ISSN:0003-2654    

DOI:10.1039/AN95378FX034

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

at any point when recording the polarogram. Two polarograms are discussed in detail in order to illustrate the application of the instrument to polarographic problems. A multipurpose polarographic cell, which allows determination of the pH of the solution and of m and t for the electrode without removal of the solution from the cell, is also described. The dimensions of this cell allow for it to be used in a substantially unmodified Cambridge thermostat bath. NOTICES Second International Congress on Rheology A PROVISIONAL programnie for the Second International Congress on Rheology, to be held in Oxford between July 26th and 31st, 1953, has been prepared. All intending participants are asked to make their reservations as soon as possible (and in any case not later than May lst, 1953) by completing a Final Form of Application, which is attached to the provisional programme.The programme, giving full details and the form of application, can be obtained from the Hon. Organising Secretary, Dr. G. W. Scott-Blair, The University, Reading, England. The Association of Clinical Biochemists THE Inaugural General Meeting of the Association of Clinical Bi0chemist.s was held at the Postgraduate Medical School of London on March 28th, 1953. The Association will be both scientific and professional in the scope of its activities. The interim committee is acting as a provisional council, with Dr. A. L. Tgrnoky, Royal Berkshire Hospital, Reading, as Honorary Secretary, from whom further details of the Association’s future activities can be obtained. MEETINGS OF THE ROYAL SANITARY INSTITUTE Wigan Sessional Meeting, Friday, June 5th, 1953 AT this meeting the following papers will be presented: “Staphylococcal Food Poisoning in the Manchester Area,” by M.T. Parker, M.B., B.Ch., Dipl-Bact., Director, Public Health Laboratory, Manchester, and “The Changing Pattern of Refuse Disposal and its Effect on Vehicle Design,” by Clive Walker, Director, Walker Bros. Ltd., Engineers. Walsall Sessional Meeting, Thursday, July 2nd, 1953 AT this meeting the following papers will be presented: “Land Use in Walsall, with Special Reference to Slum Clearance and/Reclamation of Derelict Land,” by M. E. Habershon, O.B.E., M.Eng., M.I.C.E., M.I.Mun.E., Borough Engineer and Surveyor, Walsall, and ‘Some -4venues to a Better Environment,” by James Green, Deputy Chief Sanitary Inspector, Walsall.. London Sessional Meeting, Wednesday, July 15th, 1953 AT this meeting, to be held at 2.30 p.m. at the Royal Sanitary Institute, the following papers forming a Symposium on “Salvage and Utilisation of Food Waste for Animal Feeding” will be presented: (a) “Collection and Processing,” by John Stephen, M.Inst.P.C., Director of Public CZeansing, Luton, and (b) “Distribution and Utilisation,” by Major A. McD. Livingstone, C.I.E., M.C., M.A., B.Sc., Adviser on Agricultural Matters to the Waste Foods Branch, Ministry of Agriculture and Fisheries. Enquiries about these meetings should be addressed to the Secretary, The Royal Sanitary Institute, go, Buckingham Palace Road, London, S.W.1. - PRINTED BY W. HEFFER & SONS LTD.. CAMBRIDGE

ISSN:0003-2654    

DOI:10.1039/AN953780X036

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BX037

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BP085

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

JULY, 1953 THE ANALYST Vol. 78, No. 928 PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS AN Ordinary Meeting of the Society, organised jointly by the Microchemistry Group and the Scottish Section, was held a t 7.15 p.m. on Wednesday, May 6th, 1953, in the Chemistry Department, University of Glasgow, Gilmore Hill, Glasgow. The Chair was taken by the President, Dr. D. W. Kent-Jones, F.R.I.C. The following papers were presented and discussed : “Geochemistry and Microchemistry,” by David T. Gibson, D.Sc. ; “Micro-analysis of Silicate Rocks. Part IV. The Determination of Alumina,” by Miss Christina C. Miller, Ph.D., D.Sc., F.R.S.E., F.H.-W.C., and Robert A. Chalmers, B.Sc. ; “Microchemical Determination of Sulphur in Organic Compounds,” by William H.Massie, BSc., Ph.D., A.R.I.C. The meeting was preceded by an afternoon visit to the Clydebridge Steel Works of Colvilles Ltd. A s Ordinary Meeting of the Society was held at 7 p.m. on Wednesday, May 20th, 1953, in the Meeting Room of the Chemical Society, Burlington House, London, W.l. The Chair was taken by the President, Dr. D. W. Kent-Jones, F.R.I.C. The following papers were presented and discussed : “The Determination of Ergosterol in Yeast. Parts I, 11, I11 and IV,” by W. H. C. Shaw, Ph.C., F.R.I.C., and J. P. Jefferies, B.Sc., A.R.I.C. ; “The Estimation of Micro Quantities of Calcium,” by G. E. Harrison, Ph.D., F.Inst.P., and W. H. A. Raymond; “The Ultra-Violet Spectrophotometric Estimation of the Quality of Mineral Oils Extracted from Bread,” by M. A.Cookson, B.Sc., A.R.I.C., J. B. M. Coppock, B.Sc., Ph.D., F.R.I.C., and R. Schnurmann, MSc., Dr.Rer.Nat. NEW MEMBERS Dennis n’orman Davies ; Manse1 Charles Finniear, B.Sc. (Wales), A.R.I.C. ; Robert Greenhalgh, BSc. (Lond.) ; Stanley George Markham, BSc. (Lond.) ; Geoffrey Middleton, B.Sc. (Lond.), F.R.I.C.; Trevor David Rees, B.Sc. (Lond.), A.R.I.C.; Charles England Rhodes, BSc., Dip.Ed. (Leeds), F.R.I.C. ; Alan Turner, A.R.I.C. ; Victor Barry Wright, BSc. (Lond.), A.R.I.C., A.M.Inst.Pet. PHYSICAL METHODS GROUP THE Forty-first Ordinary Meeting of the Group was held at 6 p.m. on Friday, May 8th, 1953, in the Oriental Cafd, Ipswich. This was a joint meeting with the East Anglian Section of the Royal Institute of Chemistry, and the Chairman of the Group, Dr. J. Haslam, F.R.I.C., was in the Chair. The following papers on “Emission Spectroscopy” were presented and discussed : “Semi- quantitative Techniques in Spectrochemical Analysis,” by R. L. Mitchell, BSc., Ph.D., F.R.I.C. ; “Some Techniques of Presentation of Sample to the Spectrograph,” by A. H. C. P. Gillieson, BSc., Ph.D.; “Applications of the Porous Cup Technique,” by L. G. Young. The meeting was preceded by an afternoon visit to the Research Laboratories of B.X. Plastics, Ltd., Manningtree. 389

ISSN:0003-2654    

DOI:10.1039/AN9537800389

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

390 HASLAM AND WILKIXSON: THE DETERMINATION OF [Vol. 78 The Determination of Arsenic by B. S. Evans’s Method With Some Observations on the Separation of Arsenic and Antimony BY J. HASLAM” AXD N. T. WILKINSON An improved method for the volumetric determination of arsenic in solutions after its preliminary reduction by means of hypophosphorous acid is described. I t is shown that the precipitated arsenic can be filtered on to a medium containing a large proportion of oxy-cellulose and that, after the addition of an excess of standard iodine solution to the arsenic - oxy-cellulose mixture, the excess of iodine can be directly titrated with standard arsenite solution. The end-point in the titration is determined by the disappearance of the yellow colour of the iodine. The method has been applied successfully to the determination of arsenic in arsenite solutions and, also, to the determination of arsenic in arsenite solutions containing appreciable proportions of antimony.In addition, it is shown that the antimony in a solution containing arsenic and antimony can be determined after a preliminary separation of the arsenic. Details are given of the application of the method to the determination of arsenic in pig-iron. RECESTLY we have carried out a considerable amount of work on the volumetric determination of arsenic. The method we have used is based on previous work by Evans’; it depends on the reduction of arsenic from either its tervalent or quinquivalent condition by means of hypophosphorous acid in hydrochloric acid solution, and on the iodimetric titration of the resulting elementary arsenic.A study of Evans’s original paper leaves little doubt that he experienced considerable difficulty in titrating the precipitated arsenic; this arsenic was filtered on a pulp filter and arsenic plus pulp were treated with a measured amount of standard iodine solution and water. Evans used the so-called titrated water, which contains a few drops of starch solution and enough 0.01 N iodine solution to impart to it a faint permanent blue colour. First, sodium bicarbonate was added to the arsenic, pulp, iodine and titrated-water suspension, and then a slight excess of standard arsenite solution ; after addition of further bicarbonate the excess of arsenite mas titrated with standard iodine solution until the blue end-point was reached.In this way, Evans claimed to get low “blank” values in experiments with reagents and paper pulp alone. However, in a subsequent paper,2 Evans drew attention to the fact that owing to a change in the manufacture of the filter-paper the results in “blank” experiments on the paper pulp were unsatisfactory. For this reason he changed his method of determining precipitated arsenic. In his second method the precipitated arsenic was filtered off on filter-paper pulp that had previously been submitted to a vigorous oxidation with bromine and hydrochloric acid. The arsenic and paper pulp were treated with a slight excess of standard iodine solution, benzene and sodium bicarbonate were added, and then standard arsenite was added in slight excess with intermediate addition of potassium iodide.Further bicarbonate and water were added and the excess of arsenite titrated with standard iodine solution with starch solution as indicator. Our experiments led us to conclude that the two methods of titrating precipitated arsenic proposed by Evans were unsatisfactory and unnecessarily complicated. First we prepared a white material containing a large proportion of oxy-cellulose by treating filter- paper with bromine and hydrochloric acid; this was thoroughly washed. Using this prepared material as our filtering medium, we filtered the precipitated arsenic, treated the arsenic plus pulp with an excess of standard iodine solution in a stoppered bottle and, after adding potassium iodide and sodium bicarbonate, titrated the excess of iodine with standard arsenite solution; the end-point was indicated by the change from the yellow iodine solution to the white of the oxy-cellulose filtering medium.The end-point was exceedingly sharp, and Garden City, Herts. * Present address: Imperial Chemical Industries Limited, Plastics Division, Black Fan Road, WclwynJuly, 19531 ARSENIC BY B. s. EVANS’S METHOD 391 “blank” values in experiments on the prepared pulp alone were low. By this method of titrating the precipitated arsenic, results were excellent in determinations of arsenic in sodium arsenite solutions. These arsenite solutions contained various amounts of arsenic between the limits 0.00038 g and 0.1049 g. Full details of the method appear below and the results of this set of experiments are given on p.392. In a second set of experiments, the results of which are shown in Table I, the arsenic was determined accurately in sodium arsenite solutions containing appreciable proportions of potassium antimony tartrate. Various combinations of arsenic and antimony solutions were examined, the amount of arsenic being varied up to 0.1049 g and the antimony between the limits 0.0050 g and 0.05 g. In addition, it was shown that, after precipitation of the arsenic in these determinations, the antimony in the filtrates could be determined with high accuracy. This antimony was precipitated as sulphide, the sulphide was dissolved and the antimony in the solution was reduced to the tervalent condition before its deter- mination by oxidation to the quinquivalent condition by means of bromic acid.The method fur determining antimony is given on p. 392, and the results are shown in Table 11. The method was extended to the determination of arsenic in the presence of ferric salts. \\‘ith copper as a catalyst to assist in the speedy reduction of ferric iron by means of hypo- phosphorous acid, it was shown that between 0.00075 g and 0.0524 g of arsenic could be readily determined in the presence of as much as 5 g of ferric sulphate. The principle of this method was used in the determination of arsenic in pig-iron (p. 393). THE DETERMIKATIOS OF ARSESIC IN ARSENITE SOLUTIOXS PREPARATION OF THE STANDARD ARSENITE SOLUTIONS- Exactly 4.9455 g of AnalaR* As,O, dried at 105” C were dissolved by heating in a solution containing 2.5g of pellets of sodium hydroxide (made pure from sodium) in a little water.The arsenic solution was cooled, 66 ml of N hydrochloric acid solution were added, and then 10 g of AnalaR sodium bicarbonate. The solution so prepared was diluted to 1 litre with distilled water. This arsenite solution was standardised against iodine solution by the method given in a paper on the standardisation of volumetric sol~tions.~ Twenty millilitres of the arsenious solution were equivalent to 20 ml of 0.1 N iodine solution; hence 1 ml of this arsenite solution contained 0.003746 g of arsenic. A 0.01 N arsenite solution was prepared from the 0.1 N solution by dilution. PREPARATION OF THE STANDARD IODISE SOLUTION- given in the paper referred to above.3 PREPARATIOX OF THE FILTERING MEDIUM- About twenty 11-cm Whatman No.40 filter-papers were digested on a water-bath for about 4 hours with about 400 ml of water containing 4 ml of concentrated hydrochloric acid solution and 2011-11 of saturated bromine water. The pulped material thus prepared was stored in a glass bottle and small amounts of it were washed thoroughly with water before use in the filtration of precipitated arsenic. PROCEDURE- A known volume of the standard arsenite solution was diluted to 65ml with distilled water. To this solution, in a 600-ml conical flask, were added 10 ml of 3 M sulphuric acid solution and 75 ml of concentrated hydrochloric acid. After the addition of 4 g of B.D.H. sodium hypophosphite the solution was warmed gently to effect preliminary reduction ; the temperature was not allowed to rise above 50” C in this operation.A further 10 g of sodium hypophosphite were then added and the mouth of the flask was closed by a cork carrying a length of glass tubing, approximately 80 cm long and 0.5 cm in internal diameter, which acted as a reflux condenser. The solution was then cooled and the precipitated arsenic filtered on to a pad of the previously prepared * AnalaR A%O, contains small amounts of Sb,O,, b u t we have found the proportion, as determined spectrographically, to be of the order of 0.1 per cent. only; this amount is too small to cause interference. Standard 0.1 N and 0.01 N iodine solutions were prepared according to the methods The solution was brought to the boil and boiled gently for 15 minutes.392 HASLAM AND WILKINSOX : THE DETERMISATIOK OF p o l . 58 filter pulp.With large amounts of arsenic it is better to mix a little well-washed filter- paper pulp with the precipitate in the conical flask before filtering, in order to keep the precipitate in a finely divided condition so that it will readily dissolve in the iodine solution used for the titration. The precipitated arsenic was washed with an acid hypophosphite solution prepared by adding 2 g of sodium hypophosphite to 100 mi of a solution of diluted hydrochloric acid (1 + 3), and the hypophosphite was removed by repeated washings with 5 per cent. ammonium chloride solution. The arsenic precipitate and paper pulp filter were transferred to a wide-mouthed 16-02 glass-stoppered bottle to which was added about 50 ml of distilled water.After thorough disintegration of the precipitate and addition of 2 g of potassium iodide and 2 g of sodium bicarbonate, an excess of 0.1 N or 0.01 N iodine solution was added. In practice it was found that the particular iodine solution used was decided by the amount of precipitate. With either solution, an approximately 10-ml excess was used. After the bottle had been thoroughly shaken to ensure complete solution of the arsenic precipitate, the glass stopper was removed, washed with water, and the excess of iodine immediately titrated with either 0.1 AT or 0.01 AT arsenite solution according to the concentration of iodine solution used. The end-point, i.e., the point a t which the yellow of the solution of iodine in potassium iodide changed to the white of the oxy-cellulose pulp, was quite sharp. The relevant equations are : for the preliminary reduction of quinquivalent arsenic- 2AsC1, + 2H3POz -+ 2H20 = 2AsC1, + 4HC1+ 2H3P03 for the final reduction of tervalent arsenic- for the iodine titration- 2AsC1, + 3H3P0, + 3H20 = 2As + GHC1+ 3H3P0, 2As + 51, + 5H20 = As,O, + lOHI An experiment on the reagents carried out by the above method, but omitting the known volume of standard arsenite solution, gave a “blank” value of 0.05 ml for a 0.01 A‘ solution.The following results were obtained by the application of the above method to known arsenite solutions- Arsenic added, g . . . . 0.00038 0.0011 0.0030 0,0094 0.0524 0.1049 Arsenic found, g . . . . 0.00032 0.0010 0.0030 0.0092 0.0524 0,1048 APPLICATION TO ARSENITE SOLUTIOXS CONTAIKING APPRECIABLE AMOUNTS OF ANTIMONY- A standard antimony solution was prepared from antimony potassium tartrate ; known The Results volumes of this solution were added to known volumes of standard arsenite solution.arsenic in these solutions was then determined by the method described on p. 391. are shown in Table I. TABLE I RECOVERY OF ARSENIC IN PRESEXCE OF ASTIMONY g g g Experiment No. Arsenic added, Antimony added, Arsenic found, 1 nil 0.005 Not detected 2 nil 0,050 3 0.00075 0.050 0.00071 4” 0.0026 0.010 0.0026 5 0,00375 0.005 0.00383 6* 0.0054 0.010 0.0056 7* 0.0075 0.0 10 0.0074 8* 0,0107 0.005 0.0103 9 0.0262 0.010 0.0262 10 0.0524 0.050 0.0521 11 0,1049 0.002 0.1046 12 0.1049 0.010 0.1047 * In these experiments the amounts of arsenic and antimony present were unknown to the operator.THE DETERMINATION OF ANTIMONY IN FILTRATES OBTAISED AFTER PRECIPITATING ARSENIC The filtrate from the arsenic precipitation was first boiled down to about 350 ml. Fifty millilitres of concentrated ammonium hydroxide solution were then added in order to neutralise WITH HYPOPHOSPHOROUS ACID-July, 19531 ARSENIC BY B. s. EVANS’S METHOD 393 most of the acid present, after which the antimony was precipitated by the passage of hydrogen sulphide. The precipitated sulphide was filtered off and washed with water saturated with hydrogen sulphide until the washings were free from chloride. After a hole had been pierced in the bottom of the filter-paper, the bulk of the antimony sulphide was washed into a clean beaker.The trace of antimony sulphide remaining on the filter was then dissolved by pouring on to the filter-paper a hot mixture of 20 ml of concentrated hydrochloric acid, 4 ml of 10-volume hydrogen peroxide and 10 ml of bromine water. The filter was washed with water and the mixture of antimony sulphide and acid, and so on, was brought to the boil. A mixture of 2 ml of 10-volume hydrogen peroxide solution and 5 ml of bromine was added to the boiling solution. The antimony precipitate was completely dissolved by two more additions of peroxide and bromine. The solution was cooled and diluted to approximately 250 ml with distilled water, 5 g of AnalaR sodium sulphite were added and dissolved by gentle mixing, and the solution was set aside for about 1 hour.The solution was then boiled to remove sulphur dioxide, 10 ml of concentrated hydro- chloric acid and 1 drop of methyl orange indicator were added and the hot solution was titrated with 0.02 N potassium bromate solution until the colour of the methyl orange indicator was just removed. 1 ml of 0.02 N potassium bromate solution = 0.001218 g of antimony. The results were as shown in Table 11. TABLE I1 RECOVERY OF ANTIMONY IN PRESENCE OF ARSENIC The excess of bromine was then removed by boiling. Expcriment So. Arsenic added, Antimony added, Antimony found, p. 6 6 11 (continued) 0.1049 0.0020 0.0023 3 (continued) 0.00075 0.0500 0.0500 0 (continued) 0.0262 0~0100 0.0102 THE DETERMISATION OF ARSENIC I N THE PRESENCE OF FERRIC SALTS- Five grams of B.D.H.ferric sulphate were added to a known amount of standard arsenite solution. Half a gram of AnalaR copper sulphate (CuSO4.5H,O) was added to catalyse the reduction of the ferric salt by the hypophosphorous acid. The subsequent procedure was similar to that given in the method on p. 391 ; the results obtained were as follows- Arsenic added, g . . . . . . nil 0,00075 0.0524 Arscnic found, g . . . . . . not detected 0.00067 0.0518 THE DETERMINATION OF ARSESIC IN PIG-IRON One gram of copper sulphate (CuS0,.5H20) was added to a 5-g sample of pig-iron, and the whole dissolved in a mixture of 30 ml of 3 M sulphuric acid solution and 15 ml of con- centrated nitric acid. The insoluble carbon was filtered off and 2 ml of a saturated aqueous solution of potassium permanganate were added to the filtrate. The liquid was boiled for 5 minutes, after which a little sulphurous acid solution was added in order to remove the excess of permanganate. The solution was evaporated, first on a water-bath and then on a sand-bath, until all the nitric acid was removed. I t was necessary to maintain the beaker in constant motion during the evaporation on the sand-bath in order to prevent “bumping.” The pasty mass of ferric sulphate was dissolved in water and the arsenic determined as previously described except that no preliminary reduction was carried out and the final reduction was accomplished by the addition of 14g of sodium hypophosphite. In this experiment the hypophosphite solution was boiled for half an hour in order to reduce the arsenic completely. The figure for arsenic found, 0.447 per cent., compared well with the figure of 0.44 per cent. found by the ferric chloride distillation procedure. REFEREXES 1. 3. Evans, B. S., Analyst, 1929, 54, 523. Analytical Chemists’ Committee of I.C.I. Ltd., Ibid., 1950, 75, 577. 2. -, Ibid., 1932, 57, 492. IMPERIAL CHEMICAL INDUSTRIES RESEARCH DEPARTMENT ALKALI DIVISION WINNINGTON, NORTHWICH CHESHIRE January 12fh, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800390

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

394 SEYFANG AND SMALES: TNE DETERMINATION OF URANIUM-235 I N [VOl. 78 The Determination of Uranium-235 in Mixtures of Naturally Occurring Uranium Isotopes by Radioactivation BY A. P. SEYFANG ASD A. A. SMALES A method previously used for determining uranium in minerals by neutron irradiation followed by measurement of the separated fission-product barium has been extended to the determination of uranium-235 in admixture with uranium-234 and uranium-238. With microgram amounts of uranium-235, short irradiations in the Harwell pile give ample sensitivity. Precision and accuracy of better than & 2 per cent. have been achieved for a range of uranium-235 contents covered by a factor of more than lo5. THE most frequently used method for measuring isotopic abundance-mass spectrometry -has been applied to uranium by Nier, Inghram and Neyl and many other workers.Other methods studied have included fission-fragment counting2 and a spectrographic tech- nique involving isotopic line shift. Derham and Fenning3 have described a method depending on the measurement of the growth of P-emitting daughters from the chemically separated uranium, but this seems to be suitable mainly for samples containing much uranium-235 and little uranium-234. Alpha-counting methods have also been used.4 is essentially a method for measuring isotopes rather than elements, and it has been used for determining individual coppeiG and chlorine' isotopes. The application of radioactivation to the determination of uranium in minerals was recently reported by Smaless and, as this method gave a measure of the uranium-235 content of the mineral by relying on the constant isotopic content of uranium in nature to give the total uranium, it was a logical step to iisc it as a basis for the direct determination of uranium-235 in mixtures of the naturally occurring uranium isotopes.Radioactivation analysis, as already pointed APPLICATIOX OF RADIOACTIVATION TO DETERMINATION OF URANIUM-235 Natural uranium contains 99.274 per cent. of uranium-238, 0.72 per cent. of uranium-235 The reactions of these isotopes with slow neutrons and 0.0061 per cent. of uranium-234. can be briefly summarised as follows- 238 239 (i) 9 , ~ -I- tn + 'iiu P-+ 'i: r \ ; p A + 94 PU (ii) 92 U + An .j Fission Products (iii) 92 u + ;?z -+ *;: u 235 23 4 The production of uranium-235 from uranium-234 is negligibly small in the short irradia- tion periods discussed in this work, owing to the long half-life (8.9 x 108 years) of uranium-235. From this it follows that the amount of a particular fission-product, measured by counting, is directly related to the uranium-235 content of the sample.Further, if the sample and a standard of normal uranium (known to contain 0.72 per cent. of uranium-235) are irradiated, chemically treated and counted under the same conditions, the following ratio will hold- Corrected count of standard - IVeight of 235U in standard Corrected count of sample - \;l'eight of 235U in sample As the choice of barium-140 as a suitable radio-nuclide for isolation and measurement has been fully discussed elsewhere,s it is only necessary to mention briefly the following points in its favour.(1) High fission yield (6.2 per cent.), convenient half-life (12.8 days) and p- energy (maximum 1.0 MeV). (2) Good chemical separation properties from other radioactive fission products. One valency state in solution, which favours quick and complete exchange with the inactive carrier barium. Good final weighing and counting form (barium sulphate).July, 19531 MIXTURES OF NATURALLY OCCURRING URANIUM ISOTOPES 395 The effect of fast neutron fission of uranium-238 will be considered in a later section, but it may be noted here that the effect is negligible for ratios of 235U to 238U greater than 1 to 1O00, in the “self-serve” positions of the Harwell pile. In this work only pure uranium oxide or its solution is irradiated; the presence of barium as an impurity, possible in ores, isnot amajorproblem here.Barium-139 (half-life 85 minutes; ,&energy 2.3 MeV), which is also a product of fission, may therefore be used instead of barium-140, with certain advantages arising from its shorter half-life (the fission is about the same for both). Fig. 1. Decay of fission-product barium. Fifty mg of U,O, irradiated for First ,j minutes a t 9.30 a.m. count a t noon. Barium sulphate precipitated a t 11.30 a.m. Curve A, count through absorber (197 mg per sq. cm) Hence for a given short irradiation time of a few minutes, a much greater relative activity of barium-139 is attainecl. Also, a decay curve, which can give a valuable check on radio chemical purity, can be obtained more quickly. But apart from a decay curve for the 85-minute barium-139, shown in Fig. 1, all the work described in this paper was carried out with barium-140 because it is more suitable for use in laboratories not conveniently near an atomic pile.NEUTRON SOLJKCE- The Hanvell pile was used as the neutron source for all the work described here except for one experiment, which was carried out with a 0.5-curie radium - beryllium source. In corifirmation of the calculated value, which assumed a neutron flux of lo4 neutrons per sq. cm per second, the barium-140 activity after a 13-day irradiation in this source was only 10 counts per minute per g of natural uranium. The Hanvell pile flux is about 1012 neutrons per sq. cm per second, which gives ample sensitivity for much shorter irradiations. A useful nomogram for calculating fission-product activities is given by Stang and Hance.9 STANDARDS- The comparative method of activation analysis,6 in which the sample is compared with a standard, was used.For this work, analytical reagent grade uranyl nitrate of natural isotopic composition was further purified by partition chromatography and converted to U,O,, which was used as standard. S E L F SHIELDING- If a material undergoing irradiation has a high capture cross-section, the flux at the centre of a mass of such material will be less than at the surface. For solid U30,, of natural isotopic composition and density of 3.5 g per cu. cm, irradiated in cylindrical polythene tubing of 2 mm internal diameter, the self-shielding effect, calculated from the formula of Seaborg, English, Wilson and Coryel1,lo can be shown to be negligible. This conclusion was checked experimentally by irradiating 50-mg portions of U,O, diluted uniformly with 0, 1, 2, 5, 10 or 20 volumes of analytical reagent grade magnesium oxide (acapt.<0.05 barns) in sealed396 SEYFASG AND SMALES: THE DETERMINATIOX OF UKANIURI-235 I N [VOl. 78 short lengths of polythene tubing for 5-minute periods. In this way identical amounts of U,O, were “dispersed” by increasing amounts of magnesium oxide. The separated barium- 140 activity, calculated as counts per minute per milligram of U,O,, was identical in all irradia- tions within the experimental error. For samples containing significantly greater amounts of uranium-235, in which the total neutron cross-section might be tens or even hundreds of barns, significant effects could appear, which would make irradiation of dilute aqueous solutions of the sample in sealed silica tubing necessary.The effective neutron cross-section of the solution, mainly water, would then be minute. CHEMICAL SEPARATION The aim of the chemical separation after irradiation is to isolate a radiochemically pure barium compound, i.e., one free from all other radioactive species derived either from trace impurities in the uranium or from the other products of neutron irradiation; these include neptunium-239 as well as other fission products. The chemical separation eventually used is described in full in the procedure on p. 397, but it is similar to that used on the Manhattan Project and described by Coryell and Sugarman12; it is also largely that used by Smaless for application to minerals.As a known quantity of inactive barium carrier is added there is no necessity for quantitative isolation (except insofar as a loss of sensitivity follows from a low chemical yield). It is desirable to have similar yields from standard and samples, so that corrections for self absorption, and so on, in the counting become negligible. REAGENTS- METHOD Magnesium oxide-Analytical reagent grade. Nitric acid, sp. gr. 1.42. Barium chloride solution-Dissolve 18 g of BaC1,.2H20 in water and make up to 500 ml. Lanthanum nitrate solution-A 1 per cent. w/v solution of La(NO3),.6H,O. Ammonium hydroxide, sp. gr. 0.880.Strontium carbonate solution-A 2 per cent. w/v solution. Hydrochloric acid - diethyl ether reagent-A mixture of 5 parts of concentrated hydro- Sodium tellurate solution-A 0.4 per cent. w/v solution. Zinc metal 9owder. Methyl orange indicator. Potassium iodide solution-A 1 per cent. w/v solution. Sodium hypochlorite solution-A commercial solution containing 10 per cent. of available Hydroxylamine hydrochloride. Ferric chloride solution-A 1 per cent. w/v solution. Sul~huric acid-A 20 per cent. v/v solution. IRRADIATIOS- Solids-As mentioned in the discussion on self-shielding, samples containing not much more uranium-235 than natural uranium (say, up to three times more or 2 per cent.) may be irradiated as solid; this is usually U,O,. For these cut a 5-cm length of 2-mm polythene tubing and seal one end by warming and pressing.Introduce freshly ignited analytical reagent grade magnesium oxide to form a compact layer 4 to 5 mm in height at the sealed end of the tube. Add a further similar layer of magnesium oxide on top of the U,O, and then seal the open end. Leave a free space about 1 cm long between the top of the magnesium oxide layer and the seal, for ease of opening. Treat standard and samples similarly. Place the tubes either in a special polythene bottle for irradiation in the pneumatic “rabbit” of the pile or in a 3-inch aluminium can for irradiation in the “self-serve” holes in the pile. Irradiation is carried out for any required time; usually it is about 5 minutes. After irradiating, place the containers in lead shielding for about 15 hours.After this period, tap down the contents of the polythene tube away from one end and carefully cut off the top. Empty the contents into a 50-ml centrifuge tube. (The plug of magnesium oxide serves to “rinse” the sample tube as it is emptied.) Add 2 ml of concentrated nitric acid (sp.gr. 1*42), gently warm to dissolve, and chloric acid, sp. gr. 1.18, and 1 part of diethyl ether. chlorine. Weigh the tube and contents, add about 50 mg of U,O, and re-weigh.July, 19531 MIXTURES OF NATURALLY OCCURRING URANIUM ISOTOPES 397 finally boil off the nitrous fumes. Add 5.00 ml of a barium solution to act as carrier (a solution of 18 g of barium chloride, BaC1,.2H20, in 500 ml of water). Liquids-For more highly enriched samples or when the amount of sample available is small, solutions containing weighed quantities of solid sample must be irradiated in small silica ampoules.The ampoules, which have a capacity of about 1 ml, are prepared from silica tubing. After one end of each has been sealed, the ampoules are weighed, the sample solution added from a fine-pointed glass dropping-tube and the ampoules re-weighed. Pack the ampoules, after sealing the open ends, in cotton wool in a 3-inch aluminium can and irradiate them in the “self-serve” position of the pile. The time of irradiation necessary can be calcu- lated from the usual activation formula; as an example, 1 pg of uranium-235 irradiated for 24 hours in a flux of 1012 neutrons per sq. cm per second gives about 6000 counts per minute of barium-140 at 5 per cent.counting efficiency, 24 hours after the irradiation. After removing them from the pile, place the samples and standards in lead shielding for about 15 hours; the main activity is due to silicon-31. Transfer the ampoules to 100-ml tall-form beakers containing a few millilitres of water and 5.00 ml of barium carrier solution, carefully break off both ends of each ampoule and warm to ensure thorough mixing. Decant into centrifuge tubes and wash out the ampoules and beakers with further small portions of water. CHEMICAL SEPARATION- Evaporate the solution containing the irradiated uranium and barium carrier to 5 to 6 ml and add two drops of 1 per cent. lanthanum nitrate solution. Warm if necessary to dissolve any barium nitrate that may have crystallised, add concentrated ammonium hydroxide dropwise until a permanent precipitate is obtained and then two drops in excess.Centrifuge and decant into another centrifuge tube. Add methyl orange indicator, and then hydrochloric acid until the solution is acid. Add 2 drops of 2 per cent. strontium solution, about 25 ml of hydrochloric acid - diethyl ether reagent, mix thoroughly, centrifuge and decant. Wash the precipitate with 5 ml of reagent, centrifuge and decant. Dissolve the barium chloride precipitate in 3 to 4 ml of water, re-precipitate it by adding 20 ml of reagent, centrifuge and decant. Dissolve the precipitate in about 5 ml of water, add 6 drops of lanthanum solution and 6 drops of the 4 per cent. tellurate solution and then about 3 mg of zinc metal powder.When the effervescence ceases, make the solution just ammoniacal to methyl orange, centrifuge and decant into another tube. Add 4 drops of 1 per cent. potassium iodide solution and 2 drops of sodium hypochlorite solution. Acidify with about 1 ml of hydrochloric acid, and add about 0.1 g of hydroxylamine hydrochloride. Boil under a hood until all the iodine appears to be removed and the volume is reduced to 5 to 6 ml. Add 2 drops of strontium solution and 2 drops of lanthanum solution and repeat the double barium chloride precipitation and washing, as above. Dissolve the precipitate in about 5 ml of water, add 6 drops of lanthanum solution, and 6 drops of 1 per cent. ferric chloride solution. Make ammoniacal to methyl orange, add half a crumbled Whatman accelerator tablet, and heat just to boiling. Filter through a 7-cm Whatman No.30 filter-paper into a centrifuge tube, wash twice with 2 to 3-ml portions of water. Dilute the filtrate to about 20ml and make slightly acid with hydrochloric acid. Heat nearly to boiling and add dropwise 2 ml of 20 per cent. v/v sulphuric acid. Allow the precipitate to settle, decant, wash with 10 ml of water, centrifuge, decant and repeat the washing procedure to complete removal of the excess of acid. Transfer as much as possible of the precipitate, by means of a dropping tube and a few drops of water, to a tared aluminium counting tray. Dry under an infra-red lamp and finally heat in a mufRe furnace at 500” C for 15 minutes. COUNTING TECHNIQUE- The counting equipment for this work consists of (i) a power unit (type 1082A or 200 is suitable), (ii) scaling unit (type 200 or 1009R), (iii) time accessory unit (type 1003B), (iv) probe unit (type 200B or 1014A).Time pulses can be obtained from a master electric clock serving several units. A mica end-window Geiger - Muller counter (2 mg per sq. cm), of type EHM2, is suitable; it is mounted in a lead castle with a Perspex lining and shelves. Check the counting equipment in the normal fashion with a suitable beta-emitter, such as natural uranium oxide in equilibrium with UX, and UX,. Place the sample to be counted Wash with 5 ml of reagent, centrifuge and decant. Warm and set aside for 2 minutes. Cool, weigh and reserve for counting.398 SEYFANG AND SMALES: THE DETERMINATION OF URANIUM-235 I N [VOl.78 in a Perspex carrier and insert it in a shelf at a suitable distance from the Geiger - Muller tube to attain a counting rate of 2000 to 3000 counts per minute. Count for a sufficient time to obtain a t least 10,000 counts for each barium sulphate precipitate, counting the pre- cipitates one after another without undue delay. Correction for growth of lanthanum-140 is unnecessary if samples and standards are counted within, say, 60 minutes of each other, provided the barium sulphate precipitations are carried out on each nearly simultaneously. CALCULATION OF RESULTS- Correct all counts for background, coincidence loss and chemical yield and express as the results in counts per minute. Weight of 236U in standard - Corrected count of standard Corrected count of sample - Then Weight of 2 % ~ in sample x 100 = percentage of uranium-235 in sample.Weight of 235U in sample and Weight of sample "DECOSTAMIKATION" EXPERIMENTS Some idea of the separation requirements, or decontamination necessary, can be obtained from the calculated activities of the fission products of uranium after various decay times. A useful paper by Hunter and Balloull gives these details; Table I shows a few relevant figures taken from the graph in that paper. TABLE I RELATIVE ACTIVITIES OF SOME NUCLIDES PRODUCED BY SLOW NEUTRON FISSION OF URANIUM-235 AS PERCENTAGE OF ORIGINAL TOTAL FISSION-PRODUCT ACTIVITY One day after irradiation A f > Relative Nuclide activity lroBa 1.25 1sXe 12.5 97Nb 9.8 O7Zr 9.0 S S Y 7.5 7.5 6.8 6.8 4.7 4.7 4.1 3.0 3.0 2.7 2.7 1.5 1.5 1.5 Ten days after irradiation r Relative Nuclide activity lJa13a 12.5 lroLa 12.5 l33Xe 11.5 1r3Pr 10.0 181 [ 7.0 A > 7.0 6.5 5.5 6.5 4.8 3.4 3.3 3.0 2.6 To test the efficiency of the separation scheme for relevant fission product elements whose behaviour was not certain, experiments were carried out to determine decontamination factors as follows.(In these experiments the radio-nuclide may not necessarily be in exactly the same chemical form as in the actual fresh fission products present in a sample after irradiation. Nevertheless they are of some value.) A solution of a suitable radio-nuclide was prepared, an aliquot added to 50 mg of U,O, dissolved in nitric acid and, after barium carrier had been added, the chemical separation was carried out.The activity of the solution, before the final barium sulphate step, was measured in a type M6 Geiger - Muller tube. After precipitation, the activity of the barium sulphate was measured with an end-window Geiger - Miiller tube of type EHM2. Details of the nuclides used and results obtained are given in Table 11. Decontamination was satisfactory for zirconium, molybdenum and ruthenium - rhodium and for iodine when a specific removal step was included in the procedure. Earlier workersJuly, 19531 MIXTURES OF NBTURALLY OCCURRING URANIUM ISOTOPES 399 have found that fission-product iodine is liable to be present in different forms; so to ensure exchange with added carrier an oxidation reduction step was necessary. Oxidation by sodium hypochlorite in ammoniacal solution followed by reduction to elementary iodine by hydroxyl- amine satisfactorily overcomes this difficulty.TABLE I1 Tracer and mode of preparation 06Zr ( I Z , ~ ) 97Zr (zir- 1311 carrier-free fisslon conium nitrate) 9 w 0 ( H , ~ ) @@Mo (MOO,) product 106Rn / Rh (separated 88Sr (n,y) 89Sr (“Spec- fission product) pure” SrCO,) Remarks “DECONTAMINATIOS FACTOR” EXPERIMENTS Decon- Activity -4ctivity Activity tamination added, of solution, of BaSO,, factor counts per counts per counts per minute minute minute 2.3 x 106 11 3 > 105 1.6 >: 106 3 1 N 106 1.1 x 106 4 to 12 20 - N 106 No specific iodine 1.1 x 108 - 1 to 4 2i 106 Iodine removal step removal step included 2.02 x 106 1 1 1: 2 x 10’ 0.8 X lo6 8000 3000 N lo* For strontium it was found that 99 per cent.of the activity was removed by the first double barium chloride precipitation step, but little more by the remaining steps. Examina- tion of the decay and absorption curves of the residual activity of the barium sulphate showed it to be due almost certainly to barium-135 (half-life 29 hours) and barium-131 (half-life 12-0 days) derived from trace barium impurity in the “Specpure” strontium carbonate irradiated. This trouble will not arise in the usual determination of uranium-235, and there is little doubt that the decontamination from strontium is quite adequate. SEPARATION FROM OTHER RADIO-NUCLIDES- Sodittnz-Aily contamination of the uranium sample before irradiation could give rise t o radio-nuclides; niic possibility is sodium-24. A decontamination factor of 2 x 106 was obtained for this nuclide, on starting with 2.7 x lo6 counts per minute.Impurity in analytical reageiat grade magnesium oxide-The activity accompanying the barium sulphate from 10 mg of magnesium oxide after irradiation for 58 hours amounted to about 30 counts per minute 18 hours after irradiation; again it was shown to be from barium impurity in the magnesia. As the normal irradiation time used for samples of nearly natural isotopic composition is 5 minutes any interference from the 20 mg of magnesium oxide used in packing the U,O, in the polythene tubing can be neglected, but a check should be made on possible interference from each fresh batch of magnesium oxide used. For samples con- taining little uranium-235, for which long irradiation periods are necessary, the magnesium oxide should be dispensed with.Barium-The interference from barium has been thoroughly discussed by Smales.s Only serious contamination, such as from several per cent. of barium, can affect the determina- tion of uranium-235 when the latter is present at almost normal isotopic composition. A simple preliminary separation is then desirable. A check on the over-all radiochemical purity of the isolated barium-140 can be made by plotting the growth and decay curve, but experience has shown that, if the chemical separation described in the procedure is carried out correctly, this check is usually superfluous. A typical growth - decay curve is shown in Fig. 2. PRECISIOX- as described above. described. taken, which gave a standard deviation, on the counting only, of about 0.7 per cent.RESULTS Eight accurately weighed portions of the standard U,O, were irradiated for 5 minutes After 4 days each portion was chemically treated and counted as already For each a total count of about 20,000 was It The results are shown in Table 111.400 SEYFANG AND SMALES: THE DETERMINATION OF URANIUM-235 IN [VOl. 78 can be seen that under favourable conditions a precision of better than 2 per cent. at the 66 per cent. confidence level is possible. PRECISION TABLE I11 OF PROPOSED METHOD WITH STANDARD U,o, Time after RaSO, precipitation U,O, portion 1 n L 3 4 5 6 7 8 Standard deviation , . Coefficient of variation 4 clays, counts per minute per mg of U,O, 136 136 141 138 137 141 141 140 & 2.3 1.7% 23 days, counts per minute per mg of U,O, 160 159 161 157 160 163 159 159 & 1.8 1.1% 7 28 days, counts per minute per mg of U,O, 122 122 123 121 123 125 121 121 f 1.4 1.2% ENRICHED SAMPLES- Sample N o .1-About 12 mg of a sample of U,O, were available, whose uranium-235 content, according to a mass spectrometric determination, was about 14.7 per cent. of the total uranium. I t was examined in solution by the activation method after igniting, weighing and dissolving the oxide and diluting to 250 ml. The standard was a solution of IO.0OC 9000 8000 700C 6000 SOOC u E L .- 400C W n a Y) Y 3000 -0 Y r 8 V 200c lo00 Days Fig. 2. Typical growth -decay curve for lroBa - lroLaJuly, 19531 MIXTURES OF K.4TURALLY OCCURRING URANIUM ISOTOPES 401 0.5000 g of the chromatographically purified U,O, of normal isotopic composition (uranium- 235,0.72 per cent.of total uranium) diluted to 250 ml. Portions of between 0.2 and 0.4 ml of each were irradiated for 11 hours; the results are shown in Table IV. TABLE IV RESULTS BY THE ACTIVATION METHOD FOR SAMPLE NO. 1 Uranium-235 according to mass spectrometric method = 14.7 per cent. Standards Sample 1 -~ A_- rpA--.--..- 7 Amount of 23sU equivalent to 1 count per minute Ba activity, Portion 2 T J , 01 PLR 1'0 0.000317 0~000315 0.000316 Mean 0.000316 1 15.0 ? 15.3 3 15.2 4 14.9 5 14.9 Mean 15.0 rir. 0.2 Sample No. 2-This material had been analysed by Derham and Fenning3 by their p-counting method, by which it gave 80.7 ( i 0 . 4 ) per cent. of uranium-235; by Palmer's mass spectrometric methodl3 it gave 81.2 & 0.1 per cent.After ignition and nitric acid leaching, which separated uranium from a small amount of a red insoluble material (possibly Fe,O,), the solution was diluted to 250ml; it contained 51.4mg of the U,O,. Approxi- mately 0.2-ml portions were irradiated in silica for 4 hours; the results are shown in Table V. TABLE V RESULTS BY ACTIVATION ON HIGHLY ENRICHED SAMPLES Standards Sample I' A > - 1 count per minute Ba activity, Portion a W , Amount of 23sU equivalent to t% 0.00348 0.00335 0,00336 Mean 0.00340 1 2 3 4 % 82.4 82.4 82.6 81.2 5 80.8 Mean 81.9 f 0.8 Synthetic samples-Less enriched samples were also prepared by diluting samples No. 1 and 2 with natural uranium. For these, portions of the solution were mixed with the standard solution so that each mixture contained about 50 mg of U,O,; those from sample 1 contained exactly 1.40, 1.07 and 0.87 per cent.of uranium-235 and those from sample 2 contained 0.89 and 1.05 per cent. of uranium-235, according to the determined figures of 15.0 per cent. and 81.9 per cent. for samples 1 and 2, respectively. The solutions were then treated with ammonium hydroxide, and the precipitates filtered and ignited to U,O,. Each precipitate was divided into two approximately equal portions, which were weighed accurately and irradiated for about 5 minutes in polythene tubes as described above. Results are shown in Table VI. TABLE VI RESULTS BY ACTIVATION ON SYNTHETIC SAMPLES PREPARED BY DILUTION OF SAMPLES NUMBERS 1 AND 2 WITH NATURAL MATERIAL Synthetic oxide Nominal amount of a55U, 23sU found, Yo % .... 1.40 From samule 11; : : .. .. 1.07 0.87 0.89 1.05 1.38 1-08 0.86 0.90 1.07402 SEYFANG AND SMALES : THE DETERMINATION OF URANIUM-235 I N [VOl. 78 DEPLETED MATERIAL- So far it has been assumed that the neutron flux in the pile is composed entirely of slow neutrons and that only the uranium-235 isotope undergoes fission. Nevertheless there is a proportion of fast neutrons at almost any point in the pile. Fast neutrons cause fission of the uranium-238 isotope and so give rise to barium-140 that cannot be differentiated from the barium-140 arising from uranium-235 ; this may therefore cause spurious results. Fast fission * First series of tests Y C a 5 i al \ , , 0 01 \ 1 I\\.\ __ I I 0 1 0 001 0 01 0 1 ssaU found, yo Fig.3. ”Spurious” 233U effect due to fast fission of 238U of uranium-235 can be neglected ( i ) because it is a small effect compared with slow fission, and (ii) because its effect is compensated for in the standard. The calculation of such a “spurious” contribution is made difficult because the fast-fission cross-section of uranium-238 is energy dependent; the neutron energies in the pile cover a wide “spectrum” and consequently the term “fast neutron flux” is difficult to define and measure. Fortunately, however, the practical determination of the “spurious” contribution is simpler, provided a sample contain- ing little uranium-235 is available. A sample of uranium oxide stated to contain 0.00065 per cent. of uranium-235 was used for an examination of the extent of this fast fission effect and portions of this material were mixed (via solution and precipitation) with natural U,O, to give a range of uranium-235 contents below 0.1 per cent.These samples were then irradiated in a self-serve position of the pile (E1/7) together with the natural U,O, standard; the results are shown in Table VII. The constancy of the “spurious” uranium-235 figure that arises from fast fission of uranium-238 is satisfactory and serves not only to help measure this effect but also to confirm the analysis figure of 0.00065 per cent. for uranium-235 given with the sample, although it should be noted that since the actual uranium-235 content of the sample is only about one- tenth of the “spurious” effect, the value obtained for the latter is not entirely dependent on the original analysis.The significance of this “spurious” figure is best seen in Fig. 3, where the experimental and actual values are plotted; it shows that as the uranium-235 content becomes greater than 0.1 per cent., the fast fission effect becomes negligible. If, however, the analysis of samples at or below the 0.1 per cent uranium-235 level becomes necessary, and irradiation is to be carried out in the “self-serve” or “rabbit” positions in the pile, it will be necessary to irradiate, along with the samples, not only a standard of naturalJuly, 19531 MIXTURES OF TATCR;\LLY OCCURRING UR..\NIUZZ ISOTOPES 403 isotopic composition uranium, but also one that contains very little uranium-235 so that the “spurious” effect due to fast fission can be measured in the actual position of the samples in the pile.Even so it is clear that determinations on samples containing much less than 0.01 per cent. of uranium-235 will not be satisfactory, owing to the magnitude of the correction. Fortunately, use may be made of the thermal column of the pile in which, although the slow neutron flux is less than in the “self-serve” or “rabbit” positions and thus sensitivity is lower, the relative decrease in the effective fast neutron flux is much greater (possibly by a factor lo4 compared with a decrease of 10 to 100 for slow neutron flux).14 TABLE VII EFFECT OF FAST FISSIOX OF 2 3 W T J present z36U found (assuming value (assuming value Irradiation Length of of 0*00065 per cent. of 0.72 per cent. date irradiation for initial material), in standard), Spurious V J , hours % % % 22.5.52 50 0.00065 0.00675 0.0061 0.001415 0.00775 0,0063 0.00652 0.0130 0*0066 5.8.52 6 0.00359 0*0100 0.0064 0*01038 0,01676 0.0064 0.01608 0.0221 0.0060 0.0327 0.0400 0.007 A 128-hour irradiation of the sample mentioned above, which was said to contain 0.00065 per cent.of uranium-235, was carried out in the thermal column (T.E.14, 10 feet down) together with the standard U,O,; the results for uranium-235 were 0.00059 and 0.00061 per cent. in the duplicate experiments. This satisfactory result shows that the fast fission effect must be extremely small under these circumstances. HEALTH ASPECTS- The decay of the unshielded total y-activity of 0.1 g of normal U,O,, together with 20 mg of magnesium oxide, irradiated in polythene tubing for five minutes in the “self-serve” position of the Harwell pile, is illustrated in Fig.4. If the normal radiation tolerance is taken as 0.06 roentgens per day, it can be seen that the radiation at one foot from a single sample has dropped to below this tolerance dose after a decay period of only 1 hour. Since most of the remaining activity is removed in the first ammonium hydroxide precipitate the radiation dosage is small, particularly when the normal overnight decay period is included in the procedure. During the whole of the experimental work described no excessive radiation dose, as shown by the standard film badge, has occurred. (All exposures of over one-fifth tolerance in any week would have been reported.) If aluminium cans are irradiated, appreciable activity is emitted for a short time after their removal from the pile, but these cans are normally transferred by experienced pile oper- ators to lead castles to allow the 2.3-minute aluminium-28 to decay.Little activity is found after irradiation of polythene. As mentioned earlier the silica ampoules after irradiation are quite active, but emit only P-activity (with a 170-minute half-life) and may be handled with tongs or behind Perspex, especially after 15 hours decay. It is advisable to wear surgical gloves at least until after the completion of the double barium chloride step. The opening of irradiated polythene and silica tubing should be done carefully because slight pressure builds up during the irradiation. This step, together with that of the initial dissolving in nitric acid, and that of removing free iodine, should be carried out in a fume cupboard.-4 portable monitor should always be used under the actual working conditions, in order that a safe working time may be decided upon. COXCLUSIOXS The activation method described has been shown to have high sensitivity and to give generally an accuracy and precision of better than $_2 per cent. This might be improved by paying particular attention to increasing weighing and, particularly, counting accuracy.404 Even as it stands the method is of value, possibly not so much at close to the natural uraniuni- 235 level, but because the precision quoted should be attainable over a wide range of uranium- 235 contents; in fact, if the thermal column of the pile is used for the lower range, the precision quoted should be attained for samples containing from 0 to 100 per cent.Nevertheless at high uranium-235 contents, it may be preferable to determine the uran- ium-238 content directly, e.g., by separating and counting uranium-239 or neptunium-239. This and similar applications of the method to isotopic analysis of the heavy elements, such as uranium-238 in uranium-233, are a t present under consideration. SEYFASG 4 N D SMALES : THE DETEKMISATION OF UHAXIUhI-235 I N [Vol. 7 8 Time from radiations, hours Fig. 4. Decay of total y-activity of 0.10 g of U3O8 Curve A, at 1 foot; curve B, irradiated for 6 minutes. at 3 inches; P, radiation aftcr 24 honrs a t 3 inches Apart from the use of the Harwell pile, in which facilities may be made generally avail- able, no special apparatus other than that found in a modern laboratory is required.The actual manipulation and technique can readily be acquired by any competent laboratory chemist even without previous radiochemical experience. Our thanks are due to Dr. R. H. Dawton, Mr. P. Egelstaff and Mr. F. W. Fenning for providing some of the samples used; to Mr. Fenning for valuable discussion: to the Chief Chemical Inspector for giving leave of absence from Springfields to one of us (A.P.S.) during the period January to June, 1952, in order that the experimental work could be carried out a t Harwell; to Chemical Inspectorate, Springfields, for the provision of the purified normal U,O,; to members of the Pile Operating Group for their unfailing assistance; and finally to the Director, A.E.R.E., for permission to publish. REFEKEXCES 1. Nier, A . O., Inghram, M. G., and Key, E. P., U.S. Atomic Energy Commission Report r1313, October, 1942; rejerred to by Quill, L. L., in Rodden, C. J., (Editor) “Analytical Chemistry of the Manhattan Project,” McGraw-Hill Book Co. lnc., New York, 1950, p. 727. Whittaker, A,, Atomic Energy Research Establishment lieport N/R665, H.M. Stationery Ofice, 1951. Derham, J., and Fenning, F. W., Atomic Energy Research Establishment Report Ii/R834, H.M. Stationery Office, 1951. 2. 3.July, 19531 MIXTURES OF NATURALLY OCCUHRIKG URANIUM ISOTOPES 405 4. Clark, F. L., Spencer-Palmer, H. J., and Woodward, R. N., British Report BR431, H.M. 5 . Smales, A. A., Ann. Rep. Prop. Claenz., 1949, 46, 285. 6. Swartout, J . A, quoted by Boyd, G. E., and Hume, D. N., i n Rodden, C. J., (Editor) “Analytical Chemistry of the Manhattan Project,” McGraw-Hill Book Co. Inc., New York, 1950, Chapter XII. 7. Kennedy, J. W., and Seaborg, G. T., Phys. Rev., 1940, 57, 843. 8. Smales, A. L4., Analyst, 1952, 77, 778. 9. Stang, L. G., jun., and Hance, P. D., III, Nucleonics, 1952, 10, 48. 10. Seaborg, G. T., English, S. G., Wilson, V. C., and Coryell, C. D., U S . Atomic Energy Commission 11. Hunter, H. F., and Ballou, N. E., Nucleonics. 1951, 9, c2. 12. Coryell, C. D., and Sugarman, N., (Editors) “Radiochemical Studies. The Fission ProductsJJJ 13. Palmer, G. H., Private communication. 14. Stationerv Office, May, 1944. Report MDDC763, 1946, p. 15. Book 3, McGraw-Hill Book Co. Jnc., New York, 1951, pp. 1460 and 1657. Fenning, F. M7., Private communication. ANALYTICAL CHEMISTRY GROUP XTOMIC ENERGY RESEARCH ESTABLISHMENT H.~RwBLL, BERKSHIRE October 24th, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800394

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

July, 19531 MIXTURES OF NATURALLY OCCUHRIKG URANIUM ISOTOPES 405 The Amperometric Titration of Traces of Ammonia with Hypobromite at the Rotated Platinum Wire Electrode Application to the Deterininatioii of Nitrogen in Organic Coinpouiids BY I. M. KOLTHOFF, W. STRICKS ASD L. MORREN Current - voltage curves of sodium hypobromite solutions at the rotating platinum electrode as indicator electrode have been determined in buffer solutions of pH values between 8.3 and 13.0, both in the absence and presence of oxygen. In air-saturated sodium bicarbonate solutioii the current nieasured a t +Om2 volt with respect to the saturated calomel electrode is proportional to the concentration of hypobromite. Use of this is made in the amperometric titration of arsenic trioxide and ammonia. Procedures are given for the rapid titration of arsenite, of ammonia and of nitrogen in organic compounds after a Kjeldahl digestion.Ammonia can be determined at con- centrations between G s and 4 x A4 with an accuracy and precision better than 2 per cent. At high dilutions the amperometric titration methods with hypobromite are simpler, more rapid and more accurate than previous methods. The methods described in this paper should find application to the ainperometric titration of many substances that react stoicheiometrically with hypobroniite. SODIUM hypobroinite is a well-known \701umetric reagent, but it has never been used in amperometric titrations. In this paper the voltamrnetry of hypobromite a t the rotated platinum electrode is described and conditions are established under which traces of ammonia can be titrated rapidly, simply and accurately with hypobromite by the amperometric technique.Analytical use has been made of the oxidation of ammonia by hypobromite for almost 30 years. Artman and Skraball and, independently, Rupp and Rossler2 showed that ammonia and urea can be titrated with hypobromite. ,4 critical study of these titrations has been reported by Kolthoff and Ti trations with hypobromite found wide application to the determination of nitrogen in biological materials, so obviating a distillation of the Kjeldahl digest (for examples see references 4 to 10). The determinations were carried out by adding an excess of hypobromite and titrating with tliiosulphate after addition of iodide or by potentiometric titration. From current - voltage curves for hypobroinite solutions at the rotating platinum electrode it ~7as inferred that it should be possible to carry out titrations with hypobromite amperometrically.In practice it was found that, with the rotated platinum electrode as406 KOLTHOFF, STRICKS AND hIORRES THE XMPERO.IIETRIT [Val. 78 indicator electrode, traces of ammonia can be titrated more simply, rapidly and accurately than by any other titration method. The end-point in amperometric titrations with hypo- bromite can be detected very sharply. The method has been successfully applied in this laboratory to routine determinations of protein nitrogen in blood sera and their albumin and globulin fractions. EXPERIMESTAL MATERIALS USED- Water was made ammonia-free by redistillation from dilute sulphuric acid medium in an all-Pyrex glass apparatus, air being excluded.Glutathione in the reduced state was a Pfanstiehl product. The purity of this product was 99 per cent. as determined by titration with cupric copper.ll The stock solution used was 0.024 M in glutathione, which corresponded to 1 mg of nitrogen per 1 ml of solution. The water content of this product was found to be 6.0 per cent. by heating a t 110" C to constant weight. A 10-3 M stock solution of albumin was prepared, containing 1.02 mg of nitrogen in 0.1 ml. Stock solutions of ammonium chloride and ammonium sulphate were 0.003 M . Stock solutions of sodium hypobromite (0.06 M in hypobromite and 0.1 M in sodium hydroxide) and of arsenious oxide (0.025 M) were prepared according to conventional procedures.l2S13 The stock solution of hypobromite was stored in a dark bottle in a refrigerator a t 4" C.Under these conditions the titre of this solution was found to decrease by about 10 per cent. within three months. Stock solutions of sodium bicarbonate and of borax were 0.6 !M and 0.1 Jf, respectively. The titrations were carried out with solutions prepared from stock solutions by Crystallised bovine plasma albumin was an Armour product. appropriate dilution with ammonia-free water. EXPERIMESTAL METHODS- Current - voltage curves were prepared with a Heyrovskgj self-recording polarograph and with a manual apparatus and circuit,14 which was also used for amperometric titrations. Potentials were measured against the saturated calomel electrode. Amperometric titrations were also carried out with the simplified apparatus and circuit described by Kolthoff and Harris.15 In the present work with this apparatus a reference electrode was used that has a potential of +O.lS volt against the saturated calomel electrode.The electrolyte of the reference half-cell is a saturated mercurous chromate solution in M potassium chromate. Mercurous chromate was prepared as described by Gmelin and Krautle by precipitation from an excess of a slightly acid mercurous nitrate solution with potassium dichromate. The product was filtered, washed with water and suspended in a M potassium chromate solution. In order to protect the titration mixture from contamination with chromate, a vessel filled with saturated potassium nitrate was interposed between each half-cell and the titration vessel, the solutions in the two half-cells being connected by means of an agar - potassium chromate and an agar - potassium nitrate bridge, respectively.An unused platinum electrode must be cleaned with concentrated nitric acid and rinsed with water. If not in use, the electrode is kept in distilled water. One- arid two-miliilitre semi-micro burettes graduatcd at each 0.01 ml were used in the t iti-at ions. The pH was measured with a Beckman pH meter, Laboratory Model G. Oxygen was removed from hypobromite solutions by a stream of pure nitrogen, which was passed through two wash-bottles containing hypobromite solutions of the same com- position as that in the test mixture.A layer of mercury serves as the electrode of the half-cell. A motor provided rotation a t 1800r.p.m. for the platinum electrode. CURREST - VOLTAGE C ~ R V E S AT ROTATISG PLATINUX WIRE ELECTRODE- Current - voltage curves were prepared for hypobromite solutions of various concen- trations of sodium hypobromite and pH, both in the absence and presence of oxygen. Hypobromite is reduced a t the rotating platinum electrode and gives a reduction wave which, under proper conditions, exhibits a well-defined diffusion current region in the absence of oxygen. Current - voltage curves of hypobromite a t different concentrations in air-freeJuly, 19531 TITRATIOS OF TRACES OF AMMONIA 407 medium at pH 10.3 are shown in Fig. 1. Current -voltage curves for each concentration were plotted from measurements of the current while the potential was varied first from positive to negative values and then in the reverse direction.It is seen from Fig. 1 that the curves plotted in these two ways are different in shape, which indicates that the electrode 28 24 20 Potential, V Fig. 1. Current - voltage graphs of hypobromite in the absence of oxygen, a t different concentrations in a carbonate - bicarbonate buffer solution (pH 10.35). Curve A, 1.45 x 10-4 M sodium hypobromite; curve B, 0.65 x 10-4 M sodium hypobromite Potential, V Fig. 2. Current - voltage graphs of 0.65 x lO-'M hypobromite in air-free solutions a t various pH values. Curve A, 0.24 M bi- carbonate solution (pH 8.3) ; curve B, 0.05 M borax solution (pH 9.2) ; curve C, 0.15 M carbonate - 0.05 M bicarbonate solution (pH 10.35) ; curve D, approximately 0.1 M sodium hydroxide solution (pH 13.0) has become polarised during the electrolysis.The curves plotted in the usual way (from positive to negative potentials) show diffusion currents that are proportional to the concen- tration of hypobromite. M solutions, the diffusion currents are 12.5 and 28.2 PA, respectively, which corresponds to a current of about 19.3 PA for a 10-4 M hypobromite solution. Fig. 2 shows current - voltage curves for hypobromite in air-free solutions at various pH values. It is seen that the hypobromite wave is shifted to more negative potentials as the pH of the medium is increased. It can be seen from Fig. 1 that, for 0.65 x lo-* and 1-46 X408 KOLTHOFF, STRICKS AND MORREN : THE AMPEROXETRIC [Vol.78 Hence a t pH 8.3 (bicarbonate), 9.2 (borax), 10.35 (carbonate -bicarbonate), 13 (0.1 31 sodium hydroxide) the waves start a t f0.72, +060, +0.50 and +0.20 volt, respectively. As liypobromite in bicarbonate medium is reduced at the rotating platinum electrode a t more positive potentials than is oxygen, it should be possible to detect the reduction wave of sodium hypobromite in air-containing media also. This is demonstrated in Fig. 3, which shows the cathodic waves of hypobromite in a 0.24 -1I sodium bicarbonate solution (pH 8.3) in the absence and presence of air, and the current - voltage curve of the air-saturated Potential, V Fig. 3. Current - voltage graphs of 0.55 x M hypo- bromite solution in 0.24 M sodium bicarbonate solution (pH 8.3).Curve A, in presence of air; curve B, in absence of air; cur\.e C, vith supporting electrolyte (0.21 M sodium bicarbonate soluticin, no lixrpobromite, air saturated); curve D, current a t V against amount of 3.1 x 10-3 M sodium hypobromite soliltion added to 30 in1 of 0.24 ;M sodinm bicarbonate solution (pH 8.3) supporting electrolyte. It is of interest that, in the presence of oxygen, the hypobromite wave (Fig. 3, curve -4) starts a t a more positive potential and is steeper than that in the absence of oxygen. This efiect has not been subjected to a detailed study because of the poor reproducibility of the liypobromite wave a t the platinum electrode, the ascending part of the wave being greatly affected by the pre-treatment of the electrode. The diffusion current in the presence of oxygen is not well defined.From the point of intersection of the steep and slowly rising part of the current - voltage curve it is estimated that the current between +0.25 and 0.20 volt corresponds to the diffusion current. In the absence of air, a well-defined diffusion current is found and is equal to the current a t +0.2 volt in the presence of air (compare curves A and B, Fig. 3). The straight line in Fig. 3 (curve D), which was constructed by plotting the current ac +O+ volt against the volume of a 3.1 x &' hypo- bromite solution added to 30 ml of an air-saturated bicarbonate solution, is evidence that the current for sodium hypobromite at +0.2 volt is strictly proportional to the hypobromite concentration. Use of these observations was niadc in the development of amperometric titrations with hypobromite. AMPERONETRIC TITRATION IVITII IITPORROMITT: Amperometric titrations with liypobromite were carried out in carbonate - bicarbonate and borax buffers a t pH values varying from 8.2 to 10.0.The end-point was found to be best defined in a 0.2 M sodium bicarboink solution (pH 8.2)), and this buffer was used in most of the titrations described belon.July, 19531 TITRATION OF TRACES OF AMMOSIA 409 TITRATION OF ARSEXIOUS OXIDE- Arsenious oxide reacts rapidly with hypobromite according to the equation- This reaction can be used for the standardisation of the hypobromite solution. The titration of 3 ml of a 5 x lo-* M arsenite solution in 30 ml of bicarbonate with hypobromite is shown graphically in Fig.4, curve A. It is seen that arsenite gives a small anodic current at the rotated platinum electrode a t +0.2 volt against the saturated calomel electrode. This current decreases upon addition of hypobromite and is zero before the end- point is reached. On further addition of hypobromite a cathodic current is observed, and this increases along a curved line at first and then linearly with the volume of hypobromite .. * * (1) As,O, + 2NaOBr = As,O, + 2NaBr . . .. . . Amount of hypobromite added, mi M As,O, solution in 30 ml of 0.2 14" sodium bicarbonate solution with an unknown hypobromite solution a t + 0.2 V (S.C.E.) a t the rotating platinum wire electrode; curve B, blank titration of 30 ml of 0.2 M sodium bicarbonate solution with the same hypobromite solution Fig.4. Curve A, titration of 3 ml of 5 x added. The end-point is at the intersection of the straight reagent line with the zero current line, which corresponds to the residual current measured at +O-2 volt with the supporting electrolyte in the absence of arsenite. The curved part of the reagent line, which is observed after the zero current is reached, does not correspond to the reaction between arsenite and hypobromite but to a slow reaction between hypobromite and impurities (most likely ammonia) in the bicarbonate solution. This is substantiated by Fig. 4, curve B, which represents a blank titration of the supporting electrolyte (in the absence of arsenite). The blank value is equal to the amount of hypobromite that corresponds to the curved part of Fig. 4, curve A.In a mixture of arsenite and ammonia, the hypobromite apparently reacts first with arsenite and then slowly with ammonia. TITRATION OF AMMOSIA- Ammonia reacts with hypobromite according to the equation- 2NH, + 3NaOBr = N, + 3NaBr + 3H,O . . .. .. . * ( 2 ) This reaction is slow and therefore, in the classical titration method, an excess of hypobromite must be added and titrated back. Amperometric titrations can be carried out directly and rapidly in spite of the sluggishness of reaction ( 2 ) . From Fig. 5 , curve A, which represents an amperometric titration of ammonium chloride in a 0.2 M bicarbonate solution, it is seen that the first increments of hypobromite added give rise to a considerable cathodic current, which decreases on standing and so indicates a slow reaction.On further addition of reagent the current first increases slowly and then decreases. The reaction is complete after a slight excess of hypobromite has been added, when further addition of hypobromite gives a current that is stable and increases linearly with the volume of reagent added. The intersection410 KOLTHOFF, STRICKS ASD MORRES : THE AXPEROMETRIC [Vol. 78 of this line with the zero current line or original residual current gives the end-point. For the performance of a titration it is, therefore, only necessary to measure the current before addition of hypobromite and to record this as the residual current, and then to add reagent until a stable current is observed. A few more increments are then added and the current measured.Fig. 5 , curves B and C, shows the excess reagent lines in ammonia titrations in bicarbonate - carbonate buffers of different pH values. It is seen that a pH higher than 8.2 leads to higher results and also a decrease in the slope of the reagent lines. The results were corrected for the blanks determined for supporting electrolytes of the same pH as those Amount of 3.19x IO~’M sodium hypobromite added, ml Fig. 5. Xniperometric titration at the rotating platinnm wire electrode a t +0.2 V (S.C.E.) of 30 ml of 4.43 x 10-5 M ammonium chloride in: ;1, 0.2 M sodium bicarbonate solution (pH 8.23); B, 0.2 M sodium bicarbonate solution and 0.01 M sodium hydroxide solution (pH 8.78); C, 0.2 M sodium bicarbonate solution and 0.1 M sodium hydroxide solution (pH 9.98), with 3.19 x M sodium hypobromite used in the titrations.Because the hypobromite wave is shifted to more negative potentials at higher values of pH, titrations were carried out at -0.2 as well as at +0.1 volt. At both potentials results were high at the higher pH values. A pH of 8.6 & 0.3 is recommended for the titration. The amperometric titration of ammonia has been applied to the determination of nitrogen in glutathione, bovine plasma albumin and human sera. The Kjeldahl digestion was carried out with concentrated sulphuric acid and persulphate in a way similar to that suggested by Scott and Myers,17 and by Willard and Cake.* This Kjeldahl digestion was applied to known amounts of ammonium sulphate. It was found that ammonia is not oxidised by persulphate if the procedure is carried out under proper conditions.Thus in one instance 0486mg of ammonium sulphate was subjected to the Kjeldahl digestion and the amount recovered, as found by titration with hypobromite, was 0,899 mg, which corresponds to a positive error of 1-5 per cent. METHOD REAGESTS- 10 ml of sulphuric acid (use litmus as indicator). Sodium arsenite solutioiz-Dissolve 1.2363 g of arsenious trioxide (primary standard) in sodium hydroxide and make the solution slightly acid by addition of 0.5 M Transfer the solution to a 250-ml calibrated flaskJuly, 19531 TITRhTIOS OF TRACES OF .\MVONIA 41 1 and fill it to the mark with ammonia-free water. Dilute part of this solution to one-tenth of its strength to produce a 2.5 x 10-3M standard solution of arsenious trioxide.Sodium hypobromite solutiom---Slowly add 20 g of bromine to a solution of 12 g of sodium hydroxide in 500ml of water with constant thorough shaking. Make the solution up to 2 litres. Dilute this solution, which is about 0.06 M in sodium hypobromite and 0.1 M in sodium hydroxide, to half strength and standardise it with sodium arsenite solution. STAKDARDISATIOX OF HYPOBROMITE AGAISST ARSENITE- Place 10 ml of 0.6 M sodium bicarbonate solution and 15 ml of ammonia-free water in a 100-ml beaker. Immerse the rotating platinum-wire electrode and a glass tube with a sintered-glass bottom that is covered with a potassium nikate - agar plug. Into this tube insert the tip of the potassium nitrate - agar bridge, which, by way of a saturated potassium nitrate solution, is connected to a mercurous chromate half-cell.Connect the two electrodes directly through a microammeter. A spotlight galvanometer or a direct-reading Leeds and Northrop microammeter can be used. Record the value of the current, which usually is nearly zero. M standard sodium arsenite solution and titrate with a hypobromite solution that is approximately 3 x 10-2M in sodium hypobromite. As long as the hypobromite is not present in excess, the current is negative or zero. After the end-point, the cathodic current increases rapidly on further addition of hypobromite. When the ammeter indicates that the end-point has been passed, measure the current after the addi- tion of a few more increments of hypobromite. Plot the ammeter readings against the volume of hypobromite added.Draw a horizontal line through the point corresponding to the residual current of the supporting electrolyte (before the addition of sodium arsenite). The point of intersection between this line and the line drawn through the points recorded in presence of an excess of reagent corresponds to the end-point. Carry out a blank titration with the supporting electrolyte (10 ml of 0.6 ill sodium bicarbonate plus 15 ml of ammonia-free water) under the same conditions as for the standardisation. Deduct the blank from the volume of hypobromite used in the standarisa- tion. The molarity of the sodium hypobromite is given by- Add 5 ml of 2.5 x 2 x (volume of arsenite) x (molarity of arsenite) volume of hypobromite The hypobromite solution should be standardised daily.PROCEDURES- Tifration of ammonia-Place bicarbonate solution and water in a 100-ml beaker and measure the current through the supporting electrolyte in the cell as described in the standardisation. to G x 10-4M in ammonia. Titrate with a hypobromite solution of suitable concentration (3 x As long as the hypobromite is not present in excess, the current increases a t first and then decreases in the last stage of the titration before the end-point. As soon as the end-point is reached, the current increases regularly on the addition of reagent. Determine the end-point as described for the standardisation and subtract the blank determination. Add a volume of the sample that will make the solution 4 x to 3 x 10-2M in sodium hypobromite). 1 ml of 3 x 10-2 M hypobrornite solution = 0.34 mg of ammonia (NH,) or 0.28 mg of nitrogen Determination of nitrogen in organic materials-Heat 0.1 to 1 ml of the sample, containing approximately 1 mg of nitrogen, with 1 ml of concentrated sulphuric acid in a 100-ml Kjeldahl flask until the escaping vapour is transparent.Continue heating for 15 minutes and then allow the flask to cool. To the cold solution add 150mg of dry potassium persulphate through a long-stemmed thistle funnel, taking care to prevent the powder from sticking to the neck of the Kjeldahl flask. The solution as well as the persulphate must be water- free, otherwise some ammonia will be oxidised. Heat the flask gently for about 1 minute, at the end of which time the mixture should be colourless.If not, add more persulphate. Generally the amount of persulphate required is not more than 10 times the weight of organic matter in the sample. Boil the colourless solution for 5 minutes to destroy the excess of persulphate. Boil the solution again to remove sulphur dioxide and then transfer it completely to a 25-1111 calibrated flask and allow it to cool. Neutralise it carefully with sodium hydroxide, a t first partly Allow the solution to cool and add 5 ml of ammonia-free water.412 KOLTHOFF, STRICKS AND MORKES : THE AMPEROMETRIC [Vol. 78 with 10 &' sodium hydroxide and then with a 0.1 A4 solution, using a drop of bromocresol green as indicator. Make up the cold neutral solution to the mark with ammonia-free water. Titrate 5 ml of this solution amperometrically with approximately 3 x 10-2 M hypobromite solution as described for the titration of ammonia.Make a blank Kjeldahl digestion and titration with ammonia-free water in place of the sample. RESULTS The results of standardisation of hypobromite solutions of different concentrations It is seen that the results are more reproducible against arsenite are summarised in Table I. TABLE 1 STA?;DARDISATIOli OF HYI'OBROMITE AGAISST ARSEYITE Supporting electrolyte: 30 ml of 0.2 -11 sodium bicarbonate if not otherwise stated Results are corrected for blank Total volume of Hypobromite Concentration Concentration hypobromite used in o f hypobromite -w ml ml M of arseriite, used, blank, found, Remarks Fresh hypohromite solution 3.32 x 0.670 0.043 3.32 x 10-5 (a) 0.695 0.060 3.32 x 10-5 0.688 0.060 3.18 x 10-3 3.32 x 10-6 5.00 x 10-5 5.00 x 5.00 x 10-3 5.00 x 10-6 (a) 2.05 Y 10-4 4.18 x 10-4 4.18 x 10-4 4.18 x 10-4(b) 4-18 x 10-4(4 8.25 x 10-4 6.00 x 10-5 2.Oi x lo-' 0.692 1.005 ...I .005 0.986 0.984 1.000 0.837 0,837 0.853 0.952 0.953 0.970 1.607 0.060 0.060 0.060 0.053 0.053 0.055 0.008 0.008 0.005 0.006 0.005 0.011 0.005 Hypobromite solution one day old 3.17 x 10-3 3.17 x 10-3 3.21 x 1.51 x 1.51 Y lo-? 2.95 x 10-2 3.64 x 10-2 2.61 x 10-2 2.92 x 10-2 (a) Supporting electrolyte: 0.05 M borax (pH 9.2). ( b ) Supporting electrolyte : 0.2 M sodium bicarbonate, approximately 0.01 A t in sodium hydroxide (c) Supporting electrolyte : 0.2 M sodium bicarbonate. approximately 0.1 M in sodium hydroxide (pH 8.68). (pH 9.87).TAELE I1 AMPEROMETRIC TITRATIOSS OF AMBIOSIA WITH HYPOBROMITE Supporting electrolyte : 30 ml of 0.2 M sodium bicarbonate Results are corrected for blank Approximate Titre of molarity in solution, salt added mixture, ammonia, added, found, Error, hypobromite Ammonium ammonia of Blank of Ammonia Ammonia ;M M mg mg "fi % 3.19 x iYH4C1 4.43 x 10-5 0.0016 0.0227 0.023 + 1.3 3.19 x 10-3 (NH4)$04 4.46 x 10-5 0.0016 0.0228 0.023 + 0.9 1.52 x lo-* NH4C1 2.77 x 10-4 0.0017 0.1419 0.147 + 3.4 1.51 x 10-2 (NH4)2S04 2.22 x 10-4 0.0014 0.1140 0.116 + 1.8 1.51 x (NH4)ZSOd 3.22 x 10-4 030014 0.1140 0.117 + 2.9 2.97 x 10-2 NH,Cl 5.60 x 10-4 0.001 7 0.2840 0.286 + 0.5 0.2840 0.287 + 1.0 2.99 x (NH,),SO, 5.80 x 10-4 0.0017 0.2855 0.288 + 0.9 2.99 x 10-2 (NH,)ZSO, 5.60 x 10-4 0.0017 0.2855 0.290 + 1.4 2.64 x NH4C1 5.60 x 10-4 0.0018 0.2840 0.287 + 1.0 2.97 x 10-2 NH4C1 (a) 5.60 x 1.0-4 0.0017 0.2855 0.291 + 1.7 2.97 x 10-2 NH,Cl ( b ) 5.60 x 10-4 0.0135 0.2855 0.300 + 4.9 2.64 x XH4CI (b) 5.60 x 10-4 0,0033 0.2840 0.300 + 5.5 2.97 x 10-2 XH4C1 5.60 x 10-4 0.0017 (a) Supporting electrolyte: 0.2 M sodium bicarbonate, 0.01 M in sodium hydroxide (pH 8.78).( b ) Supporting electrolyte: 0.2 M sodium bicarbonate, 0.1 M in sodium hydroxide (pH 9.98).July, 19531 TITRATION OF TRACES OF 9MMOKI.4 413 in titrations with more concentrated solutions to 3 x 10-2M sodium hypobromite). At these concentrations the blank correction is less than 1 per cent. of the total volume of hypobromite used. From Table I it is also seen that the results of titrations in a borax buffer (pH 9.2) and in a carbonate buffer of about pH 9.9 are from 1 to 2 per cent.lower than those found with bicarbonate as supporting electrolyte. Hypobromite solutions kept for one or two days in a refrigerator show hardly any change in titre. Table I1 shows results for amperometric titrations of ammonia. It is seen that 23 pg of ammonia at a concentration of about 4 x M can be determined with an error of less than 1 per cent. The accuracy and precision of the ammonia titrations is better than 2 per cent. All errors are positive. This must be attributed to formation during the titration of nitrogen oxides, as found by Kolthoff and L a ~ r . ~ At any pH markedly higher than 8.6 the result is high, as is seen from the last two experiments shown in Table 11.Table I11 gives results for the nitrogen content of reduced glutathione and albumin as found after Kjeldahl digestion and titration of the digest with hypobromite. The nitrogen content of glutathione is found to be 13.72 per cent., which compares favourably with the theoretical value of 13.68 per cent. The nitrogen content of albumin is generally assumed to be 16 per cent.,la a value which is in fair agreement with our average result of 16.22 per cent. It is also seen from Table I11 that the blank corrections are not larger than 3 per cent. of the titration results, a factor that is of importance for the accuracy of the method. TABLE I11 AMPEROMETRIC DETERMISATION OF NITROGEN IX REDUCED GLUTATHIONE AND IN BOVINE PLASMA ALBUMIN AFTER KJELDAHL DIGESTION Supporting electrolyte: 30 ml of 0.2 M sodium bicarbonate Concentration of hypobromite used: approximately 3 x 10-2 M Results are corrected for blanks Theoretical amount of Quantity of material nitrogen in used for Kjeldahl Blank the titration Sitrogen Average No.of deter- digestion, of nitrogen, mixture, found, error, minations mg mg mg mg 70 7.37 mg of glutathione 0.0058 0.2016 0.2022 + 0.3 4 6.98 mg of albumin . . 0.0035 0.2232 ( a ) 0.2264 + 1.4 2 (a) This value corresponds to a nitrogen content of 16 per cent. in albumin. Xitrogen determinations in normal and pathological blood sera and in their albumin and globulin fractions are being carried out in this laboratory by the Kjeldahl digestion and titration method described in this paper. The results of these determinations will be reported elsewhere.This investigation was supported by a research grant from the National Cancer Institute, U.S. Public Health Service. 1. > -. 3. 4. 6. 7. 8. 9. 10. 11. a. REFERENCES .irtman, P., and Skrabal, A,, Z . anal. Clzeiii., 1907, 46, 3. Rupp, E., and Rossler, E., Arch. der Pharm., 1905, 243, 104. Kolthoff, I. M., and Laur, A., Z . anal. Chem., 1928, 73, 177. Willard, H. H., and Cake, W. E., J . Amer. Chem. SOC., 1920, 42, 2646. Rappaport, F., and Geiger, G., Mikrochemie, 1935, 18, 43. Sandor, D., Ovvosok Lapja ds Nepegeszse-giigy, 1946, 2, 1505; Chenz. Abstr., 1949, 43, 5 0 6 9 ~ . Rappaport, F., and Eichhorn, F., J . Lab. Clin. Med., 1947, 32, 1034; Anal. Chim. Acta, 1949, 3, Chinard, F. P., and Sewell, D. A, J . Biol. Chenz., 1948, 176, 1449. Leipert, Th., Mikrochemie, 1949, 34, 276. Kibbrik, A. C., and Skupp, S., Amer. J . Clin. Path., 1951, 21, 881. Kolthoff, I . M., and Stricks, W., Anal. Chem., 1951, 23, 763. 674.414 GREGOR\-, MAPPER AND WOODWARI) THE MrCRO-DETERhII?jATION [VOl. 78 12. 13. 14. 15. 16. 17. 18. Kolthoff, I. &I., blenzel, H., and Furman, N. H., “Volumetric Analysis,” John M’iley and Sons Kolthoff, I. M., and Sandell, E. 13., “Textbook of Quantitative Inorganic Analysis,’’ Third Edition, Lingane, J . J., and Kolthoft, I. M., J . Anzer. Clzem. SOC., 1939, 61, 825. Kolthoif, I. M., and Harris, W. E., I n d . Eng. Chein., Anal. E d . , 1946, 18, 161. Gmelin-Kraut, “Handbuch dev anorganischen Chemie,” Carl Winter’s Universitatsbuchhandlung, Scott, L. C., and Myers, R. G., J. Amer. Chem. SOC., 1917, 39, 1044. Haurowitz, F., “Chemistry and Biology of Proteins,” Academic Press lnc., New York, 1950, p. 12. Inc., New York, 1929, Volume 11, p. 468. The MaclLlillan Conipany, New York, 1962, p. ,593. Heidelberg, 1914, Volume V, Division 11, p. 1120. SCHOOL OF CHEMISTRY UNIVERSITY OF MINNESOTA MINNEAPOLIS 14, MINNESOTA, U.S.A. Febriiavy 9th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800405

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

414 GREGOR\-, MAPPER AND WOODWARI) THE MrCRO-DETERhIINATION [Vol. 78 The Micro-determination of Traces of Gaseous Elements in Metals by the Vacuum Fusion Method BY J. N. GREGORY, D. MAPPER AND J. A. WOODWARD A vacuum fusion method is described for the micro-determination of gases in metals. The experimental work involved in the development of the method and details of the proposed apparatus are fully described. Steel and uranium can be rapidly analysed for oxygen, nitrogen and hydrogen with an accuracy of better than f l 0 p.p.m. with 50 to 200-mg samples. An unsuccessful attempt to determine oxygen and nitrogen in zirconium on a micro scale is also recorded. I s recent years much attention has been directed towards the effects of traces of gaseous elements on the properties of metals.A considerable amount of work has been carried out on various steels and, more recently, attention has been directed towards such metals as zirconium, titanium, uraniuni, molybdenum, thorium and vanadium. Slomanl and Sloman and Harvey2 have successfully determined oxygen, hydrogen and nitrogen in these metals with the aid of a relatively large-scale vacuum fusion apparatus. In Sloman’s method, metal samples ranging in weight from 5 to 25 g are required and the 2 to 10 ml of gas produced is analysed by conventional macro methods. We believed that some improvement and greater simplicity of operation could be attained by carrying out vacuum fusion analyses on a much smaller scale. There is no theoretical reason why the principles applied in Sloman’s method cannot be applied to the extraction and collection of gas volumes one-thousandth the size of those handled in the macro system.The gas could then be extracted from samples of 100mg or even smaller and analysed by one of the recently devised micro-analytical methods or by mass spectrometry. In general, the vacuum fusion method as applied to the determination of oxygen, hydrogen and nitrogen consists in introducing the metal sample into an outgassed graphite crucible at high temperature in a continuously evacuated system. The graphite is usually heated by high-frequency induction by means of an external coil. The operating temperature is determined by the particular metal and gas; it can be as high as 2200” C. The theoretical principles involved are discussed in detail by Sloman and Harvey2 and B r e ~ e r .~ In general, the oxygen is evolved as carbon monoxide containing a small proportion of carbon dioxide, and the nitrogen or hydrogen in the elemental state. These elements are often combined with the metal, so the operating temperature should be such that the dissociation pr, assure for the reactions involved would be substantially greater than the pressure maintained in the system, namely, about loe5 mm of mercury. For rapid evolution of the gas the dissociation pressure should be about 2 to 3 mm of mercuryS2 Metals that have refractory oxides and nitrides are liable to cause considerable difficulties at the high temperatures involved. I t is difficult to outgas the graphite to give a sufficiently low blank value for gas at the high temperature, and considerable evaporation of the metal can also occur.The condensation of metal on the cold walls of the furnace jacket or on the graphite crucible at intermediate temperatures can, with certain metals, cause considerableJuly, 19531 OF TRACES OF GASEOLTS ELEMENTS I N METALS 415 chemisorption losses of gas. Sloman reduces the amount of evaporation by diluting the metal in an iron bath in order to reduce its activity. This, he claims, also facilitates oxide or nitride dissociation, particularly with metals having high melting points. We have found the metal-bath technique to be of little or no advantage in micro-analytical work. Gas analysis of metals on a micro scale has several marked advantages. The small sample required lias particular value in determinations involving the rarer metals ; it also facilitates the study of homogeneity in metal specimens.However, for samples with gross inhomogeneities and inclusions, sampling difficulties may arise. Efficient heating of the large crucibles used in macro-analysis, without large and expensive induction heating units, requires careful design of the crucible and considerable skill in its construction and assembly. All this is time-consuming and is rendered unnecessary by the micro system in which simple disposable crucibles can be used in conjunction with an %kilowatt heater. The micro method involves no more difficulties than the large-scale method and it is considerably more rapid, so making it suitable for routine work. -4 number of papers have described various systems operating on more or less the same fundamental principles, but which vary somewhat in detail, specific purpose, and method of gas analysis.The determination of oxygen in uranium has been given considerable attention by a number of workers. Rice4 and Lipkin and Perlman5 have described equipment suitable for this purpose, and Seifert, Gilpatrick, Phipps and Simpson6 describe a similar system in which the gas is extracted from cupric oxide, silica, uranium trioxide, uranium dioxide or uranium nitride. Little detail is given in these papers of methods of analysis, and in some work the evolved gas is assumed to be all carbon monoxide. This assumption, as will be shown later, can lead to serious errors. Guldner and Beach' and Walters also describe in detail another variation of the vacuum fusion method on a small scale, but they use a somewhat unnecessarily complicated crucible and furnace system.In the work described below, the apparatus incorporates ideas taken from all these previous investigators, but also has new features that lead to simpler construction and facilitate speedy manipulation with no loss of accuracy. GESERAL COSSTKVCTIOS- Fig. 1 illustrates the apparatus in its finally developed form. The heavily outlined part indicates the main line along which the gas sample is transported. In this part, mercury cut-offs are used instead of taps, so that the gas sample does not come into contact with grease at any stage. The parts of the system lightly outlined are the ancillary units required for operating the cut-offs, and so on.All ground joints are sealed with black wax, which is applied lightly so that it is not extruded into the interior of the vacuum line on assembly. -411 the mercury cut-offs except D are of the normal type and are set at approximately baro- metric height above the reservoirs ; they have stainless steel balls fitting into ground seats to ensure good sealing. In cut-offs B and C the ball seal ensures collection of a constant volume of gas on the backing side of the pump, 13. The all-glass ball seals are carefully ground to withstand a pressure of at least two atmospheres when mercury- sealed. This enables a pressure of 16 lb per sq. inch more than atmospheric to be applied to the reservoir of this cut-off. By doing this, it is possible to open the system to atmosphere on either side of the cut-off without disturbing the vacuum on the opposing side.This is useful when crucibles and samples are being charged, as the furnace side alone need be opened t o the atmosphere and the subsequent outgassing of the mercury pumps is avoided. Before opening cut-off I), after having had the furnace open to the atmosphere, it is essential to evacuate the furnace side again. This can be readily done via the auxiliary lines shown, i.e., through taps G, H and J, or I< and L. Tlie whole system is evacuated by the all-glass two-stage mercury diffusion pump, 10, backed by the two-stage rotary pump, 20. The other pumps, as described later, are for gas handling only. At the furnace end of the system a small gas burette, 11, is provided, by which small amounts of gas introduced via the mercury lift, 21, can be measured.The gas is measured by a pressure, volume and temperature determination when compressed into the precision Micro-scale vacuum fusion has several applications. DESCRIPTIOS OF THE -4PPARATUS The cut-off, D, is of special wide bore to allow fast pumping.416 GREGORY, MAPPER AND WOODWARD : THE MICRO-DETERMINATION [Vol. i s bore capillary, 22, which is 2 mm in diameter. The gas can be released into the system by drawing down the mercury to the position shown in Fig. 1, and can be used for calibration purposes or for adding known amounts of other gases to the gas sample. FURNACE DESIGN AND SAMPLE-ADDIXG MECHANISM- The apparatus and graphite crucible must be thoroughly outgassed as described below before the metal sample is added.Once outgassed the graphite must be kept under high vacuum at a temperature of at least 1100" C, or the blank gas value will not remain constant. It is essential to have the metal samples already in the system, so that they can be added to the crucible when required. I t is possible to analyse successively five separate samples with one graphite crucible in this system. FIG. 1. I-acuuni fusion apparatus for micro-scale gas analysis I Silica furnace tube 2 Graphite crucible and heat shield 3 H.F. heating coil 4 Silica window for viewing with disappearing filament pyro- meter 5 Solenoid-operated trap door for protecting viewing window 6 Movable silica funnel for sample addition 7 Solenoids 8 Solenoid-operated t r i p buckets for sample storage 9 Soft-iron-in-glass counterweight for funnel (6) 10 Gold foil for protection of furnace I I Gas burette for testing purposes 12 Safety trap I 3 Two-stage all-glass mercury diffusion pump 14 McLeod gauge, double range, for measuring evolved gas I 5 Two-stage mercury diffusion pump (small silica) 16 Merwry non-return valve for Toepler pump 17 Toepler pump from mercury 18 Gas collecting thimble fur analyses in Blacet - Leighton.apparatus 19 Two-stage all-glass mercury diffusion pump 20 Two-stage rotary backing pump 21 Mercury lift for introduction of other gases 22 Gas burette measuring capillary (made from precision bore tubing) A, 8, C, D, E, Cut-offs F, Constriction G, H, I, K, L, M, TapsJuly, 19831 OF TRACES OF GASEOUS ELEMEKTS IS METALS 417 The samples are held in the hinged trip buckets, 8 (Fig.l ) , which are suspended between two glass-covered tungsten rods attached to the stopper at the top. Each of these buckets has sealed into its base a small piece of soft iron by which it can be tipped up by a solenoid, 7, so projecting the contents into the inclined tube. The solenoids are actuated through a selector switch with press-button control so that any particular sample can be chosen. The metal sample falls into the silica funnel, 6, which, during the addition, is lowered by magnetic neutralisation of its counterweight, 9, so that its tip projects into the mouth of the crucible. As soon as the sample is introduced into the crucible the funnel is raised magnetically by 0 1 2 3 4 5 6 Scale (centtmetrer) Fig.2. Graphite crucible, heat shield and silica furnace assembly. crucible: B, heat shield; D, constriction; E, shelf; F, funnel; G, 3/16-inch hole; H, sample; J, heating coil the counterweight, which is attached to the funnel by a stainless steel chain and is adjusted so that friction will hold the funnel stationary in any position. The crucible is kept at a temperature between 1100" and 1200" C while the silica is in contact with it. In Fig. 2 the detailed construction of the crucible and lower part of the clear silica furnace jacket is shown. The crucible and heat shield are made separately from best-grade fine-grained graphite; there is a good tight sliding fit at C for assembly. The stem of the crucible is held firmly in the lower constriction, D, of the silica jacket and the heat shield rests on the shelf, E.All the components described here are circular in section. The top of the crucible must be flush with the top of the heat shield. The tightly fitting cap of the crucible has a shallow recess at the top to take the funnel, F, and has a hole, G, The hole is drilled SO that the sample at H is invisible from the exterior. In tlvs way, effusive evapora- tion of metal is reduced to a minimum without seriously lvndering gas evolution. The crucible is heated by means of the high-frequency heating coil, J. To minimise high-frequency heating of the heat shield, the shield is split vertically for about 7 to 8 cm by two fine saw cuts diametrically opposite each other.In constructing a furnace of this type, care must be taken to match the work to be heated inch in diameter, drilled at approximately 45" through it.418 GREGORY, MMAl’PER AND WOOD\V.%RI) : THE MICRO-DETERMINATION [‘\.Ol. 78 and the work coil to attain maximum heating efficiency. The greater the sectional area of the coil occupied by the work the greater is the efficiency, but this consideration must be balanced against the necessity for having reasonable clearance between the walls of the silica vessel and the graphite shield. If this clearance is too little the temperature gradient in the silica becomes dangerously high. The work coil should have the requisite number of turns for maximum efficiency; this can only be finally determined by experiment.The temperature of the graphite is measured by a disappearing-filament pyrometer, which directly faces the inclined mouth of the crucible through the window, 4 (Fig. l ) , in the side-arm. The side-arm has a magnetically operated flap, 5, to prevent condensation of metal and graphite on the window. The silica vessel is cooled by four jets of water; two impinge on the area around the cone and socket joint between the silica and glass parts of the apparatus and two are directed just above the coil. The water flow must be arranged to that a continuous film of fast-flowing water covers the whole of the silica vessel. The free flow of water over the heating coil does not give rise to any difficulty with the type of heater used in this work. The work coil should have just enough clearance to enable a free flow of water around the silica jacket.It should not touch the jacket a t any point. A “Radyne” heater, with a maximum continuous rating of 8 kilowatts, operating a t 450 kilocycles and specially adapted for graphite heating, was used in this apparatus. With a specially constructed transmission line, 6 kilowatts could be dissipated in the graphite to produce a maximum temperature of 2200” C. The high-frequency heater is provided with a control unit that gives a continuous range of temperature from about 1100” to 2200” C. Fig. 3 shows a general view of the whole apparatus in operation and Fig. 4 shows the details of the furnace and sample-adding device. For safety, the furnace is surrounded on the three working sides by firmly held sheets of “Triplex” glass (front removed in Fig.4). The principle use of the trap, 12 (Fig. l), is to prevent water being sucked back into the diffusion pump in the event of a fracture of the furnace tube. It is unnecessary to cool this trap, as mercury vapour does not appear to interfere with the operation of the furnace. However, as a precaution, a strip of gold foil, 10 (Fig. l ) , was inserted to collect any mercury diffusing back. This foil is regularly cleaned by ignition. OPERATIOX OF THE APPARATUS ISITIAL EVACUATIOS . ~ N D OUTGASSIXG- After the crucible and samples have been set in position, the initial evacuation is carried out by the rotary pump, 20 (see Fig. l ) , via the auxiliary vacuum linc, through taps J, H and G and cut-off E.The reservoir taps of the cut-offs, IIcLeod gauge, Toepler pump, and so on, are all turned to this line so that the reservoirs are at the same pressure as the system. When evacuation is complete, these taps are closed, H is switched to the auxiliary vacuum pump (not shown) and 1cI is opened. At this stage, the mercury in the cut-offs is set a t the positions shown in the diagram, except for E, which is closed. The three mercury diffusion pumps are started and left until the pressure indicated by the McLeod gauge is less than With the cooling water flowing, the graphite crucible is raised slowly to the outgassing temperature and maintained at that temperature for several hours. To minimise evaporation of the graphite this temperature should not exceed 2000” C.Any gas evolved from the graphite or the pumps is extracted continuously through the main pump, 19. During this process and a t any other time when not in use, the mercury in the burette, 11, should be mdintained at barometric height, so sealing off this unit from the main system. The gas evolved from the furnace is measured by the iiicrease in pressure on the backing side of the diffusion pump, 13, with cut-offs B and C closed. The gas is pumped into a constant volume, which is determined by releasing measured amounts of gas into the system via the burette, 11, and pumping it into the measuring space. The mercury pump is capable of operating against a backing pressure of several millimetres of mercury; the increase in pressure produced by 0.03 cu. cni is of the order of 0.10 mm of mercury (on the backing side), so that the system can pump effectively several hundred cubic millimetres into the measuring space.The McLeod gauge, 14, has a double scale; one is calibrated from 0 to 0.10 mm of mercury and the other from 0 to 3 nim. The volume enclosed between traps B and C and the pump in this apparatus is 244 cu. cm, giving a calibration factor of 0.0321 cu. cm for In this way, no mercury can fly over. mm of mercury.Fig. 3. General view of apparatusFig. 4. View of furnace and sample-adding mechanismsJuly, 19531 O F TRACES OF GASEOUS ELEMENTS I N METALS 419 a pressure rise of 0.10 mm of mercury. Calibration graphs show that the accuracy of this measurement is limited only by the experimental error in reading the McLeod gauge, which is unlikely to exceed $-5 per cent.The progress of the graphite outgassing can be followed by closing B and C, setting the graphite at the operating temperature and collecting the gas during a period of about 5 minutes. The acceptable blank rate depends on the expected volume of gas from the metal sample and the time taken for it to be evolved. In general, the operating temperature should be high enough to give complete evolution of the gas in less than 5 minutes; with a gas volume of more than 10 cu. mm a blank rate of between 0.05 and 0.1 cu. mm per minute is negligible. With high working temperatures it may not be possible to produce a negligible blank rate, so a sample must be collected and analysed. The contribution of the blank value must then be allowed for in the gas sample.Time. minutes Fig. 5 . Gas evolution of uranium sample at 1600" C. Curve A, gas from uranium plus hlank value; curve R, blank value for graphite at 1600°C As the blank rate increases with temperature, whereas the time for gas evolution decreases, it is obvious that there is a temperature at which the proportion of blank in the sample is at a minimum. If practicable this temperature should be used: in general, this seems to be the temperature at which the gas is evolved in about 5 minutes. Another factor that determines the maximum operating temperature is the volatility of the metal. The maximum volatility permitted will vary with each metal and will depend on the chemisorptive properties. Some guidance can be obtained from the fact that uranium can be extracted satisfactorily at 1800" C, at which temperature its vapour pressure js about 0.01 mm of merc~ry.~ Some workers take great trouble to flame and outgas all the glass parts of their apparatus.This is unnecessary, as the final criterion of the contribution of extraneous gas sources is the blank rate produced by test. After outgassing hardly any gas is collected when the graphite is cold, so that any other effect is negligible. EVOLUTIOX AND COLLECTIOX OF GAS FROM METAL SPECIMEY- \Vhen the blank rate is reduced to a satisfactory value, all the cut-offs are opened and the whole system pumped for a few minutes. The graphite is cooled to about 1200" C and the silica funnel lowered. Cut-offs B and C are closed, the sample is immediately introduced into the crucible and the funnel raised.The temperature is then raised quickly to the operating value and the progress of the collection of the evolved gases observed at regular intervals by the McLeod gauge. Gas is collected until the rate of evolution is equal to the original blank rate. A typical gas evolution graph for a uranium sample is shown in Fig. 5. From this it is clear that all the gas in the metal is evolved in 5 to 6 minutes and thereafter only the gas blank is collected. The factor controlling the rate at which the gas appears in the McLeod system is almost always the rate of evolution. The pumping speed is high; 30 cu. mm of gas can be transferred from the measuring burette to the McLeod gauge in less than 1 minute. It is essential to have a fast removal of the evolved gas to avoid reverse reactions (see below, p.422).420 GREGORY, MAPPER .AND W'OODWARD : THE MICRO-DETERMINATION [VOl. 7 5 When gas evolution is complete, the cut-off, D, is closed, the McLeod gauge pressure and temperature are read carefully and the crucible is cooled to 1200" C to await the next sample. Once a sample has been added it is essential to maintain the crucible at the working temperature for the minimum possible time and never to exceed this temperature, otherwise excessive metal evaporation will cause chemisorption losses in subsequent samples. By closing cut-off A and opening B, the gas is then transferred completely within a few minutes to the backing side of pump 15. This portion includes the bulb of the Toepler pump, 17, which constitutes the greatest proportion of the backing volume because the line between pumps 15 and 17 is made as narrow as possible.About 3 or 4 strokes of the Toepler pump suffice to transfer over 99 per cent. of the gas to the thimble, 18; it can then be analysed Scale (centinierrer) 0 1 2 3 4 5 6 7 Fig. 6 . JIultiple breaker-seal unit (front view). Inset B,, B2, B,, B,, ( a ) , side yiew of one breaker-seal tube. U-tubes; C, E, F, H, constrictions by the Blacet - Leighton method.1° As soon as the gas is pumped through by pump 15, B can be closed and C and D opened for a few minutes to clear the system, and the whole procedure can be repeated with a second sample in the crucible. In this way it is possible to collect the gas evolved from five separate samples in less than 2 hours.The greatest delay occurs in the gas analysis and the preliminary outgassing of the graphite. There are several advantages in carrying out the extractions in rapid succession, apart from speed. If an appreciable interval occurs between two extractions, the blank rate rises, particularly if the crucible is cooled to room temperature, and considerable heating at temperatures greater than the operating temperature is necessary to lower the blank rate. This procedure can cause appreciable evaporation of the metal. Once a negligible blank rate has been established it will remain negligible if not more than 5 to 15 minutes elapse between each extraction and the temperature is maintained at about 1100" to 1200" C in this time interval.Alternatively, the gas evolved can be analysed by a mass spectrometric method, in which event a collecting bulb of about 15 ml capacity, with a breaker-seal, is attached via a con- stricted arm to the top of the Toepler pump. The gas contained in the Toepler bulb is compressed into it and the constriction is sealed off with a flame. By further heating at theJuly, 19531 OF TRA.CES OF GASEOUS ELEMENTS IN METALS 42 1 constriction, the tube containing the gas sample can be removed without breaking the vacuum. As the Toepler bulb contains over 80 per cent. of the gas, the remainder being in the line to pump 15, a considerable fraction of the gas is taken as a sample for complete analysis. To enable the five gas samples to be collected in breaker-seal tubes successively, without reducing the vacuum when replacing the tube, a special multiple breaker-seal unit has been designed; it is shown in Fig.6. This unit is fitted to the top of the Toepler pump by the A10 cone; a small tap with a capillary bleed line to atmosphere is fitted to the top socket. The five successive gas samples are collected in the breaker-seal tubes 1 to 5. After B and C (Fig. 1) are closed, the Toepler pump and this unit can still be evacuated continuously through A (Fig. 1). Just before A is closed for collection of sample 1, the mercury in the Toepler pump is allowed to rise to the level A (Fig. 6) and then drawn back again. Since the pressure is the same on either side of the small U connections, B, mercury remains behind in each of them and seals off the breaker-seal tubes from each other at the lower end.The constriction at C is then closed and separated from the upper manifold. Cut-off A (Fig. 1) is closed and the gas sample transferred via pump 15, with B (Fig. 1) open to the Toepler bulb, as described earlier. On raising the mercury in the Toepler pump to the level D (Fig. 6) nearly all the contents of the bulb are compressed into the breaker-seal tube 1, which is sealed off a t E and removed for mass-spectrometric analysis. As the total pressure of the gas in the tube is never more than 2 or 3 mm, the mercury in B, effectively seals tube No. 1 from the succeeding ones. Gas sample No. 2 should be ready for collection and is stored on the backing side of pump 13, with B and C (Fig.1) closed. Cut-off A is opened for a few minutes to flush out the residue of the previous sample. During this period the mercury must be cleared from B, to open the path to breaker-seal 2 . This is achieved by slowly increasing the pressure of air via the bleed line on the top of the unit. This blows the mercury in B, back into the Toepler bulb but does not affect B,, B, and Kj, since the pressure will be equal on either side of these U-tubes. \I'hen this is effected and all the added air has been removed via ;i (Fig. l), the constriction, F, is closed, as was C previously. Cut-off :\ is then closed and the second gas sample pumped into the Toepler pump. On raising the mercury to the level G, sample 2 is pushed into tube 2 , which can then be sealed off at H, and removed as was tube 1.In this way the gas from all the samples is evolved and collected without breaking the vacuum in any part of the apparatus at any stage and with little delay between samples. The frame holding the breaker-seal tubes is so constructed that i:ew tubes can be readilj. titted by a competent glass-blower. .~X.~LYSIS OF GAS- Mass s$ectrometry-This method is speedy and highly sensitive ; it has an over-all accuracy of about &lo per cent. Gases other than those present in the graphite can be determined at concentrations equivalent to about 0.1 part per million in 100-mg samples of metal. The main disadvantage of this method is the great difficulty in differentiating between the two major components, carbon monoxide and nitrogen, which are of equal mass.This can be overcome with some considerable loss of accuracy by determining the fraction of mass 29, which is principally the monoxide of I3C, and multiplying the value found by the 12C/laC abundance ratio (68.5) to obtain an estimate of the total carbon monoxide in the fraction of mass 28. But as the fraction of mass 29 is so small, considerable errors arise if any impurity of mass 29 is present in the gas. Impurities of mass 29 are quite likely to be met, as traces of hydrocarbons are frequently present in the graphite. Blacet - Leiglzton method-The Blacet - Leighton methodlo of gas micro-analysis has been applied with considerable success and is now generally used when the gas sample is larger than 10 cu. mm. The collection of the gas in a Blacet - Leighton thimble requires much simpler apparatus than the mass spectrometer breaker-seal ; manipulation is also easier.Nevertheless, for samples smaller than 10 cu. mm, the mass spectrometer is inore reliable. EXPERIMENTAL CRUCIBLE AKD FUI~SACE DESIGS- diameter of only 30 mm. Perlmai~.~ The procedure is repeated for each of samples 3, 4 and 5 . In the earlier work a heat shield was not used and the furnace tube had an internal This was similar to the system used by Rice4 and Lipkin and At 2000" C there was considerable evaporation of graphite on to the walls and 011422 GREGORY, BIAI'PER AND \YOOD\VARD : THE MICRO-DI:TEHhIINATION [VOl. 78 two occasions the furnace tube cracked during a temperature change. The radiant heat absorption of the graphite film probably caused intolerable thermal gradients in the silica.To eliminate the danger, the heat shield was added and the furnace tube widened to 40 mm. The heat shield also gave better thermal efficiency and partly compensated for the loss in coil efficiency at the larger diameter. The first heat shields projected about 1.5 cm above the top of the crucible, but it was noticed that the gas recovery in a batch of uranium samples consistently decreased in the third, fourth and fifth samples. This was found to be due to chemisorption by uranium, which had evaporated and become condensed on the upper part of the heat shield; its temperature (1100" to 1300" C) was ideal for rapid reversal of the reaction. a-Counting tests on parts of the heat shield showed that the evaporation was effusive, as no part of the heat shield hidden from the uranium was appreciably contaminated. A simple way of testing the system for losses of this type is to add known amounts of carbon monoxide or nitrogen via the burette, 11, while maintaining the furnace containing the metal at the operating temperature (after thoroughly outgassing the metal) and with cut-off D closed.The gas is left in contact with the system for a few minutes and then pumped off, and the amount recovered is compared with the amount added. TABLE I RECOVERY OF ADDED GAS WITH DIFFERENT TYPES OF HEAT SHIELD Temperature Volume of Volume of " C cu. mm a t cu. mm a t Type of heat shield Gas of crucible, gas added, gas recovered, S.T.P. S.T.P. Old design projecting above CO 1600 45. 7 6.0 crucible co 1600 16.3 4.5 co 20 22.2 20.2 N 1600 55.0 0.0 S, 1600 30.6 0.7 s 2 20 *55*4 60.0 with crucible top co 1600 '1.4 22.6 S , 1600 18.8 10.7 N2 1800 37.8 34.6 9, 1800 12.1 11.1 I V i t l l top of heat shield flush CO 1600 6.5 5.7 Recovery, % 15 30 90 0 2 100 90 100 60 90 90 Measured amounts of nitrogen and carbon monoxide were introduced into the system after 100 mg of uranium had been heated in the crucible for about 1 hour at 1800" C, so that considerable evaporation would have been expected.After a few minutes in contact with the furnace the gas was pumped away and measured in the usual manner. From the first part of Table I, with the projecting heat shield, it is clear that both the nitrogen and the carbon monoxide were substantially absorbed. Since it is well established that the formation of carbon monoxide from uranium oxide and graphite is rapid at 1600" C (Sloman and Harvey2), it is clear that the loss of carbon monoxide was due to chemisorption by the evaporated uranium on the heat shield, which was at about 1200" C.As it is not certain whether the decomposition of uranium nitride was complete under these conditions at 1600" C, it is probable that some of the nitrogen was lost by formation of nitride in the crucible. This is confirmed by the results at 20' C, which indicate that cold evaporated films do not cause loss and show also that the graphite when cold does not absorb appreciable amounts of the gas a t the prevailing pressure. For the rest of the experiments the heat shield described on page 417 was used.The top of the shield was flush with the crucible top and out of the range of effusive evaporation. The much better recoveries with carbon monoxide at 1600" C confirm that the loss probably occurred by the reaction with evaporated uranium, whereas the rcduced loss with nitrogen at 1600" C seems t o confirm that some of the original effect was due to reaction of the nitrogen with uranium in the crucible. However, recoveries were good at 1800°C for nitrogen, which confirmed that this temperature is essential for complete dissociation of the nitride. The results shown in the second part of Table I indicate that for efficient recovery of oxygen and nitrogen from uranium a working temperature of about 1800" C must be used and the heat shield should be designed so that it does not collect evaporated uranium. Subsequently it was found In these experiments an open graphite crucible was used.J~ily, 19531 OF TRhCES OF GASEOUS ELEMENTS I N METALS 423 that the results with uranium were improved even further by use of the crucible with the detachable cap with the inclined entry port, as described above (p.417). The conditions in these experiments with regard to concentration of gas in contact with the evaporated film and the time of exposure were much more severe than those met with in actual gas determinations, in which the gas is pumped away as soon as it is evolved and the pressure is always low compared with the static conditions in the gas addition experiments. Hydrogen is completely lost from uranium at quite low temperaturesll and its quantitative evolution should not produce any difficulties.Nevertheless, Sloman and Harvey2 mention Volume of gas recovered. cu inm a t S T P Fig. i. Gas transfer efficiency an anomalous effect with hydrogen ; its solubility in uranium increases with increase in temperature. As they use a steel bath they must maintain a temperature of at least 1550" C ; at this temperature hydrogen is appreciably soluble. In our apparatus the sample is added at about 1200" C and all the hydrogen is released and pumped away before the temperature is raised. GAS TRANSFER EFFICIEKCY- To test the gas transfer efficiency, the delivery of the Toepler pump was connected to a gas burette identical to burette 11 in Fig. 1. Measured volumes of air were introduced into the system by the burette, 11, and transferred by the pumps and Toepler pump to the collecting burette.The graph (Fig. 7 ) shows that no significant amount of gas is lost or gained during the procedure. In tests carried out without the second pump, 15, in the system, about 10 strokes of the Toepler pump reduced the pressure in the McLeod gauge to about 0-0001 mm, which is equivalent to a residue of less than 0.05 cu. mm of gas at S.T.P. To attain this efficiency it was necessary for the delivery capillary of the Toepler pump to have an internal diameter of 0.5 mm. In tests with a sample of approximately 30 cu. mm, with the system in its present form, it was shown that more than 99.5 per cent. of the gas was pumped into the McLeod gauge from the furnace in about 1.5 minutes.The second pump, 15 (Fig. l), left less than 0.1 cu. mm in the McLeod gauge after 3 minutes, and 5 strokes of the Toepler pump accounted for 99.9 per cent. of the gas accumulated on the backing side of pump 15. APPLICATION The preparation of standard samples containing known amounts of gas is an extremely difficult problem in vacuum fusion work. The obvious method of approach is to use oxides and nitrides of the particular metal, since with many of them the gases do exist in chemical combination with the metal. However, it has been found that oxides and nitrides, even in an extremely fine state of subdivision, are much more difficult to decompose than when the elements are distributed throughout the metal. Also, in micro work, unweighable amounts of oxides or nitride would be required for a comparable standard.Satisfactory oxygen standards can be prepared for uranium by a quantitative surface oxidation process and it424 GREGORY, MAPPER AND WOODWARD : THE hlIClI0-nETERhrIS.4TlON [VOl. 78 has been found that a thin surface film decomposes readily at the normal working temperature for uranium. UR.4NIUM- The apparatus used for preparing the standards is shown in Fig. 8. I t is a constant- volume system, in which the amount of oxygen absorbed by the heated uranium specimen at A is measured by the change in pressure involved. rotary To 2-stage pump& To atmosphere 5 Lnr-rV - 0 I1 - Fig. 8. .Apparatus for preparation of uranium standards. X, sample; B, Rourbn-type gauge; C, manometer; D, calibrated bulb; E, mirror; F, scale; G, lamp; H, seal; J , cathetometer The constant volume is that part closed off by the taps 2, 3 and 4 and includes the spiral of the all-glass Bourdon-type gauge, B.For maximum sensitivity this volume should be small. About 30 to 50mg of carefully cleaned uranium is sealed in at A by a join at H, and the whole system is evacuated. Pure oxygen is then introduced at a pressure of about 30 to 40cm of mercury into the small volume via 3. The pressure of the system is measured by using the Bourdon gauge (mirror, E, lamp, G, and scale, F) as a null-point instrument. The pressure in the outer jacket is adjusted by the taps 5 and 6 until the reading on the scale is zero; the pressure is determined by the open manometer C. With 2, 3 and 4 closed, the gauge is carefully balanced, the mercury level is then read by the cathetometer, J, and the uranium is carefully heated electrically on a small sand-bath.The amount of reaction is controlled by time and temperature ; with experience this can be judged semi-quantitatively by the movement of the light spot on the scale. A temperature of 180" to 200" C gives a steady and controllable reaction rate. Removal of the sand-bath stops the reaction when required and the system is carefully brought back to within 0.2" C of its original temperature. The pressure change due to oxygen absorbed is measured by the change in the manometer pressure required to return the light spot to zero. The amount of oxygen absorbed for a given pressure change is independent of the absolute pressure of the oxygen and it is only the small pressure change, which can be accurately determined by the cathetometer, J, that is quantitatively significant.It is essential to know the magnitude of the constant volume and this is found by measuring the pressure change on admitting air at atmospheric pressure to the evacuated This part is mainly constructed from 1 6 m m diameter capillary tubing.July, 19531 OF TRACES OF GASEOUS ELEMENTS IN METALS 425 volume from the calibrated bulb, D. Specimens with as little as 10 cu. mm of added oxygen could be prepared with an accuracy of Fl.0 cu. mm by this method. The gas was extracted from a number of samples prepared in this way and also from the same uranium containing no added oxygen, and was then analysed by the mass spectro- metric method.The uranium samples containing added oxygen were kept as small as possible so that the contribution of their normal oxygen content was a minimum. In Table I1 the oxygen calculated from the carbon monoxide and dioxide of the gas in excess of that obtained from the same uranium unoxidised, is compared with the amount of oxygen added. TABLE I1 RECOVERY OF OXYGES FROM PREPARED URANIUM OXIDE STASDARDS Operating temperature, 1800" C Order of analysis 1-olume of oxygen Volume of oxygen The oblique-entry crucible was used in these experiments as before. of sample added, recovered , Recovery 1 15.0 14.5 97 * 10 I 13.7 13.3 97 * 10 3 IS.0 16.7 9 3 & 8 4 "0.2 21.6 107 f 5 The recoveries were all 100 per cent. within the experimental error and it is obvious that the system, operated as previously described, is satisfactory for oxygen in uranium. However, if relatively large uranium samples are used in an analysis the adsorption losses still appear in the fourth and fifth sample.If the sample weights are greater than 150 mg, it has been found that a maximum of three analyses can be carried out accurately in one determination. Some similar experiments were carried out with an open crucible, 1 cm in internal diameter; the effect of uranium evaporation can be seen from the results shown in Table 111. Between samples 1 and 2 the crucible was kept for about 30 minutes at 1800" C ; subsequent recoveries were accordingly low. TABLE I11 EFFECT ON OXYGEN RECOVERY OF URANIUM EVAPORATIOS FROM AX OPEX CRUCIBLE Operating temperature, 1800" C O/ cn.mm at S.T.P. cu. mm at S.T.P. /Q > Order of Oxygen Oxygen analysis added, recovered, Recovery, cii. inm at S.T.P. cu. mm a t S.T.P. % 1 2 3 4 7.7 7.3 14.0 1o.s 14.3 10.9 24.2 18.0 05 & 15 77 * 10 76 & 10 74 & 10 Artificially severe conditions were applied before samples 2, 3 and 4 were analysed, but Table I11 illustrates the effect of chemisorption by evaporated uranium even when the heat shield is flush with the top of the crucible. Greater consistency in normal samples has also indicated that the oblique-entry crucible is more reliable. Many attempts were made to prepare nitrogen standards by this method, but the vacuum fusion analysis always gave a gas that was substantially carbon monoxide in spite of elaborate purification of the nitrogen.This failure was probably due to the high temperature required for the reaction of uranium and nitrogen (about 400" C), which caused considerable outgassing of oxygen from the capillary walls even when silica was used. Preliminary vacuum outgassing of the heated part of the system was useless; it caused spurious effects owing to re-absorption on admission of nitrogen followed by desorption at the reaction temperature. .\NALYSIS OF URAXIUM OF KXOWK GAS COSTENT- A few analyses were carried out with a small piece of uranium taken from the same source as some immediately adjacent samples that were examined by Sloman by the macro method. Sloman's results showed that this sample was reasonably homogeneous, so that426 GREGORY, MAPPER AXD \VOOL)\YARL) : THE MICRO-DETER)IINATIOS [I-01.7s the gas content of adjacent samples can be said to represent the true composition of the piece used in our analyses. In the micro-determination, samples ranging in size from 120 to 190 mg were used. The results by the micro method are, within experimental error, similar to the results by the well-established macro method. The slight differences may be due t o small-scale inhoniogeneities, which are more likely to influence the micro method. Table IV shows the results by the two methods of analysis. TABLE I V COMPARISON OF THE PROPOSED METHOD WITH SLOMAN'S METHOD FOR URANIUM Analysis by micro-method Analysis by Sloman's macro-method A A r I r \ Sample Sample NO. Oxygen, Nitrogen, Hydrogen, No. Oxygen, Kitrogen, Hydrogen, yo w/w yo w/w ml per 100 g yo w/w yo w/w ml per 100 g 1 0.0078 0.0018 0.9 A O ~ O O i O 0*0010 1.04 2 0,0095 0.0005 1.1 B 0.0083 0.0012 0.84 Mean .. 0.0087 0.0012 1.0 Mean . . 0.0077 0~0011 0.94 A and B are samples from either side of the micro-analysed metal. The gas evolved in these samples was analysed for carbon monoxide, carbon dioxide, hydrogen and nitrogen by the Blacet - Leighton method. STEEL SAMPLES- Steel and most other ferrous alloys have never given any real difficulties when their gas contents have been determined by the macro vacuum fusion method. The analysis of several standard steels provided by the British Iron and Steel Research Association was carried out in our apparatus and the results are shown in Table V. At 1700" C the gas evolution was complete within 5 minutes and further heating or higher temperatures produced no more gas than that expected from the graphite blank rate, which was negligible in these analyses.Steel samples of 50 to 200 nig were used in this work, according to their total gas content. Sample No. 1 2 3 Mean . . 1 2 Mean . . I 2 3 4 Xean . , R. I. S. R. A. Identification No. 784 4654 8652 TABLE IT GAS DETERMISATION IS STEEL Analysis by micro-method B.I.S.R.A. analyses Oxygen, Yo w/w 0405 0.010 0.006 0.008 0.007 0.003 0.005 0.006 0.01 0.02 0.01 0.012 Xitrogen, 0-004 0.002 0.006 0.004 0.008 0.008 0,007 0.04 0.03 0.02 0.05 0.035 % w/w Hydrogen, 0.0003 0.0004 0.0005 0.0004 04003 0~0001 0.0002 0.0002 0.0004 0.0002 0.0007 0.0004 % w/w Oxygen, Nitrogen, % w/w yo W/U' 0.010 0.0045 0,004 0*0085 0.0145 0.016 The results by the micro method shown in Table V, although showing some scatter, are regarded as satisfactorily close to the standard figures, considering the small amounts of gas involved.. Again the scatter may be due to inhomogeneities sufficiently large to give differences in samples a few hundred milligrams in weight, but small enough to have negligible effect on the 5 to 20-g macro samples commonly used in the standardisation of these ingots.July, 1953: OF TRACES OF GASEOUS ELEMENTS IK METALS 427 JVe consistently found a small percentage of hydrogen in the samples, but no figures for this gas were reported in the B.I.S.R.A.analyses. ZIRCOKIUM- Nacro-analysed zirconium was supplied by Mr. H. A. Sloman, and an attempt made to analyse this metal on the micro scale met with little success. Sloman2 reports that oxygen and nitrogen are readily evolved from zirconium at 1750” C, but in the micro system no gas could be collected at temperatures below 2000” C.At this temperature and even up to 2200” C gas evolution was slow and incomplete. As the furnace of the micro system is not designed to operate above 2000” C for prolonged periods, the work on zirconium was dis- continued. It is difficult to say why our experience with this metal differs so much from that of Sloman. The difficulties with zirconium could probably be overcome by the use of a furnace designed to operate at considerably higher temperatures. The use of an iron bath did not effect any improvement. COSCLCSIOSS The results have shown that analysis of trace amounts of gases in steel and uranium can be carried out on the micro scale with an accuracy almost equal to the conventional macro method. The apparatus is in many ways simpler to operate at the micro level, particularly in regard to furnace design and assembly, and is probably more rapid. The procedure has been reduced to a reliable routine and it has obvious advantages in application to analysis of rare and valuable metals. The authors wish to thank Mr. ,4. A. Smales for his interest and valued advice in this work. Acknowledgments are also due to Mr. G. Palmer and his group for considerable assistance in mass spectrometric analysis of gas samples, and to Mr. J. Wright for advice in the Blacet - Leighton analyses. The excellent glass blowing of Mr. B. George was a decisive factor in the success of this work. Finally, the authors wish to thank the Director, Atomic Energy Research Establishment, Harwell, for permission to publish this paper. 1. 2. 3. 4. 13. 6. 7. 8. 9. 10. 11. REFERENCES Sloman, H. A , , “Seventh Report on the Heterogeneity of Steel Ingots,” Iron and Steel Institute, Sloman, H. A., and Harvey, C. rl., J . Inst. Metals, 1951-52, 80, 391. Brewer, L., Atomic Energy Commission Document MDDC-366. Rice, C. N., Atomic Energy Commission Document MDDC-356 (LADC 143). 1944. Lipkin, D., and Perlman, M. L., Atomic Energy Commission Document MDDC-294 (LADC-142), Seifert, R. L., Gilpatrick, L. O., Pliipps, T. E., and Simpson, 0. C., i\tomic Energy Commission Guldner, W. A , , and Reach, A. G., .4?zal. Chem., 1950, 22, 366. Walter, D. I., Ibid., 1950, 22, 297. Katz, J. J., and Rabinowitch, EugBne, ‘ I The Chemistry of Uranium,” National Xuclear Energy Series, Division VIII, Vol. 5, McGraw-Hill Book Co. Inc., New York and London, 1951, p. 130. Blacet, F. E., and Leigliton, P. A., I n d . E n g . Chem., A n a l . Ed., 1931, 3, 266; 1933, 5, 277. Chiotti, P., and Rogers, B. London, 1937, Section IV, p. 52. 1944. Document (AECD-2331). Metal Pvogr., 1951, 60, 60. ANALYTICAL CHEMISTRY GROUP CHEMISTRY DIVISION A.E.R.E., HARWELL, BZRKS. October 22nd, 1952

ISSN:0003-2654    

DOI:10.1039/AN9537800414

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

428 r r, LOW tension D.C. unit Q Q Valve voltmeter - e-- I\ :El”,”, E.x,.* _t) PALMER AND VOGEL: APPARATUS FOR AlJTOMATIC CONTROL OF [Vol. 5s \ Apparatus for Automatic Control of Cathode Potential in Electro-analysis BY J. F. PALMER ASD A. I. VOGEL St,-<, ~- Error input K Full details are given of an electronic instrument that has been constructed for the automatic control to within & 1 millivolt of the potential a t the cathode with respect to a standard reference electrode in electro-analysis, The parts of the apparatus include a D.C. output unit working from the A.C. mains for the supply of current at low voltage, a valve voltmeter and a high-tension power-pack: these can be used independently and should be of value in an analytical laboratory. . J IT is now generally accepted that an apparatus for the automatic control of cathode potential is very useful in an analytical laboratory, and many instruments have been designed for this p u ~ p o s e .~ , ~ , ~ , ~ , ~ , ~ ~ ~ , ~ The instrument described has been built by the authors and was exhibited at the International Congress on Analytical Chemistry at Oxford, September, 1952. It is based on the principle originated by Harvey Dieh14 in 1948, and is believed to Mains input e Amplifier Variac unit Fig. 1. Diagrammatic scheme of thr apparatus assembly be a great improvement on previous instruments ; moreover, all the components used are obtainable in Great Britain. The prototype of this apparatus was described in the second edition of Vogel’s “Textbook of Quantitative Inorganic ,4nalysi~.”~ In the form in which it was first devised, the operation of the circuits as described in the textbook did not give complete satisfaction, and in its present form many improvements have been incorporated.The apparatus maintains the cathode potential constant at a pre-determined value in the range of 0 to 2.0 volts during electro-analysis. The current used in the electrolytic cell can vary from 10 amperes to a few milliamperes. A potentiometer that can be read directly in millivolts and a valve voltmeter are incorporated in the apparatus: both are directly calibrated against a built-in standard Weston cell. The potential difference between the cathode and a reference electrode (usually a saturated calomel half-cell) can be set and automatically maintained within 1 millivolt of the set value.The ultimate sensitivity of the apparatus depends on the efficiency of stirring, the rigidity of the electrodes, and on any other factors that may contribute to mechanical instability. This maximum sensitivity is higher than is usually necessary for most applications of controlled-potential electro-analysis ; a variation of & 10 millivolts is gcnerallj. sufficient.Fig. 2. Circiiit diagram of valve amplifier nnit (inclding potentiometer and valve voltmeter) 4 Y E430 The operation of the instrument depends on the detection of a D.C. “error voltage” between the potentio- meter and the controlled cell (e.g., cathode of the electro-deposition cell and saturated-calomel reference electrode). This “error voltage” is fed into a single-stage balanced D.C.amplifier, which has an input resistance of approximately 11 megohms ; this resistance is large enough to ensure that polarisation of the calomel-cathode cell does not occur. The output of the D.C. amplifier is fed into a polarised relay, which converts the D.C. “error voltage” into 50-cycle square-topped alternating current. The alternating voltage is passed into a mu- metal step-up transformer and then into a three-stage resistance-capacity-coupled amplifier. The output of this amplifier is connected to the grids of two 12AU7 valves, joined in parallel, the anodes (plates) of which are joined directly to the secondary winding of the centre-tapped power transformer, T, (Fig. 2). These valves therefore act as phase-discriminating power amplifiers; the phase relationship between the amplified pulse and that of the secondary winding of the transformer produces a rectified impulse in the motor winding L,.The phase relationship between this impulse and its counterpart through the motor winding L, determines the direction of rotation of the motor. The reversible induction motor, geared to 1 r.p.m., is coupled to a 6-ampere capacity “Variac” transformer controlling the A.C. input to the electrolysis unit (Fig. 3). When a D.C. “error voltage” is present, the motor (which controls the variable arm of the “Variac” transformer) will rotate slowly and so alter the potential difference between the mode and cathode until the “error input” is reduced to zero, when the motor will cease to rotate.Hence by \-ariation of the potential difference between the anode and cathode, the potential of the cathode in relation to the saturated-calomel reference electrode is maintained at a constant value. The apparatus, excluding the electro-deposition cell and associated electrodes, is coin- posed of six units- 1. A D.C. output unit, working from the A.C. mains, to supply the necessary current a t low voltage to the electro-deposition cell. 2 . A motor-driven “Variac” transformer to control the A.C. input to the D.C. output unit. 3. A potentiometer that can be set to the algebraic sum of the desired potential drop between the cathode and the potential of the reference electrode, or to any other pre-determined value. 4. A valve amplifier unit operating on a minimum stimulus of 1 millivolt D.C., which is connected to a reversible motor controlling the “Variac” transformer. 5.,4 valve voltmeter for measuring the cathode - reference cell potential (Einput; Fig. 1) and also, if required, for external use (Eext). 6. A valve-stabilised power-pack for supplying the necessary high-tension and low- tension voltages for units 4 and 5. PALMER AND VOGEL: APP.4RATUS FOR AUTOMATIC COXTROL OF [Yol. 78 A diagrammatic representation of the whole assembly is shown in Fig. 1. D.C. OFTPUT wn- The direct current required for the electrolysis is drawn from the A.C. mains by the use of the step-down transformer T, (10 amperes, 15 volts), a selenium bridge rectifier, SE, and a smoothing filter circuit. Current at 3, 6, 9, 12 or 15 volts may be drawn from the transformer secondary by setting the switch S,.The bridge rectifier can handle up to 20 volts, and the switch S, ensures that there is an adequate margin with due regard to the maximum possible input from the “Variac” transformer. The rectifier, SE (see Fig. 3), is followed by an inductance - capacity smoothing filter, consisting of a 3000-pF, 24-volt, paper condenser, C,, a 40 millihenry choke, L,, and a 9000-pF, 24-volt, condenser, C, (three 3000-pF condensers wired in parallel). Meter M, is a triple- range ammeter (0 to 0.1,O to 1.0 and 0 to 10 amperes), and meter M, is a triple-range voltmeter (0 to 1.0,O to 5.0 and 0 to 20.0 volts) ; these indicate the electro-deposition current and voltage, respectii-ely. (The low-tension unit is fed from the socket P, 011 the “Variac” transformer unit .) “I’AR1.4C” TRASSFORMER LSIT- The mains input to the low-voltage unit, Fig. 3, is controlled by the “Variac” transformer, the variable arm of which is attached to a geared-down reversible two-phase four-pole inductionJ~ily, 19531 CATHODE POTENTIAL I S ELECTRO-AKALYSIS 43 1 motor, L, and L,, Fig.4. to rotate a t 1 r.p.m. or by the amplifier. The motor, developing 5 ft.-lb torque per 1 r.p.m., is geared down Provision is made in this unit for controlling the motor either manually ’IT’ith the double-pole change-over switch, S5, in the position shown TI R2 , 53 If - - L I T” I / o(lll I I 1 Fig. 3. D.C. low-tension unit circuit in Fig. 4 and the six-pin plug from the amplifier inserted, the motor ilI will be controlled by the amplifier.With switch S, in the alternative position, and the two-way manual motor-direction control unit, Fig. 5 , plugged in a t P, in lieu of the six-pin plug from the amplifier, the “Yariac” transformer can be set manually. The safety switches, S, and S,, are fitted to the “Variac” transformer unit to prevent the moving contact arm of the “Variac” transformer from rotating too far and so damaging the carbon brush. If the automatic Fig. 4. Circuit diagram of motor-driven Variac unit control, for some reason or another, does not operate correctly and, in consequence, either of the safety swtiches S, or S, is brought into action, the motor will cease to function. To bring the moving arm of the “Variac” transformer back to its normal working range, it is then necessary temporarily to connect the manual unit, Fig.5 . With the instrument working correctly, over-running of the moving arm of the “Variac” transformer should not occur. The possibilities should be borne in mind that, if the standard reference electrode circuit is broken during the analysis owing to the presence of an air bubble, or if the switch, S,, on the transformer, Fig. 3, is set at too low a value initially, either of the safety switches may come into operation. Switch S,, Fig. 4, disconnects the windings when the motor is idle between determinations, and so prevents overheating of its windings.432 connected to the A.C. input of the low-tension unit, Fig. 3. either direction will be indicated by the movement of the needle of the voltmeter, M,.PALMER AND VOGEL: APPARATTS FOR A‘CTOMATIC CONTROL OF [Yol. 7s The A.C. output of the “Variac” transformer is brought out to a socket, P,, which is The rotation of the motor in I THE POTESTIOMETER USIT- This incorporates a direct-reading “Colvern” helical potentiometer, K,,, Fig. 2, of exactly known resistance (about 100,000 ohms). The potentiometer is divided 1 to 1000, and so reads directly in millivolts when exactly 1 volt is placed across it. A resistance network of equal value, R,, + R,, + R,,, is placed in series with the direct-reading potentiometer, R,,. The toggle switch, S,,, by bringing this network into the circuit, extends the voltage range of the potentiometer, R,,, by 1.000 volt. I 1 ! I T H E V.4LVE AMPLIFIER UNIT- The first part of this unit consists of a single-stage D.C.ainplifier with two valves in a bridge circuit, VSA and VSR, Fig. 2. The input to the D.C. amplifier is introduced via the switch Sgn, which forms one of the sections of the operation selector switch. This switch has four positions: OFF, ZERO, CALIBRATE and USE. In the first two positions the input to the D.C. amplifier is earthed through RZ6. During the initial adjustment of the amplifier, the manual zero control, R,,, is set at about the mid-point of its range and the amplifier is adjusted to zero by means of the pre-set zero control, RZ5; the latter need not be touched again unless a valve is replaced. The two valves of the bridge amplifier are shown separately in Fig. 2 for the sake of clarity, but are in reality combined in one glass envelope in the 12AX7 valve.When S, is in the CALIBRATE position, the potentiometer will be balanced against the standard Ljreston cell as previously described; in the USE position, the potentiometer will be balanced against the cathode - reference cell potential. Note on the operation selector switch., Sg-This switch consists of six single-pole 4-way sections. The first section, S,?, disconnects the potentiometer battery, B,, when it is not in use. Section SBa selects the appropriate input for the D.C. amplifier. Sections S,,, SBD and S,, are concerned with the inter-connections of the Weston cell, potentiometer unit, and the cathode - reference electrode cell. Section S,, is the input selector switch of the valve voltmeter.July, 19531 CATHODE POTENTIAL I S ELECTRO-ANALYSIS 433 The operation selector switch, S,, controls both the potentiometer - amplifier and the valve- voltmeter circuits.In the first position, OFF and V.V. E X T . , the amplifier input is connected to earth, the battery B, is disconnected, and the valve-voltmeter input leads are brought out to the valve- voltmeter external terminals on the panel. In the second position, ZERO, the amplifier input and the valve-voltmeter input leads are connected to earth, which permits the zero controls of the amplifier and valve voltmeter to be correctly adjusted. In the third position, CALIBRATE, the potentiometer is connected in opposition to the standard Weston cell; the latter is also connected to the input terminals of the valve voltmeter.In the fourth position, USE and V.V. CELL, the potentiometer is connected so as to oppose the cathode - reference cell potential, whilst the valve voltmeter indicates the magnitude of this potential. The output of the D.C. amplifier is impressed between the vibrating reed of a polarised relay and the centre-tap of a mu-metal-screened step-up transformer, T,. The condenser, C,,, serves to by-pass all higher frequencies to earth and its value is best determined experi- meritally; the optimum capacity lies between 0.01 and 0.1 microfarad. The output of the transformer, T,, is passed into a three-stage A.C. amplifier, which incorporates two stages of resistance-capacity-coupled amplification wired in cascade for which a single 12AX7 valve, V,, $- VZB, is used, followed by another resistance-capacity-coupled stage for which a 12AX7 valve, V,, strapped as a single valve, is used.A potentiometer, R,,, is inserted in the grid circuit of the final 12‘4x5 valve to ser1.e as a sensitivity control for the complete amplifier, and this can bc used to prevent “liuntiiig,” which would occur if the amplifier were over- sensit i\-c. The output of the resistance-capacity-coupled amplifier is fed to the grids, connected in parallel, of two 12AU7 valves, V4 and V5 in Fig. 2 , which act as phase-discriminating amplifiers, as their anodes (plates) are connected directly to the secondary of the centre- tapped mains transformer, T,, and so control the direction of rotation of the “l’ariac” transformer motor through L,, Figs. 2 a.nd 4. The condenser, C,,, is coupled in parallel with condenser C,, Fig.2 , when the amplifier is connected to the “Variac” transformer unit. This condenser increases the torque of the motor to a satisfactory value when it is under automatic control. T-XLVE VOLTMETER- The circuit (Fig. 2 , upper left-hand corner extending to N,) is of the Wheatstone bridge type; the two arms consist of equal resistances, R, and K,, of 30,000 ohms each, whilst the other two arms are formed by the internal resistances of the two parts, V,* + V,=, of the ll’AX7 valve. The voltage to be measured is applied to the control grid of the other part, V7B; this causes an unbalance in the circuit, which is indicated on the meter, M,, the deflection being linear to well within 1 per cent. As for the D.C.amplifier there is a pre-set and a manual zero control, and the initial adjustment to zero is carried out by setting the manual zero control, R,, to about the mid-point of its range and turning the pre-set control, K,, to give no deflection on ?f4. The manual sensitivity control, R,,, is used in the CALIBRATE position of the operation selector switch, S,, to secure a deflection of 1.018 volts on M,. The selector switch, S,,, permits the use of any of the three ranges, 0 to 1.0, 0 to 2.5 or 0 to 5 0 volts; these three ranges are controlled by the potentiometers R,,, R,, and R14. The valve voltmeter records directly to 0.001 volt on the 0 to 1.0 range. The initial calibration of the valve voltmeter is carried out by setting R,, a t about its mid-point and adjusting R,, to give a deflection of 1.018 volts on M, with S, in the CALIBRATE position.This is followed by setting the switch S,, to the 2.5 and 5.0-volt ranges in turn and making the appropriate adjustments to R,, and R,,, respectively. During the initial calibration the zero position of MA must be repeatedly checked and adjusted with R,, if necessary. A reversing switch, S,,, is incorporated in the circuit to provide for change in the polarity of the cathode in relation to the reference electrode. The thermal delay switch, S,,, in series with the meter hf,, protects the latter during the initial warming-up period of the valve 11,. POWER-PACK- The circuit for the valve-stabilised power-pack is shown in Fig. 6.* The VR105 regulators provide a constant reference voltage for the triode component, VSB, of the 12-4U7 * This is a modification of the circuit given in Fig.7-15 (p. 219) in “The Radio -4mateurs’ Handbook,” The control grid of one of the parts, V7*, is earthed. 29th Edition, .4merican Radio Relay League, 1952.434 valve. A reduction in the load current produces an increase of the anode voltage on VSu and an increase of the bias on VBL; in consequence more current passes through the 500-ohm resistance R,,, the bias on V, (6Y6G) is increased, the voltage across 11, is also increased and therefore the output voltage falls to the pre-set value of 300 volts. The high-tension supply for this stabilising circuit is provided by a full-wave rectification circuit. The mains supply is fed into the apparatus via the main on - off switch, S,,, and the fuses, and thence to the A.C.ammeter, M,. From the ammeter, mains leads are taken to the stirrer motor, to the transformer, T3, Fig. 2, to pins 3 and 5 of P, on the “Variac” transformer unit, Fig. 4, and to the input winding of the power-pack transformer, T,, and also to T,, Fig. 7 , if series heating of the valves is used. PALhIER AND VOGEL: APPARATUS FOK AUTOMATIC CONTROL OF y01. 7s 240 3 stirrer To *+---i~ A.C. P7 Mains S volt 3 amp E- , 1 S14> I I I 1, Fig. ti. High-tension power pack circuit The full-wave rectifier comprises V,, (U52) and the 450-0-450-volt secondary winding on T,. The H.T.+ supply is taken from the cathode of V,, to the thermal delay switch, S14. This switch protects the smoothing condensers and the 85A1 regulators of the D.C.amplifier and valve voltmeter, Fig. 2, during the warming up of the valves of the apparatus. After the switch S,, is a two-stage inductance - capacity smoothing filter, consisting of chokes L, and L, and condensers C,, and Cl,. The valve heaters can be supplied from a 6.3-volt winding on the mains transformer, T4, if a stability of i 5 millivolts suffices. In any event, valve V, must receive its heater supply from an independent winding isolated from the high-tension negative source. In the instrument as built, the valve heaters were wired in series and supplied by the stabilising circuit shown in Fig. 7 . This circuit makes use of full-wave rectification, followed by an inductance - capacity filter. The output is then fed into two 6Y6G series valves, a 6AU6 being used as the regulator valve.The cathode potential of the regulator valve is maintained at a constant value by a neon regulator, VR75. The resistances R,, and R,, are adjusted so that a current of 150 milliamperes passes through the heater chain, as detected by a milli- ammeter inserted in the jack, J3. The cabinet on the A photograph of the complete assembly is reproduced as Fig. 8.July, 1953: CATHODE POTEKTIAL IN ELBCTRO-ASALYSIS 435 left houses the motor and “Variac” transformer unit,* the adjacent cabinet contains the potentiometer, valve amplifier, valve voltmeter and power packs (this is kept cool by means of a “Vent-Axia” fan installed a t the back of the unit), whilst the cabinet on the right incor- porates the low-tension D.C.unit. The “Variac” transformer unit is connected by cable supplying alternating current to the low-tension D.C. unit. A 6-wire cable connects the “Variac” unit (1) and the potentiometer, valve voltmeter and power-pack unit ( 2 ) . The latter T5 Fig. ‘7. Low-tension power pack circuit V8B) V7B) V6B) VZB) unit (2) and the cell unit are connected to two wires from the D.C. input terminals (one to the cathode and the other to the reference cell) : the anode and cathode are also connected by two wires to terminals on the low-tension D.C. unit. OPERATION OF THE COMPLETE INSTRUMENT FOR CONTROLLED-POTESTIAL ELECTRO-ANALYSIS The apparatus is first earthed to a water pipe or other suitable earthing conductor; the earthing connection should be soldered. The electro-deposition cell is connected to the apparatus.Two 15-ampere single insulated wires are connected from the D.C. output terminals on the low-tension unit, Fig. 3, to the anode and cathode, respectively. Two more wires from the reference electrode and the cathode are connected to their respective terminals on the potentiometer - amplifier unit, Fig. 2. It is important to use four wires for connecting the cell to the apparatus, for, if the negative output terminal of the D.C. low-tension unit were joined directly to the cathode terminal of the potentiometer - amplifier unit and a common wire were run to the cathode, it would be found that the valve voltmeter would indicate, and the amplifier would respond to, the voltage drop through the common lead superimposed upon the cathode - reference cell potential.The apparatus is plugged into the mains and switch S,,, Fig. 6, is closed. The stand-by switch, S,, Fig. 4, and the mains switch, S,, Fig. 3, on the low-tension D.C. power unit are kept in the “off” position. The apparatus is allowed to warm up for 10 to 15 minutes. The operation selector switch, S,, Fig. 2 , is then turned to the ZERO position. The valve voltmeter is adjusted for zero by means of R,. The stand-by switch, S,, is turned on, the sensitivity control, Rdo, on the amplifier unit is advanced and the amplifier set to zero by R,,, Fig. 2 ; the zero position is indicated by the cessation of movement of the “Variac” transformer motor and of the pointer of A&, Fig. 4. The operation selector switch, S,, is now turned to CALIBRATE, and the valve voltmeter is adjusted by R,, to read 1.018 volts.The potentiometer unit is calibrated by manipulating R,, until movement of the “Variac” transformer motor or of the needle of M, ceases and M, indicates about 20 volts. The apparatus is then ready for use; should any delay occur, the stand-by switch, S,, is opened. * This description should be read in conjunction with Fig. 1.436 The tip of the saturated-calomel electrode is placed as close as possible to the cathode and at about its middle. If the cathode happens to be positive to the reference electrode, the reversing switch, S,,, Fig. 2, on the valve voltmeter must be thrown into the correct position. The ammeter switch, S,, Fig. 3, should now be set to the 10-ampere range and the voltmeter switch, S,, to the 20-volt range.The stirrer motor for the electrolysis cell is set in motion, the A.C. input switch, S,, Fig. 3, is closed, and the voltage selector switch, S,, rotated to 15 volts. The stand-by switch, S,, on the “Variac” transformer unit, is closed, and the voltage selector switch, S,, is re-adjusted so that the A.C. output of the “Variac” transformer unit, as indicated on the voltmeter M,, is between 100 and 200 volts. If M, reads over 225 volts or less than 5 volts, the safety switches, S, or S,, Fig. 4, will come into operation and switch off the motor; it will then be necessary to use the manual press-button unit, Fig. 5 , to re-start the apparatus. If the “Variac” transformer is functioning in the upper part of its range, it will exert a smoother and more accurate control on the voltage of the D.C.output of the low-tension unit. This will permit the sensitivity control, R,,, Fig. 2, to be more fully advanced without instability occurring. The voltage selector switch, S,, should therefore be set so as to keep the A.C. output of the “Variac” transformer unit as high as possible. Once electro-analysis has begun, the sensitivity control should be advanced as far as is compatible with absence of instability; any instability will be revealed by a gentle oscillation of the needle of the A.C. voltmeter, M,. The operation of the “Variac” transformer unit is intermittent owing to the dependence of the cathode over-voltage on the current and for other reasons. When the electrolysis current has fallen below 1 ampere, the range of M, is changed by S, first to 0 to 1.0 ampere and subsequently to 0 to 0.1 ampere.The electrolysis is discontinued when the current falls and remains at between 10 and 30 milliamperes. The switch S, is set at ZERO, the electrolyte is removed ( e g . , by lowering the electrolysis vessel), and the electrodes are simultaneously washed without interrupting the current. The stirrer motor is then switched off, and switches S,, S, and S,, are turned off. If another electrolysis is to be carried out immediately, the stand-by switch, S,, and the A.C. input switch, S,, are opened. The ZERO and CALIBRATION positions of the potentio- meter unit are checked again; the “Variac” transformer output should be reduced to about 20 volts during the re-calibration.This ensures that the initial cathode - reference cell potential will not exceed that required in the subsequent electrolysis. The potentiometer unit may now be set to the required value and S, to 15 volts; switches S, and S, are closed, the stirrer started, and S, re-adjusted as described above. PALMER AKD VOGEI,: APPARATUS FOR ACTOMATIC CONTROL OF [Vol. 78 The solution to be electrolysed is now placed in the electro-deposition cell. CHECKING THE VALVES- The efficiency of the valves may be checked (e.g., when a breakdown occurs) by inserting a milli- ammeter (0 to 10 mA) into the cathode leads of all the valves by means of a rotary switch. The failure of a rectifier valve will be apparent by a reduced or zero reading on all the valves supplied by it.When the milliammeter is inserted in the common cathode lead of the phase discriminators V, and V,-a parallel shunt will be required to prevent over-deflection of the meter-it will serve as a good zero point indicator, as when no signal is being fed to the phase discriminators and consequently the “Variac” transformer motor is stationary, the current reading will be at a maximum. DETAILED LIST OF COMPONENTS The details of the various components are listed below to correspond with the Figs. D.C. LOW-TENSION UNIT (FIG. 3)- The components are those used in the authors’ apparatus and are intended as a guide. R, = IIp= c, = c, = L, = T, = M, = M, = SE = s, = shunt supplied with Pullin voltmeter (M,). shunt supplied with Pullin ammeter (M,). 3000-pF, 24-volt, paper condenser (M.R.Supplies). 9000-pF (3 x 3000 pF), 24-volt, paper condenser (M.R. Supplies). 40-millihenry 10-amp. smoothing choke (M.R. Supplies). mains step-down transformer; primary windings 10-0-200-220-240 volts; secondary windings, Pullin voltmeter, CSO/IV, knife-edge pointer and mirror scale, scaled to read 0 to 1.0, 0 to ’ 0-3-6-9-12-15 volts a t 15 amp. (M.R. Supplies). . - 5.0 and 0 to 20.0 volts. Pullin ammeter, C30/100M, knife-edge pointer and mirror scale, scaled to read 0 to 0.1, 0 to 1.0 and 0 to ‘10.0 amp. (MeaGurkg Instruments (Pullin) Ltd.). or Yaxley). selenium bridge rectifier, funnel type, rated a t 15 volts, 15 amps. (M.R. Supplies). two-pole six-way six-bank double-contact rotary switch, wired as one-pole six-way (OakFig. 8. General view of complete assemblyJuly, 19531 CATHODE POTENTIAL I N ELECTRO-ANALYSIS 437 S, or S, = six-pole three-way switch, wired as one-pole three-way (Bulgin, S.438).S, A mains socket and plug (not shown in Fig. 3) was used to connect the low-tension unit with Pz on the “Variac” transformer unit (Bulgin, P.162). MOTOR-DRIVEN “VARIAC” TRANSFORMER UNIT (FIG. 4)- L, + L, = w-indings of four-pole two-phase reversible induction motor, geared down to 1 r.p.m., It was necessary to = two-pole on - off switch (Rulgin, S.267). attached t o a “Variac” transformer of 5 amperes capacity (Zenith). rewire the motor as supplied by the manufacturers to conform with Fig. 4. C, = 0*25-pF, 500-volt, paper condenser (Webbs Radio). C, = 0.4-pF condenser, supplied with motor unit (Zenith). S, S, and also S7 = safety switch supplied with motor.S, At, P, = six-pin socket (Bulgin, P.166). P, MAKUAL MOTOR-DIRECTION CONTROL UNIT (FIG. 5)- This consists of one press-button two-way switch as supplied by Zenith Electrical Co. Ltd., with the motor-controlled “Variac” transformer. I t was necessary to drill another hole a t the end opposite to the one already present; the original hole was used for the wires to the six-way plug, which was inserted into P2, Fig. 4, and the new hole carried the two wires to the A.C. mains. = four-pole change-over switch, required for alternative circuit with manual control (Bulgin, = two-pole on - off switch (Bulgin, S.267). = Pullin 0-300 volts A.C. voltmeter, R30/300V, knife-edge pointer and mirror scale (Measuring S.301).Instruments (Pullin) Ltd.). = three-pin socket, one pin not used (Bulgin, P.162). VALVE AMPLIFIER UNIT (INCLUDING POTBNTIOMETER AND VALVE VOLTMETER) (PIG. 2)- (The resistances and condensers, unless otherwise indicated, were Erie or Dubilier products.) T, = mu-metal-screened input transformer, 1 to 100 step-up ratio; primary winding, 400 turns, centre-tapped, D.C. resistance 34 ohms : secondary winding, 10,000 turns, D.C. resistance 2700 ohms; inductance about 2.15 henries (Sowter). Relay = Carpenter polarised high-speed relay, type 5PA18A (Telephone Manufacturing Co. Ltd.). (V, + VGn) = ( V 7 A + V,R) = ( V 2 A + V2B) = V, = 12AX7. V, = V, = 12AU7. Four 85A1 Neon regulators (Mullard). S, S,, S,, S,, S,, R, = 1000-ohm, 1-watt potentiometer. R, = R, = 6000-ohm, 1-watt, high-stability ( & 5 per cent.tolerance) resistance. R, = R, = 30,000-ohm, 1-watt, high-stability (&5 per cent. tolerance) resistance. R, = 1000-ohm, 1-watt, potentiometer (carbon track), R, = 10-megohm, 1-watt, high-stability resistance. R,, = 1-megohm, 1-watt, high-stability resistance. R,, = 25,000-ohm, 1-watt, potentiometer. R, = R,, = R,, = 50,000-ohm potentiometer. R,, = 10,000-ohm wire-wound potentiometer with 5000-ohm resistance R,, in series, both 1-watt. R,, = 25,000-ohm potentiometer (carbon track). R,, = 100,000-ohm potentiometer (carbon track). R,, = 100,000-ohm helical potentiometer (Colvern Ltd.). q0 = 2500-ohm pre-set potentiometer. q, = 20,000-ohm pre-set potentiometer. R,, = 90,000-ohm, 1-watt, high-stability resistance.q, = %., = 2000-ohm, 1-watt, high-stability (+5 per cent. tolerance) resistance. %, = 1000-ohm potentiometer (carbon track). q6 = 1-megohm, high-stability resistance. 112, = 10-megohm, high-stability resistance. %, = R,, = 27,000-ohm, 1-watt, high-stability (+5 per cent. tolerance) resistance. R,, = 1000-ohm, 1-watt, potentiometer (carbon track). R,, R, = 0.25-megohm, O.5-watt, high-stability resistance. R, = 15,000-ohm, O.B-watt, high-stability resistance. R, = 20,000-ohm, O.j-watt, high-stability resistance. R, = 0.25-megohm, Oej-watt, high-stability resistance. R,, = R,, = 1-megohm, 0.5-watt, high-stability resistance. R,, = 100,000-ohm, O.B-watt, high-stability resistance. R,, = 50,000-ohm, O.5-watt, high-stability resistance. R,, = 0.25-megohm potentiometer (carbon track), &, = 100-ohm, 1-watt, potentiometer (carbon track), = two-bank six-pole four-way ceramic rotary switch (Webhs Radio).= two-way single-pole toggle switch (Bulgin). = three-way single-pole rotary switch (Bulgin or Yaxley). = two-pole change-over switch (Bulgin, S.270). = thermal delay switch (Rulgin S.364). = 10,000-ohm wire-wound potentiometer with 5000-ohm resistance in series, both 1-watt.438 R4, = R,, = c, = c, = c, = $11 = PALMER AND VOGEL: APPARATUS FOR AUTOMATIC CONTROL OF [Vol. $8 0.1-megohm, 1-watt, resistance. 300-ohm, 3-watt, wire-wound resistance. 25-pF, 25-volt working, condenser. C, = 8-pF, 500-volt working, condenser. C, = C,, = 0.1-pF. 600-volt, condenser. 0.01 t o Osl-pF condenser (see text). 25-pF, 25-volt working, condenser.1-pF, 1000-volt working, condenser. C,, = O.Ol-pF, 100-volt working or higher-rated condenser. two Siemens l*J-volt T-tvDe cells. Lia = c,, = c,, = il B, = €3, = standard Weston cell. A 6-inch “Vent-..lxia” extract unit (grille model) was fitted into the back of the cabinet to control the temperature. It was found that the change in temperature during the initial warming up of the components in the cabinet caused a gradual drift in the calibration of the potentio- meter. The “Variac” transformer motor is shown in outline in Fig. 2; the numbers correspond to those on the 6-pin socket shown in Fig. 4. HIGH-TENSION POWER-PACK (FIG. 6)- mains transformer; primary windings, 10-0-200-220-240 volts; secondary windings, 450-0-450 volts a t 150 mA, 5 volts a t 3 amperes, 6.3 volts a t 3 amperes, centre-tapped, and 6.3 volts at 3 amperes, centre-tapped (optional, not required if series heating is used) (M.R.Supplies). VaB) = 12AU7 (\Vebbs Radio). 6Y6G (Webbs Radio). U52 (Webbs Radio). thermal delay switch (Bulgin, S.363). two-pole on - off switch (Bulgin). 500-ohm, 0.5-watt, resistance. R,, = 33,000-ohm, 2-watt, resistance. 0.25-megohm, O*5-watt, resistance. 9000-ohm, 1-watt, resistance. 200,000-ohm, Oej-watt, resistance. 100,000-ohm potentiometer (carbon track). 0.3-megohm resistance. C,, = 16-pF, 600-volt working, paper condenser. O.l-pF, 300-volt working, paper condenser. O.Ol5-pF, 300-volt working, paper condenser. 8-pF, 600-volt working, paper condenser. LOW-TENSION POWER-PACK (FIG. 7)- T, = RE, = Rm = Rs, = R55 = R,, = c,, = c2, = c,, = c,, = c,, = J s = v,, = v,, = VR 75/36” = neon regulator (nrimar).L6 = 10 to aO-henry, 200-mA smoothing choke (M.R. Supplies). RS6 = mains transformer; primary windings, 10-0-200-220-240 volts; secondary windings, 250-0-250 volts a t 200 mA, 5 volts a t 3 ampercs, 6.3 volts a t 3 amperes, centre-tapped, 6.3 volts a t 1 ampere, centre-tapped (1l.R. Supplies). O.S-megohm, 0.5-watt, resistance. 10,000-ohm, 3-watt, resistance. 25,000-ohm, 3-watt, resistance. lO,OOO-ohm, 3-watt, wire-wound potentiometer (Colvcrn). 5000-ohm, 3-watt, rcsistance. 1000-ohm, 25-w-att, pre-set resistance. 8-pF, 350-volt working, paper condenser. 16-pF, 350-volt working, paper condenser. O.l-pF, 350-volt working, paper condenser. O.Ol-pF, 350-volt working, paper condenser.S-pI;, 350-volt working, paper condenser. closed-circuit jack socket. U52 valve. V,, = 6Y6G valve. ~IISCELLANEOUS COMPONENTS- Low-tension D.C. unit: 22 X 15 x 17 inches, asbestos-lined (incorporating an air-cooled selenium rectifier; the size of the low-tension unit cabinet can be reduced slightly if a motor-driven blower is used to cool the rectifier). The addresses of the suppliers of the important components are- IVooden cabinets-“Variac” transformer motor unit: 7 x 1s x 9 inches. Metal cabzmt-This was a standard Imhof model. Colvern Ltd., Mawneys Koad, Romford, Essex. Alfred Imhof Ltd., 112-116, New Oxford Street, London, W.C.l. Measuring Instruments (Pullin) Ltd., Winchester Street, Acton, London, W.3. M.R. Supplies Ltd., 68, Xew Oxford Street, London, W.C.1. Dr. G. A. V. Sowter, l ~ , Head Street, Colchester, Essex.July, 19531 CATHODE POTENTIAL I N ELECTRO-ANALYSIS 439 Telephone Manufacturing Co. Ltd., Hollingsworth Works, Martell Road, West Dulwich, London, S.E.21. Webb’s Radio, Ltd., Soh0 Street, Oxford Street, London, W.l. Zenith Electric Co., Ltd., Villiers Road, Wllesden Green, London, N.W.2. Othcr components for which manufacturers’ addresses are not listed, can bc obtained from most suppliers of radio components. The authors’ thanks are tendered to the Governors of IVoolwich Polytechnic for liberal grants and excellent facilities, and to Imperial Chemical Industries Ltd. for financial assistance that has helped to defray the heavy cost incurred in the development of the apparatus to its final form. REFERENCES 1. 2. 3. 4. Diehl, El., “Electrochemical Analysis with Graded Cathode Potential Control,” G. F. Smith 5. 6. 7. 8. 9. Vogel, A. I., “Text Book of Quantitative Inorganic Analysis: Theory and Practice,” Second Hickling, A., Trans. Faraday SOC., 1942, 38, 27. Caldwell, C. W., Parker, K. C., and Diehl, H., I n d . Eng. Cltein., Anal. Ed., 1944, 16, 532. Lingane, J . J., Ibid., 1945, 17, 332. Chemical Co., Columbus, Ohio, 1948. Allen, M. J., Anal. Chem., 1950,22, 804. Lamphere, R. W., and Rogers, L. B., Ibid., 1950,22, 463. Wehner, P., and Hindman, J . C., J . Amer. Chern. SOC., 1950,72, 3911. Milner, G. W. C., and Whittern, R. N., Analyst, 1952, 77, 11. Edition, Longmans, Green & Co., Ltd., London, 1951, pp. 530-545. DEPARTMENT OF CHEMISTRY WOOLWICH POLYTECHNIC LONDON, S.E.18. October 29th, 1952 ERRATUM: May (1953) issue, p. 309. The equations on lines 33 and 35 should read:- N+H2V, + Zn’ ’ + ZnNa2V, + 2H’ where V, is [ (COO.CH,),.N .CH,.CH,.N (CH,COO),]””

ISSN:0003-2654    

DOI:10.1039/AN9537800428

出版商:RSC    

年代:1953

数据来源: RSC