing methods for the quantitative determination of sulphur in steel the combustion method1,2 is most commonly used in routine work because of its rapidity and simplicity of procedure. However, this method does not permit the complete collection of the combustion gases, and therefore has the basic disadvantage that it gives a lower value for the sulphur content of a sample than that obtained by other methods.Various aspects of this problem have been studied by other workers in the past.39495 More recently, some investigators have used tracers to determine the distribution of sulphur remaining after a sample has been analy~ed,~,~,* and as a result of their work it is now believed that the characteristic low analytical results obtained are caused by the condensation of sulphur trioxide, which is formed with the gaseous sulphur, in the combustion process. We have found that sulphur trioxide condenses on the cooler parts of the combustion tube, on the dust filter, on the piping line carrying the gas to the absorption cell and elsewhere, resulting in incomplete collection of the gaseous sulphur from the combustion gas. Most other investigators in the past have merely sought to identify the cause or causes of the low results obtained by this method, and when analysing samples they have adopted the practice of compensating for the low results by using a specific coefficient or factor.1J However, if the ambiguousness in the analytical results is attributable to the behaviour of the sulphur trioxide formed in the combustion, its mode of formation must differ, depending on the conditions, the kind of steel analysed, its sulphur content and other factors.It is unreasonable to expect a single coefficient to provide adequate compensation for all these factors, and the use of several alternative coefficients involves difficulties. To avoid these difficulties it was decided, after considering the various possibilities, to use a specially designed apparatus, with which the errors could be reduced to a minimum.By using this apparatus, we have analysed various kinds of samples and have obtained analytical results about 10 per cent. higher than those obtained with the conventional direct combustion methods, which agree well with the standard values for the samples. EXPERIMENTAL APPARATUS- Diagrams of the analyser used are shown in Figs. 1 and 2. Combustion furnace-A high-frequency combustion furnace* was modified for the purpose * Made by Kokusai Electric Co. Ltd. 0 SAC and the authors.836 KAJIYAMA AND HOSHINO: DETERMINATION OF SULPHUR I N STEEL BY [Artdyst, Vol. 96 Oxygen I Pulse Gate I generator + circuit --f I i ~~ Coulometric tit rat0 r I I Indicator I circuit I I I Pu J Circulation-type Heating furnace (30O0C) Absorption H igh-frequency induction furnace ' f ' Oxygen Fig.1. Diagram of analyser , KM-HYI! detecting supplier I I I 1 I J. I I Fig. 2. Block diagram of analyser of sulphur determination. In our apparatus, the combustion tube is of the constricted-end type and the sulphur trioxide formed is prevented from condensing by heating the combustion tube, and the dust filter assembly that follows it, to about 750 "C with an electric heater. Automatic circulation-type jiushing device-This device ensures that all gaseous sulphur formed by the combustion is carried into the absorption solution without condensing in the tubing. With the conventional method, a glass tube is simply placed into the absorption solution. In such an arrangement, the tip of the glass tube is cooled and sulphur trioxide condenses on the inner surface of that part of the tube.Also, an intermittent oxygen flow causes a part of that surface to become wet with solution, and gaseous sulphur tends to condense on that part ; therefore the glass tube should be flushed with the absorption solution.December, 19711 A MODIFIED COMBUSTION PROCEDURE AND COULOMETRIC TITRATION 837 This flushing is achieved automatically in our apparatus and, in addition, a heating furnace is used to prevent condensation of sulphur trioxide. The apparatus is made of transparent quartz, and arranged as shown in Fig. 3. As it is designed for simultaneous heating and flushing, the device is kept in a heating furnace and connected to the combustion tube with a spherical fitting.A pump constantly circulates the absorption solution, which flushes the tubing line. The combustion gas together with excess of oxygen is led via the route shown to the absorption cell. When the combustion tube is opened for sample introduction the oxygen flow line is automatically changed, and the back stopper shown in Fig. 3 ensures that only oxygen passes into the cell. Oxygen To cell Fig. 3. Automaticcirculation-type flushing device Cell assembly-As illustrated in Fig. 4, an absorption cell and an anode cell are set together with a porous diaphragm (a kind of Kibushi-nendo, or K-clay, of porosity about 25 to 30 per cent.) between them, each cell containing a platinum electrode. The absorption cell is also connected in a similar way to a silver chloride reference electrode. Separated cells are used to prevent interference from electrical, thermal or mechanical causes.The absorption solution consists of about 100 ml of 1 per cent. sodium sulphate solution containing hydrogen peroxide, the appropriate amount of hydrogen peroxide (1 to 2 drops of 30 per cent. hydrogen peroxide) being added, either automatically or manually. The anode cell is filled with about 50 ml of a saturated solution of sodium sulphate, which includes undissolved solid. Measwing unit-The measuring unit of the Coulomatic “C” instrumentg was modified for the purpose of sulphur determination. The reference electrode consists of silver - silver chloride and the measuring electrode is a glass electrode. Pulse current is used for coulo- metric titration, each pulse giving an equal quantity of electricity which is equivalent to 0.5 x g of sulphur.As a 1-g sample is used for analysis, two pulses represent a sulphur content of 0.0001 per cent. The number of pulses, that is, the analytical result, is indicated digitally on an electromagnetic counter with a six-digit capacity.838 KA JIYAMA AND HOSHINO: DETERMINATION OF SULPHUR IN STEEL BY [Arta/?J.td, VOl. 96 device Na2S04 (Saturated) To pump f- = a E , Fig. 4. Cell arrangement A = Absorption cell B = Anode cell C = Reference electrode D = Glass electrode E, and E, = Electrodes SAMPLE- steel), which contains 0.798 per cent. of carbon and 0.023 per cent. of sulphur. The sample used in the present study was the Japanese Standard sample 5d (carbon REAGENTS- experiments, were of analytical-reagent grade.The metallic tin used as combustion accelerator, and the various reagents used in the COMBUSTION CONDITIONS- The samples were subjected to Combustion under the following conditions : sample weight, 1 g; combustion accelerator, metallic tin in small grains, 1 g; and oxygen flow, 2 1 min-1. SULPHUR EXTRACTION AT HIGH FREQUENCY COMBUSTION- From observations on the changes in pH of the absorption solution, as measured with the pH meter attached, and also on the behaviour of the sample during combustion, it is noted that by turning on the high-frequency switch of the combustion furnace, combustion starts at a low temperature, resulting in extraction of the carbon. The time taken for the extraction, although depending somewhat on the shape of the sample and the kind of steel, is determined primarily by the carbon content of the sample, as shown in Fig.5. With a pig iron, explosive combustion occurs several times. Sometimes complete combustion of the carbon requires as long as 5 to 6 minutes, after which the sample becomes white hot; a t this stage extraction of the sulphur begins. The carbon dioxide produced early in the com- bustion process temporarily dissolves in the absorbing solution and changes its pH, but with agitation of the solution the carbon dioxide is released and the pH of the solution returnsDecember, 197 13 A MODIFIED COMBUSTION PROCEDURE AND COULOMETRIC TITRATION 839 to the pre-set level. Hence we consider that the measuring unit should be switched on after extraction of the carbon from the sample and the release of the carbon dioxide from the absorption solution have been completed. In our experiments, we started the titration 3 minutes after combustion began on each sample.pH OF ABSORPTION SOLUTION- Carbon dioxide is always produced early in an analysis cycle and passes into the absorp- tion solution. To facilitate the elimination of the effect of the carbon dioxide, it is preferable to use a solution with an acidic pH. Therefore, we have examined the effect of pH of the absorption solution on the analytical results and found, as Fig. 6 illustrates, that the effect of carbon dioxide is negligible when the pH is below 4 but that a clear titration end-point cannot be obtained at this pH level.If, on the other hand, the pH is 5, much of the carbon dioxide is absorbed, and more time is required to drive it off. For these reasons, we adjusted the pH of the absorption solution to 4.5. t so;! + so3 In + C 3 8 E \ c, C .- + m L c, '""I2 l I A 0 0 1 2 3 4 5 0 1 2 3 4 5 Tirne/minutes Time/minutes Fig. 5. Typical extraction patterns for Fig. 6. Extraction patterns a t various pH values: (a), pH 4.5; ( b ) , pH 4-0; and ( c ) , various carbon contents: (a), 0.015 per cent.; ( b ) , 0.80 per cent.; and (G), 3.87 per cent. of carbon pH 3.0 CALIBRATION OF APPARATUS- The proposed method is based on coulometric titration with a pulse current, and each pulse is designed to represent 0.5 x 10-6g of sulphur. To ascertain this relationship, we used a combination of an electrical method and a chemical method.In the former, a certain quantity of electricity (20.0 mA for 50.0 s = 1 coulomb = 332.2 pulses) was used to produce sulphuric acid in the absorption cell, and the number of pulses required to neutralise the acid was taken as the analysis count. In the latter, a known amount of a standard solution of sulphuric acid was introduced dropwise into the absorption cell, and the number of pulses required to neutralise the acid was measured. The results in both tests showed good agree- ment with the theoretical value, thus indicating that the apparatus is sufficiently reliable. CONDENSATION OF SULPHUR TRIOXIDE IN A TUBING LINE- We analysed samples by using a conventional tubing line. As had been expected, most of the results obtained were found to be lower than normal.We therefore examined the quantitative distribution of sulphur present in the various parts of the tubing shown in Fig. 7. Ten samples were analysed in succession, and the amount of sulphur that had condensed in each part was determined as follows: an absorption solution was used to flush the combus- tion tube and the various parts of the tubing line, the flushing solution was passed into840 KAJIYAMA AND HOSHINO: DETERMINATION OF SULPHUR IN STEEL BY [Analyst, Vol. 96 Absorpt Fig. 7. Conventional tubing line between combustion tube and absorption cell the absorption cell and the resulting increase in counts was measured. The results are shown in Table I, from which it can be seen that considerable amounts of sulphur were present in the various parts of the tubing line.We also found that the amount of sulphur that had condensed in the line increased as more samples were analysed in succession. The condensed material was identified chemically as sulphur trioxide by the barium chloranilate method and the rosaniline method.10 TABLE I SULPHUR TRIOXIDE CONDENSATION IN CONVENTIONAL Counts in Counts in cool part of dust filter and Total counts Counts per combustion tube tubing for condensation sample analysed 194 230 424 42.4 SULPHUR COLLECTION BY HEATING AND FLUSHING- TUBING LINE Condensation count Sample count 8-5 per cent. x 100 The condensate in the tubing line is apparently sulphur trioxide as this substance solidifies at room temperature and would condense on cooler parts of the line.It is reasonable to assume, therefore, that if the tubing system is heated to prevent sulphur trioxide vapour from condensing it may be wholly carried over into the absorption cell. To support this assumption, we used a constricted-end rather than a cylindrical combustion tube, and moved the dust filter into the combustion tube. In addition to heating this assembly, we also heated the entire tubing line leading to the absorption cell (Fig. 8). As a result, sulphur trioxide condensation in the heated parts decreased sharply, as may be seen in Table 11. However, much of the sulphur trioxide driven out of the heated parts of the combustion tube and the tubing line was found to be deposited again at the end of the line, Le., the final section Heater 110 or 310°C / il Heater 700°C Fig.8. Combustion tube and tubing line with heating arrangementDecember, 197 13 A MODIFIED COMBUSTION PROCEDURE AND COULOMETRIC TITRATION 841 TABLE I1 SULPHUR TRIOXIDE CONDENSATION REDUCED BY HEATING Heating temperature/’C Condensation count with ten samples Combustion Tubing Combustion Tubing Tubing ’ tube A line B tube A line B line C Total Room Room 185 13 (B + C) 198 700 Room 8 221 (B + C) 229 700 110 2 61 153 216 700 310 2 16 164 182 before the absorption cell. Thus, we have found that, in practice, heating the tubing line alone cannot completely solve the problem of sulphur trioxide condensation. Another inconvenience was that the inflow of a large amount of hot combustion gas into the absorption cell raised the temperature of the solution.Therefore, an automatic circulation-type flushing device was designed, as shown in Fig. 3, which was used in com- bination with heating and thus made possible the complete collection of the sulphur from the combustion gas. With this apparatus, ten samples of high-sulphur steel (0.1 per cent.) analysed in succession produced an amount of residual sulphur that corresponded to a count of only 2. RESULTS By using the above “Coulomatic S” apparatus, various standard samples of steel were analysed and the results obtained, which are summarised in Table 111, are sufficiently close to the standard values and generally about 10 per cent. higher than those obtained by the conventional direct combustion method, Le., without precautions being taken against condensation of sulphur trioxide.Also, as shown in Table IV, a high degree of reproducibility of analytical results was obtained with each type of steel. Sample JSS 1-d JSS 4-d JSS 2-e JSS 5-d B.C.S. 221/1 B.C.S. 290 B.C.S. 224 B.C.S. 211/1 N.B.S. 65d N.B.S. 121c N.B.S. lOle N.B.S. l l l b N.B.S. 139a N.B.S. 5k Type Carbon steel 13% Mn steel Cr - V steel 13% Cr steel Carbon steel 18% Cr, 11% Ni steel 18% Cr, 9% Ni steel Ni - Mo steel Cr - Ni - Mo steel Cast iron TABLE I11 RESULTS OF ANALYSIS Standard Present value of sulphur, method (A), per cent. per cent. 0.030 0.0342 0.024 0.0258 0.016 0.0165 0.023 0-0252 0.03 1 0.0308 0.019 0-0185 0.029 0.0295 0.032 0.0330 0.010 0.0078 0.009 0.0084 0.010 0.0083 0.015 0.0130 0.019 0-0176 0.099 0-0974 Conventional method (B), per cent.0.0311 0.0236 0.0155 0.0231 0.0278 0.0166 0.0268 0.0292 0-007 1 0.0077 0-0072 0.0116 0.0155 0.0902 TABLE IV REPRODUCIBILITY IN ONE DAY Results for sulphur content, Mean value, Standard Sample per cent. per cent. Range deviation Electrolytic 0.0069, 0.0063, 0.0063, 0.0064, 0.0061, 0.0064, 0-0008 O*OOO 23 JSS 4-d JSS 5-d 0.0252, 0.0253, 0.0243, 0.0254, 0.0244, 0.0248, 0.001 1 0.000 48 N.B.S. 82a 0.1006, 0.0991, 0.0999, 0.1000, 0.0995, 0*1005, 0.0029 0.000 97 iron 0.0065, 0.0065, 0.0065, 0.0063, 0.0067 0.0164, 0-0160, 0.0162, 0.0166, 0-0166 0.0244, 0.0244, 0.0244, 0.0252, 0.0253 0.1010, 0.1016, 0-1000, 0.1015, 0.1020 0.0165, 0.0165, 0.0165, 0.0163, 0.0167, 0.0164, 0.0007 0.000 2 1842 KAJIYAMA AND HOSHINO CONCLUSIONS Quantitative analysis of steel for the determination of sulphur by using our apparatus, as the above results demonstrate, has several important advantages, including the following : the apparatus is highly sensitive and very small sulphur contents can be determined; measure- ments are made with high precision; the analytical procedure is automatic and free from operator errors; and results of analysis can be obtained easily and rapidly.In particular, the unique flushing device with heating and automatic circulation features ensures complete collection of the sulphur from the combustion gas, which is an outstanding improvement in a practical device for determining sulphur in steel. The authors wish to thank Dr. K. Goto, President of Toyama University (formerly Professor a t Tbhoku University) for his many helpful suggestions and discussions during the present study. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Japanese Industrial Standard G 1215, 1963. A. S.T.M. E30-60TI American Society for Testing Materials, Philadelphia, 1960. Tawara, K., and Mitsui, S., “Tekko Kugaku Bunseki Zensho” (Chemical Analysis of Steel Series), Smith, T. B., Backhanse, A., and Woodward, P., J . Appl. Chem., Lond., 1954, 4, 75. Holler, A. C., Klinkenberg, R., Friedman, C., and Aites, W. K., Analyt. Chem., 1954, 26, 1658. Rooney, R. C., and Scott, F., J . Iron Steel Inst., 1970, 417. Takano, S., Shirai, T., Endo, T., and Matsushima, I., J . Iron Steel Inst. Jupun, 1961, 46, 1069. Fulton, J. W., and Fryxell, R. E., Anulyt. Chem., 1959, 31, 401. Kajiyama, R., Watanabe, M., and Mochizuki, H., Bunseki K i k i , 1965, 3, 1. “Taiki Osen no Sokutei” (Measurement of Air Pollution), Edited by National Council on Air Received December 21st, 1970 Accepted May 5th, 1971 First Edition, Volume 2, Nikkan Kogyo Shinbunsha, 1952, p. 174. Pollution, Corona-Sha, 1962, pp. 139 and 148.

ISSN:0003-2654    

DOI:10.1039/AN9719600835

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, December, 1971 , Vol. 96, pp. 843-846 843 The Determination of Uranium by Atomic-absorption Spectrophotometry BY MARGARET J. MARTIN (The Gas Council, London Research Station, Michael Road, Fulham, London, S . W.6) The determination of uranium by atomic-absorption spectrophotometry is complicated by significant interference effects and demands critical control of fuel composition and burner adjustment. For the determination of uranium in uranium-based nickel - uranium catalysts a solvent-extraction method was used to avoid the matrix effects. CATALYSTS containing nickel and uranium on a corundum (a-alumina) support are used in the gas industry for the continuous reforming of hydrocarbon feedstocks. Other metals may also be present as deliberate additions or as impurities and these can affect the accuracy of chemical and X-ray fluorescence methods for the determination of uranium.For this reason it was decided to investigate the possibilities of using atomic-absorption spectro- photometry as an independent method. EXPERIMENTAL APPARATUS- A Techtron AA4 atomic-absorption spectrophotometer was used, equipped with an AB50 50-mm slot burner and a uranium hollow-cathode lamp made by Atomic Spectral Lamps. The instrument settings were as follows. The wavelength used was either 358.5 or 356.7 nm, the slit width 25pm (0.08nm optical slit width), the lamp current 15 mA, the pressure of nitrous oxide 138 kN m-2 (9.6 1 min-l) and the acetylene flow was set to give a luminous flame, being finally adjusted to give maximum response (3.4 1 min-1).REAGENTS- All other reagents were of analytical-reagent grade. nitrate containing 10 mg ml-l of uranium. CHOICE OF FUEL AND ABSORPTION WAVELENGTH- In a series of preliminary experiments it was found that little or no absorption took place unless the acetylene-to-nitrous oxide ratio and burner height were kept within extremely close limits. An acetylene-rich flame was found to be necessary and the beam from the hollow-cathode lamp had to just clear the top of the burner. In the first instance a 50-mm (AB40) flat-topped nitrous oxide burner was used, but it was found that the fuel-rich condi- tions caused immediate carbon deposition. However, this effect was minimised by the use of the grooved AB50 burner. The most sensitive absorption lines for uranium are at 358.5 and 356.7 nm.l The 358.5 nm line is the more sensitive, but it is in a region of high CN band absorption where the light beam passing through the primary zone gives a high noise level.If it is necessary to use the instrument at maximum sensitivity it is therefore better to use the 356.7 nm line. Although other lines are available in regions of very low CN band absorption their sensitivities are too low for practical use. SENSITIVITY- The sensitivity (defined as the concentration of metal required to give 1 per cent. absorbance) of uranium at a wavelength of 3586 nm has been quoted2 as 120 pg ml-1, but in our investigations with uranyl nitrate we were able to record a sensitivity of only 250 pg ml-l. The nitric acid used to extract the uranium from the catalyst was of Aristar grade.The standard uranium solutions were obtained by diluting a stock solution of uranyl 0 SAC and the author.MARTIN : DETERMINATION OF URANIUM BY [Analyst, Vol. 96 844 STUDY OF INTERFERENCE- The effect of a number of ions and compounds on the absorption of uranium was investi- gated. The solutions contained 2-5 mg ml-l of uranium and either 10 mg ml-l of a foreign ion or amounts of various acids such that their concentrations were N. The metals were added as the nitrates and the anions as the potassium salts. The results are given in Tables I, I1 and 111. TABLE I INTERFERENCE OF METALS Foreign metal a t a concentration of 10 mg ml-l None Na K Ba Mg Foreign anion a t a concentration of 10 mg ml-l None NO,- HS0,- Mn0,- HF,- CraO,a- sop- Percentage absorption of uranium in presence of interfering ion a t a wavelength of 388-5 nm 356-7 irn 11.5 5.5 13.0 6.0 13.0 6.0 18.8 8.5 14.0 7.3 Foreign metal a t a concentration of 10 mg ml-l A1 La Pb Ff3 c o Ni TABLE I1 INTERFERENCE OF NON-METALS Percentage absorption of uranium in presence of interfering ion a t a wavelength of & 358-5 nm 356-7 nm 11.5 5.5 12.5 7.5 11.5 5.5 14.5 11.0 14.0 9.8 8.5 5.0 3.0 2.0 Foreign anion a t a concentration of 10 mg ml-1 F- c1- Br- I- Fe (CN) 63- Fe(CN) 64- CH3COO- Acid TABLE I11 INTERFERENCE OF ACIDS Percentage absorption of uranium in presence of interfering ion a t a wavelength of 1 358.5nm 356.7 nm 21.7 11.0 7.7 4.5 15.0 7.5 13.5 6.5 24.3 14.5 20.8 11.0 Percentage absorption of uranium in presence of interfering ion a t a wavelength of & 3586nm 356.7 nm 4.5 2.0 9.8 5.0 9.8 5.0 9.8 5.0 1.0 < 1.0 1.0 < 1.0 4.7 2.0 Percentage absorption of uranium in presence of interfering acid a t a wavelength of 7- 358.5 nm 356.7 nm None .... .. .. 11.5 Acetic acid, N . . .. .. 9-8 Hydrochloric acid, N . . .. 15.0 Nitric acid, N . . .. .. 11.5 Sulphuric acid, N . . .. .. 6.5 5.5 7.5 10.0 5.5 5.5 ANALYSIS OF CATALYST- Because uranium absorption is so sensitive to impurities, it was decided to isolate the uranium from its matrix by using a solvent-extraction procedure. Sufficient catalyst sample to give a uranium concentration of about 10 mgml-l was heated with 50 ml of nitric acid (1 + 1). The solution was boiled down to about 10 ml, filtered, and made up to 50 ml. The residual alumina was examined by X-ray fluorescence spectrometry to check that it contained no uranium, while the uranium was extracted from the solution into ethyl acetate by using the method described by Guest and Zimme~mann.~ (A 5-ml portion of the uranium solution is placed in a separating funnel and extracted with ethyl acetate after adding a concentrated solution of aluminium nitrate to act as a salting- out agent.)December, 19711 ATOMIC-ABSORPTION SPECTROPHOTOMETRY 845 Originally it was hoped to increase the sensitivity of the uranium absorption by spraying the ethyl acetate solution directly, but in practice the sensitivity was decreased.This de- crease was probably caused by the organic solvent altering the flame characteristics, but without secondary nitrous oxide it was impossible to vary the ratio of fuel to nitrous oxide effectively so as to verify this contention. Reasonably accurate absorption readings could be obtained by using maximum instrument sensitivity but it was essential to spray water for a prolonged period between each uranium solution to prevent a memory effect, which may have occurred as a result of evaporating the volatile solvent with consequent deposition of uranium in the nebulising system.Because a burner with secondary nitrous oxide was not available, we re-extracted the uranium into water by the method described by Guest and Zimmermann.3 In using this method, we found that three water washes were adequate rather than the five recommended. We established this by examining each washing fraction for uranium by X-ray fluorescence spectrometry. Because of the solubility of ethyl acetate in water the viscosity of a uranium solution after solvent extraction will differ from that of a solely aqueous solution.As the altered viscosity will affect the nebulisation rate it is essential to prepare the standard uranium solutions by extracting uranyl nitrate solutions with ethyl acetate and then back-extracting with water. A number of catalysts of known uranium content were analysed by the method described here. The results obtained are given in Table IV. TABLE IV DETERMINATION OF URANIUM CONTENT Uranium content found I True uranium content, - Mean of 4 readings, Standard Catalyst per cent. deviation per cent. A 5-37 B 7.02 C 8.17 D 8-70 E 10.30 0.06 0.06 0.07 0-24 0.25 5.3 7.0 8.2 8.6 10.2 DISCUSSION It can be seen from Tables I to I11 that, with the exception of nitric acid, every compound or ion that was added to the uranium solution had some effect on the uranium absorption.It is estimated that the ionisation of uranium in the nitrous oxide - acetylene flame is 45 per cent.1 It is therefore surprising that the addition of an excess of alkali metals did not suppress the ionisation of uranium sufficiently to give a considerable increase in the uranium absorption. Interference effects in the nitrous oxide - acetylene flame have been found by other workers. The metals examined include the alkaline earths,* berylli~m,~J titanium2p6 and vanadium.' Koirtyohann and Pickett4 suggest that many of the interference effects in the nitrous oxide - acetylene flame relate to the spatial distribution of the sample within the flame.Sachdev, Robinson and West7 found a number of interference effects in the deter- mination of vanadium when using the nitrous oxide - acetylene flame. Metals that form stable oxides, such as aluminium and titanium, gave a large enhancement of the absorption signal. Sachdev et al. suggest that this is due to competition for available oxygen, which shifts the following equilibrium in favour of vanadium atoms- It is possible that a similar mechanism exists for uranium- v + o + v o u,o, + x u 0 + OYsX U O $ U + O If the above reactions take place, then it should be possible to increase the sensitivity of uranium by using a large excess of, for example, aluminium.846 MARTIN CONCLUSIONS The determination of uranium requires critical control of burner height and fuel com- position, and is subject to interference from many elements and ions.Matrix effects can be avoided by extracting uranium as the nitrate with ethyl acetate and back-extracting it with water. Satisfactory results were obtained for the determination of uranium in alumina-based catalysts but the sensitivity when using a practical sample weight is insufficient for uranium contents of less than 2.5 per cent. The author thanks the Gas Council for permission to submit this paper for publication. She also acknowledges the help of colleagues at the Gas Council’s London Research Station during the preparation of this paper, especially that given by J. E. Davey, who carried out the X-ray fluorescence analysis. REFERENCES 1. 2. 3. 4. 5. 6. 7. Manning, D. C., Atomic Absorption Newsletter, 1966, 5, 127. Amos, M. D., and Willis, J. B., Spectrochim. Acta, 1966, 22, 1325. Guest, R. J., and Zimmermann, J. B., Analyt. Chem., 1955, 27, 931. Koirtyohann, S. R., and Pickett, E. E., Ibid., 1968, 40, 2068. Fleet, B., Liberty, K. V., and West, T. S., Talanta, 1970, 17, 203. Bowman, J. A., and Willis, J. B., Analyt. Chem., 1967, 39, 1210. Sachdev, S. L., Robinson, J . W., and West, P. W., Analytica Chim. Acta, 1967, 37, 12. Received Novelnber 26th, 1970 Accepted May 31.4 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600843

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

Analyst, December, 1971, Vol. 96, $9. 847-853 847 Selective Atomic-absorption Determination of Inorganic Mercury and Methylmercury in Undigested Biological Samples BY L. MAGOS (Medical Research Council Laboratories, Toxicology Unit, Woodmansterne Road, Carshalton, Surrey) A simple method for the determination of total mercury in biological samples contaminated with inorganic mercury and methylmercury is des- cribed. The method is based on the rapid conversion of organomercurials first into inorganic mercury and then into atomic mercury suitable for aspira- tion through the gas cell of a mercury vapour concentration meter, by a combined tin(I1) chloride - cadmium chloride reagent. It was found that if 100 mg of tin(I1) chloride alone were added instead of the tin(I1) chloride - cadmium chloride reagent, only the release of inorganic mercury influenced the peak deflection of the potentiometer, thus permitting the selective deter- mination of inorganic mercury in the presence of methylmercury.It was possible first to release inorganic mercury then, after re-acidification of the reaction mixture, methylmercury, by adding the tin (11) chloride - cadmium chloride reagent and sodium hydroxide. When total mercury and inorganic mercury were determined separately, the difference between results gave the methylmercury content of the sample. MERCURY and its compounds are released into the environment as a result of industrial and agricultural activities and of the weathering of mercury-bearing rocks. Because of biological methylation in aquatic organisms1 mercury accumulates as methylmercury in ecological systems.Contamination of food, particularly fish, with highly toxic methylmercury has prompted toxicological research on this compound and compelled authorities to analyse a large selection of foodstuffs for mercury. Techniques requiring the digestion of samples do not enable inorganic mercury and methylmercury to be distinguished, so that before digestion organomercurials have to be extracted with benzene and determined either by a titrimetric method,2 or by gas chromato- graphy,3 but not by atomic absorption as benzene gives falsely high readings. Gage and Warren4 avoided extraction and made use of the varying lability of methoxyethylmercury, phenylmercury and ethylmercury in the presence and absence of cysteine with and without a 1-hour digestion.For evaluation they used the atomic-absorption method of Magos and Cernik6 in which mercury is determined by aspirating the vapour through an ultraviolet absorptiometer after reduction with tin( 11) chloride. Unfortunately, methylmercury in the presence of cysteine releases mercury from the carbon bond at a rate of only 0.4 per cent. per day,6 and it is the experience of the author that not more than one third of the mercury can be released even after digestion for 1 hour in acidic cysteine solution. In experiments in which 203Hg-labelled mercury compounds are used, inorganic mercury can be selectively determined in the presence of organomercurials, either by an isotope- exchange method,s or by its selective reduction with tin(I1) chloride.' The method described here is based on the discovery that the rate of reduction of methylmercury or other organomercurials by tin(I1) chloride can be made identical to that of inorganic mercury if the amount of tin(I1) chloride is above a certain level and a cadmium salt is added to the reaction mixture.Thus, either inorganic mercury plus methylmercury or inorganic mercury alone can be released from the sample for determination by atomic absorption. The difference between the two readings gives the amount of methylmercury in the sample. Further, it is possible to release first inorganic mercury and afterwards methylmercury from the same sample. 0 SAC and the author.848 MAGOS : SELECTIVE ATOMIC-ABSORPTION DETERMINATION OF INORGANIC [Analyst, Vol.96 APPARATUS- The mercury vapour concentration meter was manufactured by the Hendrey Relays Division of Columbia Industrial Development Ltd., Slough, Bucks., but as described else- where,5 a water-pump was substituted for the original fan. The inlet of the gas cell was connected by plastics tubing successively with two midget impinges of 30-ml volume and a 200-ml Quickfit test-tube (B34 socket) fitted with a Drechsel bottle head, the inlet of which was converted into a thistle funnel. The first midget impinger was left empty and served as a liquid trap. The second impinger contained 10 ml of distilled water and was immersed in melting ice to act as a water vapour absorber. The Quickfit test-tube was the reaction vessel. When recordings were made the electric output of the mercury vapour concentration meter was connected to a Servogor Potentiometric Recorder (Goerz Electro GmbH, Vienna).METHOD PREPARATION OF S4MPLES- Homogenates were prepared in 1 per cent. saline with an Ultra-Turrax* homogeniser. When the preparation of a homogenate is difficult to achieve, as with fishmeal, small animals or skin and fur, the following preparative techniques were used. For fishmeal, 0.5g was mixed in a test-tube with 1 ml of 1 per cent. cysteine solution, 1 ml of 20 per cent. sodium chloride solution and 1 ml of 45 per cent. sodium hydroxide solution. The contents were heated to boiling-point and washed into the reaction vessel. For the whole rat, the sample was weighed and dropped into boiling40 per cent. sodium hydroxide solution (the volume in millilitres being twice the animal weight in grams).After boiling for 20 minutes the volume was made up with distilled water to give a 20 per cent. w/v solution (based on the weight of the rat). Further dilutions were made as soon as possible to avoid gelatinisation. REAGENTS- All the reagents were of B.D.H. analytical-reagent grade unless otherwise stated. Mercury standard solutions-All the standard solutions contained 0-5 mg ml-l of mercury. To prepare an inorganic mercury standard 0.6767 g of mercury(I1) chloride was dissolved in sufficient 5 per cent. sulphuric acid to give 1000 ml. From this 1 ml was taken and made up to 1000 ml with a solution of 9.0 g of sodium chloride, 0.7545 g of ethylenediaminetetra- acetic acid, disodium salt, and 0.063 g of L-cysteine hydrochloridet in distilled water.If this solution is kept in a refrigerator the mercury concentration remains unchanged for at least 6 months. To obtain a methylmercury standard 60.08 mg of methylmercury chloride$§ were dissolved in 100ml of acetone and from this solution a 1 to 1000 dilution was made with distilled water. Alternatively, 36.96 mg of methylmercury dicyandiamideg were dissolved in 500 ml of distilled water and from this solution a 1 to 100 dilution was made with distilled water. As methylmercury solutions tend to lose mercury owing to either volatilisation (from methylmercury chloride) or precipitation (from methylmercury dicyandiamide) their mercury concentrations were checked frequently by the method of the Dow Chemical Company.8 Cysteine hydrochloride, 1 per cent.wlv solution. Sodium chloride, 1 per cent. wlv solution. Sulphuric acid, 16 N. Tin(II) chloride-100-mg portions were used. Tin(II) chloride - cadmium chloride reagent-Tin(I1) chloride (25 g) and 5 g of cadmium chloride were mixed and heated with distilled water until boiling; the volume was made up to 50 ml with distilled water after cooling. Sodium hydroxide, 45 per cent. wlv solution. Silicone MS antifoam-The use of this material was necessary occasionally. With a glass rod that had been dipped slightly into antifoam a ring of antifoam was drawn on the inner wall of the reaction vessel at about middle height. * Janke and Kunkel K.G., Staufen i. Br. t Hopkin and Williams Ltd. 1 K and K, California, U.S.A.$ AB CASCO, Stockholm, Sweden.December, 19711 MERCURY AND METHYLMERCURY IN UNDIGESTED BIOLOGICAL SAMPLES 849 PROCEDURES- Approximately 30 minutes before the start of a run switch on the apparatus, adjust the air flow to about 2-5 1 min-l and set the filter control to “Sample.” Immediately before beginning the determinations select the sensitivity and adjust the full-scale deflection accord- ing to the operating instructions. Methods 1 and 2-Transfer with a pipette 1 to 20 ml of homogenate or standard into the reaction vessel. Add 1 ml of cysteine solution and make the volume up to 21 to 23 ml with 1 per cent. saline. (If automatic pipettes are used add 20 ml of saline to 1 ml of homogenate and 15 ml of saline to 5 ml of homogenate.) Add 10 ml of 16 N sulphuric acid.In the case of fishmeal, fumes formed by the action of sulphuric acid on the sample must be removed by bubbling air through the sample to avoid falsely high readings. After this procedure, or with other samples immediately after the addition of sulphuric acid, add either 1 ml of tin(I1) chloride - cadmium chloride reagent (Method 1) or 100 mg of tin(I1) chloride (Method 2) to the sample. Connect the Drechsel head with the reaction vessel, thus starting an air flow through the reaction vessel, and add 20 ml of 45 per cent. sodium hydroxide solution through the thistle funnel. Read the peak deflection and calculate the concentration either by use of an internal standard or of the established factor for the type of sample. Method 3-This is a combination of Methods 1 and 2, the procedure starting as in Method 2.Read the peak height and disconnect the air flow between 1 and 3 minutes after the sodium hydroxide addition. Add 10 ml of 16 N sulphuric acid to the reaction mixture followed by 1 ml of tin(I1) chloride - cadmium chloride reagent. Restore the air flow through the reaction vessel and add 20ml of 45 per cent. sodium hydroxide solution through the thistle funnel. Again read the peak height. PRACTICAL CONSIDERATIONS- Linearity between peak deflections on the scale of the instrument and mercury contents of up to 1-0,ug have been well established for the tin(I1) chloride reduction m e t h ~ d . ~ , ~ The area recorded under the curve can also be used for the evaluation of the mercury content. However, as recorders usually register absorption and not extinction, the use of peak area requires not only an additional step, that is, the measurement of the area, but also the preparation of calibration graphs.TABLE I PEAK DEFLECTIONS* CAUSED BY 0.5 pg OF MERCURY IN DIFFERENT BIOLOGICAL MEDIA Peak deflection caused by 0.5 pg Deflection in of mercury as relation to A r \ standard Biological media Mercury( 11) chloride Methylmercury (standard = 100) None . . .. .. .. .. .. 1 ml of 0.5 per cent. blood solution . . 1 ml of 0.5 per cent. kidney homogenate . . 1 ml of 2 per cent. liver homogenate . . 5 ml of 10 per cent. liver homogenate . . 1 ml of 2.5 per cent. brain homogenate . 324 331 312 323 325 320 265 275 199 214 6 ml of 20 per cent. tuna fish homogenate 0.5 g of fishmeal .. .. .. .. 241 236 32 1 324 313 323 310 325 265 265 209 224 141 23 1 100 97 98 82 65 72.5 184 174 160 152 52 175 174 198 20 1 61 * Values are corrected for blank.850 MAGOS SELECTIVE ATOMIC-ABSORPTION DETERMINATION OF INORGANIC [A PZdyst, VOl. 96 The rate at which mercury is released from a biological sample after reduction differs from one medium to another. It was noted that mercury(I1) chloride, when added to urine, was released at the same rate as from saline, but that the rate of release, and consequently the peak height of deflection, was considerably less if mercury was added to 1 ml of bl00d.5 Similarly, it has been found in the present work that the rate of release was hardly changed if 1 ml of 0.5 per cent. blood solution or 1 ml of 0-5 per cent.kidney homogenate was analysed, but that the rate of release was considerably decreased when 5 ml of 10 per cent. liver homo- genate, 1 ml of 2.5 per cent. brain homogenate or 5 ml of 20 per cent. tuna fish homogenate (made from tinned fish) was analysed. Consequently, when a single determination is carried out, internal standards must be analysed parallel to the sample. When a series of determinations have to be carried out on the same type of sample the average deflection given by 0.1 pg of mercury (from a few measurements) can be used for the whole series. Calculation of the factor is as follows: if, for example, the blank gives a reading of 5, the sample gives a reading of 205 and the sample with 0.5 pg of added mercury gives a reading of 455, then the response for 0.1 pg of mercury = 455 - 205 5 = 50 units.Because of its greater concentration stability, inorganic mercury was used to provide factors and inner standards in Methods 1 and 2. Naturally, in Method 3 the use of methyl- mercury as a standard is indispensable in the evaluation of the second peak. Changes in the sensitivity of the potentiometer or the emanation from source samples (e.g., fishmeal) of some material that might condense temporarily on the gas cell may affect the instrument, so it is more useful to express the factor not as an absolute value, but in relation to the deflection caused by the standard without added biological sample. Table I shows how different biological midia influence the peak deflection. - B D 400 - I- 300 t 5 7 0 Fig. 1. Release curves given by 0.5 pg of mercury as methyl- mercury [curves B, C(i) and C(ii)] and inorganic mercury (curve D).All the curves were made with tin(I1) chloride - cadmium chloride reagent (Method 1) except C(i), which was made with 100mg of tin(I1) chloride (Method 2 or the first part of Method 3) followed by the subsequent release of mercury by tin(I1) chloride - cadmium chlo- ride reagent from the same sample [curve C(ii), Method 31. Curve A constitutes a blank 400 ’ 100 50 Fig. 2. Release curves given by methylmercury from 5 ml of 10 per cent. liver homogenate. Curve A constitutes a blank. Curves B to G show the effect of adding 0, 0.1, 0.2, 0.3, 0-4 and 0.5 pg, respectively, of mercury, as methylmercury, to the homo- genate (Method l ) , curves H and J that of adding 0.5 pg of mercury as inorganic mercury with Method 1 and Method 2, and curves K(i) and K(ii), 0.5 pg of mercury as methylmercury with Method 3 RESULTS AND DISCUSSION Method 1 releases all the mercury from the sample irrespective of the presence of a Method 2 releases inorganic mercury at a fast rate and mercury The release of mercury from methylmercury carbon - mercury bond.from the methyl bond at a very slow rate.December, 19711 MERCURY AND METHYLMERCURY I N UNDIGESTED BIOLOGICAL SAMPLES 851 when Method 2 is used produces a peak or an increase in the existing peak that corresponds to less than 5 per cent. of the methylmercury present in the sample. This release is partly caused by the fact that tin(I1) chloride slightly splits the bond between carbon and mercury and partly by the presence of inorganic mercury, even in highly purified methylmercury preparations6 The results in Table I and curves B and D in Fig.1 show that there was no difference in the height of the peak caused by methylmercury or inorganic mercury. Curves H and J of Fig. 2 also show that the rate of release of mercury was the same whether inorganic mercury was released by Method 1 (Fig. 2, curve H) or by Method 2 (Fig. 2, curve J). It can also be seen from Fig. 2 that when mercury was released from 5 ml of 10 per cent. liver homogenate without added mercury (curve B) and with methylmercury added to the homogenate in amounts of 0.1,0-2,0-3, 0.4 and 0.5 pg of mercury (curves C to G) , the increase in concentration showed a linear relationship with the increase in peak height read on the scale.Methods 1 and 2 make possible the determination of total mercury or of inorganic mercury. The difference between values obtained by Method 2 and Method 1 gives the methylmercury content of the sample. The combination of Methods 1 and 2 also makes possible the separate determination of inorganic mercury and organomercury in the same sample (Method 3). In Fig. 1 curves C(i) and C(ii) were made by using Method 3 with a standard solution containing 0.5 pg of mercury as methylmercury. The reaction vessel was disconnected 1 minute after the first addition of sodium hydroxide. In Fig. 2 curves K(i) and K(ii) illustrate a similar deter- mination with 5 ml of 10 per cent. liver homogenate (air flow disconnected after 2 minutes).The height of the second reading depends on the time lapse between the first addition of sodium hydroxide and the disconnection of the air flow. If the air flow is disconnected too soon the second peak might be even larger than a peak given by Method 1 as atomic mercury is left in the sample after the first reduction. If enough time is left for aspirating out all the inorganic mercury, the second peak will be lower than it should be according to the common factor for Methods 1 and 2, as some mercury escapes from methylmercury during the first phase of the reaction. Consequently, if Method 3 is used the time taken must be standardised and a separate factor established. 100 50 Fig. 3. Analysis of 1.0 ml of 0.5 per cent. kidney homogenate [curves A, B(i) and B(ii)] and 1.0 ml of 2.5 per cent.brain homogenate [curves C, D(i) and D(ii)] of a rat treated with 33 consecutive doses of 0.85 mg kg-l of mercury as methylmercury. Curves A and C were obtained with Method 1, curves B(i), B(ii), D(i) and D(ii) with Method 3 It is shown in Fig. 3 that methylmercury added in vitro can be releasedbyuseof Method 1 not only from homogenates but also from organs of rats treated with methylmercury dicyan- diamide. Curves A, B(i) and B(ii) resulted from 1 ml of 0.5 per cent. kidney homogenate,852 MAGOS : SELECTIVE ATOMIC-ABSORPTION DETERMINATION OF INORGANIC [ A ?Zdy.d. VOl. 96 and curves C, D(i) and D(ii) from 2.5 per cent. brain homogenate of a rat treated with 33 daily doses (5 doses per week) of 0.85 mg kg-l of mercury as methylmercury dicyandiamide.Curves A and C were obtained by use of Method 1, curves B and D by use of Method 2. It can be seen from the peak heights that in kidneys approximately 75 per cent. of the mercury was in the organic form while in the brain this rose to more than 95 per cent. The recovery obtained with Method 1 from tuna fish and fishmeal was compared with results obtained by using the method described by the Analytical Methods Committee9 and carried out by the Laboratory of the Government Chemist. Table I1 shows that the differences between the two methods were not significant and that the variations were not one-sided. The recovery from whole rats of 0.85 mg kg-1 of mercury as methylmercury 24 hours after its administration was tested in four animals. Faeces and urine were also analysed and it was shown that less than 2 per cent.of the dose was excreted during this period. Recoveries of the remaining dose were 95, 98, 93 and 95 per cent. Homogenisation of the whole animal probably caused slight splitting of the methylmercury molecule as inorganic mercury usually made up from 5 to 8 per cent. of the total mercury content. TABLE I1 MERCURY CONTAMINATION OF TUNA FISH AND FISHMEAL DETERMINED BY THE METHOD OF THE ANALYTICAL METHODS COMMITTEE AND BY THE DESCRIBED METHOD I Mercury contamination, p.p.m. Sample number 1 2 3 4 5 6 7 8 9 Weight of sample analysed by Method I 1 g of tuna fish ,, , I ,* 0.5 g oifishmeal > I D, I , A.M.C. Method 0.13 0.15 0.22 0.29 0.55 0.01 1.20 1.60 3.64.0 Metho: I 0.15 0-20 0.17 0-26 0.58 0.06 0.84 1.48 3.64 The method is not specific in the sense that phenylmercury behaves like methylmercury and about 30 per cent.of the mercury is released from ethylmercury even when tin(I1) chloride alone is used. However, in food that has not been directly contaminated with phenyl- or ethylmercury the mercury contaminant is always either inorganic mercury or methyl- mercury. Consequently, mercury that is released by 100mg of tin(I1) chloride is probably all inorganic mercury while that which is released by the tin(I1) chloride - cadmium chloride reagent is either inorganic mercury, methylmercury, or a mixture of both. The situation is similar in animal experiments when methylmercury is administered. A great advantage of the method is its simplicity together with its adaptability.In serial investigations, when only those samples having mercury contamination above the safety limit are of interest, a standard sample weight can be selected to give a good deflection at the critical concentration. In this instance, if separate internal standards are not used for each sample the same factor can be used for the whole series of determinations. The error of the method is the same as that previously reported for tin(I1) chloride evolu- tion methods5J0 and depends upon the height of deflection. If the peak height is less than 20 on the scale and this deflection is caused by 0.05 pg of mercury in 5 ml of 20 per cent. tuna fish or 10 per cent. liver homogenate, the error can be 25 per cent. If the peak deflection is more than 40 the error is not more than 10 per cent.The sensitivity of the method can be increased in two ways, firstly by increasing the sample weight and secondly by decreasing the volume of saline added to the sample. Because of the increased temperature of the reaction mixture after it was made alkaline, mercury is removed more rapidly from the sample by the air flow. Although this last effect had been tested and proved, it is thought that the described procedure satisfies the demands of sensitivity for most routine and experimental investigations.December, 19711 MERCURY AND METHYLMERCURY IN UNDIGESTED BIOLOGICAL SAMPLES 853 I acknowledge with thanks The Hendrey Relays Division of Columbia Industrial Development Ltd. for lending the mercury vapour concentration meter, Dr. D. C. Abbott of the Laboratory of the Government Chemist for supplying analysed samples of tuna fish and fishmeal, Mrs. A. Green for excellent technical assistance and Mr. J. A. E. Jarvis for planning and constructing the glassware. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Jensen, S., and Jernelov, A., Nature, Lond., 1969, 223, 5207. Gage, J. C., Analyst, 1961, 86, 457. Westoo, G., Acta Chem. Scand., 1968, 22, 2277. Gage, J. C., and Warren, J. M., Ann. Occup. Hyg., 1970, 13, 115. Magos, L., and Cernik, A. A., BY. J . Ind. Med., 1969, 26, 144. Norseth, T., and Clarkson, T. W., Archs Envir. Hlth, 1970, 21, 717. Clarkson, T. W., and Greenwood, M. R., Analyt. Biochem., 1970, 37, 236. The Dow Chemical Company, Method CAS-AM-70, 1970, p. 13. Analytical Methods Committee, Analyst, 1965, 90, 515. Lindstedt, G., Ibid., 1970, 95, 264. Received June 16th, 1971 Accepted July 12th, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600847

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

854 Analyst, December, 1971, Vol. 96, PP. 854-857 A Critical Study of Brilliant Green as a Spectrophotometric Reagent : The Determination of Perchlorate Particularly in Potassium Chlorate BY A. G. FOGG, C. BURGESS AND D. THORBURN BURNS (Deeartment of Chemistry, Univenity of Technology, Loughborough, Leicestershire) A modified procedure for the extraction of perchlorate with Brilliant green into benzene is described, which gives a reproducible and nearly quanti- tative recovery of perchlorate (apparent B~~~~~ = 94 000) and low blanks ( A = 0.02 to 0.04). Studies on the use of untreated glassware, silanised glassware and polypropylene-ware have indicated that adsorption effects are not very important. Nevertheless, the standard and sample solutions should be treated in the same vessels and the absorbance measurements made in quartz cells.Brilliant green perchlorate can be extracted from aqueous solutions in the pH range 3 to 6.5, but the solutions should be buffered at a constant pH value. The procedure has been applied successfully to the determination of perchlorate in samples of potassium chlorate. IN Part I of a series on the determination of perchlorate in potassium chloratel an infrared method of determining 0.1 to 0.5 per cent. of perchlorate in samples of potassium chlorate was described. In Part 112 the colorimetric determination of down to 0.003 per cent. of perchlorate in samples of potassium chlorate was described, the coefficient of variation being 2 per cent. In this method tetrabutylphosphonium perchlorate was extracted into o-di- chlorobenzene and the perchlorate ions were replaced by iron( 111)-thiocyanate anions.The present paper describes the development of a colorimetric procedure fifteen times more sensitive than this last method and is based on the extraction of Brilliant green perchlorate into benzene. Reusmann3 outlined a procedure for the determination of traces of perchloric acid with Brilliant green. He was unable to obtain reproducible results with the original procedure of Golosnitskaya and Petrashen* because of adsorption of perchlorate on to glass surfaces and also oxidation of the reagent by dissolved molecular oxygen. These difficulties were overcome by using quartz-ware and by adding ascorbic acid to prevent oxidation of the Brilliant green. By measuring the absorbance of the benzene extracts at 638nm, Reusmann obtained a five-fold increase in sensitivity over the procedure of Golosnitskaya and Petraschen.The apparent molar absorptivity of Brilliant green perchlorate in benzene depended to some extent on the amount of Brilliant green added to the aqueous solution. When 4 mg of Brilliant green were added to 25 ml of aqueous solution and the perchlorate was extracted with 10 ml of benzene, the apparent molar absorptivity was 0.945 x lo5, and with 30mg of Brilliant green added, 1-06 x lo5. Improved procedures for the determination of errh hen ate,^ gold5 and antimony6 with Brilliant green have been described recently. The molar absorptivities of Brilliant green perrhenate in benzene, and of Brilliant green tetrachloroaurate( 111) and Brilliant green hexachloroantimonate(V) in toluene, were found to be 1.00 x lo5, 1.01 x lo5 and 1.03 x lo5, respectively.These values agree very closely with Reusmann’s maximum value for Brilliant green perchlorate in benzene. Reusmann’s procedure has the disadvantage that 2-mm absorption cells are used in order to overcome the difficulty of the high blank. In the present work, as in the previous method for rhenium, the amount of Brilliant green added was relatively small (0.5 mg) and 10-mm cells could be used, the blank being of the order of 0-02. The apparent molar absorp- tivity (94 300) obtained under optimum conditions compares well with that obtained by Reusmann, although the accuracy of his values must be affected by the high blank values he obtained.0 SAC and the authors.FOGG, BURGESS AND THORBURN BURNS 855 A further disadvantage of the previous procedure is that it involves the use of expensive quartz-ware. Procedures are given here in which the preliminary chemical operations are carried out either in glass or in polypropylene-ware and the final absorbance measurements are made in quartz cells. EXPERIMENTAL I t was shown previously5 that the Rf form of Brilliant green is most fully formed at pH 6 to 6-5 and that this was the optimum pH for the extraction of perrhenate, although quantitative extraction can be made in the pH range 2 to 7. As the perchlorate ion is also stable over this range of pH it is to be expected that the optimum pH for its extraction with Brilliant green will also be close to 6.5.Initial studies with glass apparatus in a pro- cedure analogous to the recommended procedure for perrhenate gave a rectilinear calibration graph for the determination of perchlorate, with an apparent molar absorptivity of 91 000 and a coefficient of variation of 2.5 per cent. at the 15pg of perchlorate level. The blank determination was 0.040. In subsequent work ascorbic acid was added as an antioxidant as recommended by Reusmann. When a pure sample of Brilliant green was used its addition was not essential; the increase in colour observed by Reusmann was probably caused by oxidation of the leucobase impurity. There was, however, some indication in the present work that the precision was increased slightly when ascorbic acid was included. Quartz absorption cells were used.0.21 ' I I I 2 4 6 8 PH Fig. 1. Effect of pH on the extraction of Brilliant green perchlorate (15 pg of Clod-) : x , in glassware and 0, in polypropylene-ware The effect of pH on the extraction of perchlorate was studied by using glass vessels, silanised glass vessels and polypropylene-ware (Xlon Products Ltd.) . Hopkin and Williams' Repelcote (a 2 per cent. solution of dichlorodimethylsilane in carbon tetrachloride) was used to silanise glassware. The results obtained are shown in Fig. 1; the procedure used was essentially that described later except that Britton - Robinson buffers were used to obtain a range of pH values. The silanisation process did not affect the recovery of perchlorate. The recovery of perchlorate from solutions of pH 2 to 7 (the pH range over which perrhenate can also be quantitatively recovered) was dependent on the pH and on whether glass or polypropylene equipment was used.With glassware two small maxima in the absorbance zletwus pH curve are apparent. These correspond to an apparent molar absorptivity of 94 300 at pH 3.2 and 93 300 at pH 7. A t pH 5 to 6.5 the apparent molar absorptivity is slightly lower, at 91 000. The absorbance veYszts pH curve for polypropylene-ware shows one maximum only at pH 7 with an apparent molar absorptivity of 94000. It is not completely certain whether or not the maximum apparent molar absorptivity attained (94 300) represents quantitative extraction of the perchlorate as this value for856 FOGG et al.: A CRITICAL STUDY OF BRILLIANT GREEN [Analyst, VOl.96 Brilliant green perchlorate is slightly lower than those for the other Brilliant green ion- association complexes ~ t u d i e d . ~ ~ ~ The apparent loss may be caused by a small, persistent adsorption of perchlorate even under the optimum conditions. The extraction of the per- chlorate appears to be complete, as indicated by the colour of the benzene extracts. The results of attempts to attain a higher apparent molar absorptivity by increasing the amount of Brilliant green added were inconclusive, because of the higher blanks obtained. On consideration of Fig. 1 the use of polypropylene-ware appears to have no advantage. A recommended procedure at pH 6.5 with Pyrex or hard-glass (e.g., borosilicate) glassware is given below. This procedure has been applied to the determination of perchlorate in samples of potassium chlorate.The procedure for the destruction of chlorate prior to the determination of perchlorate has already been described.2 The usual precautions should be taken when using benzene. [Unfortunately toluene and xylene cannot be used as they give higher blank absorbance values and lower apparent molar absorptivities (67 900 and 37 700).] REAGENTS- Brilliant green solution, 0.05 per cent. WID in absolute ethanol-Use a pure sample of Brilliant green, e.g., British Pharmacopoeia grade. A method for purifying impure samples was reported previ~usly.~ Bufe'er solution, pH 6.5-Dissolve 6.80 g of potassium dihydrogen orthophosphate and 5 g of ascorbic acid in water, add 13.9 ml of 0.1 M sodium hydroxide solution and dilute the mixture to 1 litre with water in a calibrated flask.Standard potassium perchlorate solution, 5 pg ml-l-Dissolve 0.0500 g of potassium per- chlorate (dried at 105 "C) in water and dilute the resulting solution to 1 litre in a calibrated flask. Dilute 10.00 ml of this solution to 1 litre in a calibrated flask as required. Store the solutions in polythene vessels. TABLE I ANALYSIS OF ANALYTICAL-REAGENT GRADE POTASSIUM CHLORATE SAMPLES AND RESULTS OF STANDARD ADDITION Weight of Potassium Total Potassium perchlorate potassium chlorate perchlorate potassium perchlorate in the potassium chlorate, Sample takenlg added/ pug found/pg per cent. A 1.029 0 32.7 0.0160 33.0 31.5 34.0 6.6 7.0 28.9 294 274 7.9 8.5 B 1.001 0 7.5 0.0035 1.000 21.4 27.4 0.0034 C 1.075 0 9.0 0.0040 PREPARATION OF CALIBRATION GRAPH- Place 0 to 6-ml aliquots of standard potassium perchlorate solution in a series of seven 100-ml glass separating funnels and dilute each solution to 10 ml with buffer solution.Add 1 ml of Brilliant green solution and extract twice with 10-ml volumes of benzene (analytical- reagent grade). Filter the extracts successively through a Whatman No. 31 filter-paper into a 25-ml calibrated flask, then wash the paper with benzene and dilute the solution to 25 ml with benzene. Measure the absorbance of the solution at 640 nm in a 1-cm quartz cell against benzene. Deduct the absorbance of the blank determination. DETERMINATION OF PERCHLORATE IN POTASSIUM CHLORATE- Place up to 2 g of sample in a suitable vessel and add up to 25 ml of concentrated hydro- chloric acid.Evaporate the solution to dryness, using an infrared lamp in the final stages. Evaporate to dryness three further times after successive additions of 5-ml volumes of waterDecember, 19711 AS A SPECTROPHOTOMETRIC REAGENT 857 in order to remove all the acid. Dissolve the residue in buffer solution (pH 6.5) and dilute to 50 ml with buffer solution in a calibrated flask. Take 10-ml aliquots of this solution and extract the perchlorate with Brilliant green and benzene as indicated for the preparation of the calibration graph. Results obtained with this procedure are given in Table I. DISCUSSION When the procedure for the determination of perrhenate5 is used, the apparent molar absorptivity of Brilliant green perrhenate remains constant at 1.00 x lo5 if the extraction is made from solutions of pH 3 to 6.5.Nearly quantitative recoveries of perchlorate are obtained from solutions in the same pH range, but the apparent molar absorptivity varies from 0.91 x lo5 to 0.94 x lo5, depending on the exact pH of the solution and on the material of the vessel from which the extraction is made. Nevertheless, precise determinations of perchlorate can be made provided that standard and sample solutions are treated in the same way and in the same glassware. The choice of buffer system used is important as certain counter anions give high blanks. when present in high concentrations. For general use Britton - Robinson Universal buffer, which gives blanks of about 0.040 in the pH range 3 to 6.5, is satisfactory. REFERENCES 1 . 2 . 3. 4. 5. 6 . Briggs, A. G., Hayes, W. P., Howling, P. A., and Burns, D. T., Mikrochim. Acta, 1970, 888. Fogg, A. G., Burns, D. T., and Yeowart, E. H., Ibid., 1970, 974. Reusmann, G., 2. analyt. Chem., 1967, 226, 346. Golosnitskaya, V. A., and Petrashen, V. I., Zh. Analit. Khim., 1962, 17, 878. Fogg, A. G., Burgess, C., and Burns, D. T., Analyst, 1970, 95, 1012. Fogg, A. G., Jillings, J., Marriott, D. R., and Burns, D. T., Ibid., 1969, 94, 768. Received May 27th, 1971 Accepted July 26th, 1971

ISSN:0003-2654    

DOI:10.1039/AN9719600854

出版商:RSC    

年代:1971

数据来源: RSC

摘要:

858 Analyst, December, 1971, Vol. 96, fip. 858-864 Rapid Polarographic Microdetermination of Dissolved Oxygen in Water with Flavin Enzyme BY JUN OKUDA, TSUNEAKI INOUE AND ICHITOMO MIWA (Faculty of Pharmaceutical Science, Meijo University, Showa-ku, Nagoya, Japan) The rapid polarographic microdetermination of dissolved oxygen in water with flavin enzyme, /3-D-glucose oxidase [EC 1.1.3.4; @-D-glucose : oxygen oxidoreductase] or D-amino-acid oxidase [EC 1.4.3.3 ; D-amino-acid : oxygen oxidoreductase] is described in detail. The method is based on the measurement of the ratio of oxygen consumed by the enzyme in the presence of a limited amount of substrate as an internal standard to the total dissolved oxygen, which can be measured by the addition of an excess amount of substrate or sodium hydrosulphite.It was found that this method requires only 1 ml of test water a t a temperature between 5 and 40 "C, and gives satisfactory results over a wide range of oxygen concentrations up to 100 per cent. saturation, and the result can be obtained in only 5 minutes. The values found are almost identical with those generally accepted. ALTHOUGH many useful methods for the determination of dissolved oxygen in water have been reported,lW6 there are very few methods by which the dissolved oxygen in a small amount (less than 10 ml) of test water can be determined accurately and rapidly. The method involving the use of an oxygen electrode is both accurate and rapid,' but it has been difficult t o determine the absolute amount of dissolved oxygen in the water because no adequate external or internal standard has yet been reported. Recently, Robinson and Coopers des- cribed the determination of the absolute amount of dissolved oxygen in water by using an oxygen electrode with N-methylphenazonium methosulphate, reduced nicotinamide adenine dinucleotide and catalase. However, they do not say whether or not the nicotinamide adenine dinucleotide method is applicable to a sample with a temperature lower than 20 "C or a pH of other than 7.4.Moreover, the method takes about 15 minutes for each sample. We have devised a convenient method with an oxygen electrode and flavin enzyme by which the absolute amount of dissolved oxygen in water can be determined in only 5 minutes. As the flavin enzyme reaction equilibrates on the products side, oxygen corre- sponding to the amount of substrate added to the test water is consumed by the enzyme.Therefore, it is possible to determine the absolute amount of dissolved oxygen in the test water by measuring the ratio of the oxygen consumed by the enzyme, in the presence of a limited amount of substrate as an internal standard, to the total dissolved oxygen, which can be measured by adding an excess amount of substrate or sodium hydrosulphite (Na,S,O,). In this paper, rapid polarographic microdeterminations of dissolved oxygen in water of pH 4.0 to 8.0 with /3-D-glUCOSe oxidase, and in water of pH 6.5 to 9.5 with D-amino-acid oxidase are described in detail. EXPERIMENTAL APPARATUS- For measurements of oxygen consumption, a polarographic oxygen analyser (Model 777, Beckman Instruments Inc., Fullerton, California, USA.) connected to a recorder (EPR-2TC, Toa Electronics Ltd., Tokyo, Japan) was used.Cell I, 1 ml in volume, was made from an acrylic plastics (see Fig. l), and fixed on a magnetic stirrer. Cell 11, 8 ml in volume and made of glass, had two small taps and an electrode (see Fig. a), and was used when hermetic introduction of the oxygen-depleted test water into the cell was required. Both cells were placed in a thermostatically controlled circulating system at times when the test water was to be kept at a particular temperature. Microsyringes, 10 and 25 pl in volume (Jintan Terumo ~CO. Ltd., Tokyo, Japan), were used for the addition of reagents to the reaction mixture. 0 SAC and the authors.OKUDA, INOUE AND MIWA 859 ~ 1 0 .4 m m 4 A Fig. 1. Cell I Oxygen electrode 1 1 : - - - ~ ~ I/ Glass plate Stirring bar B (l-ml acrylic plastics cell) REAGENTS- P-D-Glucose oxidase used in this experiment was purifiedgJ0 from crude powder (kindly supplied by Dr. K. Kusai of Nagase & Co. Ltd., Osaka, Japan) and dissolved in 0-5 M sodium acetate buffer (pH 5.6) to a concentration of 15 mg ml-1. Reagent-grade D-glucose, usually a-D-glucose, was dissolved to a concentration of 10 mg ml-1 in water containing 0.2 per cent. of sodium azide (a strong inhibitor for cata1ase)ll and 0.25 per cent. of sodium benzoate (a preservative), and allowed to stand overnight to reach mutarotational equilibrium prior to use. Holoenzyme of D-amino-acid oxidase was purified from pig kidney by Yagi and Ozawa’s method.12 The purified enzyme was dissolved to a concentration of 10 mg ml-l in 0.1 M sodium pyrophosphate buffer (pH 8.3).Reagent-grade m-alanine was dissolved to a concentration of 8 mg ml-l in 0.1 M sodium pyrophosphate buffer (pH 8.3) containing 0.2 per cent. of sodium azide. The purity of ~ ~ - a l a n i n e was checked by elemental analysis and nuclear magnetic resonance spectroscopy, and the ratio of two stereoisomers was confirmed to be exactly 1.0 from the fact that virtually no optical activity was observed. Mutarotase [EC 5.1.3.3, aldose l-epimerase], which catalyses the interconversion of D-glucose anomers, was prepared from pig kidney cortex according to the purification procedure of Lapedes and Chase.13 Sodium Stirring bar Fig.2. Cell I1 (8-ml glass cell)860 OKUDA, INOUE AND MIWA : RAPID POLAROGRAPHIC MICRODETERMINATION [Analyst, Vol. 96 hydrosulphite was dissolved to a concentration of 10 per cent. in water immediately before use. The sea water used in this study was drawn from the Ise Bay near Shinojima Island. It was filtered through powdered kieselguhr to remove traces of suspended matter, heated to 80 "C for 1 hour as described by Truesdale et aZ.14 and diluted to the required salinity with sterile distilled water. MICRODETERMINATION OF DISSOLVED OXYGEN IN AIR-SATURATED PURE WATER P-D-Glucose oxidase catalyses the following reaction in the absence of catalase- If catalase is present in the test water, one molecule of hydrogen peroxide formed in the reaction is decomposed to half a molecule of oxygen and one molecule of water, and the oxygen thus formed can be used again for the oxidation of another molecule of P-D-glucose by reaction (1).Therefore, the catalase action must be blocked by adding sodium azide as a strong inhibitor to the reaction mixture. Test water (about 1.2 ml) was carefully run into cell I with a wide-mouthed pipette and the oxygen electrode was immersed in the test water and placed securely on the top of the lower wall (see Fig. 1, B). An excess amount (about 0.2 ml) of the test water rose into the narrow space between the electrode and the upper wall. The meter reading of the oxygen analyser was set at about 90 on the recorder chart while stirring continuously with a stainless- steel stirring bar.After 5p1 of 15 mgml-1 P-D-glucose oxidase solution had been added with a microsyringe through the hole, 6 p1 of 10 mg ml-l D-glucose solution were added with a microsyringe. An oxygen consumption (see Fig. 3, A ) exactly equal to the number of moles of P-D-glUCOSe in D-glucose added was recorded. Diffusion of oxygen from the atmos- phere was not observed in the present method when this cell was used. When a large excess of 10 mg ml-l D-glucose solution (15 p1) or 10 per cent. sodium hydrosulphite solution (1Opl) was added with a microsyringe to determine the true reading corresponding to the total dissolved oxygen in the test water, the additional oxygen consumption (see Fig. 3, B-A) was recorded on the chart. /?-D-GLUCOSE OXIDASE METHOD- P-D-Glucose + 0, + H20 -+ D-Gluconic acid + H,O, .. * - (1) The oxygen concentration of the test water was calculated from equation (2). a x 1000 x 32 x B x 0.635 180.16 x b x A Oxygen concentration, p.p.m. = where a mg is the amount of D-glucose added, 32 is the molecular weight of oxygen, 0*635* is the proportion of P-D-glucose in total D-glucose (cc-anomer + p-anomer), 180.16 is the molecular weight of D-glucose, and b ml is the volume of test water. The value (9.2 p.p.m.) for the concentration of dissolved oxygen in air-saturated pure water at 20 "C obtained by the P-D-glucose oxidase method (see Fig. 3) is identical with the value, 9.2 p.p.m. at 20 "C, given by the American Public Health Association.16 D-AMINO-ACID OXIDASE METHOD- D-Amino-acid oxidase catalyses the following reaction in the absence of catalase- Dissolved oxygen in water was determined with D-amino-acid oxidase by the following procedure.Test water (about 1-2 ml) was placed in cell I and the oxygen electrode was placed on the top of the lower wall of the cell. The meter reading of the oxygen analyser was set at about 90 on the chart under continuous stirring conditions as with P-D-glucose oxidase. After 5 pl of D-amino-acid oxidase solution had been added, 5 pl of 8 mg ml-l DL-alanine solution containing 0.2 per cent. of sodium azide were added with a microsyringe. An oxygen consumption (see Fig. 4, C), exactly equal to 0.5 mole of DL-alanine added was recorded. * As fl-D-glucose oxidase oxidises exclusively fl-D-glucose, the amount, a, of D-glucose must be multiplied by the proportion's of fl-D-glucose in total D-glucose, 0.635.D-Amino-acid + 0, + H,O + 2-Keto-acid + H20, + NH, * ' (3)December, 19711 OF DISSOLVED OXYGEN IN WATER WITH FLAVIN ENZYME 861 I I I .- - I r + - I D , 0-D-Glucose oxidase D -G lucose D -G I u cose I -A-I I + Fig. 3. Tracing of the oxygen con- sumption in the determination of dissolved oxygen in air-saturated pure water (20 "C) by the fi-D-glucose oxidase method a t a pressure of 760 mm of mercury (Y is the residual current) When the oxygen consumption reached a maximum, 15 pl of 8 mg ml-l DL-alanine solution or lop1 of 10 per cent. sodium hydrosulphite solution were added to give an additional oxygen consumption (see Fig. 4, D-C). The oxygen concentration of the test water was calculated from expression (4).c x 1000 x 32 x D x 0.5 89.09 x d x C Oxygen concentration, p.p.m. = . . * - (4) . . -- - x 179.59 d x C where c mg is the amount of DL-alanine added, 32 is the molecular weight of oxygen, 0.5 is the proportion of D-alanine in DL-alanine, 89.09 is the molecular weight of D-alanine, and d ml is the volume of test water. The value (9.2 p.p.m.) for the concentration of dissolved oxygen in air-saturated pure water at 20 "C obtained by the D-amino-acid oxidase method (see Fig. 4) was identical with the value obtained by the /?-D-glucose oxidase method and that given by the American Public Health Association.16 SOLUBILITY OF OXYGEN IN AIR-SATURATED PURE WATER AT VARIOUS TEMPERATURES- The temperature of the pure water was raised from 5 to 40 "C by keeping cell I containing the water in the thermostatically controlled circulating apparatus under vigorous stirring conditions for 30 minutes to allow the water to become saturated with air, and the dissolved oxygen at this temperature was measured by the p-D-glucose oxidase method.As shown in Table I, the relationship between the temperature and the oxygen content in pure water was reasonably consistent with that given by the American Public Health Association.16 Almost the same result was obtained by using the D-amino-acid oxidase method. TABLE I SOLUBILITY OF OXYGEN IN AIR-SATURATED PURE WATER AT VARIOUS TEMPERATURES Temperature/'C 5 10 15 20 25 30 40 Present method, * mean values, p.p.m. 12.9 11.3 10.2 9.2 8.4 7.7 6.7 Certificate values for solubility,t p.p.m. 12-8 11.3 10.2 9.2 8.4 7.6 6.6 Number of determinations 7 7 7 7 7 7 7 Standard deviation, p.p.m.0.05 0.02 0.04 0.03 0.03 0.04 0.04 Values in this table were corrected to a total pressure of 760 mmHg and an oxygen partial * p-D-Glucose oxidase method. t Values determined by the American Public Health Association. pressure of 160 mmHg.862 OKUDA, INOUE AND MIWA : RAPID POLAROGRAPHIC MICRODETERMINATION [Analyst, VOl. 96 SOLUBILITY OF OXYGEN IN AIR-SATURATED SEA WATER OF VARIOUS DEGREES OF SALINITY- Sea water of various degrees of salinity (0 to 20000 p.p.m. of chloride) was saturated with air at 20 "C and the dissolved oxygen was determined in cell I. A result very close to that given by the American Public Health Associationle was obtained (Table 11). Almost the same result was obtained by the D-amino-acid oxidase method as the /%D-glUCOSe oxidase method.TABLE I1 SOLUBILITY OF OXYGEN AT 20 "C IN AIR-SATURATED SEA WATER OF VARIOUS DEGREES OF SALINITY Chloride in sea water, p.p.m. 0 5000 10 000 15 000 20 000 Present method, * mean values, p.p.m. 9.2 8-7 8.3 7.9 7.4 Certificate values for solubility, t p.p.m. 9.2 8.8 8.4 7.9 7.5 Standard Number of deviation, determinations p.p.m. 7 0.03 7 0-04 7 0.03 7 0-03 7 0.03 Values in this table were corrected to a total pressure of 760 mmHg and an oxygen partial * B-D-Glucose oxidase method. t Values determined by the American Public Health Association. pressure of 160 mmHg. DETERMINATION OF DISSOLVED OXYGEN IN OXYGEN-DEPLETED WATER PREPARATION OF OXYGEN-DEPLETED WATER- In order to prepare oxygen-depleted water, the authors principally used the /3-D-glucose oxidase method.However, it was supposed that a-D-glucose remaining in the reaction mixture after oxidation of P-D-glucose with /3-D-glucose oxidase slowly mutarotates to /3-D-glucose and is subsequently oxidised if the reaction mixture is allowed to stand for more than 7 minutes. Therefore, mutarotase was added to the reaction mixture to convert the remaining a-D-glucose rapidly into P-D-glucose and then to oxidise the total D-glUCOSe completely. Stable oxygen-depleted water was prepared as follows. A reagent bottle (about 50ml) containing a Teflon-covered stirring bar was charged with 60 p1 of 200 units ml-l mutarotase solution,1° 10 p1 of 2 per cent. sodium azide solution, and an aliquot of 10 mgml-l D-glucose solution (the volume should be calculated to give a water sample of the required oxygen concentration), and the bottle was filled with air- saturated pure water at a definite temperature.Immediately after 25p1 of 90mgml-1 /3-D-glucose oxidase solution had been added to the solution, the bottle was quickly stoppered, care being taken to exclude all air bubbles. The bottle was placed on a magnetic stirrer and the solution was stirred for 30 minutes to complete the oxidation of not only /3-D-glUCOSe but also a-D-glucose. The volume of the bottle fitted with its stopper should be precisely determined in advance. The theoretical oxygen concentration of the oxygen-depleted water is given by the following expression- 1 Oxygen concentration, p.p.m.= e - - x 3 2 x - 180.16 g = e -Lx 0.1776 .. .. . . g where e p.p.m. is the oxygen concentration of air-saturated pure water certificated by the American Public Health Association, f p g is the amount of D-glucose added, 180.16 is the molecular weight of D-glucose, 32 is the molecular weight of oxygen, and g ml is the volume of the bottle with stopper. PROCEDURE- Oxygen-depleted water, which was prepared as described above, was introduced into cell I1 (see Fig. 2) from the lower tap through a polythene tube to fill the cell with the fresh test water until the effluent from the upper tap amounted to about 20ml.December, 19711 OF DISSOLVED OXYGEN IN WATER WITH FLAVIN ENZYME 863 The oxygen electrode should previously be roughly calibrated with air-saturated pure water.After adding lop1 of 200 units ml-l mutarotase solution and 2Op1 of 15mgml-1 /3-D-glucose oxidase solution to the test water, the meter reading of the oxygen analyser was automatically set at the position roughly corresponding to the oxygen concentration of the test water while continuously stirring with the Teflon-covered stirring bar. The dissolved oxygen was determined by adding an aliquot of 10 mg ml-l D-glucose solution (the amount of D-glucose added should correspond to about 70 per cent. of the amount of the dissolved oxygen in the test water) in the same manner as described in the /3-D-glucose method. The oxygen concentration of the test water was calculated from the following expression- a x 1000 x 32 x B 180.16 x b x A Oxygen concentration, p.p.m.= where a, b, A , B, 32 and 180.16 have the same significance as in equation (2). values, as shown in Table 111. its substrate. The values thus obtained for the oxygen concentration were consistent with the theoretical The oxygen-depleted water can also be prepared by using D-amino-acid oxidase and TABLE I11 DETERMINATION OF OXYGEN CONCENTRATION IN OXYGEN-DEPLETED WATER Theoretical values,* p.p.m. 8.5 6-5 4.8 4.1 2.6 1.7 Values found, p.p.m. 8.5 6.6 4.9 4-2 2.7 1.8 Saturation of oxygen, per cent. 100 76 56 48 31 20 Standard Number of deviation, determinations p.p.m. - 2 2 5 0.03 2 2 5 0-04 - - - Values in this table were corrected to a total pressure of 760 mmHg and an oxygen partial * All the test water used in this experiment was prepared from pure water saturated with The theoretical values were calculated from the value 8.5 p.p.m.determined on pressure of 160 mmHg. air a t 24 OC. pure water at 24 "C by the American Public Health Association. DISCUSSION The determination of dissolved oxygen in water with an oxygen electrode has been carried out with pure water saturated with air as an external standard. However, this method is not very accurate for a number of reasons. First, the oxygen electrode is sensitive only to the partial pressure of oxygen in water, and does not directly measure the absolute amount of dissolved oxygen. Secondly, after the calibration of the electrode in the standard pure water, the electrode must be transferred to the test water so as to measure the dissolved oxygen. This movement of the electrode often causes a slight change in the membrane of the electrode, which influences the sensitivity.Thirdly, the electrode has some residual current (r) (see Figs. 3 and a), although this has not been considered in the determination of dissolved oxygen with the oxygen electrode. To eliminate these defects, the authors used D-glucose or DL-alanine as an internal standard for dissolved oxygen in water in the presence of the respective flavin enzyme. The transfer of the electrode is not necessary in the present method. For test water of pH 4-0 to 8.0 the P-D-glucose oxidase method is recommended, while the D-amino-acid oxidase method is recommended for test water of pH 6.5 to 9.5, these being the optimum pH ranges of the enzymes, respectively. Although some conditions that accelerate the interconversion of the anomers of D-glucose are supposed to interfere with the P-D-glUCOSe oxidase method, most of the interferences864 OKUDA, INOUE AND MIWA [Afizalyst, Vol.96 can be ignored because of the rapid oxidation of /3-D-glucose with /3-D-glUCOSe oxidase. As phosphate ions (concentration above 0.05 M) greatly accelerate the interconversion of the anomers of D-glucose, however, it is recommended that in the determination of the test water containing phosphate ions, mutarotase be added to the test water in the ,&D-glucose method as described in the determination of dissolved oxygen in oxygen-depleted water, or the D-amino-acid oxidase method be used. Hg2+, Ag+ and Cu2+ may inhibit the action of both oxidases at a concentration over 10-5 M and prolong the time taken for measurement.With the enzyme reactions used in the present methods the temperature of water samples must be in the range approximately 5 to 40 “C. The methods described can be applied to the determination of dissolved oxygen in sea water and many commonly used buffer systems, because hardly any inhibition of these enzymes was observed at the salt concentrations of such water samples. The authors emphasise that the results obtained for air-saturated pure water by both methods agree closely. I t was found that the results obtained by these methods are almost identical with those generally accepted. It became clear also that the present methods can be applied to the determination of dissolved oxygen in oxygen-depleted water. Other oxidoreductases, e.g., xanthine oxidase and catalase, and other types of oxygen electrode can also be used for the determination of dissolved oxygen in water on the same principle. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. REFERENCES Winkler, L. W., Bey. dt. chem. Ges., 1891, 24, 89. Fox, C. J. J., Trans. Faraday Soc., 1909, 5, 68. Miller, J., J . Soe. Chem. Ind., Lond., 1914, 33, 185. Richter, H. G., and Gillespie, A. S., jun., Analyt. Chem., 1962, 34, 1116. Fishman, M. J., and Robinson, B. P., Ibid., 1969, 41, 337. Tolk, A., Lingerak, W. A., Kout, A., and Borger, D., Analytica Chirn. Acta, 1969, 45, 137. Carrit, D. E., and Kanwisher, J. W., Analyt. Chem., 1959, 31, 5. Robinson, J., and Cooper, J. M., Analyt. Biochem., 1970, 33, 390. Kusai, K., Sekuzu, I., Hagihara, B., Okunuki, K., Yamauchi, S., and Nakai, M., Biochim. Biophys. Miwa, I., Analyt. Biochem., in the press. Okuda, J., and Okuda, G., Clinica Chim. Acta, 1969, 23, 365. Yagi, K., and Ozawa, T., Biochem. Z . , 1963, 338, 330. Lapedes, S. L., and Chase, A. M., Biochem. Biophys. lies. Commun., 1968, 31, 967. Truesdale, G. A., Downing, A. L., and Lowden, G. F., J . Appl. Chem., Lond., 1955, 5, 53. Okuda, J., and Miwa, I., Analyt. Biochem., 1971, 39, 387. “Standard Methods for the Examination of Water and Waste-water,” American Public Health Association, Twelfth Edition, New York, 1965, p. 409. Received March 31st, 1970 Amended June 7th. 197 1 Accepted July 14th, 1971 Acta, 1960, 40, 555.

ISSN:0003-2654    

DOI:10.1039/AN9719600858

出版商:RSC    

年代:1971

数据来源: RSC