ISSN:0003-2654    

DOI:10.1039/AN95479FX037

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

No. 21 August, 1954 THE SOCIETY FOR ANALYTICAL CHEMISTRY BULLETIN COMMUNICATIONS ACCEPTED FOR PUBLICATION IN THE ANAL YST b THE following communications have been accepted for publication in The Analyst, and are expected to appear in the near future. It is not possible to enter into correspondence about any of them. “A Rapid Method for the Determination of Lead in Steel,” by G. H. Bush. “Organic Reagents for the Analysis of the Platinum Metals,” by D. E. Ryan. “Paper Chromatography of Cations with Azo Derivatives of 8-Hydroxyquinoline, ” by “The Determination of Fluorene in Tar Fractions,” by G. A. Vaughan and D. W. Grant. “A Potentiometric Method for Macro- and Micro-determination of Thallium1 and Thallium“’ “The Determination of Potentially Ionic Fluorine in Non-aqueous Solvents,” by H.F. “Colorimetric Determination of Magnesium in Titanium and Titanium Alloys,” by H. J. G. “Determination of Iodine in Common Salt by Catalytic Reduction of Ceric Ions,” by “An Impurity-compensated Polarographic Method for the Determination of the Gamma- “The Determination of Lead in Cocoa with a Square-Wave Polarograph,” by D. J. Ferrett, “Use of 8-Hydroxyquinoline for the Determination of Zinc in Solutions containing Copper,” “The Removal of Dissolved Carbon Dioxide in the Volumetric Determination of Boron,” “The Estimation of Trace Metals, particularly Lead,” by J. G. Maltby. “A Modified Method for the Decomposition of Chromite,” by P. D. Malhotra. “The Colorimetric Determination of Acetylacetone with Ferric Iron,” by T. G. Bonner and “Amperometric Titration of 8-Hydroxyquinoline and Some Derivatives with Potassium “The Elimination of the Blank Value in the Unterzaucher Method for the Micro-determination Q.Fernando and M. de Silva. (Note.) by Oxidation with Alkaline Yermanganate,” by I. M. Issa and R. M. Issa. Liddell. (Note.) Challis and D. F. Wood. M. DubravEiC. Isomer in Technical Benzene Hexachloride ,” by J. Watt. G. W. C. Milner and A. A. Smales. by S. 2. Haider and M. H. Khundkar. by H. Jackson and R. E. Bailey. (Note.) (Note.) (Note.) (Note.) M. Thorne. Bromate,” by Q. Fernando. (Note.) of Oxygen,” by A. F. Colson. (Note.)“A Continuous Recorder for Dissolved Oxygen in Water,” by R. Briggs, G. Knowles and “A Modified Hand-operated High-pressure Hydrogen Sulphide Generator,” by P. Heath.“The Comparison of Inks and Writings by Paper Chromatography,” by B. B. Coldwell. L. J. Scragg. (Apparatus.) BRITISH STANDARDS INSTITUTION DRAFT SPECIFICATIONS A FEW copies of the following draft specifications, issued for comment only, are available to members of the Society, and can be obtained from the Secretary, The Society for Analytical Chemistry, 7-8, Idol Lane, London, E.C.3. Draft Specifications prepared by Technical Committee LBC/l-Volumetric, Mouldblown and Lampblown Glassware. CT(LBC)4845-Draft B.S. for Iodine Flasks. CT(LBC)4848--Draft B.S. for Rees - Hugill Powder Density Flask. CT(LBC)4849-Draft B.S. for Boiling Flasks. Draft Specification prepared by Technical Committee LBC/1 1-Microchemical CT(LBC)4846--Draft B.S. for Vacuum Drying Ovens for Microchemical Apparatus. Purposes (Part G2 of B.S.1428). Draft Specification prepared by Technical Commit tee GLC/3-Analysis of Glass. CT(GLC)4847--Draft B.S. Methods for the Analysis of Glass. (Part 1 : Recom- mended Procedure for the Analysis of Glasses of the Soda-Lime-Magnesia- Silica Type.) Notices LA SOCI&TE SUISSE DE CHIMIE ANALYTIQUE APPLIQUQE THE 66th Annual Meeting of this Society will be held on Friday and Saturday, 3rd and 4th September, 1954, at Flims-Waldhaus (Grisons). Members of the Society will be welcomed at this meeting, of which further information can be obtained from the Secretary of the Swiss Society, Dr. F. I;. Achermann, Bollwerk 27, Berm, Switzerland. PUBLICATIONS OF THE ROYAL INSTITUTE OF CHEMISTRY THE Royal Institute of Chemistry has reprinted the Monograph “Statistical Methods with Special Reference to Analytical Chemistry,” by D. K. Read, B.Sc., F.R.I.C., F.I.S. (Lectures, Monographs and Reports 1951, KO. l), which had been out of print for some time. Copies can be obtained from the Royal Institute of Chemistry, 30, Russell Square, London, lV.C.2, price 4s. Gd. post free. PRINTED BY HEFFER & SONS LTD., CAMBRIDGE, ENGLAND

ISSN:0003-2654    

DOI:10.1039/AN954790X039

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479BX041

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479FP103

出版商:RSC    

年代:1954

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95479BP111

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

AUGUST, 1954 THE ANALYST Vol. 79, No. 941 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY SPECIAL MEETING A SPECIAL Meeting of the Society was held at 6 p.m. on Wednesday, July 21st, 1954, in the Lecture Theatre of the Royal Institution, 21 Albemarle Street, London, W.l. The Chair was taken by the President, Dr. D. W. Kent-Jones, F.R.I.C. The subject of the meeting was “The Use of Perchloric Acid in Analytical Chemistry,” and the following papers were presented: “Perchloric Acid and Some Organic Perchlorates, ” by Professor Harold Burton, Ph.D., D.Sc., F.R.I.C., and P. F. G. Praill, B.Sc., Ph.D. “A Bomb in a Test Tube. Perchloric Acid Idiosyncrasies,” by Professor G. Frederick Smith, Ph.D. Professor Smith made a special journey from America to London to give this lecture to the society.The discussion that followed was opened by Mr. B. Bagshawe, A.Met. There was an attendance of 500 members and visitors, the large audience completely filling the lecturg theatre. Professor Burton discussed the action of perchloric acid and some organic perchlorates on certain types of organic compounds, which could lead to the formation of explosive compounds. Professor Smith demonstrated by actual experiments in the lecture room the general properties of perchloric acid and how it could be used with complete safety for wet-oxidising such substances as chromacised cat-gut and a cigar, and for the analysis of metals, notably stainless steels. He stressed that safety depended on a knowledge of the reactions involved. NEW MEMBERS ORDINARY MEMBERS Samuel Edward Qualtrough Ashley, B.S.(N.Y.) , M.A. (Princeton) ; Arnold Eric Bender, B.Sc. (Liv.), Ph.D. (Sheff.) , F.R.I.C. ; Cyril Leake Grayson; Edmund Green, M.Sc. (Lond.) ; Noel Francis Maclagan, DSc. , M.D. , F.R.C.P. (Lond.), F.R.I.C. ; William Joseph Sommerville Pringle, BSc. (Lond.) ; Muhammad Shafi, B.Sc. ; John Frederick Harrop, B.Sc. DEATHS WE regret to record the deaths of John Theodore Hewitt Richard William Woosnam. NORTH OF ENGLAND SECTION A MEETING of the Section was held at 2.30 p.m. on Saturday, April loth, 1954, at the City Laboratories, Mount Pleasant, Liverpool 3. The subject of the meeting was the Food and Drugs (Amendment) Bill, a bill to amend the Food and Drugs Act, 1938, and the Food and Drugs (Milk and Dairies and Artificial Cream) Act, 1950. The Chairman reviewed the salient parts of the Bill, and his review was followed by a discussion. In addition to the planned programme, Mr. N. Heron, F.R.I.C., kindly gave a talk on the determination of alcohol in body fluids and demonstrated the apparatus used for this deter- mination. A discussion followed. TFIE Seventeenth Summer Meeting of the Section was held at the Grand Hotel, Morecambe, from Friday, June 18th, to Monday, June 21st, 1954. On the morning of Saturday, June 19th, a paper was presented by Mr. E. Green, M.Sc., on “Cosmetics.” A motor-coach trip to Windermere was made on Sunday qfternoon. 465 Mr. T. W. Lovett, F.R.I.C., was in the Chair. The Chairman, Mr. T. W. Lovett, F.R.I.C., presided.

ISSN:0003-2654    

DOI:10.1039/AN9547900465

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

466 OBITUARY [Vol. 79 Obituary HAROLD GOVETT COLMAN HAROLD GOVETT COLMAN died on February 20th, 1954. His wide circle of friends will feel the loss of a great personality, wise counsellor and outstanding chemist. Colman was born in 1866, studied chemistry under Roscoe at Owens College, Manchester, and took his B.Sc. (Vict.) in 1885. During the same year he successfully sat for the Associate- ship of the Institute of Chemistry. He also studied at Strasbourg and Wurzburg, taking his Ph.D. at the latter University in 1888. In that year he also took the M.Sc. (Vict.), and set out to follow an academic career, becoming; private assistant to Sir Henry Roscoe and joint editor of Roscoe and Schorlemmers’ farnous “Treatise on Chemistry.” He joined Professor Tilden’s staff at Mason College, Birmingham, as demonstrator and assistant lecturer in 1891.About 1893, Professor Tilden advised the Birmingham Gas Committee to appoint a qualified and knowledgeable chemist to develop up-to-date chemical control of gas- manufacturing processes, and Dr. Colman was the chemist chosen and appointed. He spent ten years in the Gas Industry putting chemical control on a firm basis and became a Member of Council of the Institution of Gas Engineers,. Attendance at Council meetings was not encouraged by the Gas Department, and this aktitude, along with other drawbacks, caused Colman to embrace the freedom of the consulting world. He found his real life’s work in this sphere, and the gas industry is greatly in his debt for his valuable investigations into the working efficiencies of large-scale carbonising phnt and allied processes.In 1904 he took his F.I.C., and in 1910 he received the degree of D.Sc. Manchester for research. In the first World War he analysed and assessed for the Ministry of Supply the benzole and toluole recovered by the gas and allied industries for high explosives. This necessitated devising methods of analysis, and the term “1;o colmanise” in connection with such work came into everyday use in many laboratories. He was doing similar work for Midland Tar Distillers right to the end. Colman’s lectures on Gas-works Practice in the Fuel Department of Leeds University were classical, and when he gave up lecturing in 1924 it required the services of two chief chemists to carry on the work. Colman was also a tower of strength in the development of gas-works refractories for retorts and other plant, being a great supporter of Dr.J. W. Mellor, F.R.S., of Stoke, another Manchester University man. Colman was Chairman of the Joint Refractories Committee of the Gas Industry and the Refractory Manufacturers, and it was the writer’s privilege to make him a presentation on behalf of the Committee when he resigned in 1941. He carried out a considerable amount of unpaid work from sheer generosity and sometimes because of the intellectual satisfaction he got from solving some particular problem. Had he been a really keen business man he would certainly have made a considerable fortune. Dr. Colman had a delightful and charming manner, a keen sense of ethics and a profound knowledge of chemistry; he was always kindly disposed to juniors, a fact that many of us have cause to remember with gratitude. Chemistry and the gas industry have lost an outstanding personality and one whose name will be remembered with great esteem and affection. T. F. E. RHEAD

ISSN:0003-2654    

DOI:10.1039/AN9547900466

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

August, 19541 HUNTER AND GOLDSPINK 467 The Micro-determination of Bromide in Presence of Chloride BY G. HUNTER AND A. A. GOLDSPINK* (Presented at the meeting of the Society on Wednesday, May 5th, 1954) Conditions have been found for (a) quantitative conversion of micro amounts of bromide to bromate by hypochlorite and (b) quantitative formation of tetrabromorosaniline, as a basis for colorimetry, from bromine so formed by the interaction of the bromate formed in the above reaction with excess of added bromide. The colorimetric procedure is applicable in the range 0 to 5 p g of bromide in 5 ml of solution. In the same volume accurate measure- ments can be made of 0.1 p g of bromide, and of even smaller quantities by reducing the volume. The method is independent of the presence of chloride, although chloride has some effect on potential interference by chlorate and iodate.With a gross excess of chloride, about 200 equivalents of iodate or 6000 to 8000 equivalents of chlorate are necessary to cause appreciable inter- ference with 1 equivalent of bromide. The production of chlorate in relatively large amounts is, however, an integral part of the method, but experimental conditions are such as to avoid interference from it. The method is simple and readily adaptable to multiple ‘micro-determinations of bromide. THE method is based on several well known reactions. to bromate, in suitably buffered solution, by hypochloritel- Bromide is quantitatively oxidised Br- + 3C10- --f BrO,- + 3C1-. . .. .. .. .. . . .. (1) Br0,- + 5Br- + 6H++ 3Br, + 3H20 .. . . .. * . * - (2) .. .. .. - (3) Bromate quantitatively gives six equivalents of bromine by reaction in acid solution with an excess of bromide- Bromine in minute amounts readily substitutes in rosaniline decolourised by acid, with the formation of a red solution2 . . . . . . In this method reaction (2) is carried out in the presence of rosaniline, when the liberated bromine is found to react quantitatively. The method dispenses with the various oxidising agents, such as potassium dichromate and sulphuric acid,, chlorine: monopersulphuric acid,6 that have been used to generate bromine from bromide for the rosaniline reaction. As is well known, their action is difficult to regulate, and the presence of small amounts of chloride, leading to the production of chlorine, usually vitiates the rosaniline - bromine reaction.Although another reaction- . . . . . .. * * (4) BrO, + 6C1- + 6H+ -+ 3C1, + Br- + 3H20 .. also occurs, under the conditions of the present test reaction (2) takes precedence, even in the presence of a gross excess of chloride. The oxidising agent (BrO,-) is self-regulating as it uses itself up in the presence of an excess of bromide. Iodide, if present, is quantitatively oxidised to iodate in the course of reaction (1) and the reaction- also occurs, but this is much slower than (2) and can only interfere when iodate is greatly in excess of bromate. Reducing agents, such as metals in a low state of valency or sulphur dioxide, are also oxidised by hypochlorite, so that all of the bromate formed in reaction (1) is available for reactions (2) and (3) together.Presence of the chlorate ion is the most serious threat to the precision of the method. For, as is well known, when hypochlorite in slightly acid solution is heated, chlorate is formed by one molecule at the expense of the oxygen from two others- 10,- + 6Br- + 6H+ -+ 3Br2 + I- + 3H20 . . .. . . .. - * (5) C10- + 2ClO- --+ ClO,- + 2C1- . . .. .. .. .. .. * - (6) * Present address: Richard Thomas & Baldwins Ltd., “The Firs,” Whitchurch, nr. Aylesbury, Bucks.468 HUNTER AND GOLDSPINK : THE MICRO-DETERMINATION [Vol. 79 .. .. - * (7) Now the reaction- takes place in the presence of a high concentration of mineral acid at room temperature, and its velocity is increased by the presence of chloride.Reaction (2) requires for completion at least a 5 N concentration of mineral acid, but it is possible to reduce this to about 2-8 M by the use of molybdate as catalyst. This device greatly reduces reaction (7), which is not affected by the catalyst. The extent of chlorate production in the course of reaction (1) has been determined, as well as the effect of chlorate under the conditions of reaction (2). The procedure as finally developed will be described first and the evidence for its accuracy will follow. It is assumed that the material used for the test does not contain organic matter. C10,- + 6Br- + 6H+ -+ 3Br, + C1- + 3H,O . . * . METHOD REAGENTS- In a 1-litre, 2-necked Wolff bottle place 500 ml of 1-1 N sodium hydroxide and set it aside in a refrigerator to cool.Bromine-free chlorine is generated, for example, from 130 ml of analytically pure hydrochloric acid and 40 g of manganese dioxide by gently heating the mixture in a 500-ml flask fitted with a safety tube. It is washed by passing it through 100 ml of water in a smaller 3-necked Wolff bottle, and it is led into the bottle containing the cooled sodium hydroxide solution, the bottle being in a bath of ice. From time to time, 1 ml of this is removed to a small flask, 5 ml of 3 per cent. hydrogen peroxide are added to destroy the sodium hypochlorite, and the solutilon is titrated with 0.1 N hydrochloric acid, with alizarin red as indicator, until about 1.0 ml of the 0.1 N acid is required for neutralisation. The solution is stable indefinitely when stored in a refrigerator.Sodium formate, 50 per cent. w/v. Bu$er solution at pH 6.35-This is prepared by mixing 10 volumes of 40 per cent. w/v sodium dihydrogen phosphate dihydrate, NaH2P0,.2H20, 7 volumes of 2 N potassium hydroxide and 5 volumes of water. Rosaniline solution-Prepared by dissolving 6 mg of rosaniline in 100 ml of 2 N sulphuric acid. The use of either the base or an equivalent amount of acetate, sulphate or chloride has been found suitable. After prolonged exposure to light this solution ceases to react quantitatively with bromine, but it keeps for several weeks in a brown bottle. It should be tested periodically with standard bromate solution. Bromide - molybdate mixture-By dissolving 0.15 g of potassium bromide and 3.0 g of ammonium molybdate in water and diluting to 100 ml.Sulphuric acid, 14 N. tert.-Butanol containing 5per cent. v/v of absolute ethanod-The ethanol is added to prevent freezing at room temperature. Sodium hypochlorite, 1 N in 0.1 N sodium hydroxide-This is made as follows. PROCEDURE, STAGE I : CONVERSION OF BROMIDE TO BROMATE- To a test tube (16 mm x 130 mm) add 1 ml of buffer solution and then a suitable volume of bromide solution and make up with water to a total volume of 4-5 ml. To this add 0.25 ml of hypochlorite solution, mix the contents and immerse the tube in a bath of boiling water for 10 minutes. Add 0.25 ml of formate solution to destroy any excess of hypochlorite, mix the solution and replace the tube in the boiling water for a further 5 minutes. Cool the tube, add water to replace that lost (0.25 ml under our conditions) by evaporation, and mix the contents.This solution is used for the colorimetric procedure described in Stage 11. (a) For standard bromate solutions containing less than 5 pg of bromine as bromate--In a test tube, or colorimeter tube, place 0.1 ml of bromide - molybdate mixture, 0.1 ml of rosaniline solution per pg of bromine in the test solution to be used, 0.4 ml or less of water, 0.4 ml of 14 N sulphuric acid and 1 ml of the solution prepared in Stage I, in that order and mix them. The total volume at this stage is 2.0 ml, the amounts of bromide - molybdate and sulphuric acid being constant in all tests. Any volume of test solution less than 1.0ml can be used if the balance is made up with water, which is added before the test solution.Leave the reaction mixture at 20" to 30" C for 3 minutes, add 2 ml of tert.-butanol and 1 ml PROCEDURE, STAGE I1 : COLORIMETRY-August, 19541 OF BROMIDE I N PRESENCE OF CHLORIDE 469 of 14 N sulphuric acid. Mix the solution and measure the optical density in a suitable absorptiometer, after all air bubbles have disappeared. The colour is stable for many hours even in bright light. Maximum absorption by the bromorosaniline solutions is near 570 mp and all optical densities are here recorded a t this wavelength for solutions contained in l-cm cells. (b) For solutions with unknown bromine as bromate-In two test tubes place the solutions as in (a) on the assumption that 0.1 ml of test solution is to be used, and also that one test tube contains 0.05 ml and the other 0.50 ml of rosaniline.Now add 0.1 ml of the unknown solution to each test tube and continue as in (a). From the two observations one can deter- mine approximately the amount of bromine present, unless the amount taken for the determination exceeds about 250 pg. Further test portions are then taken with suitable amounts of rosaniline solution until the optical densities are found to be proportional to the test portions taken. When the liberated bromine is greatly in excess of the rosaniline necessary to absorb it, the readings will be low owing to the bleaching action of the bromine on any bromorosaniline that may be formed. RESULTS A linear relationship is found to exist between the amount of bromate present and the optical density of the bromorosaniline formed in solutions resulting from the procedure described. When the optical density is measured a t 570 mp in I-cm cells, the slope (pg of bromine in the volume of 5 ml divided by optical density) of the line that goes through the origin is 5-5.When a series of standard bromide solutions was treated with hypochlorite and formate, as described above, and suitable test portions were taken, the relationship between bromide as bromate present and the density of the resulting bromorosaniline solutions was likewise found to be linear and coincident with the curve found for the standard bromate solution. This conversion of bromide to bromate was shown to be precise from a series of 12 tests, in each of which 5 pg of bromide were used, and readings were taken for 1 ml of test solution.The extreme variation in readings was about 6 per cent. and the standard deviation 1-8 per cent. That the presence of bromide-free sodium chloride has no effect either on the conversion of bromide to bromate or on the production of a colour by rosaniline is shown by theresults of the six tests recorded in Table I. DISCUSSION AND CONCLUSIONS SODIUM HYPOCHLORITE SOLUTION- Those who have used the Van der Meulen reaction for determining bromide by titrating the iodine freed by the bromate from iodide have always found an appreciable blank.' This TABLE I OPTICAL DENSITY INDEPENDENT OF THE AMOUNT OF CHLORIDE PRESENT Six tests were carried out as described under Procedure, Stages I and 11, each tube containing 5 pg of bromide and the amounts of sodium chloride shown Sodium chloride in 5 ml, Stage I, mg .. 0 62-5 125 250 500 750 Sodium chloride in 1 ml, Stage 11, mg . . 0 12.5 25 50 100 150 Optical density x 100 . . .. . . 17.9 18.2 18.5 17.6 18.0 18.2 has commonly been attributed to the presence, either before or after a period of heating, of chlorate in the reagent. We tested several of our reagents prepared as described above, including one that had been stored in a refrigerator for over 30 months, and found that chlorate was absent. Foerster and Dolch6 have noted the very slow change of hypochlorite to chlorate in 0.1 N sodium hydroxide. Nevertheless several such reagents had appreciable blanks by the titration method. The use of the colorimetric method has disclosed that the source of those blanks in hypochlorite solutions is not chlorate, but bromate.We have prepared a bromide-free hypochlorite solution from chlorine made by the action of sulphuric acid on sodium chloride freed from bromide by recrystallisation from methanol. This solution showed no titration blank. We have determined the bromine in several commercial bleaches diluted to 1.0 N with respect to hypochlorite, by a titration method and by the present colorimetric method. If470 HUNTER AND GOLDSPINK : +THE MICRO-DETERMINATION [Vol. 79 bromine were introduced into a hypochlorite solution it would soon be changed to bromate, so that, by destroying the hypochlorite present with a formate, without any previous heating, the same value should be found as from a sample heated for 10 minutes on a bath of boiling water before addition of the formate.This is seen to be so from the observations shown in Table 11. TABLE 1.1 PRESENCE OF BROMATE IN COMMERCIAL HYPOCHLORITE SOLUTIONS Time of heating before formate added, Sample minutes 1 10 2 0 2 10 3 0 3 10 Bromine found (as mg per 100 ml a t a N sodium hypochlorite concentration Titration method Colorimetric method 7 A \ 48-0 17.7 17.6 15.4 15.5 Clearly such hypochlorite solutions are too grossly contaminated with bromate to be of use as analytical reagents. Much better comrnercial samples have been found,' but such material has proved to be unreliable. The bromate in commercial bleaches is apparently dependent on the bromine content of the chlorine from which such hypochlorite solutions are made. With some A.R.hydrochloric acids it is, however, possible to make reagents, as described above, with quite low blanks. Our current reagent contains about 0.1 mg of bromine per 100 ml of solution. In the present method a 1-ml test portion would contain 0.05 pg of bromine, which gives a just perceptible blank, but is negligible for most analytical purposes. CHLORATE- It has been noted that chlorate is formed from slightly acid hypochlorite solutions according to reaction (6). In the Van der Meu1t:n conversion of bromide to bromate under the conditions described by Hunter,' we have now found that some 60 equivalents of chlorate are produced for each equivalent of bromide converted to bromate. Indeed, the Van der Meulen principle can be applied despite the formation of clorate, because the reaction- is negligible in extent compared with the reaction- under the usual conditions of titration with thiosulphate.8 As will be shown, sulphuric acid more concentrated than 6 N is necessary to suppress the colour of free rosaniline before colorimetry. In earlier attempts to develop the method we used such an acidity for reactions (2) and (3), and, as might be expected from a considerationof Figs. 1 and 2, we found that an appreciable colouir was produced by amounts of chlorate neces- sarily formed in the process of conversion of bromide to bromate. It is clear from Fig.1 that the amount of bromine liberated by the action of chlorate on bromide is greatly decreased at the lower normalities, and that the molybdate has no effect. From Fig.2 it is seen that a t a 3 N concentration of acid, an increasing concentration of chloride enhances the oxidising action of chlorate on bromide. As will be clear later, our test solutions contain less than 0.30 mg of chlorate as potassium chlorate, and this amount has no perceptible effect in the presence of less than 50 mg of sodium chloride and only a doubtful effect in the presence of 100mg of sodium chloride. Hence it is apparfent that the amount of chlorate formed in Stage I of the procedure will have no effect 011 the colorimetric procedure described. It is of less interest, as its formation is not, unlike chlorate, inherent in the method. Iodate is more active than chlorate in freeing bromine from bromide, but, even in the presence of 50mg of sodium chloride, 27 pg of potassium iodate produces only about as much colour as 0.05 pg of bromine as bromate.That is, on an equivalent basis, bromate is 200 times more sensitive than iodate. THE CONVERSION OF BROMIDE TO BROMATE- of the following observations, C10,- + 61- + 6H+ -+ 31, + C1- + 3H,O Br0,- + 61- + 6H+ -+ 31, + Br- + 3H,O The effect of iodate is also indicated in Figs. 11 and 2. The conditions described under Procedure, Stage I, have been arrived a t on the basisAugust, 19541 OF BROMIDE IN PRESENCE OF CHLORIDE 47 1 To avoid unnecessary chlorate formation the amount of hypochlorite used should be kept ,at a minimum. The amount used, 0.25 ml, is shown to be adequate from the results in Table 111. These results were obtained by carrying out the Procedure, Stage I, but cooling the tubes after 10 minutes in the bath of boiling water.Then 0.25 ml was removed from each tube, acidified, iodide was added and it was titrated with 0.005 N sodium thio- sulphate. In blanks, without heating, 0.25 ml of hypochlorite = 2.46 ml of the thiosulphate. The bromate present does not significantly affect the titration for hypochlorite, but it catalyses to some extent chlorate formation. It is of interest that upwards of 40 per cent. of the hypo- chlorite is oxidised even in the absence of bromide and less than 60 per cent. in the presence of 200 pg of bromide. That such amounts of bromide are converted to bromate may be seen from Fig. 3, although the method is not designed for the larger amounts, in whose presence great dilution of the test solution becomes necessary before absorptiometry.It can further be seen from Fig. 3 that a minimum of 6 minutes' heating is necessary for the conversion of 10 pg of bromide, while amounts of bromide up to 200 pg are converted after 12 minutes' heating. From many other observations, we have decided that 10 minutes in the bath of boiling water is enough for the amounts of bromide for which the method is suitable, Normality of sulphuric acid Fig. 1. The effect of concentration of acid //- /.:;- 3 0 I $A0 Sodium chloside, mp Fig. 2. The effect of concentration of on thewaction of chlorate on bromide and of chloride on the action of chlorate on bromide iodate on bromide and of iodate on bromide in 3 N sulphuric acid. -0 -0 -, chlorate Curve A, 0.25 mg; curve B, 0.50 mg; curve C, --, chlorate in presence of molybdate 1.0 mg; and curve D, 2.0 mg of potassium X, iodate.chlorate. Curve E, 0-027 mg; and curve F, Curve A, 1.0 mg and curve B, 10.0 mg of 0.13 mg of potassium iodate. potassium chlorate. The amounts of iodate a t the 3 N concentration of sulphuric acid are point C, 0.7 mg; point D, 1-4 mg; and point E, 2.7 mg of potassium iodate By the use of a series of buffers, a value of pH of 6-35 has been chosen on the evidence, shown in Fig. 4 (a), that the rate of conversion of bromide to bromate is a t a maximum near this point. It may be noted that complete conversion occurs after heating for 6 minutes at this pH, but that conversion is not complete after 3 minutes heating. With a similar series of buffers, 0.25 ml of hypochlorite'was treated as under Procedure, Stage I, and then the excess of hypochlorite was destroyed with formate.From each tube, 2-ml test portions were taken and titrated for chlorate by reduction with a ferrous salt and back titration with 0.1 N potassium dichromate, with diphenylamine as indicator (see Cumming and Kayg). As might be expected, the rate of chlorate formation is roughly parallel t . that of bromate formation, but the amount of chlorate formed is much greater, about 1.35 mg as potassium chlorate. From such a test solution a 1-ml test portion thus contains 0.27 mg of potassium chlorat e. The proportion of chlorate to bromate formed under the above conditions is higher than it need be, with complete conversion of the bromide. For example, we have found that 10 pg of bromide is completely converted to bromate in the above buffer - hypochlorite472 HUNTER AND GOLDSPINK : THE MICRO-DETERMINATION [Vol. 79 mixture in about 36 hours at 4" C, with the formation of only about 0.30 mg of potassium chlorate.At room temperature in the same time, about twice as much chlorate is formed. There might be circumstances calling for such treatment, but, as 0030mg of chlorate as potassium chlorate has no effect on the absorptiometric method, the saving of time would seem to justify the method adopted of heating far 10 minutes in the bath of boiling water. The capacity of the buffer is sufficient to render unnecessary the neutralisation of the bromide solutions to be analysed unless they contain a gross excess of acid or base.The volume of the solution at Stage I allows for a volume of up to 3.5 ml of solution for bromide determination. The total volume of the resulting bromate (test) solution is 5 ml, which allows for many trial tests. Indeed, all volumes can often be reduced by half and still yield adequate material for the determination. 5 50 I I I I W 2 4 6 8 1 0 1 lime, minutcs PH Fig. 3. Time of heating required for the con- version of bromide to bromate by hypochlorite. The amounts of bromide taken for Stage I are for curve A, 10 pg; curve B, 25 pg; curve C, 100 pg; and curve D, 200 P.lg chlorate from hypochlorite Fig. 4 (a). The effect of pH and time of heating on the rate of conversion of 10 pg of bromide to bromate by hypochlorite. (b) The effect of pH on the formation of THE FORMATION OF BROMINE FROM BROMATE AND BROMIDE- According to EphraimlO this reaction is quadrimolecular and both hypobromous and bromous acid are probably among the intermediate entities.For the complete stoicheio- metrical change shown in reaction (2), a high normality of a mineral acid is necessary. Schirlowl2 observed that the related reaction- BrO,- + 61- + 6H+ -+ 31, + Br- + 3H20 is catalysed especially by solutions of iron, Chromium or molybdenum. We have found the colourless molybdate ion to be very effective as a catalyst and, as shown in Fig. 5, its presence enables the reaction to go to completion in 2.8 N acid at a temperature of about 25" C within three minutes. In the absence of the catalyst the optical density is not a t a maximum until the acid concentration is greater than 5 N.The importance of thus lowering the normality has already been noted in relation to the potential interference by chlorate. ROSANILINE AND BROMINE- Guareschi, observed that bromine vap0u.r gives a red colour with Schiff's reagent (rosaniline decolourised with sulphur dioxide), and DenigW3 observed that rosaniline solutions are also decolourised by an excess of mineral acid, thus obviating the reducing action of sulphur dioxide on the halogens. A concentration of about 6 N sulphuric acid is necessary for maximum suppression of this colour, as shown in Table IV. Here the amount of 10 pg of rosaniline is somewhat greater than the sma.11 excess of rosaniline necessarily present for the procedure, but it serves to show that a considerable error might arise from the use of low normalities of acid.At the normality of 6.5, the bromo derivative retains its colour, and its optical density may, therefore, be accurately measured without interference by a small excess of rosaniline.August, 19541 OF BROMIDE I N PRESENCE OF CHLORIDE 473 Fig. 6 shows the densities of the colours obtained with increasing amounts of rosaniline at levels of 1, 2, 3 and 4pg of bromine as bromate in the test solutions. It is seen that about 6 pg of rosaniline per pg of bromine are necessary for maximum colour development. When there is gross deficiency of rosaniline, colour production is greatly hindered and, when there is gross excess of rosaniline, there is slightly depressed colour density. 0 1 2 3 4 5 6 Normality of sulphuric acid Fig.5. The effect of concentration of acid, with and without molybdate, on the action of bromate on bromide. - 0 4 - without molybdate. --@+ with molybdate Rosaniline used per,ug of bromine, ug Fig. 6. The amount of rosaniline required for maximum optical density with a given amount of bromate. Curve A, 1 pg; curve B, 2 pg; curve C, 3pg; and curve D, 4 pg of bromine as bromate The amount of rosaniline necessary, as indicated by Fig. 6, is precisely twice that required on Guareschi's evidence of the formation of tetrabromorosaniline. TurneId suggests that a pentabromorosaniline is formed. We have prepared crystalline material by adding bromine to rosaniline in 4 N acetic acid at room temperature, drying the product and recrystallising it from hot methanol.Analysis of two preparations by Weiler and Straws of Oxford proved the substance to be tetrabromorosaniline acetate. (Calculated for C2,H,,02N3Br,': C, 38-91 per cent.; H, 2-83 per cent.; Br, 47.23 per cent. Found for preparation (a): C, 38.83 per cent.; H, 2.95 per cent.; Br, 47.4 per cent.; and found for preparation ( b ) : C, 39.21 per cent.; H, 3.12 per cent.; Br, 47.8 per cent.) Neither chlorine nor iodine substitutes in rosaniline as bromine does, but chlorine alters it to a brown water-soluble material similar to that produced by the action of light. By careful regulation, chlorine may be used to free bromine from bromide, but if there is more TABLE I11 UTILISATION OF HYPOCHLORITE IN CONVERSION OF BROMIDE TO BROMATE Bromide added, pg .. 0 50 100 150 200 Hypochlorite used, yo . . 35-8 37.0 41.5 47-0 53.3 Thiosulphate used, ml . . 1.58 1.55 1.44 1-30 1.15 TABLE IV EFFECT OF INCREASING NORMALITY OF SULPEIURIC ACID IN SUPPRESSING THE COLOUR OF ROSANIL~NE Volume of solution 5 ml, as in Stage 11. Present: 10 pg of rosaniline Normality . . . . 0.47 0.94 1-4 1.86 2.3 3.0 4.9 5.8 7.0 Optical density x 100 9.4 7.6 5.8 4.1 3.5 2-3 1.6 1.0 0-6474 HUNTER AND GOLDSPINK THE MICRO-DETERMINATIOS [Vol. 79 present in solution than required for this action, the chlorine attacks the rosaniline and bromo- rosaniline and nullifies the method. The presence of chloride greatly increases this action of chlorine, presumably through increasing its oxidation potential.ll A similar bleaching action of bromine will be found if the amount of bromine liberated in reaction (2) exceeds the amount of rosaniline needed to absorb it.This point has been noted in the Procedure, Stage 11. To get valid absorptiometric readings a slight excess of rosaniline must be present in the solution before the bromine is produced by reaction (2), for rather less colour is produced if the rosaniline is added after the test solution containing bromate. This finding would appear to be explained on the assumption that free bromine dissociates- Br, + Br+ + Br- The presence of chloride appears to catalyse reaction (2) or (3) or both, although the colour density is not affected. It is clear that some chloride is necessarily present in the test portion. It may be readily calculated that a l-ml test portion contains about 3.0mg of chloride as sodium chloride, arising from the corresponding hypoc’hlorite solution.ACTION OF teYt.-BUTANOL- Immiscible solvents, such as chloroform, pentanol- and benzyl alcohol, have been used by other workers to extract the bromorosaniline formed. A solvent is necessary, because the bromorosaniline is insoluble in the acid aqueous medium in which it is formed; indeed, there is little colour perceptible in the solution until after addition of a solvent. The addition of such solvents also suspends further oxidative changes in the solution. But we have found it more convenient to use tert.-butanol, as this is miscible with the acid aqueous medium and no separation of layers is necessary for absorpticlmetry. In the proportion stated the colour intensity is approximately the same as after extraction in 5 ml of benzyl alcohol.REFERENCE s 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Van der Meulen, J. H., Chem. Weekbl., 1931, 28, 82. Guareschi, I., 2. Anal. Chem., 1913, 52, 451. Denigbs, G., and Chelle, L., Compt. Rend., 1912, 155, 1010. Indovina, R., Biochem. Z., 1934, 275, 286. Turner, W. J., Ind. Eng. Chem., Anal. Ed., 1942, 14, 599. Foerster, F., and Dolch, P., 2. Elektrochem., 1917, 23, 137. Hunter, G., Biochem. J., 1953, 54, 42. Szabo, Z., 2. Anal. Chem., 1931, 84, 24. Cumming, A. C., and Kay, S. A., “Quantitative Chemical Analysis,” Gurney and Jackson, London, Ephraim, F., “Inorganic Chemistry,” Gurney and Jackson, London, 1948, p. 386. Schirlow, N., 2. phys. Chem., 1898, 27, 513. Denighs, G., Compt.Rend., 1912, 155, 721. 1948, p. 141. -, oP. cit., p. 168. LECTRO-MEDICAL RESEARCH UNIT MEDICAL RESEARCH COUNCIL STOKE MANDEVILLE HOSPITAL AYLESBURY, BUCKS. December 16th, 1953 DISCVSSION DR. J. HASLAM said that he was interested in the authors’ observations on the oxidation of bromide by persulphate. He thought that a similar mechanism (existed with other oxidising agents, such as chromic acid ; they would oxidise bromide to bromine quantitatively, but would not oxidise chlorides directly. If the bromide and chloride were mixed, however, oxidation both of the bromide and of some chloride took place. He confirmed Dr. Hunter’s remarks about the presence of bromine compounds in commercial hypo- chlorite. It was, however, a fact that not all the hypochlorite produced contained the same amount of bromine. Years ago, when working in this field, he used to order bromine-free hypochlorite. With that material the blank in the hypochlorite oxidation was reduced to a very small figure. He was of the opinion that the significant feature of the authors’ method was the fact that i t tolerated appreciable proportions of iodine; that would certainly be of value in industrial work. He asked whether the authors had had any experience of the application of their method to the determination of small amounts of bromide in iodides. DR. R. F. MILTON asked whether the presence of reducing substances likely to occur in biological fluids would cause an uptake of bromine a t the liberation stage and thereby affect the result.August , 19641 OF BROMIDE I N PRESENCE OF CHLORIDE 475 MISS M. CORNER said that this paper seemed to her to be the answer to a maiden’s prayer. She was working on the problem of determining small quantities of fluorine as lead fluorobromide (provided it was sufficiently insoluble for use on the microgram scale) and this would be the method she would chose for determining the bromide. DR. HUNTER, in reply to Dr. Haslam, said they had had no experience in the application of the method to bromide in iodides. In reply to Dr. Milton, Dr. Hunter stated that the method had been used with success on ashed biological fluids, and reducing materials in the ash were fully oxidised by the excess of hypochlorite in Stage I. In reply to Miss Corner, Dr. Hunter said that lead had not been tried as a catalyst, but in the presence of large amounts of chloride it appeared that molybdenum could almost be dispensed with. She asked whether lead would act as a catalyst instead of molybdenum. I

ISSN:0003-2654    

DOI:10.1039/AN9547900467

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

August , 19641 OF BROMIDE I N PRESENCE OF CHLORIDE 475 A Volumetric Procedure for the Determination of Zirconium in its Binary Alloys with Uranium BY G. W. C. MILNER AND P. J. PHENNAH Zirconium in acid solution can be readily determined by a volumetric procedure that involves the addition of an excess of a standard solution of ethylenediaminetetra-acetic acid to form a complex with the zirconium followed by the titration of unused EDTA with a standard iron solution and salicylic acid as the indicator. Under these conditions it is possible to determine up to at least 100 mg of zirconium with an accuracy of about f 1 per cent. This procedure is also applicable after preliminary separation of zir- conium from other elements with organic precipitants. This titration has proved of advantage in the rapid analysis of uranium - zirconium binary alloys after the preliminary separation of the zirconium as the insoluble mandelate €rom a perchloric acid solution of the alloy.FOR the analysis of binary alloys of zirconium with uranium, Milner and Skewiesl recently reported a gravimetric procedure that consisted in separating the zirconium from most of the uranium by extracting it, as its cupferron complex, into chloroform. After the evapora- tion and ignition of the organic matter, the residue was re-dissolved and the zirconium was finally separated from small amounts of co-extracted uranium by precipitation as the insoluble mandelate. The zirconium mandelate precipitate was ignited to the oxide, and the zirconium content of the alloys was determined from the final weight of zirconium oxide.Although this procedure gave very reliable results for zirconium, it proved rather time-consuming because of the evaporations and ignitions that were needed to remove organic matter. Investigations were started to develop a more rapid procedure for the determination of zirconium in its binary alloys with uranium. The main requirements included a quick separation procedure that was specific for the separation of zirconium from uranium, and also a rapid technique for the determination of the final zirconium concentration that avoided the ignition to zirconium oxide. Consideration was first given to the possibility of applying a physico-chemical technique to the rapid determination of macro amounts of zirconium.Mayer and Bradshaw2 recently made use of the coloured lake that is formed by zirconium with alizarin in a direct absorptio- metric procedure for the determination of small amounts of this element in magnesium-base alloys. This method was possible because the alloys contained only small amounts of zirconium and they were readily dissolved by hydrochloric acid. In contrast, some of the 'uranium - zirconium binary alloys contained substantial amounts of zirconium and this type of material could only be completely dissolved by the use of hydrofluoric acid. However, a procedure that is based on the formation of a coloured lake was not considered suitable for the determination of a main constituent of alloys and consideration was therefore given to other physico-chemical techniques.The polarographic technique seemed to be completely unsuitable for the determination of zirconium, as the only step so far reported3 for this element occurs after the discharge of hydrogen from a 0-1 N potassium chloride solution of pH 3. The half-wave potential of the step is -1.65 volts against the saturated calomel electrode. Kolthoff and Johnson4 have,476 MILNER AND PHENNAH: A VOLUMETRIC PROCEDURE [Vol. 79 however, recently reported an amperometric titration procedure for the determination of milligram amounts of zirconium. The titrating reagent is m-nitrophenylarsonic acid, which precipitates the zirconium from its solution as the insoluble m-nitrophenylarsonate. A voltage of -0.5 volt against the S.C.E. is applied across the electrode system and after the titration end-point, a diffusion current results from the reduction of the nitro group of the excess of titrant at the dropping-mercury electrode.The graph for the determination of the position of the end-point is therefore of the reversed L-type. According to these workers the titration is best carried out in a 1.5 N hydrochloric acid medium containing about 20 per cent. of ethanol. This technique did not seem applicable to the direct titration of zirconium in the presence of greater amounts of uranium, as this latter element is reported by Kolthoff and Harris5 to give a reduction step with an E, value of about -0.18 volt against the S.C.E. from 0.01 to 6 N hydrochloric acid solutions. The uranium diffusion current could be expected, therefore, to seriously interfere with the measurement of the current from the excess of m-nitrophenylarsonic acid.Ethylenediaminetetra-acetic acid has recently been used for the volumetric determination of certain elements , including magnesium ,6 calcium,G zinc,6 aluminium' J and t h ~ r i u m . ~ Milner and Woodhead's procedure' for the determination of aluminium consists in first adding a small excess of a standard EDTA solution to t'he aluminium solution, which is adjusted to pH 6.5. The solution is then heated to form the aluminium - EDTA complex and, after cooling, the unused EDTA is titrated with a staindard ferric iron solution with salicylic acid as the indicator. Cabelllo recently reported the existence of a complex that is formed by zirconium with EDTA. He observed that whereas zirconium hydroxide is normally precipi- tated from solutions of pH 2, the presence of E.DTA prevented this precipitation until pH values of about 7 were attained.It was evident from these observations that the zirconium - EDTA complex existed at pH values of less than 7 and that this behaviour can be a basis for the development of a volumetric method for zirconium. EXPERIME~NTAL DEVELOPMENT OF A RAPID METHOD FOR THE DETERMINATION OF ZIRCOXIUM- In preliminary experiments the aluminium - EDTA volumetric procedure was applied directly to suitable aliquots of a zirconium solution that was prepared by dissolving 6.362 g of Specpure zirconium chloride, ZrOC1,.8H20, in a total volume of 500 ml of distilled water that contained a small amount of hydrochloric acid to prevent hydrolysis.The standardisation of 20-ml aliquots of this solution, which was done by precipitation of the zirconium with TABLE I THE EFFECT OF pH ON THE RECOVEIRY OF ZIRCONIUM BY THE EDTA TITRATION PF:OCEDURE pH before titration Zr recovered , Recovery, mg % 2.2 73.5 99.9 4-0 73.7 100.1 5.0 73.7 100.1 6.0 73.7 100.1 ammonium hydroxide followed by filtration and ignition of the zirconium hydroxide to zirconium oxide, gave an average result of 73.6 ing of zirconium. Assuming that 1 molecule of zirconium reacted with I molecule of EDTA, the zirconium content of a 20-ml aliquot of the zirconium solution was found to be 73*9mg, by the EDTA volumetric technique. The good agreement between these results justified a Euller investigation of the titration procedure.As the pH of the solution was quite important for the aluminium titration, the effect of pH was first studied for the zirconium titration so that the best conditions for the formation of the zirconium - EDTA complex could be selected. Further 20-ml portions of the standard zirconium solution were taken and they were adjusted to suitable pH values in the required range by means of ammonium acetate - acetic acid buffers; the volume of solution containing an excess of 0.1 M EDTA just before boiling was about 150 to 200 ml. After cooling, the amount of unused EDTA remaining in the solution was determined by titration with a standard 0.1 M ferric iron solution. The results for zirconium by this technique are shown in Table I and they indicate that the recoveries can be good over a wide range of pH values.August, 19541 FOR THE DETERMINATION OF ZIRCONIUM 477 At values below pH 3, the titration end-point appeared to be sluggish, whereas at pH values of from 4 to 6.5 an excess of one drop of the iron solution was sufficient to develop the ferric salicylate colour rapidly at the titration end-point.The pH range of 5 to 6 was finally selected as the optimum range for this titration. REAGENTS- Standard iron solution, 0.1 M-Dissolve 5685 g of Specpure iron in 20 ml of hot hydro- chloric acid, sp.gr. 1.16, and then oxidise the iron by boiling the solution after the addition of a few ml of nitric acid, sp.gr. 1.42. Dilute the solution to 1 litre with water. Suitably dilute an aliquot of this solution to give a 0.02 M solution.Ethylenediaminetetra-acetic acid solution, 0.1 M-Dissolve 37.23 g of the disodium salt of ethylenediaminetetra-acetic acid in water and dilute it to 1 litre. Determine the exact molarity of this solution by the following standardisation procedure. Accurately transfer 20-ml aliquots of the EDTA solution to 400-ml conical flasks. Then to each flask add 3 g of ammonium acetate followed by about 150ml of water. Dissolve 0-2g of salicylic acid in each solution and then titrate with the standard iron solution until the colour of the solution just changes to brown owing to the formation of ferric salicylate. Suitably dilute an aliquot of this solution to give a 0-02 M solution. PROCEDURE- To a solution of zirconium in hydrochloric acid, add sufficient EDTA solution to form a complex with the zirconium and to leave a slight excess of reagent.Then to the solution, add a few drops of cresol red indicator followed by ammonium hydroxide, sp.gr. 0.88, added dropwise, to develop the yellow colour of the indicator. Dissolve 3 g of ammonium acetate in the solution, cool it, and add either ammonium hydroxide or acetic acid to adjust the pH of the solution to between 5 and 6, as shown by a direct-reading pH meter. Dilute the solution with water to give a final volume of about 200 ml. Heat the solution to the boiling- point, continue boiling it for about 2 minutes and then cool it to room temperature. Dissolve 0.2 g of salicylic acid in the solution and, with the standard iron solution, titrate the EDTA that is in excess of the amount required to form a complex with the zirconium.For solutions containing less than 30 mg of zirconium, use a 0.02 M EDTA solution and a 0.02 M iron solution for the titration; for solutions containing more than 30 mg of zirconium use the 0.1 M reagents for the titration. Calculate the weight of zirconium present as follows. Let the volume of EDTA solution added to the zirconium be V , ml of M , molar solution. Suppose that V , ml of M , molar ferric iron solution are needed for the back-titration. This is equivalent to V , x M,/M, ml of a M, molar solution of EDTA. Therefore the volume of EDTA solution needed to form a complex with the zirconium As the atomic weight of zirconium is 91.22, the weight of zirconium present METHOD = (V,- V z x Mz/Ml) ml of a MI molar solution.= 91-22 x M I x (V,-V, x IM,/M,) mg. RESULTS By the procedure outlined above, with a standard zirconyl chloride solution, the results are those shown in Table 11, and they indicate that the method is at least capable of deter- mining amounts of zirconium up to about 100 mg with an accuracy of better than 1 per cent. TABLE I1 DETERMINATION OF ZIRCONIUM BY THE EDTA TITRATION PROCEDURE Weight of zirconium taken, mg 2.42 4.85 9.69 19.28 29.07 50.5 73.6 99.6 Weight of zirconium recovered, mg 2.41 4.81 9-64 19.25 28.90 50.8 73.9 100.0 Error, % - 0.4 - 0.8 - 0.5 -0.17 - 0.6 + 0.6 + 0.4 + 0-4478 MILKER AND PHENNAH : A VOLUMETRIC PROCEDURE Pol. 79 Further titrations were carried out in the presence of sulphuric acid to study the recovery of zirconium under these solution conditions.The procedure consisted of taking suitable aliquots of the standardised zirconyl chloride solution and, after the addition of 25ml of a 25 per cent. v/v solution of sulphuric acid, evaporating each solution to fumes of the acid. On completing the EDTA titration procedure as above, the zirconium recoveries are those shown in Table 111, and it can be seen that small amounts of sulphuric acid do not interfere with this determination. In addition, the results of titrations in the presence of larger amounts of sulphuric acid, up to a maximum of 25 ml of concentrated acid in the zirconium solution, showed that it did not interfere. TABLE 1 I1 DETERMINATION OF ZIRCONIUM BY THE EDTA TITRATION PROCEDURE IN THE PRESENCE OF SULPHURIC ACID Weight of zirconium taken, mg 2-42 4.84 9.68 24.22 48.0 96.6 Weight of zirconium recovered, Error, mg Yo 2.51 + 1.8 4.88 + 0.8 9.69 SO.1 24.30 + 0.4 4 8.0 Nil 96.7 3 0 . 1 ISVESTIGATION OF METHODS FOR THE SEPARATION OF ZIRCONIUM FROM OTHER ELEMENTS BEFORE ITS TITRATION WITH EDTA Several organic acids, or their ammonium salts, have been used during the past few years for the precipitation of zirconium from solution to effect the separation of this element from other elements.These reagents include m-nitrobenzoic acid,ll ammonium benzoate,l2 fumaric acid,13 phthalic acid,l* mandelic acid15 >I6 and m-cresoxyacetic acid.17 With all these reagents, the zirconium precipitate is ignited to zirconium oxide after filtration and the zirconium content is calculated from the weight of the oxide. On attempting to weigh zirconium mandelate precipitates directly after suitable drying, Oesper and Klingen- berg18 encountered difficulties that were caused by the use of large amounts of mandelic acid in the precipitation stage. After filtration, it proved very difficult to remove mandelic acid from the precipitate without redissolving some of the zirconium mandelate.After a study of various substituted mandelic acids, the above workers recommended the use of p-bromo- mandelic acid for the direct determination of zirconium. The precipitate that is produced by this reagent can be washed with water to free it from excess of precipitant and, after drying at 100" C, it can be weighed. Belcher, Sykes and Tatlow,lS however, recently failed to substantiate the above findings and the results were reliable with this reagent only after igniting the zirconium 9-bromomandelate precipitate to zirconium oxide.In addition, recent methods that use p-bromomandelic acid for the separation of zirconium in stee120 and for the analysis of aluminium alloys21 involve the ignition of the zirconium precipitate to the oxide. Qualitative tests showed that many of the above complexes of zirconium with organic acids were soluble in strong mineral acids and that this behaviour presented the possibility of completing the zirconium determination voliimetrically with EDTA, providing that no difficulties arose from the presence of organic molecules in solution. Ammonium benzoate was the first reagent to be investigated, as previous work7 had shown that small amounts of benzoic acid did not interfere in EDTA titrations.Jewsbury and Osbornl2 reported that this reagent gave complete recovery of zirconium from hydrochloric and nitric acid solutions containing ammonium acetate starting from pH 1.0 to 1.5, whereas from sulphuric acid solutions the starting pH is given as from 2.0 to 2.5. In all the work the zirconium benzoate precipitates were ignited to zirconium oxide, which was then weighed. These results are in agreement with the complexing action of sulphuric acid with zirconium as compared with the weak or non-complexing powers of the other acids. These workers make no reference to the recovery of zirconium from perchloric acid solutions. Jewsbury and Osborn's precipitation procedure was applied directly to pure solutions of zirconium in hydrochloric, nitric, sulphuric and perchloric acids.Each acid was investi- gated in turn by taking aliquots of a standard xirconium solution in the requisite acid, and the effects of pH on the recovery of zirconium were studied. After digestion, the benzoate precipitates were filtered through Whatman No. 40 filter-papers, and then they were washedAugust, 19541 FOR THE DETERMINATION OF ZIRCONIUM 479 back into the precipitation beakers with water. Each precipitate was then dissolved in a 20-ml portion of hot hydrochloric acid, sp.gr. 1.18, first poured over the filter-paper, which was then washed with water. The zirconium content of these solutions was determined by the EDTA titration procedure and graphs were plotted showing the variations of the recovery of zirconium against pH.The results for the four acids are shown in Fig. 1 and these graphs agree reasonably well with Jewsbury and Osborn’s statements for the recovery of zirconium from the first three acids. The precipitation of zirconium from perchloric acid solutions proved very interesting from the analytical point of view as its behaviour was similar to that in the weakly complexing acids (hydrochloric acid and nitric acid). Moreover, as the pH values for the complete recovery of zirconium from perchloric acid solutions are 80 Y Key - -- 5 e 1 I - HNO,I 5 4 3 PH Fig. 1. Recovery of zirconium by precipitation with ammonium benzoate lower than the values for sulphuric acid solutions, the former acid should be of greater value for separating zirconium from other elements; the chances of contamination of the zirconium benzoate precipitate with other elements should be less from perchloric acid solutions than from sulphuric acid solutions.The precipitation behaviour from perchloric and sulphuric acids was more important, because hydrofluoric acid is often needed to complete the solution of many alloys and this reagent must then be removed by evaporating the alloy solution to fumes of either perchloric or sulphuric acid, before carrying out the analysis. Unfor- tunately, certain other elements12 are also precipitated by ammonium benzoate either com- pletely or partially under the conditions necessary for the complete recovery of zirconium ; the investigation was therefore extended to inchde a study of other possible precipitants. Derivatives of benzoic acid with higher dissociation constants were briefly investigated in the hope that they would give complete recovery of the zirconium from solutions of higher acidity than those necessary for benzoic acid. Toluic acid, with a dissociation constant roughly twice that of benzoic acid, appeared to offer very little advantage, nor did m-nitro- benzoic acid according to the results of Osborn’s investigationsll.o-Chloro- and o-nitrobenzoic acids, with dissociation constants roughly lo3 times that of benzoic acid, quantitatively precipitated zirconium from 0.2 to 0-5 N hydrochloric acid solutions. These acids are, however, very insoluble and filtration of the zirconium precipitates proved very difficult owing to the excess of reagent crystallising out of solution.9-Hydroxybenzoic acid, gallic acid, anthranilic acid and vanillic acid failed to precipitate zirconium from 0.1 N hydrochloric acid solutions and they are therefore inferior to unsubstituted benzoic acid.480 MILNER AND PHENNAH: A VOLUMETRIC PROCEDURE [Vol. 79 m-Cresoxyacetic acid1' has been recommended as a reagent for the precipitation of zirconium ; the precipitation is complete from 0.20 to 0.25 N hydrochloric acid solutions. However, on dissolving this type of precipitate in hydrochloric acid before the application of the EDTA titration of the zirconium, red ciresol products, presumably from the break- down of the reagent, were formed. This effect masked the end-point of the iron back-titration and the use of this reagent was therefore not investigated further.Phthalic acid,l* fumaric acid13 and mandelic a ~ i d l ~ ? ~ * have also been recommended as precipitants for the separation of zirconium. Phthalic acid was not studied, however, because a 2-hour digestion period 0.2 0.4 I %-%-- H,SO, ,*- HCIO, 8--8-- HCI -I I 0.8 l i J I 'L I Normality Fig. 2. Recovery of zirconium by precipitation with fumaric acid is needed to give the complete precipitation of the zirconium. In the investigation of the other two acids, some difficulty was initially experienced in the re-solution of the zirconium fumarate and zirconium mandelate precipitates in hydrochloric acid, before the EDTA titration. However, both of these precipitates were found to dissolve in boiling dilute sulphuric acid and this allowed the EDTA titration to be applied, as the results of previous I 2 3 4 5 j \ Normality Fig.3. Recovery of zirconium by precipitation with mandelic acid work (see p. 478) showed that some sulphuric acid can be present in the titration procedure without affecting the zirconium results. On re-dissolving the fumarate precipitates from 50-mg amounts of zirconium in 25-ml portions of 25 per cent. sulphuric acid before titration with EDTA, the results are those shown in Fig. 2, for the recovery of zirconium from hydrochloric, sulphuric and perchloric acid solutions. Venkakaramaniah and Rao13 reported the complete recovery of zirconium from solutions up to 0-35 N in hydrochloric acid. This is confirmed in Fig.2 and, in addition, the recovery is seen to be essentially complete from slightly more acid solutions. As with benzoic acid, the recoveries from perchloric acid solutions proved to be very similar to those from hydrochloric acid solutions and recoveries from sulphuric acid solutions were definitely inferior. According to Kumins' work, l5 zirconium is quantitatively precipitated by mandelic acid from stronger hydrochloric acid solutions (2M) than is possible with the other organic acids. Furthermore, HahnlB claims that the hydrochloric acid concentration is not critical, as his results were quantitative from solutions 0.1 M to 8 M with respect to hydrochloric acid.August, 19541 FOR THE DETERMINATION OF ZIRCONIUM 481 This behaviour undoubtedly accounts for the more specific nature of mandelic acid for separating zirconium from other elements; zirconium is quantitatively separated by this acid from Ti, Fe, V, Al, Cr, Th, Ce, Sn, Ba, Ca, Cu, Bi, Sb, Cd, Mg, Hg, Ni, U, Zn, Co and Mn.With Kumins conditions for precipitating 50-mg amounts of zirconium and allowing the solutions to stand for 60 minutes before filtration, the mandelate precipitates were dissolved in 25-d1 portions of 25 per cent. sulphuric acid and the results of the EDTA procedure for precipitations from hydrochloric, sulphuric and perchloric acids are shown in Fig. 3. Again the behaviour of zirconium in perchloric acid solutions agreed closely with that in hydrochloric acid solutions, and it indicated the advantages of the use of this acid over those of sulphuric acid in the analysis of alloys, especially for those samples where hydrofluoric acid is used to dissolve the material completely .THE ANALYSIS OF URANIUM - ZIRCONIUM ALLOYS From the results of the previous section, mandelic acid appeared to be the precipitant likely to give the best separation of zirconium from uranium before the EDTA titration procedure. In preliminary experiments to test this precipitant in the presence of uranium, synthetic solutions were prepared each containing a 10-ml aliquot of a standardised zirconyl chloride solution (48.4 mg of zirconium) and 2.5 g of uranyl nitrate dissolved in water. After the addition of 25 ml of 72 per cent. perchloric acid to each solution, they were evaporated to heavy fumes of this acid. On cooling, 25ml of hot water was added to each beaker followed by 50ml of hot 16 per cent.mandelic acid solution to precipitate the zirconium. The solutions were next heated at about 80" C for 60 to 90 minutes and then the zirconium mandelate precipitates were filtered on Whatman No. 40 filter-papers; a hot 5 per cent. mandelic acid solution in 2 per cent. hydrochloric acid was used for washing the precipitates. After the resolution of each precipitate in 25 ml of 25 per cent. v/v sulphuric acid, the EDTA titration technique gave results of 99.6, 99.2 and 100.0 per cent. recovery of zirconium. The above satisfactory results led to the development of a suitable procedure for the determination of the zirconium content of uranium - zirconium binary alloys. The most suitable sample weights required for this analysis are given in Table IV and the best solution technique consisted of attacking the sample with hot nitric acid to dissolve as much material as possible, followed by the addition of a small amount of hydrofluoric acid to complete the dissolution.TABLE IV SAMPLE WEIGHTS FOR THE ANALYSIS OF URANIUM - ZIRCONIUM ALLOYS Nominal zirconium content, Sample weight, % g 0 to 5 5 to 10 2.0 2.5 10 to 25 2.0 25 to 50 2.0 > 50 0.5 Remarks Use all the sample for Dilute solution to 250 50 ml for analysis Dilute solution to 250 50 ml for analysis Dilute solution to 250 25 ml for analysis Dilute solution to 250 50 ml for analysis analysis ml and take ml and take ml and take ml and take PROCEDURE Transfer the requisite weight of sample to a 100-ml platinum dish, add to it 25 ml of nitric acid, sp.gr.1.42, and warm. When the action of the nitric acid ceases, add dropwise the smallest amount of 40 per cent. hydrofluoric acid necessary to complete the solution of the sample. Then add 25 ml of 72 per cent. perchloric acid and evaporate the solution to strong fumes of this acid, making use of infra-red heaters. Continue the fuming for about 10 minutes. Cool the solution and dilute it to 250 ml, if necessary, and take an aliquot in accordance with Table IV. Place the sample solution in a 400-ml squat-type beaker, add 7 ml'of 72 per cent. perchloric acid for alloys containing 5 per cent. of zirconium or less and 25 ml of this acid for all other samples (to ensure a final acidity of about 3 N with respect to perchloric acid).Evaporate the solution on the hot-plate to dense fumes of perchloric acid, and heat strongly for about 20 minutes to remove the hydrofluoric acid. Cool the beaker482 MILNER AND PHENNAH [Vol. 79 and contents, add 25 ml of water, and heat the solution almost to boiling. Then slowly add 50 ml of a hot 16 per cent. mandelic acid solution and stir the solution whilst heating until precipitation of the zirconium mandelate is appreciable. Maintain the solution at above 80” C for about 90 minutes. Filter the solution through a Whatman No. 40 filter-paper, washing the precipitate several times with a hot 5 per cent. mandelic acid solution in 2 per cent. hydrochloric acid. Wash the precipitate from the filter-paper back into the precipitation beaker, pour 25 ml of hot 25 per cent.v/v sulphuric acid over the paper, and wash it with hot water. Return the beaker to the hot-plate and heat it to gentle boiling until complete solution of the precipitate is produced. Cool this solution and then complete the zirconium titration with EDTA, as described previously. The results for zirconium found by applying the above procedure to typical uranium - zirconium alloys and results for the same samples when analysed by the cupferron procedure reported earlier,l showing the good agreement between the results by these procedures, are as follows- Zirconium by gravimetric procedure, % . . 2.2 2.85 4.80 6-1 59.0 Zirconium by volumetric procedure, 7, . . 2.19 { i:;: { :$ { i::: 59.2 NOTE-After the completion of this investigation, information was received about an independent development of a titration procedure for zirconium with ethylenediaminetetra- acetic acid by J.S. Fritz and M. 0. Fulda, Institute of Atomic Research, Iowa State College, Ames, Iowa. Details are published in Report I.S.C. 382. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. REFERENCES Milner, G. W. C., and Skewies, A., Atomic Energy Research Establishment Report C/R 1126, Mayer, A., and Bradshaw, G., Analyst, 1952, 77, 476. Laubengayer, A. W., and Eaton, R. B., J . Amejv. Chem. SOC., 1940, 62, 2704. Kolthoff, I. M., and Johnson, R. A., J . Electrochem. Soc., 1951, 98, 138. Harris, W. E., and Kolthoff, I. M., J . Amer. Ghem. SOC., 1945, 67, 1484; and 1946, 68, 1175. For summary see “Organic Reagents for Metals,” Monograph No. 10, Hopkins and Williams Ltd., Milner, G. W. C., and Woodhead, J. L., AnaZysf, 1954, 79, 363. Ter Haar, K., and Bazen, J., Anal. Chim. Acta, 1954, 10, 23. Fritz, J. S., and Ford, J. J., Anal. Chem., 1958, 25, 1640. Cabell, M., Analyst, 1952, 77, 859. Osborn, G. H., Ibid., 1948, 73, 381. Jewsbury, A., and Osborn, G. H., Anal. Chim. A4cta, 1949, 3, 642. Venkataramaniah, M., and Raghava Rao, Bh. S. V., AnaZyst, 1951, 76, 107. Purushottam, A., and Raghava Rao, Bh. S. V., Ibid., 1950,75, 684. Kumins, C., Anal. Chem., 1947, 19, 376. Hahn, P., Ibid., 1949, 21, 1579. Venkataramaniah, M., and Raghava Rao, Bh. S. V., Ibid., 1951, 23, 539. Oesper, R., and Klingenberg, J. J., Ibid., 194!), 21, 1509. Belcher, R., Sykes, A., and Tatlow, J. C., Anal. Chim. Acta, 1954, 10, 34. Klingenberg, J. J., and Papucci, R. A., Anal. Chem., 1952, 24, 1861. Papucci, R. A., Fleishman, D. M., and Klingenberg, J. J., Ibid., 1953, 25, 1768. H.M. Stationery Office, 1953. Chadwell Heath, Essex, 1953. ANALYTICAL CHEMISTRY GROUP ATOMIC ENERGY RESEARCH ESTABLISHMENT HARWELL, NR. DIDCOT, BERKS. February 23rd, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900475

出版商:RSC    

年代:1954

数据来源: RSC

摘要:

August, 19541 MILLER AND HUNTER 483 A Study of the Separation of Zinc from Certain Other Elements by Means of Anion Exchange BY CHRISTINA C. MILLER AND JOHN A. HUNTER Five to 50 mg of zinc, in 2 N hydrochloric acid solutions containing a maximum of 100mg of metal ions, can be quantitatively adsorbed on a 15-cm column prepared from 3 g of the strong-base anion-exchange resin, Amberlite IRA-400 (Cl). After the percolation of 2 N hydrochloric acid to a total of 50 ml, aluminium, magnesium, copper, cobalt, nickel, manganesen, chromiumm, ironm, thorium, zirconium, titaniumfv, uraniumV*, beryllium and calcium are found almost entirely in the eluate. Cadmium, tinm, antimonym and bismuth behave similarly to zinc, and some lead and indium are held. Zinc, cadmium and indium are eluted by means of water and 0.25 N nitric acid, which also remove 20 per cent. of tin, small amounts of antimony, bismuth and lead, and small residual amounts of other metals.If water alone is used, various small quantities of zinc are tenaciously retained by the resin. Methods are proposed for the separation of zinc from solutions free from cadmium and indium and for its determination with 8-hydroxy- quinoline. In the analysis of a few alloys containing 2.5 to 34 per cent. of zinc, the results have been promising and of moderate precision. The quantitative separation of 0.5 mg of zinc from 100mg of aluminium or magnesium has been achieved. THERE is marked activity in the field relating to the analytical chemistry of zinc, which presents difficulty in analysis because of the lack of specificity in the reactions available for its determination. Polarographic methods have been applied in the analysis of light alloys.1 Bishop and Liebmann2 separated zinc from numerous other elements by cellulose chromato- graphy before determining it polarographically.Recently, Kinnunen and Wennerstrands used the disodium salt of ethylenediaminetetra-acetic acid for determining zinc that had first been separated from other metals by extracting it as the thiocyanate from an acid solution by means of an organic solvent. Flaschka," who used potassium cyanide as a complexing agent, successfully determined zinc and cadmium in the presence of large amounts of cobalt, nickel, copper and mercury. Cation exchange was utilised by Brown and Hayes5 for separating most of a large amount of magnesium from zinc before determining the latter with disodium e thylenediaminet et ra-ace t at e.Earlier, the selective extraction of zinc by means of an organic solvent mixture from a cation-exchange resin on which zinc, cobalt and nickel had previously been adsorbed was reported. We were primarily interested in the determination of small to large amounts of zinc in association with other metals that might normally be expected to interfere. Moore and Kraus,' after studying the anion-exchange behaviour of iron, cobalt, nickel, copper, manganese and zinc in hydrochloric acid solutions on the strong-base anion-exchange resin, Dowex 1, indicated that the separation of these elements from one another should be possible on a resin column of moderate length.All the metals except nickel were adsorbed from concen- trated hydrochloric acid solution. By examining the behaviour of the various elements in different concentrations of hydrochloric acid, they found that only zinc was very strongly adsorbed from 2 N acid. Elution of zinc was effected readily with 0-005 N hydrochloric acid. If a 2 N hydrochloric acid solution containing zinc and other elements was placed in a resin column and certain of the elements were washed out with an excess of the same acid, it seemed possible that the zinc could then be eluted with water. With the anion-exchange resin Amberlite IRA-400, in the chloride form, we have succeeded in effecting the separation of 5 to 50mg of zinc from a considerable number of elements in solutions containing a maximum of about 100 mg of metals.We have also studied the behaviour of elements whose separation is incomplete. Certain irregularities that occurred in the use of the resin are detailed. Methods adopted for the separation and determination of zinc have been applied tentatively to several alloys, and preliminary experiments have been made on the separation of half-milligram amounts of zinc.484 MILLER AND HUNTER: STUDY OF SEPARATION OF ZINC FROM CERTAIN EXPERIMENTAL [VOl. 79 THE DETERMINATION OF ZINC WITH 8-HYDROXYQIJINOLINE- As ion-exchange separations might be incomplete, a method of fairly general applica- bility, namely precipitation with 8-hydroxyquinoline, was chosen for the determination of zinc. In buffered acetic acid medium, zinc is separa.ble from small amounts of lead, manganese and magnesium, because the 8-hydroxyquinoline complexes of these metals begin to form at a pH above that required for quantitative deposition of the zinc complex.But in acid solutions, zinc cannot be separated from iron, cobalt, nickel, aluminium, copper, tin, bismuth, and so on. In alkaline tartrate solutions, cadmium, copper, magnesium, ironr1 and a few other elements, if present, are precipitated with the zinc - 8-hydroxyquinoline complex, whereas moderate amounts of bismuth, antimony, tin, arsenic, ironrrr, aluminium, chromiumlIr, manganeseI1, cobalt and nickel are probably tolerable. Berg'ss methods were slightly modified to give the following reasonably reliable gravimetric procedures for all subsequent determinations of zinc in this work.The first method is simpler and, when it is applicable, is preferred to the second. METHOD APPARATUS AND REAGENTS- Use Pyrex or similar glassware throughout arid reagents of recognised analytical grade. Prepare 6 N ammonium hydroxide from cylinder a.mmonia and store it in a polythene bottle. PROCEDURE 1- Precipitation in acid soZut~ort-To 60 ml of an approximately 0.25 N hydrochloric acid solution containing 20 to 50 mg of zinc, add 6 N ammonium hydroxide until the solution is neutral to methyl red indicator, and then add 0.1 ml of glacial acetic acid and 6 ml of 30 per cent. w/v ammonium acetate solution. Heat the solution to 60" C and add a 2 per cent. w/v solution of 8-hydroxyquinoline in 0.8 N acetic acid at the rate of 1 drop per second, untit an excess of 1 ml has been added.Heat the mixture gradually just to the boiling point and maintain it there for 2 to 3 minutes. Cool the mixture in running water for 30 minutes, then remove the precipitate on a weighed sintered-glass crucible of No. 4 porosity, using 50 ml of hot water for transferring the precipitate and washing it. Stir up the precipitate on the filter during washing, and dry it at 160" C to constant weight. For 5-mg amounts of zinc proceed as above, but replace the crucible with a filter-tube provided with another as a tare, and give both tubes the same heating and cooling treatment. If only 2 to 3 mg of zinc are present, reduce all volumes to one-half and use a semimicro- balance. When 0.5-mg amounts of zinc are dealt with, have initially 3 ml of solution in a 6-ml beaker, carry out the precipitation on a miicro-scale, with one-twentieth of the usual amounts of reagents, suitably diluted if necessary, and filter the precipitate through a sintered- glass filter-stick of No.4 porosity. Proceed with the wiping, heating and cooling of the beaker and filter and the associated counterpoise in the manner described by Miller and Chalmers,g and weigh them on a semimicro-balance. Dissolve the precipitate in a concen- trated hydrochloric acid - ethanol mixture (1 + 2, v/v), remove the solution and wash, dry and weigh the beaker and filter-stick. The zinc - 8-hydroxyquinoline complex contains 18.49 per cent. of zinc. Apply a correction of +Ow3 mg to all results for about 510-mg quantities of zinc.Notes on procedure 1-Ammonium acetate had to be added before 8-hydroxyquinoline. If it were added after, with perhaps a little ammonium hydroxide to adlust the pH of the solution, or if ammonium hydroxide was added to increase the pH of the solution after the zinc - 8-hydroxyquinoline complex had been precipitated as described, difficulty was experienced in attaining constancy of weight. Allthough, according to Borre1,lo precipitation is complete within the pH range 4-5 to 12, increasing the pH from 4.7 to 5.6 increased the negative error in the determination of 50 mg of zinc, probably because the zinc was not entirely in the correct form. Increasing the excess of 8-hydroxyquinoline used in precipitating 50 mg of zinc led to an increase in the weight of the precipitate, despite a fall in the pH of the solution. A constant volume excess of reagent was eventually used, although, as is shown in Table I, there was a consistent negative error in determining 50 mg of zinc.Experiments were made with weighed aliquots of standard solutions of zinc chloride prepared from Hilger's "H.H.P." (99.99 per cent.) zinc and AnalaR concentrated hydrochloric acid.August, 19541 OTHER ELEMENTS BY MEANS OF ANION EXCHANGE 485 TABLE I DETERMINATION OF ZINC IN BUFFERED ACETIC ACID SOLUTION Zinc pH of Zinc pH of Zinc pH of taken, Error, filtrate taken, Error, filtrate taken, Error, filtrate mg mg mg mg mg mg 52-75 - 0.29 4-7 20-90 -0.10 5.1 5-68 +0-09 5.4 52.75 - 0.29 4.7 21-05 +0.05 5.1 6-07 +O-Ol 5.4 47.18 - 0.24 - 17-60 +0.02 - 5.67 f0.02 - 48-63 - 0.37 - 18.29 +Om03 - 5.62 +0.04 - The infizdence of other substances-Only small amounts of tin and antimonyIII, but more lead, may be present, as is shown by the following results on the determination of 5 mg of zinc in the presence of these elements- Tin Lead Antimony 7 - w Weight of other metal, mg .. 0.3 0.5 3 5 2 2 0-5 0.75 Error on zinc, mg . . . . + O . l l $0.12 $0.17 $0.19 +Om03 +0*08 f0.14 +0*05 - AntimonyV was so slowly hydrolysed that 10mg of it did not affect the determination of 5 to 50 mg of zinc. The addition of 0 4 g of ammonium nitrate to solutions containing 5 to 50mg of zinc was without significant influence on the results. PROCEDURE 2- Precipitation in alkaline tartrate solution-To 60 ml of an approximately 0.25 N hydro- chloric acid solution containing 5 to 50 mg of zinc, add 1 g of tartaric acid and 0.05 ml of 0.002 per cent.w/v o-cresolphthalein solution, and then 5 ml more of carbonate-free 2 N sodium hydroxide than is required to make the indicator pink. If the amount of zinc is near the maximum, increase the excess of 2 N sodium hydroxide, if necessary, to 6-6 ml in order to ensure complete solution. To the cold solution add 3 to 5 mg of sodium tauroglycocholate, which prevents the precipitate adhering to the beaker and assists filtration, and then add, at the rate of 1 drop per second, with stirring, a fresh 2 per cent. w/v solution of 8-hydroxy- quinoline in 0.75 per cent. w/v sodium hydroxide solution, until an excess of 0.5 ml has been TABLE I1 DETERMINATION OF ZIKC IN ALKALINE TARTRATE SOLUTION Standard amount of 2 N sodium hydroxide Excess of taken, hydroxide, Error, mg ml mg 47.67 6-6 + 0.09 A r 7 Zinc 2 M sodium 48.40 6-6 + 0.03 18.15 5 +om12 17.9 1 5 - 0.09 5.35 5 - 0.04 5.37 5 - 0.04 Effect of additional 2 N sodium hydroxide Excess of taken, hydroxide, Error, *g ml mg 48.25 8 +0*19 48.57 8 +0.26 52-75 15 - 0.50 52-75 25 - 1.86 17.54 8 - 0.40 18.92 8 - 0.33 5.18 8 - 0.63 4.58 8 - 0.38 r h \ Zinc 2 N sodium added.For amounts of zinc under 10 mg (precipitation may not take place in the cold) add 2.5 ml of the precipitant. Digest the mixture at 70" to 80" C for 30 minutes in order to induce the precipitation of small amounts of zinc and to flocculate the precipitate already present. Filter the hot mixture through an unweighed sintered-glass crucible (the alkaline solution causes loss of weight) , and continue the determination as described in procedure 1.Finally dissolve the precipitate in a concentrated hydrochloric acid - ethanol mixture (1 + 2, v/v). Wash, dry and weigh the crucible. Notes on firocedure 2-The method differs from Berg's in that 8-hydroxyquinoline is dissolved in a minimum excess of sodium hydroxide solution instead of in ethanol or acetone, both of which markedly inhibited the precipitation of small amounts of zinc. As was found486 MILLER AND HUNTER: STUDY OF SEPARATION OF ZINC FROM CERTAIN [VOl. 79 by Berg, variations in the excess of sodium hydroxide used had an influence on the results, and with small amounts of zinc it was necessary to restrict the quantity of alkali used.More latitude was permissible with large amounts of zinc. In the determination of 20 mg of zinc, the 8-hydroxyquinoline could be present in a 10 to 50 per cent. excess without significant effect. With small amounts of zinc, precipitation did not occur readily in the cold and there- fore a fixed quantity of precipitant was added for all amounts below 10mg. Berg did not heat the solutions until a precipitate appeared, but heating them immediately after the addition of the precipitant did not have an adverse: influence. The results in Table 11, which are for the same standard solutions as were used in procedure 1, show that the method is adequate for the present purpose. The injztuence of other metak-As the following results show, large amounts of tin could be tolerated, except perhaps with the smallest amounts of zinc, when contamination by hydrated stannic oxide occurred and made filtration very slow- Zinc taken, mg.. .. . . 46.97 48.99 18-15 17.70 4.71 5-86 Error, mg .. .. .. +0.33 $0.05 -0.03 $0.07 +0.16 +0*11 Tin taken, mg . . .. .. 50 50 80 80 100 100 Even 100mg of ironlI1 gave no precipitate with 8-hydroxyquinoline in cold or hot solutions. Five milligrams of cobalt gave some precipitation in hot solutions, so that only a small amount could be tolerated. PRELIMINARY EXPERIMENTS ON ION EXCHANGE- In the preparation of materials containing zinc for analysis, it will often be possible to arrange to have a solution consisting mainly of chlorides of metals in hydrochloric acid. It was proposed to apply to a column of resin in the chloride form, in equilibrium with 2 N hydrochloric acid, about 5 ml of a 2 N hydrochloric acid solution of the metals, then to wash the column with sufficient of the same acid t o remove unadsorbed and weakly adsorbed ions, to elute zinc by means of water, and then t o determine it with 8-hydroxyquinoline. Moore and Kraus used Dowex-1, a strong-base anion-exchange resin, but, as this was not readily available, we used Amberlite IRA-400 resin (Analytical Grade) , and preferred it to the comparable product De-acidite FF, because of its somewhat greater uptake of zinc.Qualitative experiments, which were made with 1. to 5 mg of metals on columns containing 0.25 g of resin, provided information for testing more exactly the separation on larger columns of 5 to 50 mg of zinc from other metallic elements in solutions containing a maximum of 100 mg of metals.With these small columns, 5 mg of nitrate, sulphate, arsenate, phosphate or molybdate, present initially with 2.5 mg of zinc, were found only in the hydrochloric acid eluate, which was free from zinc. All other experiments to be described were made at about 17" C on resin beds, 15 cm deep and about 7 mm in diameter, containing about 3 g of resin. When several millilitres of a 2 N hydrochloric acid solution containing 50 mg of zinc were allowed to pass on to the chloride form of the resin, which was in contact with the same acid, and the column was then further washed with the acid, the first 50 ml of eluate were found to contain less than 0.01 mg of zinc.Probably 75 ml of acid could be used without significant loss of zinc. It was thought that EiOml would effectively remove maximum amounts of copper, cobalt , nickel, manganese'I, chtromium'II, aluminium , thorium, zirconium , titanium, uraniumv1, beryllium, magnesium and calcium. Much ironlI1 was readily eluted, but a little was retained by the resin. Cadmium behaved like zinc, and the removal of indium was incomplete. All of 100mg of tin'' and of 50mg of antimonyIII or bismuth were thought to remain on the column, but when antimony was in the quinquevalent form 25 per cent. of it was found in the eluate. Lead chloride, in the amount required to saturate the small volume of acid solution added to the column, was held by the resin. Qualitative tests were similarly made in order to find how much water would be required to elute 50 mg of zinc from the resin and the effect of that volume on other adsorbed elements.Twenty millilitres of water removed over 80 per cent. of the zinc, and, after 60 ml had been used, two further 20-ml portions showed the presence of 0.02 mg and less than 0.01 mg, respectively. Of the other metals that were adsorbed from 2 N hydrochloric acid solutions, 100 mg of tinIV yielded 20 mg to the water eluate, 50mg of antimonyIII yielded 2 mg, 50mg of antimony" yielded IOmg, a few mg of lead yielded 1 mg and 50 mg of bismuth yielded nothing. Most, if not all, Sixty millilitres of water were considered adequate.August, 19541 OTHER ELEMENTS BY MEANS OF ANION EXCHANGE 487 of 50 mg of cadmium was also removed, as was residual indium.Traces of iron and copper were in the eluate. When elution was effected first with 20 ml of water and then with 40 ml of 0.25 N nitric acid, instead of entirely with water (see p. 486), the amount of antimonyIII accompanying zinc was doubled, and there were 2 to 3 mg of bismuth when 5 to 50 mg were present initially. Contamination by iron and copper was greater. By direct quantitative test, 100 to 5 mg of copper, in the presence of some zinc, were found to yield 0.2 per cent. of the amount applied to the solution containing zinc. Similarly, 100, 10, 5 and 0.5 mg of ironIII, the last three in association with 10mg of zinc, gave 0.6, 0.07, 0.04 and 0.015mg, respectively. A blank test with 10 mg of zinc and no added iron gave 0.002 mg of iron.On occasion, however, larger amounts of iron were noted than would be predicted from these results, indicative of some erratic effect. None of the foregoing elements when present in the amounts quoted, except cadmium and indium, would be expected to interfere seriously with the determination of zinc with 8-hydroxyquinoline in an alkaline tartrate solution. The copper would be quantitatively precipitated, but a correction seemed readily applicable. It was therefore concluded that it should be possible to evolve a method for the determination of zinc originally in association with a large number of other metals, excluding cadmium and indium. QUANTITATIVE EXPERIMENTS WITH ZINC ALONE AND ALSO IN ASSOCIATION WITH OTHER The first experiments on the determination of zinc that had been eluted with water from a resin column on which it had previously been adsorbed were sufficiently satisfactory to justify a test of the behaviour of mixtures.In experiments involving the separation of the smaller amounts of zinc from metals that were not held on the resin, there were a few rather large negative errors. In general, resin columns that could be freed of adsorbed ions when eluted with water had been regenerated with 2 N hydrochloric acid and used again, but fresh columns were also needed at intervals. As the more significant errors were associated with the latter, a thorough investigation was made of the behaviour of zinc on fresh resin and on regenerated resin, in an attempt to locate the source of the irregularities.The results, shown in Table 111, indicate that larger negative errors were associated with resin that was in use for the first time. With fresh resin the amount of zinc lost from 50mg originally taken ranged from 0.1 to 0.7 mg, and that lost from 5 mg of zinc was from 0.04 to 0.4 mg. A different batch of resin showed comparable irregularities. Decreasing the flow-rate of solutions through the columns did not effect any improvement. Lost zinc was not found in the acid eluate, and in an experiment in which 50 mg of zinc had been used and 0.34 mg had been lost, only 10 per cent. of this amount was recovered from a further 40 ml of eluting water. As slight hydrolysis might have caused loss of zinc, eluting with 0.005 N hydrochloric acid instead of with water was tried, but there was no improvement.ELEMENTS- TABLE I11 DETERMINATION OF ZINC ALONE, AFTER ADSORPTION ON AND' ELUTION FROM THE ANION-EXCHANGE RESIN AMBERLITE IRA-400 (c1) Elution with 60ml of water Approximate r L \ zinc taken, Error, Error, weight of Resin not previously used. Regenerated resin. 10-3 g 10-5 g 10-6 g 50 -12, -31, -72, -34, + 2l -248 -15, -34 -30, -11 20 -12, -17 -5, -8 5 -41, -15, -18, -23 -61 3-7 -33*, - 2 5 -4 * Slower rate of flow. Elution with 20 ml of water and 40 ml of 0.25 N nitric acid. Error, 10-5 g +18, -17, +30, -11, +6 +al +9 +6, +3. + S , +3 THE USE OF RESIN PRE-SATURATED WITH ZINC- As zinc was evidently sometimes retained on the resin, experiments were made on a resin that had been previously treated with an excess of zinc chloride. Resin, in the chloride form, was soaked overnight in an excess of a solution of zinc chloride in 2 N hydrochloric acid,488 MILLER AND HUNTER: STUDY OF SEPARATION OF ZINC FROM CERTAIN [VOl.79 and a column was prepared from the suspension. The resin was washed with 2 N hydro- chloric acid and then with 60 ml of water. With jurther 20-ml water washings, a significant amount (0.05 mg) of zinc was found only in the first eluate. The resin was left in contact with water overnight and then 50 ml of 2 N hydrochloric acid were passed through it and tested for zinc, when 0.05 mg was found. Further treatments with 60ml of water and 50 ml of acid yielded 0.1 and less than 0.01 mg of zinc. A final wash with water, after the resin had been in contact with acid for 3 days, ga.ve 0.05 mg of zinc.These results showed how tenaciously a small amount of zinc was held by the resin. Two columns of zinc-saturated resin were repeatedly treated alternately with water and 2 N hydrochloric acid before they were used for testing the recovery of 5-mg amounts of zinc, which were applied to them in the usual way. But the negative errors still persisted and were similar to the maximum error previously recorded for this amount of zinc. This showed that the resin, originally saturated with zinc, had for all practical purposes been freed from zinc by the successive acid and water treatments and was reclaiming some from the test solution. Water was left in both columns overnight acid then 50 ml of 0.005 N hydrochloric acid were passed through, with removal of about 0.05 nig of zinc.Further washing of one column with 20 ml of 0.25 N sulphuric acid yielded 0.05 mg of zinc, and of the other with two 20-ml portions of 0.25 N nitric acid yielded 0.1 and 0-05 mg of zinc. I t was therefore confirmed that the manner of attachment of some of the zinc to the resin may be different from that of the rest. As the amount retained will presumably be related to the period of contact of the resin with water, it seemed that zinc might be determined correctly in the test solutions if the zinc-treated resin were so handled beforehand that none of the firmly held zinc was detached. Accordingly, columns prepared from zinc-saturated resin were, after being washed with 20 ml of 2 N hydrochloric acid and 100 ml of water, immediately treated with 50 ml of 2 N hydro- chloric acid and used for experiments with 50 and 5-mg amounts of zinc.The recovery of zinc was quantitative; the errors for 50-mg amounts were +0.01 and +O.OS mg, and for 5-mg amounts were nil and +O.Ol mg. As resin columns prepared in this way had been found to yield an amount of firmly held zinc similar to that detached from columns through which 5 mg of zinc had been passed and also to the average amount that had been lost in the passage of 50mg, it was evident that colurrms to which 5 to 50-mg quantities of zinc had been applied and from which the bulk was afterwards removed with water, should, if regenerated immediately with 2 N hydrochloric acid, give similar results. Two experiments with mixtures containing 50mg of zinc and 50mg of aluminium, one made on a column suitably prepared from zinc-saturated resin, the other on resin regenerated after passage of zinc in the ordinary way, gave similar small positive errors.In Table IV (columns a and b) are recorded all the results hitherto determined for mixtures, with fresh resin and with regenerated resin, but without special attention being paid to the period during which the resin had stood in contact with water before regeneration. In addition, results (marked 5) are included for determinations since made on resin immediately regenerated after passage of at least 5 mg of zinc, in experiments or in trials. With reference to the results for mixtures that left negligible amounts of metals on the column, it is evident that there is a lack of precision throughout.Investigation of the manner of regenerating the resin has effected no important improvement. There was no evidence that repeatedly used columns settled down and gave better r e d t s . There seemed to be an irregularity associated with the resin, and we can offer no explanation for this. THE ELUTION OF ZINC FROM RESIN COLUMNS BY MEAPU'S OF DILUTE NITRIC ACID- Since the attempt to counteract the error due to the retention on the resin of a variable residue of zinc had only partially succeeded, a means of ensuring complete removal of zinc was next sought. As 0.25 N nitric acid eliminated strongly attached zinc much more readily than water and more effectively than 0-25 N sulphuric acid (above), its use for breaking down and removing the zinc chloride complex from the resin was examined.Elution with water had led to retention of bismuth and much of the tin and antimony owing to the hydrolysis of their salts. Therefore 20 ml of water, which eluted 80 per cent, of 50 mg of zinc, were applied first, and then sufficient nitric acid to elute the rest of the zinc; 40 ml of acid provided a safe excess. Under these conditions the results were reasonably accurate (last column of Table 111) when various amounts of zinc were adsorbed on the resin, eluted and determined with 8-hydroxyquinoline in acid solution. A tendency for positive errors to occur suggested thatAugust , 19541 OTHER ELEMENTS BY MEANS OF ANION EXCHANGE 489 traces of metallic impurities, most likely iron and copper, were intruding.Precipitation with 8-hydroxyquinoline in alkaline tartrate solution will avoid the effect of iron, but not that of copper, which requires the presence of some potassium cyanide.ll When the method was applied to mixtures of zinc with other elements, the same trends as were noted for zinc alone were indicated by the results, which are shown in the last column of Table IV. Where sufficient experiments have been done, the results show greater precision than those in which water alone has been used for elution. Excellent separations of zinc from TABLE IV DETERMINATION OF ZINC IN MIXTURES AFTER ADSORPTION ON, AND ELUTION FROM, THE ANION-EXCHANGE RESIN AMBERLITE IRA-400 (C1) Elution with water A r Approximate weight of zinc taken, 10-8 g "3 5 ' 5 Other metals cu Co, Ni and Mn FeIII, Be, TiIv and Ca SbIII, Bi, Th, Zr and CrIII Th, Zr, UVI and CrI'I (1: 2: 2) (1: 1: 1: 1) (1:l: 1: 1:l) (1: 1: 1: 1) SnIV Bi Sb Approximate weight of other metals taken, 50 80 10-3 g 100 50 80 100 50 80 100 50 80 100 20 20 20 25 25 25 20 20 20 50 80 100 50 50 50 50 50 50 (4 Resin not previously used.Error, 10-5 g - - - -16, -4 -1 -5, -4 - - -38, -38 -9, -7 -19, -12 -15, -8 - - - - - - - - - - - - - - - 1 ::} SbV - 31 Regenerated resin. Error, 10-5 g -16, +4 -14, +1; +w -19§ -12, -2; +23,§ -1l§ -13, +14 +1 - +19, -25 -3, -7 -1, -2 +28,§ -35s +7,§ -5* - 8 4 +9§ -5, 3-12 -11, +20 +6, + l o - 27 - 24 - 17 -8,* +9* -23,* -13* -34,* -5* - 63 - 26 - 12 $ :F) SbIII - 9* Elution with water and 0.25 N nitric acid. Error, 10-5 g +3, +1 -11, +3 - -12, -11* - 3* +4, (--43*); +4,* +3* - 1st +Gt -21, +3 +2, +6 +11, +8 O* +%$ +3* +5,t +7* I +4 +5 + 13 - 54* - 30* - 26* - 4* - 12* - - -} 12 SbV z*} $&I11 +4 - 2* * Precipitation with 8-hydroxyquinoline in alkaline tartrate solution. t Corrected for copper, see p.487. $ Corrected for iron, 0.05 and 0.18 mg, respectively. Immediate regeneration of resin after passage of a zinc-containing solution, see p. 488. aluminium and magnesium have been achieved. If these elements had been present, the aluminium - 8-hydroxyquinoline complex would have been precipitated in acid solution and the magnesium complex in alkaline tartrate medium, so causing positive errors. With mixtures containing iron, which tended to behave a little erratically, precipitation in alkaline tartrate solution is desirable: As traces of iron were frequently noted, it may be thatthey are at least partly responsible for the small positive errors found in some other experiments.A correction has been made for copper in accordance with the results given on p. 487. The490 MILLER ASD HUNTER: STUDY OF SEPARATION OF ZINC FROM CERTAIN [VOl. 79 possibility of avoiding its interference in the final determination of zinc has not yet been examined. As indicated on p. 487 , iron and copper seemed to behave like zinc ; small amounts of these elements are held on the resin in the elution with water, but are liberated by treatment with nitric acid. Nitric acid separates zinc more satisfactorily from bismuth than does water, whereas the reverse holds for tin,' which evidently traps some zinc in the resin.One could easily reduce this error by eliminating part of the tin as stannic halide before effecting the separation on the resin. ANALYSES OF ALLOYS- In the course of this investigation some alloys were analysed and, when necessary, additional zinc was added to raise the content to the minimum under consideration. Latterly, however, it was desired to find whether the separation of less than 5 mg of zinc would be practicable, and a bronze and an aluminium alloy yielding half this amount were included, the final determination of zinc being effected in the customary manner, but on an appro- priately reduced scale (p. 484). Because of the high tin content, the determination of zinc in the bronze had to be done in alkaline tartrate :solution. As the risk of faulty adjustment of the alkalinity was greater on the small scale , precipitation of the zinc - 8-hydroxyquinoline complex in acid solution was preferred for the aluminium alloy, which contained an insignificant amount of tin.The precipitates were afterwards examined quantitatively for iron and copper and corrections applied, namely, for 6 pg of iron and 10 pg of copper in the first precipitate, and for 8 pg of each element in the second. All results are shown in Table V. Those referring to elution with water are to be com- pared with the results given in column ( b ) , Tab1.e IV. The results in the last columns of Tables IV and V are similarly comparable. Although some of the results in the second from last column are apparently good, the method re1ai:ing to elution with water and nitric acid is believed to be more reliable, and it is therefore recommended.Smaller amounts of zinc than 5 mg can obviously be dealt with, but it would be preferable to have a simpler means of determining them than precipitation of the zinc - 8-hydroxyquinoline complex in alkaline tartrate solution. TABLE 'V DETERMINATION OF ZINC IN ALLOYS Zinc found rY After I 7 elution with water and Approximate composition of alloys A Alloy Cu, Sn, Fe, Pb, Al, Mn, Mg, Ni, Sb, Zn, with water, nitric acid. Manganese brass No. 179 . . Brass No. 37b . . Bronze No. 183 with added zinc} Bronze No. 207. . Aluminium alloy Aluminium alloy No. 181 with added zinc Aluminium alloy No. 181 . . % % % % % % Yo- % % 5 9 2 1 1 2 1 - 1 - 70 1 0.2 1 - - - (1 - 83 10 0.1 2 - - - - 1 87 10 0.1 <1 - - - (1 <1 1 - 0.3 - 89 1 3 - - Essentially as below 5 <1 0.5 2 87 - 1 2 - % 33.95 2 7-,09 I [ (1) 5-01s (2) 6.185 2.535 5.807 (1) 5-26s (2) 5-539 2.375 % ' % 33.5, 33.6 33*8,*t 33*7* 26-9, 27.2 - - 4.95* 5-26* 2.28,* 2-24' 3*46,*t 2*40* 5.70, 5.78 5.97, 5.81 - 5.24 5.60 - 2-40,: 2.41: * Precipitation with 8-hydroxyquinoline in alkaline tartrate solution. t Corrected for copper (0.2 per cent.of the original copper content of the alloy), see p. 487. 5 British Chemical Standard certificate value plus a.dded zinc where required (1.86 per cent. of zinc 11 U.S.A. Bureau of Standards certificate value. 7 Value supplied by the British Aluminium Company. Corrected for copper and iron by direct quantitative test.in bronze No. 183). RECOMMENDED PROCEDURE FOR THE SEPARATION OF ZINC BY ION EXCHANGE IN ALLOY ANALYSIS- Preparation of the resin column-For the glass column take a piece of Pyrex tubing, 16 cm long and of 6 to 7 mm internal diameter, and to the top seal a cup 7 cm long and of 18 mm diameter, and to the bottom a piece of glass tubing, 3 cm long and of 3 mm internalAugust, 19541 OTHER ELEMENTS BY MEANS OF ANION EXCHANGE 491 diameter. Above the lower joint, place a small pad of glass wool as a support for the resin. Attach a short length of polythene tubing with a screw clip to the bottom of the column. Slightly grind the dry Amberlite IRA-400 (OH) resin, as supplied, and collect the solid that passes a 50 and is held on a 100-mesh sieve. Soak the resin overnight in slightly warm 2 N hydrochloric acid, remove fine particles, and then, after filling the glass column with 2 N hydrochloric acid and leaving the screw-clip open, transfer sufficient of the slurry of resin to give a 15-cm bed when the resin has packed down.Allow 20 ml of 2-0 N hydrochloric acid to percolate through the resin, and adjust the rate of flow to 3 to 4 ml per square cm per minute. On top of the resin pack tightly a thin pad of cotton wool to prevent the column from running dry. Finally pass 50 ml of 0.25 N nitric acid and 50 ml of 2.0 N hydrochloric acid through the column, which is then ready for use. The adsorption artd elution of zinc-Transfer to the column about 5 ml of 2 N hydrochloric acid containing up to 50 mg of zinc and a total of not more than about 100 mg of metallic elements, and allow the solution to percolate at the rate specified.Rinse the container five times with 1-ml portions of 2.0 N hydrochloric acid, allowing the column to drain between additions, and then add more 2.0 N acid to bring the total volume of solution added to 50 ml. Now elute the zinc by allowing 20 ml of water, followed by 40 ml of 0.25 N nitric acid, to percolate through the column. Collect the solution and proceed with the determination of zinc by the appropriate method (p. 484). Precipitation of the zinc - 8-hydroxyquinoline complex in alkaline tartrate solution is generally applicable. Precipitation in acid solution is simpler and, with reference to the information given on p. 485 and p. 487, may be used for materials that contain no bismuth, a limited amount of iron and less than 2 per cent.of tin or antimony. Whichever method is used, a correction is required for copper, and when precipitation in acid solution is adopted, it is desirable to examine precipitates for the presence of iron. For 50-mg amounts of zinc, precipitated in acid solution, a correction of +O-3 mg is required. Reject the resin if the original solution contained metals such as lead, antimony, bismuth and tin, which are retained by it. Otherwise prepare the column for further use by immediately passing through it 50 ml of 2-0 N hydrochloric acid. Preparation of alloy solzdions-Disintegrate 100-mg amounts of alloys that contain not more than 50 per cent. of zinc with hydrochloric and nitric acids.Remove the excess of nitric acid by evaporation in the presence of an excess of hydrochloric acid, and dissolve the residue in a minimum volume of 2 N hydrochloric acid. If necessary, remove insoluble matter and lead chloride (after cooling to 0" C) by filtration, and then transfer the filtrate and washings to the resin column. Proceed with the adsorption and elution of zinc as indicated. EXPERIMENTS ON THE SEPARATION OF 0 6 m g AMOUNTS OF ZINC FROM OTHER METALS BY As the separation of 5-mg amounts of zinc from an excess of other elements was reasonably satisfactory and a few analyses of alloys gave good recovery of 2.5 mg, it was of interest to investigate the recovery of much smaller quantities of zinc and to see how the irregularities associated with the columns would affect them.The 8-hydroxyquinoline procedure for the determination of zinc in acid solution was adapted to the micro-scale, and, after trial deter- minations of 0-5 mg of zinc, the water - nitric acid eluates from normal blanks on two resin columns were evaporated and added to the same amounts of zinc, which were then determined. Half-milligram quantities of zinc, first alone and then in association with 100 mg of aluminium or magnesium, were next adsorbed on the resin and eluted with water and nitric acid in the usual way. The eluates were evaporated to small bulk, transferred by a capillary tube to 6-ml beakers and evaporated to dryness in the presence of hydrochloric acid before deter- mination of the zinc with 8-hydroxyquinoline in acid solution as described on p.484. Precipi- tation was effected in acid solution in order to ensure the recovery of likely adventitious impurities. In all experiments, except those in the direct determination of zinc, the precipi- tates were disintegrated with concentrated nitric and sulphuric acids and prepared for the colorimetric determination of iron and copper with potassium thiocyanate and a chloroform solution of diethylammonium diethyldithiocarbamate, respectively. Both metals were found in amounts that tallied well with the positive errors recorded throughout. When corrections for them were made, it was evident, as is shown in Table VI, that the recovery of 0.5 mg of zinc was virtually complete, even in the presence of a 200-fold excess of aluminium or magnesium. Complications may arise, however, when more complex mixtures are examined. a ION EXCHANGE-492 MILLER AND HUNTER TABLE VI DETERMINATION OF APPROXIMATELY 06-mg AMOUNTS OF ZINC UNDER VARIOUS CONDITIONS Remarks Direct determination of zinc Direct determination of zinc after addition of the blank from the resin . . .. .. .. Zinc put through the ion-exchange 100 rng of magnesium present . . { { procedure . . .. .. -4 100 rng of aluminium present Weight of zinc taken, Pg 491 615 541 476 458 463 478 472 465 486 551 495 Weight of zinc found, Pg 477 618 547 47 7 488 503 516 518 488 492 570 520 Weight found in precipitate & Fe, cu, CLg PQ 12 6 15 8 lost 21 10 15 8 7 3 7 7 12 8 Corrected weight of zinc found, Pg - - - - 463 471 - 474 456 478 552 493 [Vol. 79 Error, PQ - 14 +3 +6 +1 +5 + 8 +2 -9 -8 +1 -2 - The separation of zinc and cadmium is now under consideration and a simpler means of determining small amounts of zinc is being sought. We are indebted to the Trustees of the Moray Fund for a grant, and thank the British Aluminium Company for a sample of an analysed alloy. One of us (J. A. H.) is grateful for a maintenance allowance from the Shell Petroleum Company and for the award of an Edinburgh University Studentship. 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. REFERENCES Stross, W., Analyst, 1949, 74, 285. Bishop, J. R., and Liebmann, H., Ibid., 1953, 78, 117. Kinnunen, J ., and Wennerstrand, B., Chemist- AnaZyst, 1953, 42, 80. Flaschka, H., 2. anal. Chem., 1953, 138, 332. Brown, E. G., and Hayes, T. J., Anal. Chim. Acta, 1953, 9, 408. Department of Scientific and Industrial Research, “Chemistry Research,’’ 1951, p. 53. Moore, G. E., and Kraus, K. A., J . Amer. Chem. Soc., 1950,72, 5792; 1952,74, 843; 1953,75, 1460. Berg, R., 2. anal. Chem., 1927, 71, 171. Miller, C. C., and Chalmers, R. A., Analyst, 195B, 78, 686. Borrel, M., Thesis, Universite de Lyons, 1952. Robertshaw, A., Analyst, 1942, 67, 259. CHEMISTRY DEPARTMENT THE UNIVERSITY, EDINBURGH, 9 March 3rd, 1954

ISSN:0003-2654    

DOI:10.1039/AN9547900483

出版商:RSC    

年代:1954

数据来源: RSC