ISSN:0003-2654    

DOI:10.1039/AN95782FX005

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BX007

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782FP017

出版商:RSC    

年代:1957

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95782BP023

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

FEBRUARY, 1957 THE ANALYST Vol. 82, No. 971 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY ORDINARY MEETING AN Ordinary Meeting of the Society, organised by the Microchemistry Group, was held at 7.15 p.m. on Friday, January 25th, 1957, in the meeting room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. K. A. Williams, A.Inst.P., M.Inst .Pet ., F.R.I.C. The subject of the meeting was “Micro-volumetric Analysis,” and the following papers were presented and discussed: “Apparatus and Technique,” by D. W. Wilson, M.Sc., F.R.I.C.; “Primary Standards,” by R. Belcher, Ph.D., D.Sc., F.R.I.C. (presented on his behalf by J. H. Thompson, B.Sc., Ph.D., A.R.I.C.) ; “End-point Location,” by E. Bishop, BSc., A.R.T.C., A.R.I.C. NEW MEMBERS ORDINARY MEMBERS Thomas Edmund Alston, A.M.C.S.T.; Henry Reason Ambler, O.B.E., Ph.D.(Lond.), F.R.I.C. ; John Granville Baber, B.Sc. (Vict.), A.R.I.C. ; Desmond Goble Brown, M.Sc. (N.Z.) ; Florence Emma Calladine, B.Sc. (Lond.) ; Alfred Lorraine Cochrane, A.R.I.C. ; William Douglas Duffield; Leslie Ernest Harrison, B.Sc. (Birm.), A.R.I.C. ; Ronald Arthur Neale, B.Sc. (Lond.), A.R.I.C. DEATHS Nelson Trafalgar Foley William Herbert Miles. WE record with regret the deaths of NORTH OF ENGLAND SECTION AN Ordinary Meeting of the Section was held at 2.15 p.m. on Saturday, December 8th, 1956, a t the City Laboratories, Mount Pleasant, Liverpool 3. The Chair was taken by the Chairman of the Section, Mr. J. R. Walmsley, A.M.C.T., F.R.I.C., F.P.S. A lecture on “Some Applications of the Weisz Ring-oven” was given by W.I. Stephen, BSc., Ph.D., A.R.I.C. SCOTTISH SECTION AN Ordinary Meeting of the Section was held at 7.15 p.m. on Monday, December loth, 1956, at the Department of Forensic Medicine, University of Glasgow. The Chair was taken by the Chairman of the Section, Dr. F. J. Elliott, M.Sc., F.R.I.C., F.R.S.E. A lecture on “Problems and Techniques in Forensic Analysis” was given by Edgar Rentoul, M.B., Ch.B., LL.B. WESTERN SECTION THE second Annual General Meeting of the Section was held at 12 noon on Saturday, December 15th, 1956, at the Royal Albert Grill, Newport, Mon. The Chairman of the Section, Mr. P. J. C. Haywood, B.Sc., F.R.I.C., presided. The following appointments 7374 PROCEEDINGS [Vol. 82 were made for the ensuing year :-Chairman-Mr.P. J. C. Haywood. Vice-Chairman- Mr. S. Dixon. Honorary Secretary and Treasurer-Dr. G. V. James, Western Counties Laboratory, 45 Colston Street, Bristol, 1. Members of Committee-Messrs. R. G. H. Boddy, H. J. Evans, R. S. Morris, A. Pickard, G. F. Price and R. Stephens. Mr. R. E. Coulson and Dr. 2. Hybs were appointed as Hon. Auditors. The Annual General Meeting was followed by an Ordinary Meeting of the Section and the following paper was presented and discussed : “The Co-ordination of Analytical Techniques in Industrial Research.” MIDLANDS SECTION AN Ordinary Meeting of the Section was held at 7 p.m. on Tuesday, December l l t h , 1956, in the Gas Showrooms, Nottingham. The Chair was taken by the Chairman of the Section, Mr. J. R.Leech, J.P. The following paper was presented and discussed : “Aspects of the Application of Chromatography to the Quantitative Analysis of Inorganic Substances,” by F. H. Pollard, R.Sc., Ph.d. AN Ordinary Meeting of the Section was held a t 7 p.m. on Thursday, January loth, 1957, in the Mason Theatre, The University, Edmund Street, Birmingham, 3. The Chair was taken by the Vice-chairman of the Section, Dr. R. Belcher, F.R.I.C., F.1nst.F. The following paper was presented and discussed: “The Analytical Chemistry of Some Newer Insecticides and Herbicides,” by K. Gardner, B.Sc., F.R.I.C. MICROCHEMISTRY GROUP THE eighth London Discussion Meeting of the Microchemistry Group was held at 6.30 p.m. on Wednesday, January 9th, 1957, in the restaurant room of “The Feathers,” Tudor Street, London, E.C.4. The Chair was taken by the Chairman of the Group, Dr, G.F. Hodsman, A.1nst.P. This was a review meeting, and an informal discussion took place on each of the subjects covered in previous meetings. PHYSICAL METHODS GROUP THE twelfth Annual General Meeting of the Group was held at 6.30 p.m. on Wednesday, November ZSth, 1956, in the meeting room of the Chemical Society, Burlington House, London, W.1. The Chair was taken by the Chairman of the Group, Dr. J. E. Page, F.R.I.C. The following appointments were made for the ensuing year :-Chairman-Dr. J. E. Page. Vice-Chairman-Mr. R. A. C. Isbell. Hon. Secretary and Tveasurer-Mr. L. Brealey, Boots Pure Drug Co. Ltd., Standards Department, Station Street, Nottingham. Members of Com- mitfee-Dr. Bella B. Bauminger, Messrs. R. A. Chalmers, H. J. Cluley, A. G. Jones, H. Liebmann and G. W. C. Milner. Dr. D. C. Garratt and Mr. C. A. Bassett were re-appointed as Hon. Auditors. The Annual General Meeting was followed at 6.50 pm. by an Ordinary Meeting of the Group. A lecture on “Optical Rotations in the Study of Organic Structures” was given by W. Klyne, M.A., B.Sc., Ph.D. This was followed by a brief description and demonstration of a prototype model of the Bellingham and Stanley Photo-electric Polarimeter.

ISSN:0003-2654    

DOI:10.1039/AN9578200073

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

Feb., 19571 SMALES, MAPPER AND WOOD 75 The Determination, Radioactivation, of Small Quantities of Nickel, Cobalt and Copper in Rocks, Marine Sediments and Meteorites* BY A. A. SMALES, D. MAPPER AND A. J. WOOD Methods are described for the determination of microgram or smaller quantities of nickel, cobalt and copper; these methods involve neutron activation of the samples in the Hanvell Pile, followed by radiochemical separation of the individual elements using carriers and, finally, comparison of the radioactivity due to these elements from the samples with that from known amounts of the elements that were irradiated simultaneously. The accuracy of the method has been checked for each element by analysing standard samples of steels, and the method can, in fact, be very useful for steel analysis when the levels of the three elements are lower than is convenient for existing methods.The results for terrestrial and oceanic rocks, marine sediments and meteorites are briefly examined with a view to the information they can give on the possibility of meteoritic origin of the nickel in marine sediments. Results are also quoted for the standard granite G1 and diabase W1, for some other rocks, commercial pure irons and, finally, for seaweeds. IN a recent communication Smales and Wisemanl discussed the significance of the nickel to cobalt, nickel to copper and copper to cobalt ratios determined in a number of samples of deep-sea sediments, oceanic rocks and meteorites as a means of yielding information on Yettersson and Rotschi’s suggestion2 of the possible meteoritic origin of the nickel found in Pacific deep-sea cores. It was apparent that the method of radioactivation analysis was very suitable for the simultaneous determination of nickel, cobalt and copper, especially as methods for these elements had already been developed in this laboratory. As was stated by Smales,3 “the particular advantage of the radioactivation method in this case was its sensitivity, which made it possible to determine all three elements at the level of a few parts per million and upwards with adequate precision on no more than a few milligrams of sample -the sample being particularly limited in the case of the marine sediments.” In all, 7 3 samples, including 5 igneous rocks and 3 manganese nodules were analysed for nickel, cobalt and copper, and in addition the cobalt contents of seven samples of seaweeds were determined.A number of standard steels was also analysed with a view to giving information on the accuracy of the method. In this paper the analytical methods used, and the detailed results obtained, are described. ,4 number of other rock samples, from the Skaergaard intrusion, East Greenland, was also analysed, and the results, which are of more geological interest in showing the distribution of the three elements in rocks formed at different stages in the crystallisation from the original magma, will be published in a separate paper. The principles, methods and limitations of the radioactivation method have been fully discussed by numerous ~ o r k e r s ~ ~ ~ ~ 6 ~ ~ ~ s and will not be dealt with here.NUCLEAR DATA- The nuclear characteristics of importance in this work for cobalt, nickel and copper are shown in Table I. Because of their short half-lives, ll-minute sOmCo and 5-%minute 66Cu were not used for determinations of cobalt and copper. The fact that the half-lives of 65Ni, G4Cu and 6oCo are 2.56 hours, 12-8 hours and 5.2 years, respectively, made it possible to deal adequately with the successive radiochemical separations in that order. After the separation of nickel and copper from cobalt, the cobalt can be determined at any convenient time. The nickel separation must be done as soon after the irradiation as possible, because of the comparatively short half-life of 65Ni and the already * Presented at the XVth International Congress on Pure and Applied Chemistry (Analytical Chemistry), Lisbon, September 8th t o 16th, 1956.76 SMALES, MAPPER AND WOOD THE DETERMINATION, [Vol.82 rather poor sensitivity, for radioactivation, for that element. Much more latitude in respect of speed is available with copper. TABLE I NUCLEAR DATA FOR COBALT, NICKEL AND COPPER Abundance Isotopic Target in natural activation nuclide element , cross-section, % barns 59C0 14 loo { 20 MNi 1 3.0 63Cu 69 4.3 65Cu 31 2.1 Product on neutron irradiation a c u Radiation MeV and energy, Half-life /3- 1-5 11 minutes /3- 0-32 5.2 years y 1.17, 1.33 p- 2.1 2.56 hours Y p+ 0.66 12.8 hours /3- 0.57 Y p- 2.6 5-2 minutes Y In Table I1 are shown the activities due to each radionuclide after irradiations of 1 pg of each element for 15 hours, 2 days and 5 days, respectively, in a flux of 1012 neutrons per sq.cm per second. TABLE I1 ACTIVITIES FOR 1 pg OF 65Ni, 64Cu AND 60Co AFTER VARIOUS IRRADIATION TIMES Activity of s6Ni, Activity of Wu, Activity of 60C0, Irradiation disintegrations disintegrations disintegrations time per minute per minute per minute 15 hours 2 days 5 days 1.7 x 104 1-7 x 10‘ 1-7 x 104 9.2 x 105 2.7 x 103 1.6 x lo6 8.5 x 103 1.7 x lo6 2.2 x 1w For most of the samples dealt with in this paper, a 15-hour irradiation was adequate, but sometimes, when this gave inadequate sensitivity for cobalt, longer times were necessary. Working sensitivity limits that allow reasonable time for the radiochemical separation of nickel (2 to 3 half-lives) and for sample irradiation for cobalt (up to a week), can be put a t 10-7 g for nickel, EXPERIMENTAL IRRADIATION- The samples were sealed in short lengths of polythene tubing, which were packed, together with the standards, in a 3-inch x l-inch aluminium can and irradiated in the “self-serve” position in the Harwell Pile (except for some of the cobalt determinations in steels, when the thermal column was used, see p.88). STANDARDS- For nickel and copper, 10-mg strips of the “pure” and AnalaR metal foils, respectively, were used as standards, but for cobalt a mild steel, sample “A,” was used containing 0.016 per cent. of cobalt, accurately determined by neutron activation by standardisation against a variety of standard cobalt-containing materials (see Table 111), and also by an absorptio- metric m e t h ~ d .~ It was considered athisable from the aspects of safety, disposal and contamination to avoid unduly large quantities of 5-2-year 6oCo. The nickel and copper were dissolved after irradiation, and suitable aliquots at the 1 to 100-pg level were taken. As 50 mg of the steel sample contained about 10 pg of cobalt, it was unnecessary to take an aliquot. All duplicate standards were irradiated and checked against each other. As relatively very large quantities of the elements to be determined were used as standards, it was essential to avoid any possibility of cross-contamination of the samples, and all the chemical operations were performed with this consideration in mind. Neutron self-shielding was unimportant, i.e., it caused errors of less than 5 per cent., in this work, as would be expected for nickel and copper from a consideration of the total g for cobalt and 10-10 g for copper.Feb., 19571 BY RADIOACTIVATION, OF NICKEL, COBALT .4ND COPPER 77 neutron cross-sections concerned,1° although it was also proved experimentally for the standards used for all three elements, by measuring the specific activities produced by irradiation of‘ 10 to 20-mg quantities of the solid standards simultaneously with aqueous solutions containing only microgram amounts of the same elements.The samples had, as major constituents, elements of low neutron-absorption cross-section, such as calcium, aluminium and iron, and, therefore self-shielding was unimportant for them also. TABLE I11 COBALT CONTENT OF THE MILD STEEL “A” Approximate Amount mass of Cobalt of standard cobalt in the Sample content of Standard used irradiated standard weight, steel “A,” mg p.p.m.Cobalt anthranilate . . ’ .. .. 10 mg 100 163,172 30 151, 167 7 mg 5 mg 1000 156 500 151,156,161 100 mg 25 mg 100 165,175 6 mg 100 152,162 Cobalt sulphate solution . . . . 0.1 ml 30 158,142 5 B.C.S. steel No. 233 (23.7% of Co) B.C.S. steel No. 241 (5.84% of Co) 100 mg 8 { 20mg . . . . Average cobalt content = 159 DISSOLUTION OF SAMPLES- Between 10 and 100mg of the samples were used for the determinations, depending primarily on the quantity of sample available, and after irradiation the samples were dissolved, after addition of carriers, in suitable acid mixtures in glass beakers. Most of the samples, such as the sediments, dissolved readily in hydrochloric acid, but for the meteorites and some of the rocks it was necessary to use a mixture of hydrochloric, nitric and perchloric acids, and a few drops of hydrofluoric acid to obtain complete solution.The excess of hydro- fluoric acid was removed by boiling and the solution was evaporated to fumes of perchloric acid. OUTLINE OF THE RADIOCHEMICAL SEPARATIONS- The nickel was first precipitated from the ammoniacal solution as the dimethylglyoxime complex, the copper and cobalt remaining in solution. The copper was then separated as the copper thionalide, the cobalt remaining in solution for later determination. The nickel and copper compounds were dissolved separately and, after radiochemical purification, the metals were precipitated as the nickel - dimethylglyoxime complex and cuprous thiocyanate, in which forms they were counted.The cobalt was later purified, re-precipitated as potassium cobaltinitrite and counted. The details of the chemistry involved are described later, but it might be mentioned here that, with most of the samples handled, the initial level of radioactivity was sufficient to necessitate shielding precautions being taken by the operators. MEASUREMENT OF RADIOACTIVITY- As carriers, 10 mg of nickel, 20 mg of copper and 10 mg of cobalt in solution form were used, and the final yields, determined on the precipitates prepared for end-window beta counting, were usually about 50 to 90 per cent. The maximum P-energy of 65Ni is 2.1 MeV, so that no correction for self-absorption was necessary.Radiochemical purity was confirmed by plotting decay curves, automatic equipment now available being used. Similarly the copper, as cuprous thiocyanate, was checked for radiochemical purity by plotting a decay curve and by measuring the maximum beta-particle energy. A correction for self-absorption is necessary in this case and values for a correction curve were obtained in the conventional manner by precipitating different amounts of cuprous thiocyanate with a fixed amount of radioactive copper. A further test on the final purified copper precipitate was provided by the gamma-ray spectrometer, 64Cu being identified by the peak at 0.51 MeV due to the annihilation radiation produced by the 0-65-MeV positron (/I+) emission. For cobalt, how- ever, it was generally possible to check the radiochemical purity only by measurement of the maximum /3-energy of 0.32 MeV, although with steel samples the gamma-ray spectrometer78 SMALES, MAPPER AND WOOD THE DETERMINATION, [Vol. 82 was also used for measuring the 1.17 and 1.33-MeV gamma-ray emissions of 6OCo. As with copper, it was necessary to prepare a self-absorption correction curve when counting beta particles from 6oCo.METHOD PROCEDURE- After irradiation, transfer the samples to 150-ml beakers and add 10.0mg of nickel in the form of nickel nitrate solution, 20.0 mg of copper as copper sulphate solution, 10.0 mg of cobalt as cobalt nitrate solution and 5 ml of hydrochloric acid, sp.gr. 1.18. Heat gently on a hot-plate until the solids have dissolved.As mentioned earlier, this treatment is adequate for the dissolution of the marine sediments, but for meteorites and rocks more drastic methods are required. To these samples add 5 m l of nitric acid, sp.gr. 1.42, and 10ml of perchloric acid, spgr. 1-70, and evaporate to fumes of perchloric acid. Allow to cool, and then add a further 5 ml of nitric acid and 10 to 20 drops of 40 per cent. w/w hydrofluoric acid, evaporate to fumes of perchloric acid, cool, and add 5 ml of aqua regia, and again evaporate to fumes of perchloric acid; if necessary, repeat this treatment should any insoluble residue remain at this stage. Allow the solution to cool and then dilute it to 80 ml with water; add 100 mg of ferric iron, as ferric ammonium sulphate solution, 10 ml of a 10 per cent.solution of sodium nitrate, 5 ml of 40 per cent. w/v solution of ammonium citrate and ammonium hydroxide in slight excess. Precipitate the nickel by adding slowly 10ml of a 1 per cent. w/v solution of dimethylglyoxime in methanol, and collect the precipitate on a Whatman No. 541 filter-paper. The nickel precipitate is dealt with as described below. To the filtrate add dilute nitric acid until the solution is just acid, and then heat to 80" C and add 10 ml of a freshly prepared 1 per cent. solution of thionalide in methanol. Place the beaker on a hot-plate, stirring until coagulation of the precipitate is complete, and then collect it on a M7hatman No. 541 filter-paper. The copper precipitate and the filtrate containing cobalt are treated as described below.Procedure ~ O Y nickel-Dissolve the precipitate of the nickel - dimet hylglyoxime complex by the dropwise addition of 5 ml of hydrochloric acid, sp.gr. 1.18, and to the solution add 1 mg of copper, in solution form, and 1 ml of nitric acid, sp.gr. 1.42, heating gently on the hot-plate. After cooling the solution, just neutralise it with ammonium hydroxide solution, re-acidify with hydrochloric acid, add 2 ml of 10 per cent. sodium nitrate solution and precipitate the copper with 2ml of the thionalide solution, as described above, removing the precipitate by means of a 9-cm Whatman No. 541 filter-paper. To the filtrate add 5 mg of bismuth in the form of bismuth nitrate solution and then a slight excess of ammonium hydroxide, and remove the precipitate by means of a Whatman No.541 filter-paper, again retaining the filtrate. Now add 5 ml of 1 per cent. dimethylglyoxime solution in methanol and precipitate the nickel as described previously. Filter off the precipitate, wash it well and dissolve it in hydrochloric acid. To the solution add 1 ml of nitric acid, sp.gr. 1-42, and boil it gently on a hot-plate. Add a further 10mg of copper as nitrate or chloride solution and add ammonium hydroxide until the solution is just acid. Then add 2 ml of a saturated solution of sodium sulphite, followed by 2 ml of a 10 per cent. aqueous potassium thiocyanate solution. Filter off the precipitate of cuprous thiocyanate and discard it, but retain the filtrate. After making the filtrate just ammoniacal, add 5 ml of the 1 per cent.dimethylglyoxime solution and filter off the nickel precipitate. Re-dissolve the precipitate in hydrochloric acid and to the solution add 10 mg of manganese, as manganous chloride solution, solutions of calcium, strontium and barium carriers equivalent to 2 mg of each element, 2 mg of ferric iron, 2 ml of nitric acid and finally 5 ml of saturated bromine water, and boil the solution until the excess of bromine is removed. Add ammonium hydroxide in slight excess, and then 5 ml of a 20 per cent. w/v aqueous solution of ammonium carbonate, boil and, using a 9-cm Whatman No. 40 filter-paper, remove the precipitate and retain the filtrate only. Finally, to the filtrate add 5ml of the dimethylglyoxime solution, spin in a centrifuge, and wash the precipitate thoroughly twice with hot water, discarding the washings.Then add a few drops of ethanol and transfer the precipitate to a weighed aluminium counting tray; dry it under an infra-red lamp and, when it has cooled, determine the chemical yield by weighing the nickel - dimethylglyoxime complex.Feb., 19571 BY RADIOACTIVATION, OF NICKEL, COBALT AND COPPER 79 For the standards dissolve the irradiated nickel foil in nitric acid solution, take a suitable aliquot (containing about 100 pg of nickel), add 10.0 mg of nickel as nickel nitrate solution and proceed as described for the samples. Procedure foy co@er-Wash the precipitate of copper thionalide into a 150-ml beaker, using the minimum amount of water. Add 10 ml of nitric acid, sp.gr. 1.42, and 3 drops of sulphuric acid, spgr.1.84, and boil on a hot-plate until nitrous fumes are no longer evolved. Cool the solution, add 20 ml of perchloric acid, sp.gr. 1.7, and heat until the perchloric acid begins to fume strongly. After cooling, dilute with water to 40ml and add ammonium hydroxide until the solution just turns blue. Add sufficient hydrochloric acid to discharge the blue colour and warm the solution on a hot-plate before adding 2 ml of saturated sodium sulphite solution and 2 ml of 10 per cent. potassium thiocyanate solution. Spin in a centri- fuge, and discard the supernatant liquid. Dissolve the cuprous thiocyanate in 2 to 3 ml of nitric acid, sp.gr. 1.42, and heat until nitrous fumes cease to be evolved. Then transfer the solution to a 50-ml beaker, add 10 mg of ferric iron and 10 mg of manganese, in solution, and 5 ml of saturated bromine water, and then boil until the excess of bromine is removed.Now make the cooled solution slightly ammoniacal, boil and filter through a Whatman No. 541 filter-paper, retaining the blue filtrate. Discharge the blue colour by adding dilute hydrochloric acid and from the slightly acid solution re-precipitate the cuprous thiocyanate as before. Spin in a centrifuge, washing the precipitate twice with water, very thoroughly. Then, by slurrying the precipitate with a little water, transfer it to a weighed aluminium counting tray, dry it under an infra-red lamp and determine the chemical yield by weighing the cuprous thiocyanate. For the standards, dissolve the foil in nitric acid, take a suitable aliquot, add 20.0 mg of copper, in solution, and precipitate the cuprous thiocyanate as for the last stage of the sample treatment.Procedure for cobalt-To the filtrate from the copper thionalide precipitation add 20 ml of nitric acid, sp.gr. 1-42, and evaporate to dryness. Then add 10 ml of sulphuric acid, sp.gr. 1.84, and evaporate to fumes of sulphuric acid. Add further quantities of nitric acid and heat strongly to destroy completely all the organic matter. After cooling the solution, dilute it to 250 ml with water, add a suspension of zinc oxide in water until precipitation occurs and then a slight excess, and set aside for 10 minutes before removing the precipitate by means of a 15-cm Whatman No. 541 filter-paper. Heat the filtrate, and add dropwise 5ml of a 10 per cent.solution of 1-nitroso-2-naphthol in glacial acetic acid, and boil for 2 minutes. Collect the precipitate on an 11-cm Whatman No. 31 filter-paper, and wash it thoroughly with hot water. Ignite the crucible and its contents at 800” C (in a muffle furnace) and dissolve the cooled oxide residue in 5 ml of hydrochloric acid, sp.gr. 1.18, warming if necessary. Make a “scavenging” precipitation by adding a few milligrams of ferric iron, precipitating with ammonium hydroxide and filtering. To the filtrate add 10 ml of a 40 per cent. w/v solution of potassium hydroxide, and carefully boil the solution until all the ammonia is expelled. Spin in a centrifuge and discard the supernatant liquid. Dissolve the precipitate in 5 ml of 3 M hydrochloric acid solution, transfer it to a 150-ml beaker, and add 15 ml of water, 5 ml of glacial acetic acid and 5 ml of 60 per cent.w/v solution of potassium nitrite. Set the solution aside for 5 minutes and then spin it in a centrifuge. Wash the precipitate well, first with water and then with ethanol, and, by slurrying it with a small quantity of ethanol, transfer it to a weighed aluminium counting tray, dry it under an infra-red lamp and determine the chemical yield by weighing the potassium cobaltinitrite. Dissolve the irradiated steel, used as standards, in hydrochloric and nitric acids. To the solution add 10.0 mg of cobalt in solution and treat as for the samples. Count all samples and standards on suitable radiometric equipment, make any corrections necessary for background, self-absorption and chemical yield, and calculate the nickel, copper and cobalt content.Check the radiochemical purity of the samples either by decay, by beta-absorption curves, by gamma-ray spectrometry, or by all the methods or any pair. Transfer the precipitate to a silica crucible, and discard the filtrate. ACCURACY AND PRECISION- The precision of the methods described in this paper has already been illustrated to some extent in Table 111, from which the cobalt content of mild steel “A” was found to80 256, 258, 218, 212, 220, 241, SMALES, MAPPER AND WOOD : THE DETERMINATION, TABLE IV NICKEL RESULTS ON STANDARD STEELS B.C.S. Nickel average B.C.S. range content by nickel for nickel activation B.C.S. No. and steel type content, content, analysis, % Y O O / / O low-alloy steel .. .. . . 0-18 0.165 t o 0-210 0.149, 0.151 low-alloy steel . . .. . . 0.048 0.042 t o 0.055 0.048, 0.047 carbon steel . . .. . . 0.170 0.162 t o 0.190 0.174, 0-182 alloy steel .. .. . . 0.04 0.035 to 0.065 0-031, 0.035 alloy steel .. .. . . 0.15 0.13 t o 0.16 0.134, 0.139 TABLE V COPPER RESULTS ON STANDARD STEELS B.C.S. No. and steel type alloy steel (Cr, V, W, Co, Mo) . . 220, alloy steel (W, 7%; Mo, 4%) 218, carbon steel . . .. 212, alloy steel .. .. 211, alloy steel (Cr, 13%) . . 150, carbon steel . . .. 206, cast iron (high Si and P) 251, low-alloy steel . . .. 252, low-alloy steel . . .. 253, low-alloy steel . . ,. 254, low-alloy steel . . .. 255, low-alloy steel . . .. 256, low-alloy steel , . .. 257, low-alloy steel . . .. 258, low-alloy steel .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. B.C.S. average B.C.S. range copper for copper content, content, % % 0.15 0.13 to 0.165 0.14 0.125 t o 0.145 0.16 0.148 t o 0.163 0.10 0.095 t o 0.101 0.08 0.07 t o 0.084 0.07 0.054 t o 0.079 0.02 0.010 to 0.023 0.090 0.083 t o 0.096 0.110 0.105 t o 0.120 0-495 0.48 t o 0.508 0.1 10 0.100 to 0.112 0.240 0.23 t o 0.244 0.230 0.215 to 0.240 0.305 0.295 to 0.320 0.185 0.165 t o 0.195 TABLE VI COBALT RESULTS ON STANDARD STEELS Copper content by activation analysis, 0.151, 0.160, 0.154 0.135, 0-144, 0.140 0.156, 0.159 0-108, 0.109, 0.109 0.081, 0.081 0.072, 0.066 0.025, 0.023 0.098, 0.099, 0.095, 0.094 0.115, 0.113 0.521, 0.515 0.117, 0.113 0.238, 0-244 0-249, 0.250 0.310, 0.315 0.190, 0.186 % B.C.S. No. and steel type 218, carbon steel .. .. .. 212, alloy steel .. .. .. 251, low-alloy steel (Ni, 5.15%) . . 252, low-alloy steel (Ni, 4.1%) . . 253, low-alloy steel (Ni, 2.9%) . . 254, low-alloy steel (Ni, 2.1%) . . 255, low-alloy steel (Ni, 0.57%) . . 256, low-alloy steel (Ni, 0.18%) . . 257, low-alloy steel (Ni, 044%) . . 258, low-alloy steel (Ni, 0.05%) . . Steel “A” used as standard in radio- activation work . . .. .. B.C.S. average cobalt content formerly, % < 0.01 0-02 0.018 0.015 0.012 0.010 0-006 0.016 0.010 0.012 B.C.S. average cobalt content revised, % - - 0.06 0.02 (5) 0.02(5) 0.02 O.Ol(5) 0-02 ( 5 ) 0.02 0.02 Cobalt content activation by analysis, % 0.014 0.021 0.070 0.043 0-031 0-027 0.019 0.03 1 0-023 0.028 [Vol. 82 Average nickel content by activation analysis, % 0.150 0.048 0-178 0-033 0.137 Average copper content by activation analysis, 0.155 % 0.140 0.157 0.109 0-081 0.069 0.024 0.096 0.114 0-518 0,115 0.242 0.250 0.313 0-188 Cobalt content by tetraphenyl- arsonium chloride absorptiometric method, % - 0,073 0.043 0.030 0.027 0.020 0.031 0.023 0.032 0.016, 0.016, 0.014Feb., 19571 BY RADIOACTIVATION, OF NICKEL, COBALT AND COPPER 81 be 159 p.p.m.with a standard deviation of 2 8 p.p.m. In fact such precision is much better than is required for many of the samples dealt with as some of these were rather heterogeneous, particularly the meteorites. A further idea of the precision for all three elements in finely ground rocks can be obtained from the information in Table XVII (p. 86), in which results on the granite G1 and diabase W1 are given.Even here there is a sampling problem, as the mass of sample used in the determination is only 10 to 100mg. Much more important than precision, in this work on trace elements, is the accuracy of the methods. It is not easy to determine this for meteorites and rocks, because of the lack of reliable analysed samples, and the best we have been able to do is to analyse so called “standard” steels. Even here there is some difficulty and, although satisfactory agreement has been obtained for nickel and copper figures with the published values, as can be seen from Tables IV and V, there was much more difficulty with cobalt. In fact, on a number of steel samples cobalt results by the radioactivation method were at complete variance with the originally published British Chemical Standard values and, although as a result of our work the latter have been re-determined and brought more nearly into line with our own figures, there is still a lack of agreement. Nevertheless, we have satisfied ourselves of the general accuracy of our cobalt figures on these steels by carrying out not only extensive experiments on possible interferences (see p.87), but also absorptiometric determinations by the tetraphenylarsonium chloride method of Pepkowitz and S h i r l e ~ . ~ For B.C.S. pure iron No. 149 one independent absorptiometric result is available,ll of value of 0-010 & 0.001 per cent., and this is in agreement with figures obtained by radioactivation of 0.011 per cent. (see Table XVIII, p. 86). TABLE VII NICKEL, COPPER AND COBALT IN GLOBIGERINA OOZE Samples from a short pilot core, collected by Dr.B. Kullenberg on Prof. Pettersson’s Swedish deep-sea expedition (1947-1948), and separated into slices 4- cm thick Sample No. P2 P8 P28 P30 P37 P38 P39 P44 P45 P46 P49 P50 P54 P57 P5 8 Depth of mid-point, cm 1.09 4-89 15.27 16.43 20.20 20.78 21-37 24.13 24.67 25.21 26-87 27.46 29.65 3 1.30 31.85 A Carbon dioxide content corrected for sea-salt, 38.89 39.29 39.53 39.56 38.01 37.77 37-74 36.23 36.17 35-58 33-27 32-02 3 1.96 31.91 31-74 % Nickel found, p.p.m. 11.0 11.9 9.1 12-7 13.7 14.0 15 15 28 26 24 26 27 28 33 26 23 25 30 31 65 62 55 57 63 61 62 68 66 Copper found, p.p.m. 27 25 31 30 36 36 37 39 49 39 40 47 58 56 79 74 66 70 65 66 Cobalt found, p.p.m. 4.3 4.1 3.5 3-2 4.9 5-0 4-6 4.6 7.6 7.4 6-0 8-2 7.1 6.9 8.0 6.8 5.9 6.0 7.4 7.3 10.0 10.1 14-6 11.7 14.6 15-3 12-5 12-7 12.0 11.882 SMALES, MAPPER AND WOOD: THE DETERMINATION, [Vol.82 The comparison of our values for cobalt with the official B.C.S. results can be seen in Table VI. TABLE VIII NICKEL, COBALT AND COPPER IN GLOBIGERINA OOZE Samples from a long core number 241, station 343, collected by Dr. B. Kullenberg on Prof. Pettersson's Swedish deep-sea expedition (1947-1948). The core, of length 14.402 metres, was collected in the Equatorial Atlantic Ocean on the north-east side of the Mid-atlantic Ridge (1" 10' N, 19" 50' W) at a depth of approximately 4350 metres Sample NO. 13 42 82 107 109 115 142 152 210 269 288 289 Depth, cm 59.7 204.7 404-7 529.1 539.7 569.7 704.7 749.7 1044.7 1334.7 1429.7 1434.7 Carbon dioxide content, uncorrected for sea-salt, 28.57 35-14 26.29 24.3 19-84 22.94 37.16 20.64 37-19 35.18 7.74 35.51 Yo Nickel found, p.p.m.34 34 21 21 45 48 34 37 48 43 55 51 15 11.6 91 90 11.7 11.6 25 24 82 75 12.8 13.9 Copper found, p.p.m. 98 91 53 54 48 61 17 15 22 20 44 53 75 75 Cobalt found, p.p.m. 33 33 5.0 4.9 8.4 8.4 10-8 10.9 10-4 10.1 10.0 11.7 2.6 2-3 20 19 6.5 6.5 13.3 13.3 17.7 18.1 4.0 3.1 TABLE IX NICKEL, COPPER AND COBALT I N MISCELLANEOUS SAMPLES OF GLOBIGERINA OOZE AND I N A SAMPLE OF RADIOLARIAN OOZE Sample No. (British Museum of Natural Depth Location History No.) M367, St. 296 . . 1825 fathoms S.E. Pacific, 38" 6'S, 88" 2' W M355, St. 283 . . 2075 fathoms S. Pacific, 26" 9' S, 145" 17' W M352, St. 280 . .1940 fathoms S. Pacific, 18" 40' S, 149" 52' W M275, St. 216a . . 2000 fathoms N. Pacific, 2" 56' N, 134" 11' E St. 167 . . 4042 metres Indian Ocean, 7" 1.1' 48" t o 7" 11' 24" N, 63" 05' 36" to 62" 59' 30" E N. Atlantic, 51" 55' N, 23" 03' W BM 1953, 164 (6) (rich in cocco- li thop horide) 1710 fathoms S t . 128 .. - Indian Ocean, 5" 31' 42'' to 5" 30' 0'' S Radiolarian ooze- M334, St. 266 . , 2750 fathoms N. Pacific, 11" 7' N, 152" 3' W Nickel found, p.p.m. 286 270 247 255 76 77 112 110 67 68 16 21 44 38 35 1 343 Copper found, p.p.m. 168 178 243 65 117 117 54 53 28 30 59 60 329 319 Cobalt found, p.p.m. 39 41 122 27 26 32 22 9.9 9.4 7.7 13.2 8.8 8-8 226 222Feb., 19571 BY RADIOACTIVATION, OF NICKEL, COBALT AND COPPER RESULTS AND DISCUSSION 83 MARINE SEDIMENTS AND ROCKS- Results obtained for samples of globigerina ooze are shown in Tables VII, VIII and IX; for a single sample of radiolarian ooze also in Table IX; for red clays in Tables X and XI; for manganese nodules in Table XI1 and for some oceanic rocks in Table XIII.TABLE X NICKEL, COPPER AND COBALT IN ATLANTIC RED CLAY Clay from core No. 230, collected by the Swedish deep-sea expedition Sample Nickel No. Depth, found, cm p.p.m. 2 371-5 to 374.5 90 88 3 621-5 to 624.5 73 72 4 896.5 to 899.5 111 116 6 1496.5 to 1499.5 133 131 Copper found, p.p.m. 128 128 143 135 120 105 115 114 Cobalt found, p.p.m. 38 36 28 28 22 21 38 39 TABLE XI NICKEL, COPPER AND COBALT I N MISCELLANEOUS RED CLAYS Sample No. (British Museum of Natural Depth History No.) M343, St. 275 . . 2610 fathoms M309, St. 244 .. 2900 fathoms M309, St. 244 . . 2900 fathoms M315, St. 251 . . 2950 fathoms (sounding tube) (pumice) M344, St. 276 . . 2350 fathoms St. 166 . . 4793 metres 35230, St. 178 . . 1525 fathoms Location S.E. Pacific, 11" 20' S, 150" 30' W N.E. Pacific, 35" 22' N, 169" 53' W N.E. Pacific, 35" 22' N, 169" 53' E N.E. Pacific, 37" 37' N, 163" 26' W S.E. Pacific, 13" 28' S, 149" 30' W Indian Ocean, 6" 55' 17" to 6" 48' 54" N, 67" 11' 18" to 67" 14' 0" E N. Pacific, 16" 47'S, 165" 20' E M273, St. 215 . . 2550 fathoms 3. Pacific, 4" 19' N, 130" 15' E M285, St. 221 . . 2650 fathoms S. Pacific, 0" 40' N, 148" 41' E Nickel found, p.p.m. 458 445 264 268 63 61 77 70 305 293 188 203 59 60 209 199 172 176 TABLE XI1 NICKEL, COPPER AND COBALT IN MANGANESE NODULES Sample No.(British Museum Nickel of Natural Depth Location found, History No.) p.p.m. M313, St. 248 . . 2900 fathoms N. Pacific, 37" 41' N, 177" 4' W 7020 7200 M344, St. 176 . . 2350 fathoms S. Pacific, 13" 28' S, 149" 30' W 4310 4520 St. 166 (41A) 4793 metres Arabian Sea, 6" 55' 18"N. 67' 11' 18"E 6630 6590 Copper found, p.p.m. 409 409 31 1 324 144 138 149 154 361 376 140 149 150 24 1 253 304 301 Copper found, p.p.m. 4190 4200 4580 4330 1924 2392 2635 2800 Cobalt found, p.p.m. 308 304 56 27 51 50 182 188 30 25 43 42 33 34 48 38 Cobalt found, p.p.m. 1125 1050 3780 3660 1140 111084 [Vol. 82 SMALES, MAPPER AND WOOD : THE DETERMINATION, TABLE XI11 NICKEL, COPPER AND COBALT IN OCEANIC ROCKS Nickel Copper Cobalt Sample No. Depth K1, St. 133 (S), basalt with augite 3385 metres R2, St.133 (15), hornblende augite 3385 metres R3, St. 133 (5), variolitic basalt . . 3385 metres and oligoclase dolerite R4, St. 133 (12), variolitic augite 3385 metres basalt Location found, p.p.m. 87 90 74 72 55 1" 25' 54'' S, 66" 34' 12" E 1' 25' 54" S, 66' 34' 12" E 1" 25' 54"S, 65" 34' 12" E 1" 25' 54" S, 65" 34' 12" E 81 86 R5, St. 166 (6), variolitic basalt . . 4793 metres 6" 55' 18" N, 67" 11' 18'' E 53 44 R6, olivine augite basalt . . 2270 metres 8" 32' N, 94" 10' E 138 136 R7, basaltic agglomerate , . 1361 metres Near Providence Reef 174 176 B.M. 1954 (128) basalt . . .. - 11" 25' N, 162" 10' E 266 (Eniwetok Atoll, E. Pacific) 269 B.M. D.T. . . .. . . . . 367 fathoms 42" 05' S, 0" 06' 2'' E 67 (Discovery Table Mount, 70 S. Atlantic) found, p.p.m.60 60 24 102 64 101 100 208 214 57 57 60 55 65 95 101 84 84 TABLE XIV NICKEL, COPPER AND COBALT IN STONE AND STONY-IRON METEORITES Nickel Sample Location found, p.p.m. M1 Khor Temiki. White chondrite , . approximately 16" N, 36" E 68 103 109 109 Fell 8 April, 1932 [1934, 2801 M2 Merua. Grey bronzite chondrite . . M3 Gilgoin. Crystalline bronzite chondrite Fell 30 August, 1920 [1924, 1341 Found 1889 [1927, 12791 M4 Futtehpur. Veined white chondrite . . M5 Ochansk. Brecciated spheroidal bronz- Fell 30 November, 1822 [33757] ite chondrite Fell 30 August, 1887 [63549] M6 Bremervode. Brecciated spherical bronzite chondrite Fell 13 May, 1855 [337392 25" 29' N, 81" 59' E 30" 23' S, 147" 20' E 18,600 18,900 8920 8920 12,100 9800 10,350 12,800 12,400 13,500 25" 57' N, 80" 49' E 57" 47' N, 55" 16' E 53" 4' N, 9" 1' E 8680 7830 M7 Crumlia.Grey chondrite . . , . 54" 37' PJ, 6" 13' W 5420 7420 5720 7660 M8 Estherville. Stony-iron mesosiderite . . 43" 25' N, 94" 50' W 17,000 20,500 M9 Mangwendi. Intermediate hypersthene 17" 39' S , 31" 36' E 9340 chondrite 8450 Fell 13 September, 1902 [86115] Fell 10 May, 1879 [53764] Fell 7 March, 1934 r1934, 8391 METEORITES- Copper found, p.p.m. 8.4 7-9 7-8 7-6 8-3 7.8 109 101 88 86 99 96 111 79 88 99 72 76 225 231 80 82 found, p.p.ni. 47 44 34 33 35 34 45 42 30 31 40 42 50 4s 60 63 48 Cobalt found, p.p.m. 3.3 4-54 4.6 4.8 847 961 730 511 539 527 376 582 576 785 433 318 104 198 203 255 1100 1135 490 465 Although there is a good deal of information about the nickel, copper and cobalt content of iron meteorites, in which the levels are comparatively high, less information is availableFeb., 19571 BY RADIOhCTIVATION, OF NICKEL, COBALT AND COPPER 85 about the stony meteorites, so it was thought worth while to examine a few of these t o supplement the existing data. It should be pointed out that such samples are rather hetero- geneous and, with the small sample weight used in the radioactivation method, agreement between replicates is not expected to be very good.Nevertheless, as the nickel, copper and cobalt were determined on the same replicate sample, values for the ratios of these elements may still be useful. From the results in Tables VII to XIV, the ratios of nickel to cobalt, nickel to copper and copper to cobalt have in each case been calculated and are surnmarised in Table XV, together with average values for igneous rocks and for meteorites taken from Rankama and Sahama.12 As stated by Smales and Wiseman,l “it would appear that there is little significant contribution of meteorite material to the nickel content of deep sea sediments, unless some remarkable differential behaviour of the three elements nickel, copper and cobalt has taken place.” This statement must not, of course, be taken to rule out local deposition of meteoritic matter.Results for nine such samples are given in Table XIV. TABLE XV RATIOS INVOLVING NICKEL, COPPER AND COBALT Ratio of nickel t o cobalt Ratio of nickel t o copper 7 7 -7 Ratio of copper t o cobalt A Number Number Number of Average of Average of Average JIaterial samples ratio Range samples ratio Range samples ratio Range Red clay .. 13 3-4 1.4 t o 7.1 13 0.8 0.4 t o 1.4 13 4.4 1.3 to 7-3 Manganese no- dules ,. 3 4.5 1.2 t o 6-5 3 2.0 1.6 t o 2.4 3 2.3 0.6 to 4.0 Oceanic rocks.. !a 2.4 1-4 to 4.3 9 1.7 0.2 t o 3.1 9 2.2 0.7 to 6.9 Average for ig- Meteorites Globigerina ooze 34 4-0 1.0 t o 7.0 24 0-8 0.4 t o 1.6 24 5.4 2.0 t o 10.9 neous rocksx2 3.5 1.1 3.0 (stones and stony-irons). . 9 21.8 16.8 to 34.5 7 104 82 t o 133” 7 0.22 0.14 to 0.39* meteorites12. . 13.1 92 0.14 Average for * Oniitting Khor Temiki, which gave a nickel to copper ratio of 12.2 and a copper to cobalt ratio of 1.9. The nickel to copper and copper to cobalt ratios for the Khor Temiki meteorite have been omitted from Table XV and are clearly quite unusual.I t has been pointed out by Mr. E. P. HendersonI3 that some stone meteorite specimens may have been treated with copper sulphate solution by geologists. Whether the particular specimen of Khor Temiki analysed has received such treatment is not known, but it is the unexpectedly high copper figure that upsets the ratios mentioned, and it might be interesting to analyse other specimens of this meteorite. TABLE XVI NICKEL, COPPER AND COBALT IN MISCELLANEOUS SPECIMENS Nickel found, Sample Location p.p.m. B.M. 44434, iron from basalt . . . . Ovifak, Greenland 15,850 16,100 16,200 23,000 1750 Paul’s Rocks, Equa- 1790 torial Atlantic) 28, quartz dolerite . . . . . . Knob Head, South Vic- 88 toria Land 88 B.M. 1927, 1248 (3), werlite - dunite . . 1’ 0’ N, 29” 30’ W (St.Copper found, p.p.m. 1074 1095 1141 1181 13 14 206 249 Cobalt found, p.p.m. 4670 4230 4540 4340 118 113 40 4586 SMALES, MAPPER AND WOOD : THE DETERMINATION, [Vol. 82 MISCELLANEOUS GEOLOGICAL MATERIALS- Three specimens of some general interest are reported in Table XVI, while in Table XVII are given the results on a granite G1 and diabase W1, some of which had already previously been reported briefly.14 Included in Table XVIII are the figures obtained by other methods, chemical and spectrographic, summarised by Ahrens,lS and polarographic, used by Smythe and Gatehouse.lG The accumulation of values by a number of different techniques on these two rocks is very desirable. TABLE XVII NICKEL, COPPER AND COBALT IN “STANDARD” GRANITE AND DIABASE Proposed Spectrographic and chemical Polarographic radioactivation method results quoted by Ahrens15 results16 Nickel Copper Cobalt Nickel-t N- Sample found, found, found, Analyst found, found, found, found, found, found, p.p.m.p.p.m. p.p.m. No. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. p.p.m. Wl diabase from Centerville, Virginia G1 granite from Westerly, Rhode Island 76 71 73 78 78 63 1.5 1.3 1.0 1-0 11 1.5 111 111 108 113 119 110 109 8.7 11.1 8.8 9.1 9.5 9-6 54 1 52 2 46 3 45 4 50 5 46 6 7 8 9 47 130 35 64 f 8 100 & 11 23 5 1 80 90 25 - 44 30 50 80 140 140 55 60 - 26 70 100 - 150 80 20 80 130 40 I “Recommended” 90 value 1.8 1 5 2.4 2 n.d. 2.1 3 2.1 4 n.d. 5 < 2.5 6 < 10 7 8 60 9 ._ __ __ 110 < 8 15 5 6 trace 10 20 - - 36 n.d.* 3 n.d. < 2 ( 5 10 3 - - - 2 + 2 1 7 + 1 6 0 “Recommended” (5) 11 (4) value * n.d.= not determined. PURE IRON- - - In of pure method addition to the steel samples mentioned earlier, some commercially available samples iron were examined, the results being given in Table XVIII. The sensitivity of the for cobalt is of value for some of these samples in which the cobalt content is as low as 1 p.p.m. TABLE XVIII NICKEL, COPPER AND COBALT IN “PURE” IRON Nickel Sample found, p.p.m. B.C.S. No. 149, pure iron . . 32, 37, 39, 39 B.C.S. No. 260, pure iron . . 96, 94, 101, 103 Specpure iron, batch 5092 . . 71, 70 Specpure iron, batch 5123 . . 79, 81 Copper Cobalt found, found, p.p.m. p.p.m. 3.3, 5.1, 3.4 107, 113, 113, 115 49, 49 26, 27, 26, 28 2.9 1.1, 1.2 2-7, 2-7 0.8, 0.8 SEAWEEDS- are given in Table XIX.The cobalt contents of several samples of dried seaweed were determined and results Again the sensitivity of the method for cobalt is valuable.Feb., 19571 BY RADIOACTIVATION, OF NICKEL, COBALT AND COPPER TABLE XIX COBALT IN SEAWEED Wet weight Seaweed Dry weight Ascophyllum nodosum 4.78 Fucus serratus . , . . 4-13 Fucus vesiculosus . . 4.48 Laminaria digitata . . 8.64 Laminaria saccarina . . 11.31 Porphyra umbilicalis . . 6-42 Rhodymenia palwtata . . 10.62 Ash as per- centage of dry weight 25.1 19-7 20-3 34.6 36.4 20.8 38.7 Sample weight, fz 0.1043 0.1097 0.1039 0-1006 0~1009 0.1008 0.1023 0.1077 0.1019 0.1098 0.1122 0.1023 0.1022 0*1010 Cobalt found, Pg 0.132 0.122 0.308 0.249 0-185 0.160 0-023 0.019 0.026 0.026 0.059 0-050 0.053 0-048 Dry weed, p.p.m.1.26 1.11 2.97 2.48 1.83 1.59 0.22 0-18 0.26 0.24 0.53 0-49 0.52 0-48 87 Wet weed, p.p.m. 0.26 0.23 0.72 0.60 0.4 1 0.36 0-026 0.021 0-023 0,021 0-082 0.076 0.049 0.045 POSSIBLE INTERFERING ELEMENTS As with other applications of radioactivation, the possibility of conflicting nuclear processes must be borne in mind. Hence, G5Ni can be formed by an n,p reaction on 65Cu or by an n,u reaction on 68Zn, as well as by the n,y reaction on 64Ni used for the determination. Fortunately, for the samples under consideration in this paper, these conflicting reactions are unlikely to be troublesome, as the nickel content was of the same order as, or much greater than, the copper and zinc contents, and the n,p and n,a cross-sections are likely to be much lower than the n,y cross-section.Similar arguments apply to the possible formation of 64Cu from 6*Zn by an n,p reaction. The presence of significant amounts of radionuclides of nickel and copper other than 65Ni and 64Cu, respectively, would be recognised when measuring the decay curve and could readily be allowed for. The picture is not quite so clear cut with cobalt, as a decay curve cannot readily be drawn, owing to the long (5-2-year) half-life of 6OCo. Although “spurious” formation of from an n,p reaction on 60Ni or from an n,u reaction on 63Cu is unlikely to be important in the samples dealt with, particularly as the n,y cross-section for formation of 6OCo from 5 9 C ~ is so high, it is known that 5 8 C ~ can be formed from 58Ni in surprisingly high yield by neutron irradiation in a Pi1e.l’ If counting is done without energy discrimination, e.g., by end-window beta counting, then any 5 8 C ~ will of course be counted as if it were 6OCo and give rise to a spuriously high cobalt figure.From the known cross-section and Pile fluxes the extent of the error can be calculated as follows- Assume beta counting with the same efficiency for both 6oCo and 5 8 C ~ , and irradiation for periods of not more tha.n 70 days, Le., where the rate of growth of the two radio- nuclides is linear with time. Then the ratio of measured beta activity from 5 8 C ~ and 6OCo from equal masses of nickel and cobalt, respectively, irradiated in the Harwell Pile for the same time, will be- where A = measured beta activity, (T = activation cross-section, ffast = fast neutron flux, fthermal = thermal neutron flux, R = beta branching ratio, T = half-life, 0.03 x 2 x loll X 0.15 x 5-2 x 365 - 34 x 1 x 10l2 x 1 x 70 = 0*0007.This means that for equal quantities of nickel and cobalt in a sample, the error involved in the ordinary radioactivation method for determining cobalt, assuming irradiation in the88 SMALES, MAPPER AND WOOD [Vol. 82 Harwell Pile and adequate radiochemical separation of the cobalt, would amount to 0.07 per cent. This is clearly negligible, and even for one meteorite sample in which the ratio of nickel to cobalt is 34.5 (Sample M7, Table XIV), the error is 2.5 per cent. and can be neglected owing to the heterogeneity of the sample. In the case of some materials, however, the error could be serious, as for example, in steels for which the ratios of nickel to cobalt are often as high as 100 or more.Fortunately, any serious error would be apparent from the beta-absorption curve as the maximum beta energy of 5 8 C ~ is 0.47 MeV (positron) and that of 6oCo is 0-32 MeV (p-). Nevertheless, an accurate correction cannot readily be made from a beta-absorption curve and methods of avoiding the error are desirable. Fortunately, there are at least two ways of overcoming the difficulty- (a) instead of using end-window beta counting, by comparing the activities of the cobalt isolated from sample and standard by measuring the areas under the 1-17 and 1-33-MeV peaks of ‘j0Co on a gamma-ray spectrometer; and (b) by irradiating the samples in the thermal column of the Harwell Pile instead of in the usual irradiation position.Some sensitivity is lost by such an irradiation, as the thermal flux is lower possibly by a factor of about ten. (This could be offset by increasing the irradiation time.) The fast neutron flux is very consider- ably lower still, and as can be seen from the above calculation this decreases the relative contribution of 5 8 C ~ . The results for the cobalt content of steels in Table VI were obtained by using both the above methods. Results not quoted here, obtained by the usual procedure of irradiation in the Pile in the normal place, and final activity measurement of the isolated cobalt activity determined by end-window beta counting, confirmed the calculation made above. Summarising the position on possible interfering elements, particular care has to be taken in attempting to use neutron-activation methods to determine small amounts of one element in the presence of major amounts of another element of mass number differing only by one or two.Specifically for elements dealt with in this paper, large amounts of zinc or copper might interfere with the nickel determination, of zinc with the copper determination, and of nickel and copper with the cobalt determination. The techniques described to over- come the interference of nickel in the determination of cobalt may also be applicable to the other metals, but these have not been considered in detail here, as they do not occur in the samples examined. The specimens of marine sediment, rocks, meteorites, etc., mentioned in Tables VII to XIV were kindly provided by Dr. J. D. H. ‘Wiseman of the Department of Mineralogy, British Museum (Natural History), and he also supplied the figures for the carbon dioxide content of the marine sediments described in Tables VII and VIII. We thank Mr. L. Salmon for considerable assistance in the gamma-ray spectrometry work. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. REFERENCES Smales, A. A., and Wiseman, J. D. H., Nature, 1955, 175, 464. Pettersson, H., and Rotschi, H., Geochim. Cosmochim. A d a , 1952, 2, 81. Smales, A. A., “Peaceful Uses of Atomic Energy,” Proceedings of the International Conference in Geneva, August, 1955, United Nations, 1956, Volume XV, p. 73. Boyd, G. E., Ana2. Chew,., 1949, 21, 335. Smales, A. A., Atomics, 1953, 4, 55. Meinke, W. W., Science, 1955, 121, 177. Plumb, R. C., and Lewis, J. E., NucZeonics, 1955, 13, No. 8, 42. Jenkins, E. N., and Smales, A. A., Quart. Rev. Chem. Soc., 1956, 10, 83. Pepkowitz, L. P., and Shirley, J. L., Anal. Chew., 1955, 27, 1330. Hughes, D. J., and Harvey, J. A., “Neutron Cross Sections,” BNL 325, U.S. Government Printing Bagshawe, B., private communication. Rankama, K., and Sahama, Th. G., “Geochemistry,” University of Chicago Press, 1950, p. 39. Henderson, E. P., private communication. Smales, A, A., Geochim. Cosnzochim. Acta, 1955, 8, 300. Ahrens, L. H., “Quantitative Spectrochemical Analysis of Silicates,” Pergamon Press, London, Office, Washington, D.C., July lst, 1955. 1954. 16. 17. Smythe, L. E., and Gatehouse, B. M., Anal. Chemz., 1965, 27, 901. Mellish, C. E., and Payne, J. A., Nature, 1956, 178, 275. ANALYTICAL CHEMISTRY GROUP ATOMIC ENERGY RESEARCH ESTABLISHMENT HARWELL, NR. DIDCOT, BERKS. September 27th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200075

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

Feb., 19571 FOREMAN, RILEY AND SMITH 89 The Absorptiometric Determination of Microgram Quantities of Uranium with the Thoronol Complex of Quadrivalent Uraniumd: BY J. I<. FOREMAN, C. J. RILEY AND T. D. SMITH In dilute acid solution, quadrivalent uranium forms a red complex with thoronol [the red sodium salt of l-(o-arsonophenylazo) -2-naphthol-3 : 6-disul- phonic acid]. The complex is stable in aqueous acetone solution and affords a sensitive and precise method of determining microgram quantities of uranium. Several elements interfere and to overcome this a method of sep- aration has been developed in which uranium is extracted as its sodium diethyldithiocarbamate complex with chloroform. Cupferron and ethylene- diaminetetra-acetic acid are used to hold back impurities. COLOURED complexes formed by the uranyl ion with a number of reagents, for example, S-hydroxyquinoline,l dibenzoylmethane,2 re~acetophenone~ and thiocyanate ion,* have been used for determining small amounts of uranium.However, for the accurate and precise absorptiometric assay of uranium at the microgram level, a method of greater sensitivity than those at present available is required. To this end, complexes of quadrivalent uranium were examined as an alternative to those of sexavalent uranium, which have already received extensive study. The uranyl ion is conveniently and quantitatively reduced to uraniumxV by using a lead red~ctor,~ a procedure having the advantage that further reduction to uraniumU1 does not occur. Thomason, Perry and Byerly6 have described a sensitive method for the determination of thorium by means of the reagent thoronol [the red sodium salt of 1-(o-arsonopheny1azo)- 2-naphthol-3 : 6-disulphonic acid] and noted that uraniumxv reacted with it to yield an unstable deep red complex.The properties of this complex have now been studied in greater detail, and, based on its formation, a simple, precise and sensitive method for determining uranium is described. Details are also given of a separation that is specific for uranium and can be used in conjunction with the proposed method. EXPERIMENTAL Reagents used in the investigation included the following- Uranium solution-Standard uranium solutions were prepared at the mg per ml level by dissolving weighed amounts of pure uranoso-uranic oxide, U30,, to a final acidity of 3 A' hydrochloric acid.Sub-standards in the p g per ml region were prepared from these solutions by appropriate dilution with 3 N hydrochloric acid. Thoronol solution-The commercial product obtained from L. Light & Co. Ltd. was recrystallised from dilute acetic acid, washed and dried at 110" C for 15 minutes. The material was used as a 0.1 per cent. aqueous solution. Lead shot-Lead shot obtained from The British Drug Houses Ltd. was cleaned thoroughly before use by vigorous shaking, first with concentrated hydrochloric acid and then with 3 N hydrochloric acid, until a high metallic lustre was observed. When not in use, the material was stored in 3 N hydrochloric acid. According to Cooke, Hazel and M~Nabb,~ the reduction of uraniumV1 to uraniumIV For studies of the stability of the Fifteen grams of freshly cleaned lead shot were introduced into a 10-ml stoppered flask, and then a suitable aliquot (containing about 20pg of uranium) of uranium solution in 3 N hydrochloric acid was added. These were shaken together for 1 minute and the uranium solution was transferred to a 25-ml calibrated flask and diluted to 5 ml * Presented at the XVth International Congress on Pure and Applied Chemistry (Analytical Chemistry), THE URANIUM~' - THORONOL COMPLEX- proceeds most efficiently in 3 N hydrochloric acid. complex, it was prepared as follows- Lisbon, September 8th to 16th, 1956.90 FOREMAN, RILEY AND SMITH THE ABSORPTIOMETRIC [Vol.82 with distilled-water washings of the lead shot. Then 1 ml of thoronol solution was added, and the solution was diluted to the mark with water.Optical-density readings were taken by means of a Spekker photo-electric absorptiometer, with use of Ilford No. 605 filters, the absorption peak of the complex being at 535mp. Acidity conditions for the formation of the complex are not critical; a final acidity of up to 0.2 N hydrochloric acid may be used for full development of the colour. Although reproducible results could be achieved by making the optical-density measure- ments immediately after development of the colour, the complex is nevertheless unstable and over a period of 15 minutes the optical density decreased by about 5 per cent. It was shown that uraniumIV is unstable under the conditions used by preparing a number of uraniumIV solutions of identical concentration and developing the thoronol complex after various intervals of time. Various reducing agents, for example, stannous chloride, sodium hypophosphite, sodium sulphite and hydrazine hydrochloride, were incorporated in an attempt to stabilise the complex, but these either interfered with the colour development or did not prevent fading.Liquid zinc amalgam and solid cadmium amalgam were investigated, but they offered no advantage over the lead reductor. The development of coloured complexes in non-aqueous or mixed aqueous - non-aqueous media to impart stability and improve sensitivity has frequently been exploited ; in particular acetone has been used by Crouthamel and Johnson7 in the determination of uranium as the thiocyanate complex, by Marriott and Wolf* and Winsorg in the thiocyanate determination of iron, and by Crowther and Largelo in the determination of ammonia by sodium phenoxide and sodium hydroxide.Similarly, by preparing the uraniumIV - thoronol complex in aqueous acetone media a stable and more sensitive coloured complex resulted. The sensitivity varied with the amount of acetone present, 80 per cent. of acetone being the optimum composition. , 450 500 6( Wavelength, rnp Fig. 1. Absorption spectra: curve A, thoronol in aqueous solution; curve B, thoronol in 80 per cent. acetone; curve C, uraniumIV - thoronol complex in aqueous solution ; curve D, uraniumIV - thoronol com- plex in 80 per cent. acetone. Final volume ef solution, 25 ml; cell length, 4 cm Volume of uranium solution ( x ) , mI Fig.2. Continuous-variation plots for the uraniumIv - thoronol complex: curve A, 500 mp; curve B, 530 mp; curve C, 560 mp. Final volume of solution, 10 ml; cell length, 4 cm Fig. 1 shows the spectra of the thoronol reagent and the uraniumIV - thoronol complex in both aqueous and aqueous acetone solution recorded with a Uvispek spectrophotometer. The introduction of acetone does not produce any spectral shifts, but merely decreases the optical density of the reagent while simultaneously increasing that of the complex. Other solvents, notably dioxan, ethyl methyl ketone and ethanol were less effective than acetone with respect to both sensitivity and stability of the complex.Feb., 19571 DETERMINATION OF MICROGRAM QUANTITIES OF URANIUM 91 Typical figures for the improved sensitivity when 80 per cent.acetone was used as solvent are given in Table I ; this represents a calibration in which the quantities of reagents used were reduced in the appropriate proportions to give a final volume of 10 ml. Optical- density measurements were then made against a reagent blank, with use of a Spekker absorptiometer in conjunction with 4-cm Uvispek cells contained in a simple holder designed to fit the cell carriage of the Spekker instrument. TABLE I COMPARISON OF OPTICAL DENSITIES OF CERTAIN URANIUMI~ COMPLEXES Uranium taken, pg 1.25 2-5 5.0 7.5 10.0 12.5 Optical density Optical density of uraniumIV - of uraniumlv - thoronol complex thoronol complex in 80 per cent. acetone in aqueous solution 0.033 0.020 0.072 0.041 0-135 0.079 0.203 0.111 0.276 0.155 0.343 0.188 Optical density of uraniumIv - dibenzoylmethane complex 0-014 0.025 0.047 0.077 0.102 0.130 The sensitivity compares favourably with that of existing absorptiometric methods.Included in Table I are calibration results obtained by the dibenzoylmethane procedure for uranyl ion as given by Yoe, Will and Black,2 which is amongst the most sensitive methods available. Again the final volume was 10 ml and Uvispek spectrophotometer cells were used with the Spekker absorptiometer, the filters being Ilford No. 601. The coefficients of variation at three uranium concentrations, fifteen determinations being carried out at each level, were as follows- Uranium concentration, pg per 25 ml . . 10 20 30 Coefficient of variation, yo .. .. .. 3-3 3.0 2.5 The sensitivity of the thoronol method for thorium may be similarly enhanced by preparing the complex in aqueous acetone; this is shown by the results given in Table 11. TABLE I1 COMPARISON OF OPTICAL DENSITIES OF THE THORIUM - THORONOL COMPLEX I N VARIOUS SOLVENTS Thorium present, Optical density in Optical density in pg per 25 In1 aqueous solution 80 per cent. acetone 0 0.044 0.023 10 0.122 0.154 20 0.197 0.285 30 0.266 0-415 MOLE RATIO OF THE URANIUM’” - THORONOL COMPLEX- Information about the empirical formula of the complex was obtained by applying the method of continuous variations.ll ,12 Equimolar solutions (0.000726 M) of uranium” in 3 N hydrochloric acid and aqueous thoronol were mixed in the proportions x to 1 -x ml, respectively, and made up to 25ml to give solutions containing 80 per cent.of acetone. Their optical densities were measured at several wavelengths by means of the Uvispek spectrophotometer. Typical plots of corrected optical density of the complex against the parameter x are shown in Fig. 2, and it is evident that the complex contains two molecules of thoronol to one of uranium. For thorium Byrd and Banks13 find the complex is pre- dominantly 2 moles of thorium to 3 moles of thoronol. In Fig. 3 is shown the extent of complex formation as the mole ratio of thoronol to uranium is increased for both aqueous and 80 per cent. acetone media. In addition to demonstrating the enhanced sensitivity resulting from the use of acetone, the curves show that the uranium” - thoronol complex is less dissociated in the mixed solvent.By preparing solutions of the complex containing an excess of thoronol, corresponding to the plateau in Fig. 3, the molar optical density of the complex may be measured. Values of this property for the complex and pure reagent at any two wavelengths together with92 FOREMAN, RILEY AND SMITH: THE ABSORPTIOMETRIC v o l . 82 optical-density measurements a t these wavelengths for uraniumm - thoronol mixtures of known composition permit the apparent stability constant, K , of the complex to be calculated, assuming the formation is in accordance with the following equation- UIV + 2 thoronol + UIV(thoronol),, whence [Urv(thoronol) ,] [UIV] [thoronol]?' A number of determinations of this quantity gave a value of K = 9.2 3- 0.5 x lo9.K=- Mole ratio thoronol to uranium Fig. 3. Composition dependence of extent of complex formation : curve A, in Final volume of solution, 80 per cent. acetone; curve B, in aqueous solution. 10 ml; cell length, 4 cm INTERFERENCES- Foreign ions may interfere at the colour-development stage and also at the lead reductor, where anions capable of strongly complexing the uranyl ion will reduce the efficiency of the reduction, as also will the presence of reducible cations that are deposited on the surface of the lead. A number of cations and anions has been examined with respect to over-all interference by applying the method to fixed amounts of uranium in the presence of successively increasing quantities of the ion being studied. The results are depicted in Figs.4, 5 and 6, the Spekker absorptiometer and 4-cm cells being used. It is evident that complexing anions represent the most serious interference, and that iron, chromium, vanadium, zirconium, copper, mercury, manganese, cerium and particularly molybdenum also interfere. Strontium, aluminium and nickel have no effect at the levels studied. PRELIMINARY SEPARATION OF URANIUM- It is evident from the previous paragraph that when uranium is a minor constituent of a sample its separation before being determined by the thoronol method is essential. Uranyl ion forms only a weak complex with ethylenediaminetetrs-acetic acid,14 so permitting it to be separated from a large number of cations by its solvent extraction as a suitable complex. For efficient retention of cations by ethylenediaminetetra-acetic acid an extraction pH above about 4 is desirable.Hence sodium diethyldithiocarbamate was used, since its uranium complex is readily extracted into chloroform in the range pH 5 to 7 ~ 5 . l ~ This combination of reagents has been used successfully by PohP for the separation of copper from iron, cobalt and nickel, and by Cheng, Bray and Melstedl' for the extraction of bismuth from impurities. The only elements that will accompany uranium through the separation are those that form a solvent-extractable sodium diethyldithiocarbamate complex that is more stable than the corresponding ethylenediaminetetra-acetate. IronIU, bismuth and copper are in this category and their presence in the sample aliquot leads to low results for uranium.These elements may be removed by a prior extraction of their cupferrates into chloroform.Feb., 19571 DETERMINATION OF MICROGRAM QUANTITIES OF URANIUM 93 Fig. 4. Effect of cationic impurities on the uraniumw - thoronol method: curve A, 30 pg of uranium + aluminium; curve B, 25 pg of uranium + zirconium ; curve C, 25 pg of uranium + vanadium; curve D, 20 pg of uranium + ironI" ; curve E, 20 pg of uranium + molybdenum; curve F, 15 pg of uranium + chromium". Final volume of solution, 25 ml; cell length, 4 cm 0-3 I------- 0 1200 Amount of foreign metal in sample,pg Fig. 5 . Effect of cationic impurities on the ursniumIV - thoronol method: curve A, 30 pg of uranium + copper; curve B, 30 pg of uranium + manganese; curve C, 30 pg of uranium + mercuryI1; curve D, 25 pg of uranium + strontium; curve E, 15 pg of uranium + ceriumm; curve F, 15 pg of uranium + nickel.Final volume of solution, 25 ml; cell length, 4 cm P 0 lo-' I O - ~ lo-) I O - ~ 10' Concentration of anion, M Fig. 6. Effect of anions on the uraniumIV - thoronol method: curve A, 25 pg of uranium + perchlorate; curve B, 25 pg of uranium + nitrate; curve C, 25 pg of uranium + sulphate; 15 pg of uranium + phosphate. Final volume of solution, 25 ml; cell length, 4 cm The performance of the separation procedure (which is detailed in the next section) may be judged from Tables I11 and IV. Table I11 shows the reproducibility of a uranium91 FOREMAN, RILEY AND SMITH: THE ABSORPTIOMETRIC [Vol. 82 calibration, and Table IV lists the amounts of various ions that were shown not to affect the recovery of uranium.The concentrations quoted refer to the aqueous solution at the cupferron extraction stage. Whilst the figures quoted for the four anions represent safe working limits, this is not so for the cations, for which the concentrations quoted are merely levels at which the separations have worked satisfactorily. The uranium concentration was 4 x 10-6M. TABLE I11 REPRODUCIBILITY OF A URANIUM CALIBRATION TJraniuni Corrected optical density taken, 3 I@ Sample 1 Sample 2 Sample 3 5 0.051 0.050 0.049 10 0.098 0.101 0.097 20 0.197 0.205 0.198 30 0.300 0.302 0.303 TABLE IV LEVEL O F CONCENTRATION O F VARIOUS INTERFERING IONS AT WHICH SEPARATION OF URANIUMI~ CAN BE SATISFACTORILY ACHIEVED Ion Concentration, Ion Concentration, Ion M M so,2- 2-5 A13 + 0.3 Mn2+ c20,2- 1-2 Hg2+ 0.003 Ni2 + F- 0-5 cu2+ 0.05 coz-k ~ 0 ~ 3 - 0.2 Th4+ 0.003 Sr2+ Fe3 i- 0.01 Ce3+ 0.02 La3 + Sn2 f Concentration, AiT 0.01 0.0 1 0.01 0.005 0.003 0.005 METHOD The sample should be in dilute nitric or hydrochloric acid solution.Dilute 1 ml of the sample to 5 ml with 1 1M hydrochloric acid (up to 1 M nitric acid can be tolerated at this stage; the concentration of complexing anions should not exceed the limits quoted in Table IV). Add 2ml of ii per cent. aqueous cupferron solution and extract with 15 ml of chloroform. Add a further 2 ml of cupferron solution and re-extract with chloroform. Separate the aqueous phase and add to it one drop of phenolphthalein solution and then 20 ml of saturated ethylenediaminetetra-acetic acid solution.Make just alkaline with sodium hydroxide and buffer to a.bout pH 6 by adding 2 ml of 20 per cent. sodium acetate solution and 0.5 ml of 20 per cent, acetic acid. Add 5 ml of 5 per cent. aqueous sodium diethyldithiocarbamate solution and then 20 ml of 0.3 M calcium nitrate solution (if appreciable quantities of oxalate are present, magnesium nitrate must be used) to bind the excess of ethylenediaminetetra-acetate. Re-adjust to pH 6 as described above, if necessary. Extract the aqueous phase with three 15-ml portions of chloroform, combine the chloro- form solutions and wash once with water containing Lml of 5 per cent. sodium diethyldithio- carbamate buffered with acetate to about pH 6 as described above and made 0.05 M in, calcium ethylenediaminetetra-acetate.Extract the chloroform phase with two 5-ml portions of 10 per cent. ammonium carbonate solution, combine the extracts and evaporate to dryness in a silica dish. Ignite the residue over a bunslen burner for half a minute, then dissolve it in 2 ml of concentrated hydrochloric acid and evaporate to dryness. Dissolve the residue in 1 ml of 3 N hydrochloric acid and transfer this solution, together with a further 1 ml used for washing the dish, to a 10-ml flask containing 15 g of freshly cleaned lead shot. Shake for 1 minute, transfer the solution, together with about 1 ml of the distilled-water washings, to a 25-ml calibrated flask, add 20 ml of acetone, 1 ml of 0.5 per cent. thoronol solution and make up to volume with water. Measure the optical density of the solution against a blank, prepared by carrying out the method in the absence of uranium, using the Spekker absorptiometer in conjunction with 4-cm cells and Tlford No.605 filters. If it is desired to increase the sensitivity of the method by using 4-cm Uvispek spectro- photometer cells in the Spekker absorptiometer with a final volume of 10m1, the quantityFeb., 19571 DETERMINATION OF MICROGRAM QUANTITIES OF URANIUM 95 of 3 N hydrochloric acid used at the lead-reduction stage must be reduced to give a final acidity not greater than 0*2M. We are indebted to Mr. E. F. Kemp for valuable discussions and t o the Managing Director, United Kingdom Atomic Energy Authority (Industrial Group), for permission to publish this work. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Silverman, C., Moudy, L., and Hawley, D. W., Anal. Clzem., 1953, 25, 1369. Yoe, J. H., Will, F., and Black, R. A., Ibid., 1953, 25, 1200. Urs, M. K., and Neelakantam, K. J., Sci. I n d . Res. B, India, 1952, 11, 79. Currah, J. E., and Beamish, F. E., Anal. Chem., 1947, 19, 609. Cooke, W. D., Hazel, F., and McNabb, W., Ibid., 1950, 22, 654. Thomason, P. F., Perry, M. A., and Byerly, W. M., Ibid., 1949, 21, 1239. Crouthamel, C. E., and Johnson, C, E., Ibid., 1952, 24, 1780. Marriott, W., and Wolf, C., J . Biol. Chem., 1905, 1, 451. Winsor, H. W., Ind. Eng. Chem., Anal. Ed., 1937, 9, 453. Crowther, A. B., and Large, R. S., Aflalyst, 1956, 81, 64. Job, P., Ann. Chim., 1928, 9, 113. Vosburgh, W. C., and Cooper, G. R., J . Amer. Clzem. SOC., 1941, 63, 437. Byrd, C. H., and Banks, C. V., U.S. Atomic Energy Commission, Document ISC 456. Cabell, M. J., Analyst, 1952, 77, 859. Bode, H., 2. anal. Chem., 1955, 144, 165. Pohl, H., Anal. Chirn. Acta, 1955, 12, 54. Cheng, K., Bray, R., and Melsted, S., Anal. Chew., 1955, 27, 24. U.K. ATOMIC ENERGY AUTHORITY WINDSCALE WORKS SELLAFIELD, CALDERBRIDGE CUMBERLAND August 24th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200089

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

Feb., 19571 DETERMINATION OF MICROGRAM QUANTITIES OF URANIUM 95 Chromatographic Separations in Phenol - Methanol - Hydrochloric Acid Solvents, with Special Reference to the Alkali Metals BY ROBERT J. MAGEE" AND JAMES B. HEADRIDGE A phenol - methanol - concentrated hydrochloric acid mixture (57.5: 22-5: 20 per cent. w/v/v) has been used to separate the chlorides of lithium and sodium, and of potassium, rubidium, caesium and ammonium. Zinc uranyl acetate has been employed for the detection and determination of 0.25 to 10 pM (micromoles) of lithium and 0.1 to 10 pM of sodium, and sodium lead cobaltous hexanitrite for 0.1 to 10 pM of potassium, rubidium and caesium and 0.25 to 10 pM of ammonium. It was found that 10-pM amounts of any of the group IA metals, singly or combined, did not interfere with the detection and determination of the minimum amounts of any particular metal.A phenol - methanol - concentrated hydrochloric acid mixture (50 : 20 : 30 per cent. w/v/v) has been used to separate small amounts of aluminium, gallium, indium, thallium and zinc, and titanium, zirconium m d iron. SEPARATION OF THE ALKALI METALS THE separation of lithium, sodium and potassium is not difficult and references in the literature to this separation are numerous ; usually the chlorides and nitrates have been used and these are generally separated with solvents consisting of one or more alcohols. For example, these three cations and the ammonium ion may be separated with a metha- nol - n-butanol mixture (70 : 30 per cent. v/v) .l The separation of potassium, rubidium and caesium is not easy, and only two solvent mixtures have been found to accomplish this ~eparation.~,~ Miller and Magee2 were able to separate 5 pg of any of these ions from a total weight of 1000 pg, but could not use the solvent mixture to separate lithium, sodium and * Present address : Chemistry Department, The Queen's University of Belfast.96 MAGEE AND HEADRIDGE : CHROMATOGRAPHIC SEPARATIONS [Vol.82 ammonium from them. Steel3 claimed a separation of 50-pg amounts of ammonium, potassium, rubidium and caesium ions with the phenol-rich layer from liquid phenol - dilute hydrochloric acid solution (1 + 2, by volume), but did not mention the effect of this solvent mixture on lithium and sodium chlorides. A good separation of the alkali metals on. one chromatographic strip was required, as it was later intended to determine the amount of each of the separated salts by micro- analytical methods.This separation was not found to be possible, but a solvent mixture giving a reliable separation of lithium and sodium salts, and of potassium, rubidium, caesium and ammonium salts is described. EXPERIMENTAL Most previous investigators have used the salts of inorganic acids, although the acetates4 and citrates5 of lithium, sodium and potassium have been separated. To investigate the influence of an organic anion on the solubility and distribution in organic solvents of the salts of group IA metals, in particular potassium, rubidium and caesium, n-butyrates were prepared. An ion-exchange resin, Amberlite IRA-400 (50 to 100 mesh), was used in their preparation.A column of the resin (10 cm long and diameter 1 cm) was regenerated with 2 N sodium hydroxide, washed free from alkali, converted to the butyrate form with 2 N butyric acid and finally washed with water to remove excess of acid. The alkali-metal chlorides (20 mg of cation) were separately changed to butyrates on this column. Of the solvents tried, the alkali-metal butyrates were found to be readily soluble only in water, methanol and anhydrous butyric acid. Lithium, sodium and potassium butyrates were separated by using certain mixtures of simple alcohols, e.g., methanol - a-butanol mixture (80 to 20 per cent. v/v), with which 200-pg amounts- of the cations gave the following R, values: lithium, 0.91 & 0-06, i.e., “0.85 to 0.97”; sodium, 0.52 & 0-06; and potassium, 0-23 0.07.With these and other solvent mixtures containing butyric acid, no separation of potassium, rubidium and caesium was achieved. From the chromatograms produced it was apparent that butyrates were reacting in a similar manner to chlorides and other inorganic salts. Further investigations were made with chlorides. Miller and Magee2 used concentrated hydrochloric acid - methanol - n-butanol- isobutyl methyl ketone mixture (55 : 35 : 5 : 5 per cent., by volume) to separate potassium, rubidium and caesium; 100-pug amounts of these cations and of lithium, sodium and ammonium were chromatographed separately, with use of this solvent mixture. The R, values for the mid- points of each spot are given in the first line of Table I.Sodium and potassium, ammonium and caesium were not separated. Chromatograms were prepared by using solvent mixtures similar to that described above, but these contained the organic base symm.-collidine. The R, values found are shown in Table I. TABLE I X, VALUES FOR THE ALKALI METALS WITH VARIOUS SOLVENTS Composition of solvent mixture, per cent. by volume Concentrated isoButy1 hydrochloric symm.- methyl. acid Methanol n-Butanol Collidine ketone A r 7 5 55 35 5 - 10 - 35 55 40 55 - 5 - 5 70 25 - - 5 95 85 - 10 5 - 65 20 10 5 - - - - - Na+ I<+ 0.45 0.47 0.38 0.28 0.36 0.25 0.58 0.58 0.75 0.79 0.65 0.70 0.52 0.51 Rp values for Rb+ Csf NH,+ Li+ 0.51 0.60 0.63 0.68 0.28 0.24 0.51 0-68 0-60 0.69 0.69 0.76 0.80 0.83 0-83 0.80 0.72 0.78 0.77 0-76 0.55 0.63 0.66 0.73 A > 0.25 0.33 - 0.75 It can be seen from Table I that no separation of all six cations is achieved.In fact, no solvent mixture gives a separation of potassium, rubidium, caesium and ammonium on one strip and so permit a mixture of these cations to be resolved and identified. Steel3 gives the order of the separated salts from the starting line as ammonium, potas- sium, rubidium and caesium. Although several chromatograms were made with chloridesFeb., 19571 I N PHENOL - METHANOL - HYDROCHLORIC ACID SOLVENTS 97 under similar conditions, these results were not reproducible. Potassium, rubidium and caesium were separated, but the ammonium and rubidium ions were always in the same position. However, it appears that the flow of the ammonium ion is at times erratic and is dependent on the anion present.6 The phenol-rich layers obtained when other proportions of dilute hydrochloric acid and phenol were used gave almost identical chromatograms to those obtained with Steel's solvent mixture.Lithium coincided with and sodium was above the potassium position. In attempts to get lithium below potassium, rubidium and caesium, methanol was introduced into the mixtures of phenol and concentrated hydrochloric acid. The relative positions of the cations with a number of these solvent mixtures and with others containing derivatives of phenol are given in Table 11, Whatman No. 1 filter-paper being used. The R, values refer to the mid-points of each band. TABLE I1 SEPARATION OF l-pM AMOUNTS OF ALKALI-METAL AND AMMONIUM IONS Solvent RE values for mixture Solvent mixture 7 r A A I , Li+ Na+ K+ Rb+ Cs+ NH,+ Phenol saturated with 20 per cent.hydro- Of composition- chloric acid . . . . .. . . 0.11 0.07 0-12 0.18 0.43 0.20 Concentrated hydrochloric Phenol, Methanol, acid, g ml ml 33 42 20 50 25 40 50 40 60 33 16 30 30 50 50 40 40 20 33 42 50 20 25 10 10 20 20 0.55 0.24 0.21 0.28 0.42 0.42 0.42 0.27 0.31 0.43 0.55 0.31 0.70 0.44 0.43 0.50 0.64 0.64 0.37 0.14 0.10 0.16 0.30 0.30 0.57 0.19 0.11 0.14 0.23 - 0.51 0.11 0.05 0.08 0.15 - 0.37 0.06 0.04 * 0.06 0.13 - 0.45 0.12 0.08 0.12 0.21 - 0.27 0.11 0.12 0.19 0.38 0.24 Cresol-rich layer from m-cresol- concen- trated hydrochloric acid (90: 10 per Catechol- methanol - concentrated hydro- chloric acid (45 : 35 : 20 per cent.v/v) cent. v/v)* . . .. .. . . 0.09 0.01 0.04 0.09 0.31 - Unsatisfactory chromatograms owing t o the dark backgrounds produced with the spraying reagents Resorcinol - methanol - concentrated hydrochloric acid (50 : 30 : 20 per cent. v/v) .. .. .. .. . . As with solvent mixture No. 4. * Whatman No. 4 filter-paper was used with this solvent mixture, With a methanol content of 30 to 35 per cent. in solvent mixtures containing phenol and concentrated hydrochloric acid (mixtures No. 2, (c) and ( d ) ) , potassium, rubidium, caesium and lithium were separated in that order, but the sodium could not be removed from the vicinity of the potassium. The separation of potassium and rubidium could not be effected when the methanol content exceeded 35 per cent. From these investigations it was evident that all the alkali metals and the ammonium ion could only be separated by using two-dimensional chromatography, with, for example, a methanol-n-butanol mixture (70 to 30 per cent.v/v) for the separation of potassium, sodium, ammonium and lithium, and a solvent mixture consisting of phenol, methanol and concentrated hydrochloric aFid for potassium, rubidium and caesium. However, for the purpose of detection and rough estimation, a solvent mixture able to separate the groups potassium, rubidium, caesium and ammonium, and lithium and sodium would be sufficient, since selective spraying reagents could be used for the detection of the cations in each group. The best spacing of potassium, rubidium, ammonium and caesium is achieved with a phenol - methanol - concentrated hydrochloric acid mixture (57.5 : 22.5 : 20 per cent.w/v/v), with which 1.7-pM amounts of the alkali-metal and ammonium ions (10 pM total) in a run of 50 cm on Whatman No. 1 filter-paper gave the following RF values: potassium, 0.08 to 0.14;98 MAGEE AND HEADRIDGE : CHRClMATOGRAPHIC SEPARATIONS [Vol. 82 sodium, 0-09 to 0.13; rubidium, 0.15 to 0.20; ammonium, 0.23 to 0-27; lithium, 0.27 to 0.32; and caesium, 0.33 to 0.40. The separations are better on the fast Whatman No. 4 filter-paper. This solvent mixture gives bands superior to those obtained with the solvent mixtures used by Miller and Magee, and by Steel. METHOD APPARATUS AND REAGENTS- Descending chromatography was employed. Chromatograms were produced in a Shandon 9-inch Universal-strip glass " Chromatank," having an over-all height of 22 inches and a solvent trough 7$ inches long.This apparatus accommodates sheets of chromato- graphic paper 16 cm wide and permits the solvent-front to travel a distance of 50 cm. The temperature of the room was maintained at 18' to 20" C. Preliminary investigations with the alkali metals were carried out with use of VVhatman No. 1 and No. 4 filter-papers, but in later work Whatman No. 41 filter-paper was used. For all experiments with other metals Whatman No. 1 filter-paper was used. In each test a 0.01-ml portion of the solution was applied as a spot to the paper from capillary tubes calibrated by means of an Agla micrometer syringe. Four such spots could be accommodated by placing them 4 cm apart, across the paper.After their removal from the tank, the strips were dried by being heated in an oven at 110" C for 10 minutes. The strips were held taut and sprayed by means of a simple compressed-air atomiser. Molar stock solutions of the group IA cations were prepared as follows-, Lithium-Prepared from anhydrous lithium chloride (obtained from The British Drug Sodium and potassium-Prepared from AnalaR sodium and potassium chlorides. Rubidium and caesium-Prepared from ordinary grade rubidium and caesium carbonates (obtained from Johnson, Matthey & Co. Ltd.), dissolved in dilute hydrochloric acid. While preparing standards, it was noticed that the rubidium solution contained some caesium (about 0.25 p M per 10 pM), and for further work a molar solution prepared from Specpure rubidium chloride was used.The following reagents were used in the detection of the cations on the paper strip by spraying- Lithium and sodium-Zinc uranyl acetate reagent prepared as described by Barber and Kolthoff .7 Potassium, rubidium, caesium and ammonium-Sodium lead cobaltous hexanitrite reagent prepared as follows- Dissolve 12.5 g of cobaltous acetate, Co(CH3C00),.4H,0, and 19.0 g of lead acetate, Pb(CH3C00),.3H,0, in about 100ml of water. To this solution add a solution of 20.7 g of sodium nitrite in about 100 ml of water. Mix thoroughly and dilute to 250 ml. Set aside for 1 hour and then filter. The reagent corresponds to the formda N a ,P b Co II (N 0 2) 6. This reagent is similar to that used by Sergienkos for the quantitative determination of potassium, but cobalt and lead acetates are substituted for cobalt and lead nitrates. Houses Ltd.).PROCEDURE- Place the solvent mixture in the solvent trough 1 hour or more before the experiment is started. On the appropriate Whatman filter-paper strip place 0.01 ml of the test solution and allow the strip to dry before inserting it in the chromatographic vessel. Allow the solvent to run down the paper for 16 to 18 hours (with Whatman No. 41 filter-paper the solvent-front reaches the foot of the paper within this time and is allowed to drip off evenly by notching the bottom edge before inserting the paper). Remove the paper from the apparatus and dry it. DETECTION AND DETERMINATION OF LITHIUM AN:D SODIUM- Spray the paper with the zinc uranyl acetate reagent and hang it in the air for 1 hour to ensure complete drying.Lithium and sodium fluoresce brilliantly (turquoise-green) . For determination purposes compare the bands produced with a set of standards, prepared from 01-1, 0.25, 0.5, 1, 2-45, 5 and 10-pM amounts of the cations on Whatman No. 41 filter-paper. Examine it under ultra-violet light.Feb., 19571 I N PHENOL - METHANOL - HYDROCHLORIC ACID SOLVENTS 99 DETECTION AND DETERMINATION OF POTASSIUM, RUBIDIUM, CAESIUM AND AMMONIUM- Spray the paper with ethanol and then immediately after with sodium lead cobaltous hexanitrite reagent. Leave the paper in the air for 10 minutes to ensure the complete development of the ammonium precipitate and then wash off the excess of reagent, which itself forms a pale yellow background.Allow the paper to dry and compare it with a set of standards prepared as before. Potassium and ammonium appear as grey bands, rubidium as a brown band and caesium as a yellow-brown band. The best results are obtained when a freshly prepared reagent is used. RE s ULT s By using the above procedure, 0.1 ph!I of sodium, potassium, rubidium and caesium can be readily detected. The minimum amounts of lithium and ammonium detectable are 0.25 pM. The minimum amount of any of the metals can be detected in the presence of 1 0 p M of any other or total of the others. In an overnight experiment lasting 16 hours a mixture of 1 ,uM of each of the six cations travelled the following distances: sodium, 5.4 to 8-4 cm; potassium, 6.4 to 9.1 cm; rubidium, 10-2 to 13-3 cm; ammonium, 15.2 to 17.7 cm; lithium, 18.1 to 20.6 cm; and caesium, 21.3 to 25-3 cm.The first and second figures give the distances to the back and front of the spots, respectively. As a test of the validity of the proposed scheme, 10 mixtures of unknown composition were submitted to one of us (J.B.H.) for analysis. The results are shown in Table 111. Mixture KO. 1 2 3 4 5 6 7 8 0 10 Lithium found, 0 0 2.5(3) 0.25 0 0.25 0.25 0.6 PM 9(10) 4(3) TABLE I11 APTALYSIS OF UNKNOWN MIXTURES Sodium found, P M 0.1(0-20) 0.25(0.1) 0.26(0.1) 0*25(0) 0.1 0 9 0.1 0 0.1 (0) Potassium found, 9 0 0 PM 0.5 (0.1) 0.25(0* 1) 0.5 (0- 1) O%( 0.2) 0-25( 0.2) 5 0 Rubidium found, 0.25 (0.2) 0.25 (0- 1) 0 0.1 0 0*4(0-1) 0-3 (0-2) PM 5(10) 0.25 (0.1) 0*25(0.2) Caesium found, PM 0.5 0.25 (0.2) 0.1 2(7) 0.1 (0.2) 8(;0) 0 0*8( 0.3) 10 Ammonium found, 0.5 0.5(0-25) 0-5(0.3) 0-5( 0.3) 0 P-cM 9(:0) 3(5) NOTE-The figures in parentheses are the amounts present when they differ from the amounts found.In mixtures No. 4 and No. 10, a small amount of sodium was found, although none had been added to these mixtures. I t was observed that this sodium was present in mixtures containing large amounts of caesium and, by producing a chromatogram with 10 pM of caesium and spraying it with zinc uranyl acetate reagent, it was confirmed that the caesium stock solution contained a small percentage of sodium as an impurity. The results are, however, considered satisfactory. APPLICATIOK TO THE ANALYSIS OF A SILICATE ROCK The success of the method of separation described suggested its application to the analysis of rocks containing rubidium and caesium, such as lepidolite and pollucite. The method was therefore applied to the determination of the alkali metals present in a lepidolite that had been previously analysed by Miller and Travesg and contained 1.6 per cent.of lithium, 0.1 per cent. of sodium and 5.8 per cent. of potassium. A 50-mg sample of the finely ground lepidolite was subjected to a Lawrence-Smith fusion.10 The fused mass was digested with water for 30 minutes on a steam-bath and the insoluble material was filtered off. The calcium in the solution was precipitated as the carbonate, which was dissolved in dilute hydrochloric acid and re-precipitated. The final solution was evaporated to dryness in a platinum vessel and the residue was ignited to remove the am- monium salts.The residue was dissolved in 0.25 ml of dilute hydrochloric acid and 0.01-ml portions were spotted on to the paper strip. The separated salts were detected in the usual manner and it was estimated that 2 mg of the rock contained 5 pM of lithium, 0.1 pM of100 MAGEE AND HEADRIDGE CHROMATOGRAPHIC SEPARATIONS [Vol. 82 sodium, 2 pM of potassium, 0.1 pM of rubidium and no caesium. These figures correspond to 1.7 per cent. of lithium, 0.1 per cent. of sodium, 3.9 per cent. of potassium and 0.4 per cent. of rubidium. The use of the proposed method for separcating the alkali metals with a view to their subsequent quantitative determination by micro-technique+ is at present under investigation.SEPARATION OF CERTAIN OTHER METALS The RF values for 100-pg amounts of certain other metals with a phenol - methanol - concentrated hydrochloric acid mixture (57.5 : 22.5 : 20 per cent. w/v/v) on Whatman No. 1 filter-paper are given in Table IV. TABLE IV RF VALUES OF SOME METALS WITH THE SOLVENT MIXTURE USED Metal R F Copper 0.22 & 0.04 MercuryII 0.34 & 0.04 Cadmium 0.30 0.05 Bismuth 0.20 k 0.04 AntimonyIII 0-44 0.05 TinIV 0-37 k 0.04 Iron111 0.39 k 0.05 Metal Chromium111 Aluminium Gallium Indium ThalliumIII Manganese11 Cobalt RF 0.12 4 0.04 0.10 +_ 0.05 0.66 & 0.06 0.20 f: 0.05 0-53 & 0.05 0.16 k 0.04 0.16 4 0.04 Metal RL? Nickel 0.14 0.04 Zinc 0.35 & 0.05 Beryllium* 0-23 & 0.10 Magnesiumt 0-17 & 0.04 Calcium? 0.70 & 0-04 Strontiumt 0.03 f: 0-03 Barium7 0 * Sulphate taken.7 Nitrate taken. It can be seen that, in addition to the group IA metals of the Periodic Table, this solvent mixture separates many elements, including those of group IIIB, vix., aluminium, indium, gallium and thallium. The separation of these elements with aqueous hydrochloric acid and aliphatic-alcohol solvent mixtures has been mentioned before.12 ,I3 Improvements in the separation of the group IIIB metals were made by varying the ratios of phenol, methanol and hydrochloric acid. They are separated with solvent mixtures containing 20 to 40 per cent. of concentrated hydrochloric acid and 15 to 20 per cent. of methanol. The best separation achieved was with a phenol - methanol - concentrated hydrochloric acid mixture (50 : 20 : 30 per cent.w/v/v) . METHOD APPARATUS AND REAGENTS- The apparatus was the same as that used for the alkali metals. Whatman No. 1 filter- Solutions of the metals were prepared by dissolving the chlorides The following reagents were used in the detection of the cations on the paper strips- Aluminium, gallium, indiam, zinc and iron-A 5 per cent. w/v solution of 8-hydroxy- This reagent paper was used throughout. in water or dilute acid. The concentration was l o g of metal per litre. quinoline in a methanol - chloroform - water mixture (85 : 10 : 5 per cent. v/v). was used by Miller and Magee2 for the detection of magnesium and calcium. Thallium-A 5 per cent. w/v aqueous solution of potassium iodide. Titanium-A 1 per cent. w/v aqueous solution of chromotropic acid. Zirconium-E t hanol saturated with alizarin.PROCEDURE- The procedure was the same as that for the alkali metals (p. 98). DETECTION OF ALUMINIUM, GALLIUM, INDIUM, ZINC AND IRON- taining a beaker of concentrated ammonia solution. the paper to dry and examine it under ultra-violet light. zinc fluoresce a brilliant yellow. Spray the paper with 8-hydroxyquinoline reagent and hang it in a closed vessel con- Allow Aluminium, gallium, indium and Iron appears as a black band. DETECTION OF THALLIUM, TITANIUM AND ZIRCONIUM- titanium as a red-brown band and zirconium as a purple band. Spray the paper with the appropriate reagent. Thallium shows as a yellow band,Feb., 19571 I N PHENOL - METHANOL - HYDROCHLORIC ACID SOLVENTS 101 RE s u LT s The R, values of 100-pg amounts of a few metals obtained with the phenol - methanol - concentrated hydrochloric acid mixture (50 : 20 : 30 per cent.w/v/v) in a 16-hour experiment are: zirconium, 0.09 & 0.08; aluminium, 0.13 0.05; indium, 0-28 & 0.04; titanium, 0-34 t_ 0.06; zinc, 0.40 0.05; thalliumIII,O.58 0.06; iron, 0.64 & 0.06; and gallium, 0.73 0.07. It was found that l-pg amounts of aluminium, gallium, indium and zinc were readily detected with 8-hydroxyquinoline and 1 pg of gallium could be separated from 1000 pg of aluminium with this solvent mixture. These results suggest that the solvent mixture might be of use in the qualitative analysis of certain aluminium alloys. REFERENCES 1. 2. 3. Steel, A. E., Nature, 1954, 173, 315. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Lederer, M., Anal. Chim. Acta, 1951, 5, 185. Sakaguchi, T., and Yasuda, H., J . Phavm. SOC. Japan, 1951, 71, 1469. Miller, C. C., and Magee, R. J., J . Chew. Soc., 1951, 3183. Erlenmeyer, H., von Hahn, H., and Sorkin, E., Helv. Chim. Acta, 1951, 34, 1419. Burma, D. P., Anal. Chim. Acta, 1953, 9, 513. Steel, A. E., private communication. Barber, H. H., and Kolthoff, I . M., J . Amer. Chew Soc., 1928, 50, 1625. Sergienko, P. S., Ukvain Khem. Zhur., 1932, 7, 36. Miller, C. C., and Traves, F., J . Chem. Soc., 1936, 1390. Hillebrand, W. F., and Lundell, C. E. F., “Applied Inorganic Analysis,” Second Edition, J. Wiley Duval, C., and Doan, M., Mikvochim. Acta, 1953, 200. Arden, T. V., Burstall, F. H., Davies, G. R., Lewis, J. A., and Linstead, R. P., Nature, 1948, & Sons Inc., New York, 1953, p. 925. 162, 691. CHEMISTRY DEPARTMENT KING’S BUILDINGS THE UNIVERSITY EDINBURGH First submitted, Deceinber 6th, 1956 Amended, August 20th, 1956

ISSN:0003-2654    

DOI:10.1039/AN9578200095

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

Feb., 19571 IN PHENOL - METHANOL - HYDROCHLORIC ACID SOLVENTS WVPfTT IPP 101 The Identification of Nylon and Related Polymers by Paper Chromatography BY M. CLASPER, J. HASLAM AND E. F. MOONEY A simple procedure has been developed for the identification of com- mercial samples of nylon and related polymers. It is based on earlier work carried out by Zahn and Wollemann and by Stuhlen and Horn. After hydrolysis of the sample with acid, the solution is evaporated to dryness and the hydrolysis products thus obtained are dissolved in ethanol and separated by paper chromatography, a mixture of n-propanol, ammonia and water being used as the developillg solvent. The various products are identified by observation of the chromatogram under ultra-violet light and by using ninhydrin and methyl red - borate buffer solutions as spray reagents.IN 1951, Zahn and Wolf,l stimulated, as they said, by work carried out by two of US,^ described a method of differentiating between polyhexamethylenediamineadipamide (nylon 66), polyhexamethylenediaminesebacamide (nylon 610), polycaproamide (nylon 6) and poly- urethane (Perlon U) , as well as a mixed condensate of caprolactam with 5 per cent. of nylon 66. Their method involved the hydrolysis of the sample with hydrochloric acid, followed by the preparation of a solution of the hydrolysis product buffered to a pH of 7-5 and a paper- chromatographic test of this solution. Two chromatograms were prepared. The first chromatogram was used for the detection of amine components, a sec.-butanol - formic acid - water mixture being used as the developing solvent.The fluorescence of the dried paper was observed, as well as its behaviour after having been sprayed with the prepared ninhydrin reagent; Sanger’s reagent and Pauly’s diazo reagent were also tried. The second chromatogram was prepared with use of a mixture of isobutanol, ammonia and glycol as developing solvent. After the paper had been dried, it was sprayed with bromothymol blue102 CLASPER, HASLAM AND MOONEY: THE IDENTIFICATION OF [Vol. 82 in order to obtain evidence of the presence of adipic acid and sebacic acid. I t is obvious, however, from Zahn and Wolf’s paper1 that their separation of the two acids was unsatisfactory . Realising this, Zahn and Wollemann3 carried out further work and showed that, if such acids as adipic and sebacic were first extracted from the hydrolysis products of the nylon with ether, these mixed acids as their sodium salts could then be much more satisfactorily resolved by use of a developing solvent consisting of Iz-propanol, ammonia and water.Indeed, it was suggested in principle that it might be possible to separate all the hydrolysis products by using this solvent mixture. Stuhlen and Horn4 examined the procedures adopted by Zahn and Wolf1 and Zahn and Wollemann3; they used a mixture of n-propanol, ammonia and water as developing solvent in an effort to develop a simple method of differentiation of commercial polymers. They found no difficulty in detecting the amine components with Pauly’s diazo reagent, but encountered considerable difficulties in detecting the acid components and in particular sebacic acid by the bromothymol blue spraying method of Zahn and W~llemann.~ As a result of their investigation, Stuhlen and Horn4 put forward a procedure for the differentiation of commercial polyamides.This involved- (a) the determination of the melting point of the original polyamide, (b) the hydrolysis of 0.4 g of the polyamide with 6 N hydrochloric acid; observations were made of the appearance of the hydrolysis product before and after cooling, and (c) the separation by filtration of any crystals formed from the hydrolysis product and the determination of their melting point. If necessary, mixed-melting-point determinations were carried out with adipic and sebacic acids. This procedure permitted them to differentiate quite readily between nylon 66, nylon 610, nylon 6 and polyundecanoamide (nylon 1 l), but a satisfactory differentiation between the copolymers nylon 66/6 and nylon 66/6JPACM 6 was not achieved.They found it necessary to filter off any acid precipitated as a result of the hydrolysis, to evaporate the filtrate to dryness and then to chromatograph the residue on paper, with a n-propanol - ammonia- water mixture as developing solvent. Subsequent use of Pauly’s diazo reagent permitted them to differentiate satisfactorily between the two copolymers. Although we were aware of the work of Ayers5 on the direct chromatographic separation of polyamides, it seemed to us that both of Zahn’s paper~l.~ and Stuhlen and Horn’s paper4 contained many useful points, some of which could be embodied in a simple procedure for the examination of commercial polyamides. Such a simple method, in which only 25 mg of substance are used, has been worked out and is based on the following procedures- ( a ) hydrolysis of the polyamide with 50 per cent.v/v hydrochloric acid: with obser- vations of the behaviour of the polyamide in this hydrolysis, ( b ) after evaporation of the hydrolysis product to dryness and removal of the excess of hydrochloric acid by further evaporation with water, the residue is dissolved in ethanol and observations are made on the solution process, and the ethanolic solution is chromatographed in duplicate, with n-propanol - ammonia - water mixture as the developing solvent: one of the chromatograms is dried and examined under ultra-violet light, after which it is sprayed with ninhydrin reagent ; the other chromatogram is sprayed with a methyl red - borate buffer solution.The methyl red - borate buffer reagent arises from work that was being carried out in this laboratory on the separation of adipic and sebacic acids. Kalbe6 in an investigation of the paper chromatography of aliphatic dicar boxylic acids reached the conclusion that a reagent prepared so as to contain 0.03 per cent. of methyl red in 0.05 M borate buffer of pH 8.0 was an extraordinarily sensitive reagent for the detection of organic aliphatic acids on paper chromatograms. A reagent of slightly different concentration, which had proved to be particularly satisfactory in chromatographic work on adipic and sebacic acids, proved to be particularly valuable in this chromatographic examination of the hydrolysis products of polyamides.This particular spray reagent, on comparative testing, proved to be im- measurably superior to the bromothymol blue spray reagent of pH 10.0 used by Zahn and Wollemann.3 From a reading of Zahn and Wollemann’s paper it appeared that their spray reagent of pH 10.0 was adapted from the work of Brown,’ who used a similar bromothymol ( c )Feb., 19571 NYLON AND RELATED POLYMERS BY PAPER CHROMATOGRAPHY 103 blue spray reagent of pH 7.5; the latter reagent was also found to be unsatisfactory. The methyl red - borate buffer spray gave trustworthy information about the acids present in the hydrolysis products and, moreover, about such amine components as hexamethylene- diamine and p-diaminodicyclohexylmet hane.The complete procedure incorporating the principles enumerated above has permitted us to differentiate quite readily between commercial polyamides and copolymers. Full details of the tests are given below, followed by information about the behaviour of various polymers in the test. METHOD APPARATUS- A imev Universal chromatographic tank. Filter-paper, Whatman No. 1, for chromatography, 10 inches x 10 inches, with Glass tubes, prepared from glass tubing having an internal diameter of about 6 mm; corner holes. the length of each tube is approximately 100 mm. REAGENTS- Developing solvent-Mix 6 parts of n-propanol, 3 parts of ammonia solution, sp.gr. 0.880, and 1 part of water, by volume.Methy2 red - borate bu$er reagent-Dissolve 12.368 g of analytical-reagent grade boric acid and 14.912 g of analytical-reagent grade potassium chloride in water in a 1-litre calibrated flask, then add 35.0 ml of N sodium hydroxide solution and dilute to the mark with water. Check the pH of this solution, which should be 8.0 & 0-1. Dissolve 0.03 g of methyl red indicator in 100 ml of the buffer solution, shaking frequently to ensure complete solution. Ninhydrin reagent-Dissolve 0.3 g of ninhydrin in a mixture of 95 parts of n-butanol and 5 parts of 2 N acetic acid, by volume. Hydrochloric acid, diluted (1 + 1). PROCEDURE- Weigh 25 mg of the substance into a glass tube and add 0.5 ml of diluted hydrochloric acid (1 + 1). Seal the glass tube carefully in a gas flame, then place it in an upright position in an oven at 120" C and leave it overnight.After removing the tube from the oven, make observations on the appearance of the solution; allow it to cool and make further observations. On opening the tube, note any odour that may be evolved, and then transfer the contents of the tube to a 10-ml beaker, washing with 1 to 2ml of water. Observe any change that may take place on the addition of water. Evaporate the contents of the beaker to dryness once or twice, with small additions of water to remove the hydrochloric acid completely, and finally dry in an oven at 105" C for 15 to 20 minutes. Dissolve the residue in 1 ml of ethanol and warm to effect solution. Now spot in duplicate 3 to 5 p1 of the ethanolic solution on to a Whatman No.1 10-inch x 10-inch filter-paper having corner holes on a line drawn 18 inches from the edge of the paper. Six spots can be placed along this line, so that at least three samples can be dealt with a t the same time. After spotting, fit the paper into the frame of the Aimer Universal chromato- graphic outfit. Put about 200 ml of the n-propanol - ammonia - water developing solvent into the trough in the chromatographic tank and, after allowing the solvent to equilibrate with the atmosphere in the tank for about 1 hour, place the frame in position in the tank and allow the chromatograms to develop for about 7 hours. At the end of this time remove the frame and allow the paper to dry in the air. When the paper is dry, place it in an oven at 105" C for 15 minutes and then examine it under ultra-violet light.Xow separate the duplicate chromatograms and spray the first chromatogram with the ninhydrin reagent, and, after allowing the paper to dry in the air, place it in an oven a t 105" C for 5 minutes. Observe the position and colour of the spots obtained. Spray the duplicate chromatogram with the methyl red - borate buffer reagent and dry it by placing it between sheets of filter-paper. After 10 minutes make observations on the position and colour of the spots that develop, and repeat these observations after a further 20 minutes. Note if there is any insoluble material.104 CLASPER, HASLAM AND MOONEY: THE IDENTIFICATION OF [Vol. 82 TABLE 11: EXAMINATION OF COMMERCIAL 1. Observation in hydrolysis tube- (a) while still hot ( b ) after cooling 2.Odour on open- ing hydrolysis tube 3. Solubility in ethanol 4. Examination under ultra-violet light Nylon 66 Complete solution Tendency for crystal formation; white precipitate on adding water No characteristic odour Completely soluble Spot due to hexa- methylenediamine dih ydrochloride (B, 0.81 to 0.86). Also slight evidence of adipic acid spot (R,0.28 to 0.36) 5. Examination Purple spot due after spraying to hexamethylene- with ninhydrin diamine dihydro- reagent chloride (RF 0.81 t o 0.86). Also faint brown spot due to adipic acid (R, 0-28 to 0.36) Nylon 610 Complete solution Heavy whit.e crystalline precipi- tate No characteristic odour Completely soluble Spot due to hexa- methylenedi amine dihydrochloride Also slight evidence of sebacic acid spot (R, 0.81 to 0.86).(R, 0.59 to 0.64) Nylon 6 Complete solution Complete solution No characteristic odour Completely soluble Spot due to 5- aminocaproic acid hydrochloride (R, 0.4 to 0.52). Nylon 11 Small amount of insoluble matter Sets solid owing to mass of fine white crystals No characteristic odour Completely soluble Large spot in same place as for hexa- methylenediamine dihydrochloride (R, 0.81 to 0.86) Purple spot due Purple spot due Large purple spot t o hexamethylene- t o &amino- in same place as diamine dihydro- caproic acid hexamethylene- chloride (R, 0.81 t o hydrochloride diamine dihydro- 0.86). Also f,aint (B, 0.47 to 0.52) chloride sebacic acid brown spot clue to (RF 0.81 to 0.86) (RF 0.59 to 0.64) 6. Examination after spraying with methyl red - borate buffer reagent- (a) after initial Pink spot due to Pink spot due to Pink spot due to Pink spot due to 10 minutes; adipic acid sebacic acid ammonium ammonium colour of paper: (22, 0.28 to 0.36).(R, 0-59 to 0.64). chloride chloride yellow Pink spot due to Pink spot due to (R, 0.47 to 0-52) (R, 0.47 to 0.52) ammonium ammonium chloride chloride (RF 0.47 to 0.52) (RE 0.47 to 0.52) (b) after 30 Pink spots become minutes ; more pronounced. colour of paper: Yellow spot yellow turning develops due to to pink hexamethylene- diamine dihydrochloride (RF 0.81 to 0-86) Pink spots become Pink spot becomes Pink spot becomes more pronounced. more pronounced. more pronounced. Yellow spot No evidence of No evidence of develops due to of any further of any further diamine dihydrochloride (R, 0.81 to 0.86) hexamethylene- spots spotsFeb., 19571 NYLON AND RELATED NYLON POLYMERS AND COPOLYMERS Igamide U Insoluble matter present No change.Same insoluble matter present Sweet odour Insoluble matter present Spot due to hexa- methylenediamine dihydrochloride (I?, 0.81 to 0.86) Purple spot due to hexamethylene- diamine dihydrochloride ( R F 0.81 to 0.86) Pink spot due to ammonium chloride ( R F 0.47 to 0.52) Pink spot becomes more pronounced. Yellow spot develops due to hexamethylene- diamine dihydrochloride (RF 0-81 to 0.86) Nylon 66/6lO (40 : 60) Complete solution Heavy white crystalline precipitate No characteristic odour Completely soluble Spot due to hexa- methylenediamine dihydrochloride Also slight evidence of adipic acid spot (RF 0.28 to 0.36) and sebacic acid ( R F 0.81 to 0.86).spot ( R F 0.59 to 0.64) Purple spot due to hexamethylene- diamine dihydrochloride Also faint brown spots due to adipic acid and sebacic acid (RB 0.59 to 0.64) ( R F 0.81 to 0.86). ( R p 0.28 to 0.36) Pink spots due to adipic acid (RF 0.28 to 0.36) and sebacic acid Pink spot due to ammonium chloride (R, 0.47 to 0.52) Pink spots become more pronounced. Yellow spot develops due to hexamethylene- diamine dihydrochloride (RF 0.59 to 0.64). (R, 0.81 to 0.86) POLYMERS BY PAPER CHROMATOGRAPHY 105 Nylon 66/6 (60 : 40) Complete solution No characteristic odour Completely soluble Spots due t o hexa- methylenediamine dihydrochloride and 5-amino- caproic acid hydrochloride Also slight evidence of adipic acid spot (RE 0.81 to 0.86) ( R F 0.47 to 0.52).( R F 0.28 to 0.36) Purple spots due to hexamethylene- diamine dihydrochloride and 5-amino- caproic acid hydrochloride Also faint brown spot due to adipic acid (HF 0.81 to 0.86). ( R F 0.47 to 0.52). ( R F 0.28 to 0.36) Pink spot due to adipic acid Pink spot due to ammonium chloride (R, 0-47 to 0.52) ( R F 0.28 to 0.36). Pink spots become more pronounced. Yellow spot develops due to hexamethylene- diamine dihydrochIoride ( R F 0.81 t o 0.86) Nylon 66/610/6 (40 : 30 : 30) Complete solution No characteristic odour Completely soluble Spots due to hexa- methylenediamine dihydrochloride and 5-amino- caproic acid hydrochloride Also slight evidence of adipic acid spot (R, 0-28 t o 0.36) (RF 0.81 to 0.86) (RF 0.47 to 0.52).Purple spots due to hexamethylene- diamine dihydrochloride (R, 0.81 to 0.86) and 5-amino- caproic acid hydrochloride Also faint brown spots due to adipic acid (H, 0-28 to 0.36) and sebacic acid (RF 0-59 to 0.64) ( B F 0.47 to 0.52). Pink spots due to adipic acid (RF 0.28 to 0-36) and sebacic acid Pink spot due to ammonium chloride Pink spots become more pronounced. Yellow spot develops due to hexamethylene- diamine dihydrochloride (R, 0.81 to 0.86) ( R E 0.59 t o 0.64). ( R F 0.47 to 0.52) Nylon 66/6/ PACM 6 ( 1 : l : l ) Complete solution No characteristic odour Completely soluble Spots due to hexa- methylenediamine dihydrochloride (I?, 0.81 to 0.86) and 5-amino- caproic acid hydrochloride (R, 0.47 to 0.52). Also evidence of a p-diaminodicyclo- hexylmethane dihydrochloride spot (R, 0.98) and an adipic acid spot (RF 0.28 to 0.36) Purple spots due to hexamethylene- diamine dihydrochloride (RF 0.81 to 0.86) and &amino- caproic acid hydrochloride (R, 0.47 to 0-52) and p-diamino- dicyclohexyl- methane dihydrochloride (H, 0.98).Also faint brown spot due to adipic acid (RF 0.28 to 0.36) Pink spot due t o adipic acid (R, 0.28 t o 0.36). Pink spot due t o ammonium chloride ( R , 0.47 to 0.52) Pink spots become more pronounced. Yellow spot develops due to hexamethylene- diamine dih ydrochloride (R, 0.81 to 0.86) and 9-diamino- dicyclohexyl- methane dihydro- chloride (R, 0.98)[Vol. 82 RESULTS Chromatograms from typical commercial nylon samples after being sprayed with ninhydrin and methyl red - borate buffer are shown in Figs.1 and 2, respectively. Observations made on the behaviour of the various substances obtained on hydrolysis of nylon-type materials when they are subjected to the conditions of the test described above are shown in Table I. Examination of the Table shows that a spot of R, value 0.48 to 0.55 is common to all the hydrochloride makerials This spot does not fluoresce under ultra-violet light and gives no colour with the ninhydrin reagent; it does, however, give a pink colour with the methyl red - borate buffer reagent. The R, values of this spot and the spot given by 5-aminocaproic acid hydrochloride are similar, but these spots can easily be differentiated, as the latter gives a fluorescence under ultra-violet light and a purple colour with ninhydrin reagent.It has been proved that this spot is due to the presence of ammonium chloride, formed by the reaction of the ammonia in the developing solvent with the chloride ion in the base hydrochlorides. 106 CLASPER, HASLAM AND MOONEY THE IDENTIFICATION OF TABLE I EXAMINATTOE OF THE SUBSTANCES OBTAINED ON HYDROLYSIS OF SYLON-TYPE POLYMERS Effect of methyl red - borate buffer reagent added A Effect of -7 Examination under ninhydrin after after Substance RF value ultra-violet light reagent 10 minutes 30 minutes Adipic acid Sebacic acid 0.28 t o 0.36 Faint fluorescence Faint Pink spot Pink spot 0-59 to 0.64 Faint fluorescence Faint Pink spot Pink spot brown spot brown spot Hexamethylenediamine 0.47 to 0.52 No fluorescence No spot Pink spot Pink spot dih ydrochloride 0.81 to 0-86 Strong fluorescence Purple spot Faint Definite yellow spot yellow spot p-Diaminodicyclohexyl- 0.47 to 0.52 No fluorescence No spot Pink spot Pink spot methane 0.98 Strong fluorescence Purple spot Faint Definite dihydrochloride yellow spot yellow spot 5-Aminocaproic acid 0.47 t o 0-52 Strong fluorescence Purple spot Pink spot Pink spot hydrochloride 10-Aminoundecanoic acid 0.47 to 0.52 No fluorescence No spot Pink spot Pink spot hydrochloride 0.81 t o 0.86 Strong fluorescence Purple spot No spot No spot Various commercial nylon polymers and copolymers have been examined by the test described above and details of the results obtained are shown in Table 11.We have found that the small purple spots that are sometimes observed after use of the ninhydrin reagent are not due to the normal base hydrochlorides.These spots are particu- larly noticeable with copolymers. We have no satisfactory explanation as yet for this behaviour, although it may be due to small proportions of impurities in the commercial samples. For the time being such spots are disregarded and have not been described in Table 11. The test for the identification of nylons described in this paper has been found to be useful when applied to various types of nylons that are not strictly speaking commercial samples, for example- (i) Examination of a sample of polyheptoamide (nylon 7) showed that a spot due to 6-aminoheptoic acid hydrochloride was obtained; it had an R, value of approxi- mately 0.72. This spot behaved in the same manner in the test as that due to 5-aminocaproic acid hydrochloride.I t was also found by applying the test that the presence of nylon 7 in nylon 6 could quite readily be detected. (ii) Examination of a sample of nylon 66/610/6 containing 14 per cent. of Santiciser 8 (a mixture of N-ethyltoluene-o-sulphonamide and -9-sulphonamide) as plasticiser gave results that were exactly similar to those of unplasticised nylon 66/610/6 (see Table 11); there was no evidence of any interference by the plasticiser.Fig. 1. Chromatograms of commercial nylon polymers and copolymers : ninhydrin spray reagentFig. 1. Chromatograms of commercial nylon polymers and copolymers : ninhydrin spray reagentFeb., 19571 NYLON AND RELATED POLYMERS BY PAPER CHROMATOGRAPHY 107 (iii) Examination of an interpolymer of nylon 66/6 containing 5 per cent. of nylon 6 showed that this amount could readily be detected; the spot due to the 5-amino- caproic acid hydrochloride in the hydrolysis products from the nylon 6 was quite apparent under ultralviolet light and with the ninhydrin spray reagent. NOTE-If a derivative of nylon 66, such as methoxymethylnylon, is suspected, it is obviously desirable to test whether formaldehyde is liberated on treatment with acid. If this derivative is present, a distinct odour of formaldehyde will be detected on evaporation of the hydrolysis product. REFERENCES 1. 2. 3. 4. 5. 6. 7. Zahn, H., and Wolf, H., Melliand Textilber., 1951, 32, 317. Clasper, M., and Haslam, J., Analyst, 1949, 74, 224. Zahn, H., and Wollemann, B., Melliami Textilber., 1951, 32, 927. Stiihlen, F., and Horn, H., Kunstofle, 1956, 46, 63. Ayres, C. W., Analyst, 1953, 78, 382. Kalbe, H., Hoppe-Seyl Z., 1954, 297, 19. Brown, F., Biochern. J . , 1950, 47, 598. IMPERIAL CHEMICAL INDUSTRIES LIMITED PLASTICS DIVISION WELWYN GARDEN CITY, HERTS. ,4 ugust 1:5tfi, 1 056

ISSN:0003-2654    

DOI:10.1039/AN9578200101

出版商:RSC    

年代:1957

数据来源: RSC

摘要:

Feb., 19571 NYLON AND RELATED POLYMERS BY PAPER CHROMATOGRAPHY 107 Determination of Glycollic Acid in Used Antifreeze Solutions BY H. GREEN A method is described for the determination of the glycollic acid content of used antifreeze solutions of the inhibited ethylene glycol type. Insoluble matter is removed by treatment with zinc oxide and soluble cations are removed by means of an ion-exchange resin. Separation of the glycollic acid from ethylene glycol is achieved by the use of an anion-exchange resin. The glycollic acid is then eluted with sulphuric acid and treated with 2 : 7-di- hydroxynaphthalene in concentrated sulphuric acid to give a coloured complex suitable for photometric measurement. IN connection with the analysis of used antifreeze solutions of the inhibited ethylene glycol type (D.T.D.779) taken from water-cooled automobile engines, it became necessary to deter- mine glycollic acid (or its salts) occurring as a decomposition product of the glycol. The avail- able literature indicated that the most sensitive test for glycollic acid was based on its reaction with 2 : 7-dihydroxynaphthalene in concentrated sulphuric acid at 100" C. At this temperature the initial product of the reaction is readily oxidised by air to form a pronounced red-violet co1our.l Used antifreeze solutions of the type referred to above contain various amounts of soluble colloidal and precipitated iron compounds, and Squires2 has found that dissolved iron compounds can catalyse the air-oxidation of aqueous ethylene glycol at moderately elevated temperatures. It seemed possible, therefore, that aqueous glycol containing iron salts would give rise to glycollic acid when treated with 2 : 7-dihydroxynaphthalene reagent as described above, and exploratory tests proved this to be so. In these circumstances, it was decided to try to separate the glycol from glycollic acid before determining the latter.As the pH of all the antifreeze solutions encountered lay between 4 and 8, it seemed probable that a separation of glycollic acid and ethylene glycol could be effected by the use of an anion-exchange resin, and, in practice, this proved to be so. A series of experiments was carried out in which aqueous solutions of ethylene glycol and glycollic acid were passed through the resin, which was then washed with water.The pH of the resulting solution was always the same as that of a similar volume of distilled water and ethylene glycol passed through the same resin. Tests with 2 : 7-dihydroxynaphthalene showed that the acid had not passed through the resin.108 GREEN: DETERMINATION OF GLYCOLLIC ACID [Vol. 82 With the anion-exchange resin, the separation of ethylene glycol from glycollic acid was satisfactorily achieved with synthetic solutions, but when used antifreeze solutions were passed through the resin some of the precipitated iron compounds were adsorbed and, since they might interfere with subsequent manipulation of the sample, complete separation of the iron compounds from the glycollic acid was desirable. Although the addition of a salt, such as sodium sulphate, followed by filtration was found to clear the solution of colloidal iron compounds, this procedure involved the use of a large quantity of resin to remove the added ions. This difficulty was overcome by treating the used antifreeze solution with a large excess of zinc oxide, followed by centrifuging.The colloidal iron compounds were collected in the insoluble residue of zinc oxide and massive ferric hydroxide. Any iron and zinc then in solution could be removed by using a cation- exchange resin. Experimental conditions that gave concentrations of glycollic acid capable of being measured colorimetrically after reaction with the 2 : 7-dihydroxynaphthalene reagent were then established and an absorption graph was plotted. This showed a maximum absorption at 535 mp (see Fig.1). A series of standard glycollic acid solutions in aqueous glycol was I I n 1 1 1 1 I I I I 360 400 440 480 520 ( 5 Wavelength, rnp I I 5 ) 560 600 Fig. 1. Absorption curves : curve A, glycollic acid complex; curve B, reagent then prepared, and the glycollic acid was separated from the glycol by using an anion- exchange resin as described above. The glycol.lic acid in the eluate was treated with the 2 : 7-dihydroxynaphthalene reagent to convert it into the red-violet coloured complex and 2.0 / j - 1.8 - 1.6- I -4 Glycollic acid present, % Fig. 2. Typical calibration curve for glycollic acid complex 2.01 1*81 I *6 1 0.04 0.05 006Feb., 19571 IN USED ANTIFREEZE SOLUTIONS 109 the optical density of the resulting solutions was measured a t 535 mp.The calibration graph obtained (Fig. 2) showed a slight curve, but the deviation from Beer’s law was small. The absorption of the solutions was measured at intervals up to 4 hours after they had been prepared without significant changes in optical density taking place. This procedure was repeated with the same standard solutions of glycollic acid in aqueous glycol and the 2 : 7-dihydroxynaphthalene reagent, and within experimental error the same calibration graph was obtained. A second solution of 2 : 7-dihydroxynaphthalene was prepared and the calibration graph was re-determined. The results obtained with the first and second 2 : 7-dihydroxynaphthalene solutions are shown in Table I. TABLE I REPRODUCIBILITY OF REAGENT SOLUTIONS Optical density Concentration of glycollic acid, r----h------, g per 100 ml of antifreeze Reagent 1 Reagent 2 0.005 0.010 0.030 0.030 0-040 0.313 0-320 0.480 0.486 0.845 0-840 1.153 1.150 1.356 1-358 A solution of the 2 : 7-dihydroxynaphthalene reagent was taken and the calibration graph was re-determined over a period of 3 weeks to investigate the effect of ageing of the reagent. The results obtained are shown in Table 11, and indicate that up to 4 days after preparation of the reagent the calibration graph is reproducible.However, after 5 or 6 days the reagent solution gradually acquires a violet tint and the higher concentrations of glycollic acid fail to produce maximum colour intensity. The method described below is designed for contents of glycollic acid in antifreeze solutions from 0.002 to 0.040 per cent.The method determines total glycollate-ion content of the solution and is not selective for “free” glycollic acid. The range of the determination can be extended by amending the original aliquot for zinc oxide treatment, since con- centration of the glycollic acid into the required volume of 80 per cent. sulphuric acid is effected by the ion-exchange resin. TABLE I1 RELATIONSHIP BETWEEN OPTICAL DENSITY AND AGE OF REAGENT Optical density on- Concentration of glycollic acid, g per 100 ml of antifreeze 0.0025 0.005 0.010 0.020 0.030 0.040 r- 2nd day 0.182 0.313 0.480 0.845 1.153 1.356 3rd day 4th day 7th day 0.196 0.190 0.185 0-321 0.319 0-308 0-475 0.476 0-472 0.886 0.854 0.855 1.156 1.150 1.152 1.341 1.340 1.275 - 21st day 0.185 0.302 0.475 0.860 1.140 1.258 METHOD REAGENT- 2 : 7-Dihydroxynap~thaZene reagent-Dissolve 0.01 g of 2 : 7-dihydroxynapht halene in con- centrated sulphuric acid to 100ml.The solution changes colour on standing from yellow to blue, and this change, which takes about 24 hours, must be allowed to occur before the reagent is used, PROCEDURE- Put a 10-ml sample of the antifreeze solution by pipette into a centrifuge tube and add 40 ml of water, measured accurately, followed by 5 to 10 g of solid zinc oxide. Mix the solu- tion thoroughly and revolve it in a centrifuge at 2000 r.p.m. and 10 cm radius for 5 minutes. Transfer a 25-ml aliquot of the supernatant liquid t o a 100-ml beaker and add 1 to 2 g of activated cation-exchange resin Zeo-Karb 225. Stir the solution well and set it aside for a few minutes before filtering through a small pulp pad, which is subsequently washed well with cold water.The filtrate, containing the ethylene glycol, glycollic acid and possibly small amounts of phosphoric acid derived from the inhibitor, should be quite clear at this stage. Stir To this filtrate add 1 g of dry activated anion-exchange resin De-Acidite FF.110 TINSLEY AND NOWAKOWSKI : THE DETERMINATION OF [Vol. 82 the solution and set it aside for a few minutes, and then filter it through a sintered crucible of porosity 2, washing the filter well with water. Reject the filtrate, which contains the ethylene glycol. Dry the crucible containing the resin as far as possible, both inside and out, by means of a filter-paper to remove extraneous moisture.Elute the glycollic acid from the resin in the crucible, adding to it by pipette two 5-ml portions of 80 per cent. sulphuric acid and, after addition of each aliquot, gently suck the crucible dry, using a vacuum-desiccator and collecting the eluate in a small dry glass-stoppered bottle. Add 20 ml of concentrated sulphuric acid to the bottle by pipette and mix the solution thoroughly. Transfer a 5-ml aliquot to a dry test-tube and add exactly 10 ml of the 2 : 7-dihydroxynaph- tlialene reagent. After mixing, place the test-tube in a beaker of boiling water for a timed period of 10 minutes and then cool it to room temperature. After cooling, transfer the solution to a 25-ml calibrated flask, using 50 per cent. sulphuric acid for the transfer. Cool the solution and dilute it to 25 ml with the 50 per cent.sulphuric acid. Measure the absorp- tion of the solution, using a 2-cm cell and a wavelength of 535mp. Set the spectrophotometer on a blank value, by placing 10 ml of 80 per cent. sulphuric acid in a dry glass-stoppered bottle, adding 20 ml of concentrated sulphuric acid and pro- ceeding as described above. If a 20 per cent. solution of ethylene glycol in water is subjected to the entire procedure, then both relative and absolute values can be found for the glycollic acid. (NOTE-Ethylene glycol appears to contain a small quantity of glycollic acid, probably as a by-product during preparation. The accuracy of the determination is improved if the glycollic acid content of the ethylene glycol, used in the preparation of the antifreeze, can be determined. Otherwise, all results are a measure of this plus the increase in glycollic acid content of the antifreeze mixture after use.) I express my thanks to the Director and Council of the British Cast Iron Research Association for permission to publish this paper. REFERENCES 1. Eegriwe, E., Z. anal. Chern., 1932, 89, 121; see also Feigl, F., “Spot Tests, Organic Applications,” Fourth Edition, Elsevier Publishing Co. Ltd., Amsterdam and New York, 1954, Volume 2, pp. 249 and 250. 2. Squires, A. T. B. P., private communication. 3. .\LVECHURCH, BIRMINGHAM September lSZh, 1956 Feigl, F., op. cit., p. 249. BRITISH CAST IRON RESEARCH ASSOCIATION

ISSN:0003-2654    

DOI:10.1039/AN9578200107

出版商:RSC    

年代:1957

数据来源: RSC