ISSN:0003-2654    

DOI:10.1039/AN95277FX001

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

2 MILLER AND CHALMERS: A CRITICAL INVESTIGATION OF A Critical Investigation of the Use of the Silver Reductor in the Micro-Volumetric Determination of Iron, especially in Silicate Rocks BY CHRISTINA C. MILLER AND ROBERT A. CHALMERS (Presented at the meeting of the Society on Wednesday, October 3rd, 1951) Hydrogen peroxide is produced, even in small silver reductors, if air-free hydrochloric acid is not used, and this prevents complete reduction of ferric salts. The use of acid saturated with carbon dioxide reduces the error to negligible proportions. Unstable end-points obtained in the titration of ferrous iron, subsequent to fusions with potassium bisulphate in platinum crucibles, are most probably caused by the presence of platinum1', which reacts very slowly with ceric sulphate.In general, the error in the determination of iron is insignificant if, near the anticipated end-point, titration is not delayed and the first colour change of the indicator is noted. The use of gold crucibles eliminates all uncertainty. Results are shown for the determination of 0.03 to 0.6-mg amounts of iron, alone and in a number of silicate rocks. THE silver reductor is a valuable means of reducing ferric iron, without interference from titaniumw, prior to determination volumetrically with ceric sulphate. The simplicity of the process commends it for micro-analysis, but there are some sources of error. In determining small amounts of iron with the large reductor of Walden, Hammett and Edmonds,l Fryling and Tooley2 found that the production within the reducing column of various amounts of hydrogen peroxide caused large errors that could, however, be rendered insignificant if reductions were effected in an inert atmosphere and an appropriate correction was applied.Edmonds and Birnbaum3 claimed to have almost eliminated the error merely by reducing the size of the reductor, and van Nieuwenburg and Blumenda14 and Colson,6 who reduced the size still further, made no mention of peroxide. If peroxide formation occurs and is ignored, negative errors arise in the determination of iron. Van Nieuwenburg attributed negative errors to partial oxidation of the reduced solutions by air; Colson excluded air in collecting and titrating reduced solutions. Wellss washed his small reductor thoroughly with acid in order to remove accumulated peroxide and determined iron potentiometrically without excluding air.Jan., 19521 THE USE OF THE SILVER REDUCTOR 3 When large amounts of potassium bisulphate had been used to dissolve samples, Fryling and Tooley noted a tendency for the reduced form of the o-phenanthroline ferrous sulphate, used as an indicator, to return after the titration of ferrous iron.The addition of bromine to the solution before passage through the reductor, or the addition afterwards of an excess of cerous sulphate, which would lower the potential of the ceric - cerous system, was said to effect an improvement, but no adequate explanation was given. In the microchemical laboratory we have been using small silver reductors for the determination of iron in the mixed oxide precipitates (Fe,O,+Al,O~+TiO,+Mn,O,+P,O,) obtained from silicate rocks.Fusion of the oxides has been effected with potassium bisulphate and the acid extracts have been reduced, collected without exclusion of air, and titrated with ceric sulphate. Very variable “blank” determinations have been reported and unstable end-points have been general. This investigation was to find the source of these irregularities and to evolve a satis- factory procedure for the determination of 1 to 15 per cent. of ferric oxide in 5-mg quantities of silicate rocks. We have found that, even in small reductors where air is not excluded from the columns, there is a variable peroxide error that may be virtually eliminated if acid saturated with carbon dioxide is used, Platinum dissolved from crucibles during fusions is not deposited on the silver of the reductor, as we had supposed; it is presumably reduced from the quadrivalent to the bivalent form, accompanies ferrous iron and is slowly re-oxidised by ceric sulphate after the oxidation of the iron.The substitution of gold for platinum crucibles eliminates all possibility of error, but if, after the use of platinum, titrations are not made too slowly near the equivalence-point for iron and the first colour change of the indicator is recorded, errors are frequently inappreciable. EXPERIMENTAL THE REDUCTOR- A reductor like Colson’s larger model6 was inserted into a suction apparatus, which was provided with inlet and outlet tubes for carbon dioxide and contained a 10-ml beaker as a receiver.The silver was prepared by the method of Walden et aZ.,l dried and sifted, and particles of appropriate size were added gradually with gentle stirring to N hydrochloric acid contained in the reductor. The glass-wool support for the silver was placed near the stopcock. The peroxide efcct-By means of Savage’s test’ about 0-5 pg of hydrogen peroxide was found in the effluent when 2 ml of N hydrochloric acid were passed through the reductor, When, immediately after, two portions of acid containing a little iron alum were successively passed through and the effluents tested, the first gave a very definite reaction for ferric iron because hydrogen peroxide initially present in the reductor passed into the receiver in advance of ferrous iron, a little of which it then oxidised.The second contained neither ferric iron nor peroxide, because the latter was immediately reduced in the reductor by ferrous iron, and reoxidised ferrous iron was reduced again by the silver. When, however, acid follows iron in the reductor, peroxide is formed and reaches the receiver before it is destroyed. Fryling and Tooley2 corrected for this effect by adding to the titre for the ferrous iron solution the amount of oxidant required to oxidise hydrogen peroxide formed in the acid used for washing the silver after all the iron had entered the column. It is important to note that this addition is legitimate only if the acid is essentially iron-free. A 2-ml portion of the N hydrochloric acid used by us contained about 0-07 pg of iron. Other reagents must also be examined for iron independently of the reductor, since, if they contain no iron, or less than the equivalent of the hydrogen peroxide found in the wash liquid, the effluent obtained from a full blank run (“reagent blank”) will contain hydrogen peroxide and no ferrous iron, the respective amounts of the former being greater than, or less than that associated with the wash liquid alone.If, however, the reagent blank contains ferrous iron and the peroxide correction is constant, the titre for the reagent blank should be deducted directly from that for the solution under examination. Earlier papers indicate that peroxide may be partly adsorbed on the surface of the silver, the effect being more marked in freshly prepared reductors. This was confirmed.When 2-ml portions of N hydrochloric acid were passed through such a reductor they used u p successively 0.085, 0.008 and 0.004 ml of 0.01 N ceric sulphate. When a solution containing 500 pg of iron was passed through immediately after, and the reductor washed with 2 ml of acid, a deficiency of 8 pg of ferrous iron was noted, despite the fact that 0.004 ml was added4 to the titre. In a second run with a solution containing iron no deficiency was indicated. Apparently most of the hydrogen peroxide first formed was removed mechanically by washing tyith acid, but a small residual amount remained adsorbed until removed by chemical action. Reduction and stabilisation of the peroxide efect-Fryling and Tooley reduced peroxide formation to a minimum by using a somewhat complicated apparatus from which air was excluded by means of hydrogen.As a micro-adaptation did not seem practicable, we tried the simple expedient of substituting hydrochloric acid saturated with carbon dioxide for that normally used in the preparation of the reductor and in the washing. It was unnecessary to saturate solutions containing iron, since a trace of peroxide would be reduced by ferrous iron formed in the reductor. Blank runs made with successive 2-ml portions of N hydrochloric acid used 0.001 to 0.0015 ml of 0.01 N ceric sulphate, the variation in the course of a day being 0.0005 ml, which is equivalent to 0.3 pg of iron. When saturation with carbon dioxide was omitted 0.004 to 0-006 ml of ceric sulphate was generally required, but occasionally the amount exceeded 0.01 ml.All the above values have been corrected for the amount of cede sulphate required to oxidise the indicator (“indicator correction”). MILLER AND CHALMERS : A CRITICAL INVESTIGATION OF p o l . 77 RECOMMENDED PROCEDURE FOR THE REDUCTION AND DETERMINATION OF FERRIC IRON- In preparing the reductor and for all rinsing, use essentially iron-free N hydrochloric acid saturated with carbon dioxide. When a reductor is freshly prepared, or has not been used for some time, pass through it 2ml of acid containing a few milligrams of ferrous ammonium sulphate to remove adsorbed hydrogen peroxide, wash the column and carry out the reductor blank. Add to the reductor the ferric solution, made up in N hydrochloric acid, and rinse out the containing vessel, if required, with 0.5-ml portions of acid.Allow the solution to percolate at the rate of 0.05 ml in 8 seconds, but never expose the silver to the air. Finally, rinse the reductor with three 0.5-ml portions of acid. The reduced solution should be collected in an atmosphere of carbon dioxide. To the effluent add 0-02ml of 0.001 M o-phenanthroline ferrous sulphate and titrate with 0.01 N ceric sulphate, while stirring the solution with a stream of carbon dioxide. Determine the indicator correction (about 0.002 ml) and deduct it. Carry out a blank run on 1.8 ml of acid, deduct the indicator correction and add the residuum to the titre for the iron. Leave the cup of the reductor filled with acid and cover it, preferably with a ground-glass cap. Notes on the titrimetric procedure-A stock of approximately 0.01 N ceric sulphate in 0-5 N sulphuric acid was prepared and kept in the dark for some time before it was required, in order to ensure oxidation of traces of reducing impurities.It was standardised on the macro and micro scales by means of sodium oxalates (U.S.A. Bureau of Standards quality) and also, as a further check, by means of a ferrous ammonium sulphate solution that was simultaneously compared with a permanganate solution standardised with sodium oxalate. There was close agreement between all the results. The ceric sulphate solution was after- wards checked at intervals. Great care had to be taken to avoid the presence of adventitious impurity (filter-paper fibre, grease and dust) in the small portion of the stock solution removed for a day’s work.Two micro-burettes, constructed from precision-bore tubing and with long tips, were calibrated by Benedetti-Pichler’s m e t h ~ d , ~ mounted verticaIly and attached to Schilow’s pressure-controlling device.1° The total capacities (1.1 and 0.17 ml) were found by deter- mining the amounts of ceric sulphate delivered. The smaller burette served for the determination of 30 to 80pg of iron. THE SOURCE OF DISAPPEARING END-POINTS AFTER FUSIONS WITH POTASSIUM BISuLpHATE- The direct addition of 300 mg of potassium bisulphate to ferric chloride solutions did not influence the subsequent end-point behaviour, nor did bisulphate that had been fused in quartz. The end-points obtained in blank runs with bisulphate and hydrochloric acid were also stable. When, however, the bisulphate was fused in a platinum crucible, alone or in the presence of iron, and the extracts were reduced and titrated, the first colour change of the indicator was not permanent and further slow addition of about 0-04 ml of ceric sulphate during half an hour was required before a stable end-point was obtained, When iron was present the first end-point corresponded approximately to the amount taken.Potentiometric titrations were also done. Whereas with ferrous chloride alone a sharp increase in potential was recorded as expected, with a solution prepared from iron alum that had been fused inJan., 19521 THE USE OF THE SILVER REDUCTOR 5 platinum with potassium bisulphate before reduction, the potential near the equivalence- point for iron rose with each further small addition of ceric sulphate and then slowly fell.The equilibrium potential very gradually rose until oxidation was complete. When a solution of chloroplatinic acid containing 1OOpg of platinum was tested with potassium iodidell it soon gave a pink colour, whereas a similar solution that had been passed through the reductor gave a colour only after some time. Both solutions after evaporation and full oxidation were found to contain the same amount of platinum. A third solution, after passage through the reductor, required for full oxidation an amount of ceric sulphate corresponding to the conversion of about 40 pg of platinumn into platinumIV. The amount of platinum dissolved as a result of a bisulphate fusion was about 40pg.It is probable that a good deal of this appears in the effluent as platinum1', the rest being platinumIv. The matter was not further investigated. Attempts made to reduce platinum1" to the elementary form, by means of formic acid, for example, before transferring solutions to the reductor, were not a success, so an alternative to platinum crucibles for fusions with potassium bisulphate was sought. A solution of gold chloride in N hydrochloric acid, after passage through the reductor, gave the customary blank value for the reductor and a stable end-point, the goldn1 being presumably reduced to metallic gold and held on the reductor. Gold crucibles were therefore obtained and found to be excellent for the determination of iron in materials that had to be fused with potassium bisulphate in vessels that were unattacked by hydrofluoric acid.They were somewhat more heavily attacked by bisulphate than platinum crucibles but the attack was not severe. PREPARATION OF SILICATE ROCKS FOR THE DETERMINATION OF IRON- With the aid of a stoppered weighing stick weigh out about 5 mg of the dried material and transfer it to a l-ml gold crucible. Add 0.1 ml of hydrofluoric acid (40 per cent.) and then, 3 minutes later, 0.05 ml of N sulphuric acid. Evaporate to dryness on a steam-bath and cautiously expel sulphuric acid over a micro-bunsen burner. Repeat the treatment, then fuse the residue with 0.1 g of potassium bisulphate and cool. If, as often happens, the melt has crept over the top of the crucible, place the latter in a 4-ml porcelain capsule containing 1 ml of N hydrochloric acid, warm until solution of the sulphates is complete, and lift out the crucible and rinse the exterior with 1 ml of N hydrochloric acid.By means of a short-stemmed pipette transfer the contents of basin and crucible to the reductor cup. Wash the crucible four times with 0.5 ml of N hydrochloric acid, transferring each washing to the basin and then to the reductor. Continue the flow of solution through the reductor until all has passed into the column of silver, then wash the latter with three 0-5-ml portions of acid. The final volume for titration is 5.5 ml, but should transference of the crucible to the capsule not be required, it can be reduced to 4 ml. Proceed with the determination of iron as already indicated.If the reagents contain iron, carry out a full blank run on them and deduct the titre found from that for the iron (see p. 3). Except for the initial dissolution, use acid saturated with carbon dioxide. RESULTS IRON Ih' FERRIC CHLORIDE SOLUTIONS- A solution prepared from Hilger's "H.H.P." iron (impurities less than 0-08 per cent.) by dissolving it in hydrochloric acid and oxidising with chlorine, was standardised on the macro scale by reducing portions in a zinc reductor and titrating with ceric sulphate. Weighed aliquots were reduced in the silver reductor and iron determined as on p. 4. The results are shown in Table I. TABLE r DETERMINATION OF IRON IN FERRIC CHLORIDE SOLUTIONS Weight of iron taken, Pg 566.5 559 558 551 556.5 553 Error, Pg + 0-5 -1 + 1.5 + I + 1.5 - 1.5 Weight of iron taken, Pg 274 274 265 250 287-5 247.5 Error, -1 0 + 0.5 0 +1 0 Weight of iron taken, Pg 47.1 35.1 30.1 28.1 39.8 45.8 Error, fLg + 0.1 - 0.2 - 0.3 0 - 0.3 06 MILLER AND CHALMERS : A CRITICAL INVESTIGATION OF In a few experiments on the determination of 500 pg of iron, in which the reduced solutions were neither collected in an atmosphere of carbon dioxide nor stirred with the gas during titration, the average result was 0-2 per cent.lower than those shown above. In$uence of other elements-Of elements that could interfere in the determination of iron only vanadium required to be considered in connection with the silicate rocks that were analysed here. VanadiumIV that is formed in the silver reductor is slowly oxidised in N acid solutions by ceric sulphate.12 Amounts of vanadium ranging from 50 to 250pg, added to ierric solutions as ammonium vanadate, caused an error of only +3 pg in the determination of 300 to 400 pg of iron, and even this could be avoided by making the solutions to be titrated 5 M with respect to sulphuric acid.1 [Vol.77 IRON IN SOLUTIONS AFTER FUSION WITH POTASSIUM BISULPHATE- Weighed portions of the standard iron solution were evaporated with sulphuric acid in l-ml gold crucibles, the residues fused with potassium bisulphate and the iron determined as on p. 5. The results, corrected for iron in the reagents, are shown in Table 111. Those shown in parentheses refer to fusions done in platinurn crucibles. TABLE I1 DETERMINATION OF IRON IN SOLUTIONS AFTER FUSION WITH POTASSIUM BISULPHATE Weight of iron Weight of iron Weight of iron taken, Error, taken, Error, taken, Error, Pg PLg Pg Pg Pg Pg 555.5 0 283.5 + 1-5 33.1 - 0.2 551.5 0 269 - 0.5 33.6 - 0.3 56 1 + 0-5 269 - 0.5 43.4 - 0.3 (542) (+ 1) (263) ( - 0.5) (38-2) ( 1- 2.0) NOTE-values in parentheses refer to fusions done in platinum crucibles.When slow titration near the end-point is avoided and the Jirst colour change of the indicator is noted, fusion in platinum need cause no significant error, except perhaps when a very small amount of iron is under consideration. IRON I N SILICATE ROCKS AND A REFRACTORY- In order to ensure homogeneity in the samples used, powders that passed through a The total iron content of each was also found on the macro 300-mesh sieve were analysed. TABLE I11 DETERMINATION OF IRON IN SILICATE ROCKS AND A REFRACTORY Approximate Fe,O, by Fez03 by mg Yo 70 weight micro macro Silicate taken, method, method, *Flint clay No.97 . . .. . . 4.0 0.93 0.94 6-1 0-93 0-94 *Burnt refractory No. 76 . . .. 4.7 2.23 2-1 8 4.9 2-26 2.23 Phonolite . . .. .. .. 5.6 6-17 6-17 6.0 6.10 6.14 Analcite syenite . . .. .. 5.5 8-76 8-83 6.5 8.88 8-87 Kinkell tholeiite . . .. .. 4.2 15.66 15.75 3-5 15.72 15-74 4.1 15-79 3.5 15-64 (4.3) (15.61) (5.1) (0.93) (5.2) (2.3 1) (5.8) (6-16) (4.0) (8-73) * U.S.A. Bureau of Standards sample. NOTE-Values in parentheses refer to fusions done in platinum crucibles.Jan., 19521 THE USE OF THE SILVER REDUCTOR 7 scale, the Bureau of Standards materials being analysed because only about 50 per cent. of the original samples passed through the fine sieve.As on the small scale the silicates (about 0.5 g) were freed from silica and fused with potassium bisulphate. Platinum crucibles were, however, used, and from the solutions of the melts platinum was removed by means of sulphuretted hydrogen, and ferric iron simultaneously reduced to ferrous iron,13 which was subsequently determined with ceric sulphate. All the samples, except the flint clay, which was dried at 140” C, were dried at 105” to 110” C before use. The results are shown in Table 111. Those given in parentheses refer to fusions done in platinum crucibles. Once more it is indicated that fusions in platinum need cause no significant error if titration near the equivalence-point for iron is not delayed.This might be of importance if high-temperature fusions with sodium carbonate, which could not be done in gold crucibles, were required before the determination of iron. In all the analyses allowance was made for iron in the reagents. We gratefully acknowledge a maintenance grant t o one of us (R. A. C.) from the Depart- ment of Scientific and Industrial Research, and grants from Imperial Chemical Industries Limited and the Trustees of the Ritchie Bequest. We are indebted to Dr. J. B. Simpson of H.M. Geological Survey for two of the rock samples. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES Walden, G. H., Hammett, L. P., and Edmonds, S. M., J . Amer. Chem. SOC., 1934, 56, 350. Fryling, C. F., and Tooley, F. V., Ibid., 1936, 58, 826. Edmonds, S.M., and Birnbaum, N., I n d . Eng. Chem., Anal. Ed., 1940, 12, 60. Van Nieuwenburg, C. J., and Blumendal, H. B., Mikvochem., 1935, 18, 39. Colson, A. F., Analyst, 1945, 70, 255. Wells, I. C., And. Chem., 1951, 23, 511. Savage, D. J., Analyst, 1951, 76, 224. Kirk, P. L., “Quantitative Ultramicro-analysis,” Chapman & ,Hall, London, 1950, p. 131. Benedetti-Pichler, 4., “Microtechnique of Inorganic Analysis, Chapman & Hall, London, 1942, Schilow, E., 2. angew. Chem., 1926, 39, 232, 582. Snell, F. D., and Snell, C. T., “Colorimetric Methods of Analysis,” Vol. I, Chapman & Hall, Furman, N. H., J . Amer. Chem. SOL, 1928, 50, 1675. Hillebrand, W. F., and Lundell, G. E. F., “Applied Inorganic Analysis,” Chapman & Hall, London, p. 258. London, 1936, p. 419. 1929, p. 304. CHEMISTRY DEPARTMENT THE UNIVERSITY, EDINBURGH, 9 DISCUSSION THE PRESIDENT thanked the authors for their paper and said that very often methods that were satisfactory on a macro scale could not be reduced to a micro scale. Clearly, this method had been studied with great care and very good results had been attained. MR. C. H. PRICE asked whether a method for the removal of platinum, if such were available, would not simplify the procedure. He had found that a-furildioxime quantitatively precipitated traces of platinum after sodium carbonate fusions. MR. CHALMERS thanked Mr. Price for his suggested procedure. They had tried several techniques, but without success, and had had recourse to the use of gold crucibles. The results clearly showed that the presence of platinum was without significant effect, provided certain conditions were complied with in the titration. DR. K. A. WILLIAMS expressed his pleasure that Dr. Miller had been able to present her paper. He said that he used gold crucibles containing about 5 per cent. of platinum, as they did not fuse easily as would pure gold crucibles, and asked if the authors could say whether crucibles of that composition yielded an appreciable amount of platinum in this work. A gold crucible lost about 100 pg in weight per fusion with 0-1 g of potassium bisulphate; a platinum crucible lost about 40 pg per fusion. MR. CHALMERS replied that he had no experience of 5 per cent. of platinum in gold. He would expect a small loss of platinum from the 5 per cent. alloy.

ISSN:0003-2654    

DOI:10.1039/AN9527700002

出版商:RSC    

年代:1952

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95277BX003

出版商:RSC    

年代:1952

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95277BP007

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

8 KING: A TECHNIQUE TO IMPROVE THE [Vol. 77 A Technique to Improve the Efficiency of Desiccators BY J. KING (Presented at the meeting of the Society on Wednesday, October 3rd, 1951) The inefficiency of the usual type of desiccator has been noted by several workers. After a study of the usual methods, the author has greatly increased the efficiency of the desiccator by a simple modification, namely, by placing a cylinder of perforated zinc in the upper compartment of a Scheibler desiccator and filling the annular space and the base with desiccant, The most suitable desiccant is calcium carbide in lumps about half an inch in diameter. Figures showing the absorption of moisture by flour previously dried at 110’ C, samples of which were placed in the modified calcium carbide desiccator and in the usual calcium chloride desiccator of Scheibler pattern, show the improvement i n efficiency.IN a recent paper, Belcher and Mottl have again called attention to the inefficiency of the usual type of desiccator. It is surprising that this should have been necessary, in view of the previous publications by Bower2 on the comparative efficiencies of various dehydrating agents and by Booth and McIntire3 on the inefficiency of desiccators under the provocative title of “When is a desiccator?” Belcher and Mott in their paper appear to condemn the use of desiccators altogether and prefer to cool materials in the open air before weighing, but always in a covered dish of their own special design. It would, of course, be possible to evacuate the air rapidly from the desiccator immediately after introducing the apparatus, and then to allow dried air to enter before weighing, but this is not always convenient and introduces complications.Some years ago I was faced with the problem of determining moisture, ash and so on, in circumstances in which the previously heated material might have to remain overnight or for some days between heating and weighing. Under these conditions the usual pattern of Scheibler desiccator was found to be unsuitable, owing to the absorption of moisture from the atmosphere of the desiccator. When a heated vessel containing a dry material is placed in a desiccator, moist air enters, and there is an immediate competition for this between the desiccant and the material to be subsequently weighed.As the desiccant is usually in the base of the desiccator it has a poor chance of competing with the material to be weighed. The amount of moisture that enters a desiccator is considerable. On the assumption that the air of the laboratory is at a temperature of 70” F, with a relative humidity of 70 per cent., each cubic inch will contain 0-2 mg of moisture. Experiments showed that in these conditions 10 to 20mg of moisture could enter a 6-inch desiccator during laboratory manipulation involving the removal of the lid, placing a heated dish in the desiccator and replacing the lid as soon as possible. The Milk Products Sub-committee of the Analytical Methods Committee of the Society4 was aware of this difficulty in the determination of total solids of sweetened condensed milk, and prescribed the use of metal dishes having readily removable but close- fitting covers, in order to prevent exchange of atmosphere in the dish with consequent absorp- tion of moisture from the partly dried atmosphere within the desiccator.These dishes are unsuitable for the determination of ash, for which they would need to be made of platinum or other equally resistant metal. It was obvious from a study of the papers by Bower and by Booth and McIntire that even the most efficient desiccant known, barium oxide, would be of little use in a normal desiccator, and that the most that could be hoped for would be to increase the speed of removal of moisture to the maximum. This was accomplished, in the form of desiccator finally adopted, by placing a cylinder of perforated zinc sheet centrally in the upper com- partment of a Scheibler desiccator and filling the annular space, which should be at least 1 inch, and the base with calcium carbide broken in pieces sifted to about half an inch in diameter (see Fig.1). This allows free circulation of the atmosphere within the desiccator and hence a much more rapid removal of moiSture than can be attained in the usual pattern. The acetylene evolved is less troublesome than the expansion of air caused by the introductionJan., 19521 EFFICIENCY OF DESICCATORS 9 of hot apparatus into the desiccator. The formation of calcium hydroxide necessitates brushing the zinc cylinder at regular intervals to prevent fine particles falling on the apparatus, and the whole of the carbide should be withdrawn, thoroughly sifted and replaced when necessary.The desiccants referred to by Bower,2 as well as others, were considered, but calcium carbide was finally chosen as it has the following advantages: it is cheap and readily available, it has no tension of aqueous vapour either initially or during use and it allows free circulation of the internal atmosphere in the desiccator even after considerable use. In some determinations, the use of an open dish in this desiccator does not lead to any appreciable absorption of moisture; but, in general, apparatus furnished with a closure is necessary, as, for example, in the determination of an ash that is hygroscopic. Typical of this is the deter- mination of the ash of acid casein; in this determination a measured volume of standard 7 c a W e Fig.1. Diagram of modified desiccator calcium acetate is added and an excess of calcium oxide is left after ignition. The small amount of calcium hydroxide associated with the lumps of carbide is helpful in this instance in absorbing any carbon dioxide that may enter the desiccator. A normal type of platinum crucible furnished with a cover was found to be satisfactory for this operation, although the crucible was by no means hermetically sealed. The efficiency of the desiccator compared with one of the Scheibler pattern charged with calcium chloride previously heated to 250" C for 2 hours can be judged from Table I. Wheat TABLE I ABSORPTION IN DESICCATORS OF MOISTURE BY 5 g OF FLOUR PREVIOUSLY DRIED AT 110" c Calcium carbide charged desiccator described in text T_--J------- Gain in weight during each Time of storage interval in Total gain in in desiccator desiccator, weight, mg mg 30 minutes 1.5 1.5 additional 1 hour nil 1.5 39 1 hour nil 1-5 91 20 hours - 1.5 nil Calcium chloride desiccator of Scheibler pattern r Gain in weight during each interval in Total gain in desiccator, weight, mg mg 5 5 7.5 12.5 7.5 20.0 12.5 32.5 A \10 KING: A TECHNIQUE TO IMPROVE THE [Vol.77 flour of 72 per cent. extraction was dried at 110” C for 18 hours in 5-g portions in the metal dishes of the pattern recommended by the Analytical Methods Committee of the Society, the covers being adjusted before removing the dishes from the oven. Cooling before weighing was carried out in the type of desiccator described.After being weighed, each dish was placed in the test desiccator, the lid was removed for the minimum time possible and the cover was taken off the dish while it was in the desiccator. The dish cover was replaced each time before removing the dish from the desiccator for weighing, and the conditions for opening the two types of desiccator were made as nearly identical as was possible, so as to admit approximately the same amount of moisture. I t will be seen that when the dish was allowed to remain in the carbide-charged desiccator for 30 minutes, 1.5 mg of moisture was absorbed, but after 1 hour or two separate hours, no further moisture was absorbed. After 20 hours in the desiccator, this 1.5 mg of moisture was removed by the desiccant.The flour in the calcium chloride desiccator showed progressive absorption of moisture with time. The experiments were repeated after the desiccators had been in normal laboratory use for six weeks. There was little difference from the previous results, which showed that recently heated calcium chloride had little advantage over the used material. Further experiments with 98 per cent. sulphuric acid and pieces of pumice gave results of the same order as those with calcium chloride. These results obtained about fifteen years ago support the contention of Belcher and Mott that the desiccant in a normal laboratory desiccator is of little use. A simple modifica- tion in design, however, may greatly increase the efficiency. The author desires to thank the Government Chemist for permission to publish this account.REFERENCES I. 2. 3. 4. Belcher, R., and Mott, R. A., J . Ap+Z. Chem., 1951, I, 204. Bower, J. H., J . lies. Nat. Bur. Stand., 1934, 12, 241. Booth, H. S., and McIntire, L., Ind. Eng. Chew., Anal. Ed., 1936, 8, 148. “Report of the Milk Products Sub-committee to the Standing Committee on Uniformity of Analytical Methods.’’ Report No. 1, Analyst, 1927, 52, 402. THE GOVERNMENT LABORATORY CLEMENT’S INN PASSAGE STRAND, LONDON, W.C.2 DISCUSSIOK THE PRESIDENT thanked the author for presenting his paper. He said that although small differences in moisture content might not matter, there were occasions when very hygroscopic substances had to be handled, and the simple type of desiccator described was then very effective. MR.C. H. PRICE pointed out the danger of evolution of acetylene and hydrogen sulphide from com- mercial calcium carbide, sometimes with the formation of acetylides. MR. KING replied that the calcium carbide desiccant was not used for dehydrating wet materials, but only for removing moisture from the atmosphere of the desiccator. If the very minute amounts of gas evolved from commercial calcium carbide by the action of the moisture in the atmosphere in the desiccator reacted with the substance being cooled, another desiccant would have to be substituted. Such a desiccant should maintain the advantages of calcium carbide, ois., it should be cheap and obtainable in a suitable graded size, and the reaction with moisture should be irreversible, giving an end product having no tension of aqueous vapour.The use of lithium aluminium hydride had been suggested. DR. K. A. WILLIAMS, in confirming some of Mr. King’s remarks, said that he thought it was not generally appreciated how slowly the atmosphere in an ordinary desiccator became dry. He could best illustrate this by reference to work carried out with desiccators used as humidifying chambers, in which the usual desiccant had been replaced by water. In these it was found that only a layer about an inch in depth immediately above the water surface acquired 100 per cent. relative humidity, even after several days; the upper layers of air in the chamber might not rise in humidity above 75 per cent. His experience was that one should not expect evenness of relative humidity, or total dryness, in a desiccator unless the atmosphere in i t was stirred by mechanical means. MR.KING agreed, and said that he knew of a desiccator that contained a fan for circulating the air; this desiccator had a small bath of sulphuric acid in the form of a ring on the top, which seemed most dangerous. THE PRESIDENT remarked that he had seen in America a circular oven, containing 10 or 12 dishes, which could be rotated to bring each dish to the front of the oven where it could be weighed while still in the oven. This was useful in routine work.Jan., 19521 EFFICIENCY OF DESICCATORS 11 DR. A. M. MAIDEN asked why desiccators should not be made of metal, the dishes being put in by a slide arrangement (as in an automatic chocolate machine), thereby allowing only the minimum amount of laboratory atmosphere to enter.DR. J. H. HAMENCE asked the author whether in his experience, except in special cases, there was any advantage to be gained by allowing a dish or a crucible to stand for half an hour in a desiccator before weighing, as opposed to the method, used in many laboratories, in which the dish was put into the empty desiccator on top of a metal block and then weighed immediately it was cold. He had tried many types of desiccator and had concluded that in many procedures their use was completely unnecessary. MR. KING agreed with Dr. Hamence’s conclusions provided the dish or crucible were fitted with an air-tight cover. DR. WILLIAMS remarked that quick cooling could easily be achieved if one used a desiccator in which a cooling-water coil had been attached to the underside of the zinc platform. Leads to the coil could readily be passed through a side tubulure. An inert atmosphere, or vacuum, could easily be provided by means of appropriate leads through a tubulure in the cover of the desiccator. MR. KING commented that for many years it had been customary to cool metal dishes on a large metal block, which dissipated the heat extremely rapidly. THE PRESIDENT remarked that a similar effect was obtained when cooling dishes used for determining moisture in tobacco. A glass-fronted cubical metal desiccator was used ; this had three water-cooled shelves and layers of desiccant a t the top and bottom. DR. WILLIAMS pointed out that these problems arose from the Society’s Report on the Determination of Total Solids of Condensed Milk. In the prescribed method, the metal dish contained about 25 g of sand, and this took about 45 minutes to cool in an ordinary desiccator; without the use of artificial methods of cooling, this time could not be shortened. THE PRESIDENT concluded the discussion by saying that even the commonest apparatus was worth an investigation from time to time.

ISSN:0003-2654    

DOI:10.1039/AN9527700008

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Jan., 19521 EFFICIENCY OF DESICCATORS 11 Controlled Potential Electrolysis in the Analysis of Copper-Base Alloys BY G. W. C. MILNER" AND R. N. WHITTEMT (Preseded at the meeting of the Society o n Wednesday, October 3rd, 1951) Full details are given of a simple electronic instrument that has been built for automatically controlling the potential of the cathode with respect to a standard reference electrode in electro-gravimetric determination of metals. With this instrument a scheme for determining the majority of the usual alloying constituents of copper-base alloys on a total sample weight of 1.Og has been successfully tested. In addition to the electrolytic equipment, an absorptiometer is necessary for the determination of iron, manganese and nickel. Practical details of this scheme of analysis are included.THE advantages to be gained in the analysis of copper-base alloys by automatically controlling the potential of a platinum cathode at definite values with respect to a saturated calomel electrode were first noted by Diehl and Brouns.1 Their electronic controlling apparatus, however, corrected for cathode voltage changes in one hrection only.2 Lingane3 constructed an instrument that automatically corrected for positive and negative changes in the cathode potential. This instrument proved of great value in the analysis of copper-base alloys; for by maintaining the cathode a t -0-36 volt with respect to the saturated calomel electrode, it removed copper and antimony simultaneously from a hydrochloric acid solution of the alloy and made possible the yolarographic determination of lead and tin in the residual electrolyte.Similarly after electrolysing at -0.70 volt with respect to the saturated calomel electrode to remove copper, lead, tin and antimony simultaneously from solution, the polarographic determination of nickel and zinc was possible. As large amounts of copper seriously interfere in most polarographic determinations, this technique proved very valuable for removing it cleanly from solution without the addition of further chemicals, Lingane, however, did not use the deposited copper for the quantitative determination of this element in alloys. The * Present address : Chemistry Division, Atomic Energy Research Establishment, Harwell, Berks. t Officer of the Department of Supply, Commonwealth of Australia.12 MILLER AND WHlTTEM : CONTROLLED POTENTIAL ELECTROLYSIS [Vol.77 work of the above authors showed that the controlled-potential electrolytic technique could be of some value to a laboratory engaged in the inspection analysis of copper-base alloys, and so we attempted to build suitable cheap and robust electronic equipment. Full details of our instrument, which has worked satisfactorily for several months, follow. THE CONSTRUCTION AND OPERATION OF A SUITABLE INSTRUMENT A survey of the literature revealed several American papers4y5s6 9’ ,8 describing suitable instruments, but generally containing American components obtainable only with great difficulty in this country. Some of these instruments were simple in design but were of limited application, whereas others were versatile but were based on complicated electronic circuits. For the present purpose and for possible future applications we have, therefore, designed a simple yet widely applicable instrument having the following characteristics- (a) Sensitivity to voltage changes of lfiS millivolts over the range +2 to -2 volts, thereby enabling the cathode to be maintained throughout any electrolysis to within (b) Stability of k5 millivolts over a period of at least one hour, assuming that this is the approximate time required for the completion of most electrolytic separations or determinations.(c) An input resistance of 10 megohms to limit the current through the reference electrode so as to minimise polarisation and resistive effects.( d ) A speed of response suitable for correcting voltage changes within a few seconds without excessive hunting or overshooting. (e) Ease of operation. (f) Construction from locally available components, preferably standard radio parts. Full details of the instrument are shown in Fig. 1. It incorporates some of the principles used by Diehl in the construction of his single-direction control instrument. In operation the cathode reference-cell potential is opposed by a precision potentiometer consisting of the network P1, P2, P3 and B2. If, therefore, the reference potential alters in any way during an electrolysis the difference between the two potentials is amplified by the three-stage D.C. amplifier V1, V2, V3, the output of which operates the relays Ryl and Ry2.These relays in turn control the reversible motor driving the variable arm of the potentiometer, P7, supplying the current for the electrolysis, and hence the electrolysis current is automatically adjusted to correct the difference between the two potentials. In this manner, a slight positive or negative difference between the input potential (cathode reference-cell potential) and that of the precision potentiometer actuates the control circuit to increase or decrease the electrolysing current , thereby eliminating this difference. Certain parts of the instrument are considered in greater detail below. AMPLIFIER- The first stage of the amplifier contains a pentode, V1, having a very low grid current to satisfy condition (c) above. To satisfy the stability requirement (b), it proved necessary to supply the heater of this valve from a &volt accumulator.It was found that the smallest available accumulators (three 2-volt , 10-ampere hour types connected in series) could be used continuously throughout a five-day working week and then recharged each weekend. It should be noted that optimum values for the plate load and cathode resistors, R6 and R4, for this valve vary with individual valves and so must be determined empirically. The second stage of the amplifier also contains a pentode, V2, giving a reasonably high’ gain. The output stage, V3, is a conventional “paraphase” circuit giving a push-pull output operating the Post Office type relays. The relays in turn control the reversible motor driving the variable arm of the potentiometer supplying the current for electrolysis, the contacts being connected so as to cause rotation of the motor in either direction, depending on which relay is operated.If neither or both relays are energised, the motor is open circuited; by adjusting the cathode resistor P5 the extent of the “dead space” can be reduced to a minimum. The condenser C1 is incorporated to minimise any relay chatter caused by mains-frequency pick-up and stray transients. It was found necessary to stabilise the plate and screen voltage supplies to satisfy the stability requirement (b). A 1 per cent. change in the mains voltage is equivalent to a change of about 20 millivolts in the input and so the power supply and stabiliser (V4, V5 and VS) was constructed to meet this need.This is essentially a full-wave rectifier followed by a 5 millivolts of its pre-determined value.Jan., 19521 I N THE ANALYSIS OF COPPER-BASE ALLOYS 13 Fig. 1. Circuit of controller KEY Fixed Resistors: R1, 67 kR; R2, 100 kR; R3, 10 MR; R4, 2 kR; R5, 820 kR; R6, 300 kS1; R7, 370kn; R8, 1.0 M a ; R9, 2 kR; R10, 680 k n ; R11, 560 k a ; R12, 370 kR; R13, 1-OMR; R14, 20kR; R15, lOkR; R16, lOkR; R17, lOkR; R18, 10kR; R19, 30kR; R20, 5kR; R21, 390kSZ. NOTE-(^) All resistors are l-watt except R19, which is 5-watt. and V2. (2) The values of R4 and R6 are chosen to suit individual valves V1 Condensers : C1, 0.25 pF, 400 v., paper; C2, 8 pF, 600 v., paper; C3, 8 pF, 600 v., paper. Variable Resistors: P1, 25 kR; P2, 200 kR in 20 equal (fl%) steps; P3, 10 kn; P4, 50 kS1 “Helipot”; P5, 25 kR; P6, 5 kS1 set to give 400 v.across C2; P7, 15 R (“Variac” type 200 CMH). Valves : V1, EF36; V2, 6SH7; V3, ECC32; V4, 6V6GT/G; V5, 6SJ7; V6, 5Z4G. Transformers : T1, “Advance” constant-voltage transformer. T2, Secondary tappings 600-0-6OOv. a t 40milliamp., 2 x 6 . 3 ~ . a t 3amp., 5 v . a t 2 amp. B1, 6-v. accumulator (150 ampere-hour); B2, 3v., two Siemens type T cells; B3, 1-5 v., one Siemens type T cell; B4, 6-v. accumulator (10 ampere-hour); B5, 67.5 v. “Minimax” battery; B6, Weston Standard Cell. Switches : S1 is a 5-bank 4-position Yaxley-type switch with the positions labelled (1) “Off,” (2) ‘ I Zero,” (3) “ Calibrate,” (4) ” Use.” S2, S3, S4, S5 are single-pole single-throw toggle switches. M is a G.E.C. motor type 497 DA70.G is a 9000 to 1 reduction gear consisting of two “Meccano” worm drives and a friction drive. A is an ammeter with the following ranges: 0 to 10, 0 to 1 and 0 to 0.1 amperes. Ryl, Ry2 are 3000-type Post Office relays, 6500-ohm coil; fitted with tungsten change- over contacts. Batteries : Miscellaneous:14 MILLER AND WHITTEM : CONTROLLED POTENTIAL ELECTROLYSIS [Vol. 77 conventional degenerative type regulator. across C2. Further stabilisation is afforded by the constant voltage transformer T1. PRECISION POTEKTIOMETER- This is used for balancing the input voltage; P3 is a twenty-step potential divider with each step corresponding to 0-1 volt and P2 is a continuous control for fine adjustment from 0 to 0.1 volt. Before use the potentiometer is always calibrated against a Weston Standard cell (1.018 volts) by adjusting P1 and using the amplifier instead of the more usual galvano- meter to indicate when balance is obtained.A lamp is connected to each of the relays to show when they are operated and with this arrangement balance is obtained either when both lamps are on or when both are off. Before using the amplifier, however, it is necessary to set it to zero by means of the potentiometer network B3, R2, P4 by again adjusting until both lamps are either on or off. POTENTIOMETER DRIVE- The drive for the main potentiometer supplying the current for electrolysis consists of two “Meccano” worm drives giving a ratio of about 2000 to 1, followed by a friction drive consisting of a 1-inch diameter rubber stopper driving a 5-inch diameter xylenite disc mounted on the knob of the “Variac” transformer. When S3 is closed the driving unit controls voltage changes in both directions, whereas when it is open a unidirectional control only is obtained.For convenience the operating instructions for the instrument are summarised below- Adjust P6 to give 400 volts across C2 and then adjust P5 to give conditions such that the voltage span over which both relays are simultaneously on or off is reduced to a minimum. With S l in No. 2 position, adjust P4 to balance the amplifier (Le., both relays operated or both released). With S1 in No. 3 position and P2 and P3 accurately set to 1.02 volts, adjust P1 to again balance the amplifier (this step is required only occasionally, once or twice a week).With S1 in No. 4 position, set P2 and P3 to the required operating potential. The instrument is now ready for use. Preliminary investigations were directed to the determination of the major alloying constituent, copper. In agreement with Lingane,g the 1-0 to 1.5 M hydrochloric acid solution conditions recommended by Diehl and Brouns were found to give a somewhat patchy copper deposit, whereas this behaviour disappeared on keeping the hydrochloric acid concentration just high enough to prevent the hydrolysis of tin, but always less than 0.5 M . Even so, the copper deposits were generally not very satisfactory on electrolysing with the cathode potential at -0.36 volt with respect to the saturated calomel electrode. Some deposits were not adherent, being easily removed in the washing process after completion of the electrolysis, and often badly discoloured. However, by exercising very great care, reasonably quantitative deposits could be obtained, but the necessary precautions prevented the technique from being widely applicable to the inspection analysis of these alloys.In recent work, OsbornlO encountered sponginess in cadmium deposits in an electrolytic procedure for determining the percentage of this element in cadmium salts. Under the best conditions the deposits were always slightly powdery and easily detachable by rubbing. Osborn noticed, however, that cadmium deposited as a hard clean deposit from dilute sul- phuric or perchloric acid solution in the presence of a small amount of a colloid, such as gelatin. This observation prompted us to add 10mg of gelatin to the weak hydrochloric acid solutions of the alloys in an effort to improve the nature of the copper deposits.This modification proved most successful, giving good adherent deposits that were suitable for quantitative analysis. At a cathode potential of -0.36 volt with respect to the saturated calomel electrode, antimony deposited with the copper, and so for accurate copper figures for alloys containing more than trace amounts of antimony it proved necessary to correct the increase in weight of the cathode for this element, In white metal analysis Lindsey and Sandll and later Torrance12 determined the copper content of the deposit of copper and antimony by carrying out a further electrolytic separation from a fluoride electrolyte.In more recent work Linganel3 achieved the same separation from an acidic tartrate solution with a pH of about 4.5. The relative proportion of antimony In setting up, P6 is adjusted to give 400 volts Once set, these controls rarely need to be adjusted. APPLICATION TO THE ANALYSIS OF COPPER-BASE ALLOYSJan., 19521 I N THE ANALYSIS OF COPPER-BASE ALLOYS 15 and copper (of the order of 1 to 1000) in copper-base alloys is, however, such that it precludes the determination of antimony by subtraction of the re-plated copper content from the com- bined copper and antimony figure. For such deposits it was desirable to determine the anti- mony content of the sample chemically, on a separate sample weight, and then obtain the figure for copper by subtraction from the combined figure for copper plus antimony.After removal of the copper, attempts were next made to determine the combined lead and tin contents directly by continuing the electrolysis with a copper-coated platinum cathode at -0.70 volt with respect to the saturated calomel electrode. The rate of deposition of the lead and tin was very slow, however, and too incomplete for quantitative purposes, but a great improvement resulted from increasing the hydrochloric acid content of the solution. Further difficulty was caused by the deposited metals redissolving during the washing process after the completion of the electrolysis, but this was overcome by using Lindsey's technique14 of making the solution ammoniacal before washing. After determining the combined lead and tin percentage from the increase in weight of the cathode, the deposit was completely stripped off with nitric acid and the resulting solution evaporated to very low bulk to precipi- tate the tin completely as metastannic acid.After filtration, the lead content was determined by deposition as the peroxide on a small platinum-gauze anode from a nearly boiling solution of the filtrate. The increase in weight of the anode multiplied by the empirical factor15 of 0-863 gave the weight and hence the percentage of lead in the alloy. The tin percentage was obtained by difference. The solution remaining after the combined electrolysis of lead and tin was re-acidified with hydrochloric acid, diluted to a volume of 200 ml with water and suitable aliquots were taken for the determination of the other alloying constituents.In all alloys absorptiometric methods were used for the determination of iron, manganese and nickel, under the conditions and by use of the reagents recommended16 in a scheme for the determination of these elements after the removal of copper from solution by deposition on pure aluminium wire. The percentage of zinc in tin bronzes was obtained, after its selective precipitation with 8-hydroxy- quinoline along with magnesium as carrier, by titration with standard potassium ferrocyanide using naphthidine as indicator, whereas the percentage in brasses was obtained by difference after determining the total percentage of all other alloying constituents. Aluminium does not occur in tin bronzes, but the method chosen for its determination in brasses consisted of selectively precipitating with ammonium benzoate before finally estimating volumetrica1ly.l8 By applying these techniques, the determination of all the usual occurring elements in copper- base alloys was made possible on a 1.0-g sample weight and the necessary working details follow.DETERMINATION OF COPPER, LEAD AND TIN IN BRASSES AND BRONZES- Transfer 1.0 g of sample to a covered 250-ml conical beaker and add 7-5 ml of hydrochloric acid, sp.gr. 1-16, and 5ml of water. Warm gently on a hot-plate, adding the minimum amount of nitric acid, sp.gr. 1-42, dropwise to dissolve the sample. Then add 2 g of ammon- ium chloride and approximately 25 ml of water and boil for five minutes. Cool, transfer to a 400-ml squat beaker, add 2 g of hydrazine hydrochloride and about 25 ml of water containing 10mg of gelatin and dilute to about 200ml with water. Immerse the platinum-gauze electrodes (cathode 6 cm high, 6 cm in diameter; anode 3.5 cm high, 3.5 cm in diameter) in the solution, adjust an electrically-driven paddle-shaped stirrer to pass centrally through the electrodes and position a saturated calomel electrode so that the connecting bridge touches the outside surface of the cathode.Adjust the speed of the stirrer to give efficient stirring of the solution. Begin the electrolysis by controlling the cathode at -0.36 volt with respect to the saturated calomel electrode, using unidirectional control and also limiting the current in the first stages to a maximum of 4 amp.After the current has become constant (usually at about 20 milliamp., after electrolysing for 30 to 45 minutes) switch off the controller motor S4, remove the saturated calomel electrode and then lower the beaker, at the same time washing the cathode with a stream of water from a wash bottle. Switch off the stirrer, disconnect the electrolysis battery at S2 and remove the cathode. Rinse the cathode in ethyl alcohol, dry at not more than 105" C for five minutes and measure its increase in weight. If the alloy contains more than trace amounts of antimony, correct this weight for the co- deposited antimony and the final weight then corresponds to the copper content of the sample. Add 10 ml of hydrochloric acid, spgr. 1.16, to the electrolyte remaining after the copper determination and immerse the electrodes as before, but this time use a copper-plated cathode,16 MILLER AND WHITTEM : CONTROLLED POTENTIAL ELECTROLYSIS [Vol.77 Prepare this electrode by plating about 50 mg of copper from a conventional sulphuric - nitric acid solution, washing with water and with alcohol, drying at not more than 105” C and weighing. Add water to the solution until the cathode gauze is completely immersed in the solution and then electrolyse with the cathode maintained at -0.70 volt with respect to the saturated calomel electrode. Continue the electrolysis for 45 minutes, as this time has been found to be sufficient for the deposition of the lead and tin in most types of copper-base alloys with the above electrodes, whereas the final value of the current has proved an unreliable indication of the completion of the electrolytic process.Switch off the controller and neutralise the electrolyte by adding 30 ml of diluted (1 + 1) ammonium hydroxide solution and immediately lower the beaker whilst washing the electrodes with water. Rinse the cathode in alcohol, dry it and determine the increase in weight to give the combined percentage of lead and tin in the sample. Make the remaining electrolyte acid by adding hydrochloric acid, sp.gr. 1-16, transfer to a 1-litre conical beaker and boil down to a volume of less than 200 ml. Cool, accurately dilute to 200 ml in a volumetric flask and reserve for the determina- tion of the remaining alloying elements. StriR the deposit from the cathode with 25 ml of nitric acid, sp.gr.1.20, in a 400-ml squat beaker and finally wash the cathode with water. Evaporate the resulting solution almost to dryness, then cool and add a further 25 ml of nitric acid, sp.gr. 1-20. Digest hot for a time and then filter the metastannic acid on a paper-pulp pad, and wash it about four times with hot water. Collect the filtrate and washings in a 400-ml squat beaker and dilute the resulting solution to about 100 ml with water. Heat the solution to boiling and electrolyse while it is as hot as possible with a small platinum-gauze anode and a current of 4 t o 5 amp. Electrolyse until the deposition of the lead is complete, which generally takes about five minutes. Remove the anode, wash it with water and then dry and weigh it as before. Calculate the percentage of lead from the weight of lead dioxide by using the empirical factor of 0.863 and determine the tin content by subtraction from the combined tin and lead percentage. DETERMINATION OF IRON, MANGANESE, NICKEL, ALUMINIUM AND ZINC IN BRASSES- ThiogZycoZZic acid reagent for iron-Prepare this reagent by diluting 25 ml of “Thiovanic acid”* (76 per cent.) to 100ml with water.Then add 100ml of ammonium hydroxide, sp.gr. 0.880, dilute to 500 ml with water and mix thoroughly. Determination of iron-Filter a quantity of the reserved solution in dry apparatus and transfer 4 ml to a 125-ml conical beaker. Add 1 ml of 50 per cent. citric acid, make the solution just ammoniacal and proceed according to the iron content- (i) For less than 1-20 per cent. of iron, dilute to 50 ml in a graduated flask, add 10 ml of the thioglycollic acid reagent and mix thoroughly. (ii) For amounts of iron between 1.20 and 4.5 per cent., dilute to 100 ml in a graduated flask, add 10 ml of the thioglycollic acid reagent and mix thoroughly. Measure the extinctions of the solutions with a Spekker absorptiometer with Ilford 604 filters, 4 and 2-cm cells and an instrument setting of 1.0.Calculate the percentage of iron by referring the extinction to calibration graphs. Potassium periodate reagent for manganese-Prepare this reagent by dissolving 20 g of potassium periodate in 560 ml of diluted (1 + 3) sulphuric acid. Then add 240 ml of phos- phoric acid, sp.gr. 1.75, and dilute to 1 litre with water. Determination of manganese-With a pipette, transfer 10 ml of filtered solution to a 125-ml conical beaker, add 5 ml of diluted (1 + 1) sulphuric acid and evaporate to fumes of this acid to remove hydrochloric acid.Cool slightly, add a few millilitres of nitric acid, sp.gr. 1-42, to oxidise hydrazine hydrochloride, and fume again. Cool, add 20 ml of water and boil gently to give complete solution. Add 4 ml of potassium periodate reagent, boil gently until the permanganate colour starts to develop and continue boiling for a further three minutes. Then digest hot for a further five minutes to ensure the full development of the colour. Add 10 ml of water to prevent the crystallisation of salts and cool the solution to room temperature. Dilute to 50ml with water in a graduated flask and mix thoroughly. Measure the extinction with the Spekker absorptiometer with Ilford 604 filters, a suitable size of cell and an instrument setting of 1.0, Add 1 or 2 drops of a 2 per cent.solution of sodium nitrite to the solution remaining in the flask to completely reduce the manganese * “ Thiovanic acid ” is high-grade thioglycolic acid supplied by Evans Chemicals Ltd., Boreham Wood, Herts.Jan., 19521 I N THE ANALYSIS OF COPPER-BASE ALLOYS 17 colour, and measure the extinction of this compensating solution in the same cell and with the same conditions as for the previous operation. Obtain the manganese content by referring the extinction difference reading to appropriate calibration graphs. DimethyZgZyoxime reagent for nickel-Prepare this reagent by dissolving 1 g of dimethyl- glyoxime in 1 litre of diluted (1 + 1) ammonium hydroxide, allow the solution to settle, and filter before use.Determination of nick&-Transfer 4 ml of filtered solution to a 125-ml conical beaker, add 1 ml of diluted (1 + 1) sulphuric acid and evaporate to fumes of this acid. Cool slightly, add a few millilitres of nitric acid, sp.gr. 1.42, to oxidise hydrazine hydrochloride and fume again. Then add, from burettes, 5 ml of 50 per cent. ammonium citrate solution, 5 ml of 0.1 N iodine and 20 ml of 0.1 per cent. dimethylglyoxime reagent, mixing thoroughly after each addition. Dilute the resulting solution to 50 ml with water. Measure the extinction of this solution in a suitable size of cell within 30 minutes of the addition of the reagents; use Ilford 604 filters and an instrument setting of 1-0.Prepare a blank solution by taking another 4-ml aliquot and applying the above procedure but adding 20 ml of diluted (1 + 1) ammonium hydroxide instead of the 20 ml of dimethylglyoxime reagent. Measure its extinction under the same conditions and obtain the percentage of nickel by referring the extinction difference reading to prepared calibration graphs. Determination of aZurninium-Take one of the following suitable aliquots of solution, depending upon the approximate aluminium content of the alloy: 100 ml for less than 2 per cent., 50 ml for between 2 and 4 per cent. and 20 ml for greater than 4 per cent. Adjust the volume of the aliquot taken to about 75 ml either by evaporation or by addition of water. Carefully add diluted (1 + 1) ammonium hydroxide until the solution is just alkaline and then add just sufficient diluted (1 + 4) hydrochloric acid dropwise to redissolve any precipi- tate, Add 70ml of a pH 5 buffer (sodium acetate and hydrochloric acid) and 15ml of 5 per cent.hydroxylamine hydrochloride solution and heat the solution to boiling. Boil for one minute, then remove from the hot-plate and add 20 ml of 10 per cent. ammonium benzoate solution to selectively precipitate the aluminium. Digest the solution hot for about 15 min- utes to enable the aluminium benzoate precipitate to settle and the supernatant liquid to become perfectly clear. Filter the precipitate on a paper-pulp pad, redissolve in hot ammon- iacal tartrate solution and complete the determination volumetrically after precipitating the aluminium with 8-hydroxyquinoline.Determination of zinc-Determine the percentage of this element by difference. DETERMINATION OF IRON, NICKEL AND ZINC IN BROWES- Iron and nickeZ-Determine the percentage of these elements by using the same pro- cedures as those outlined for brasses. Zinc-With a pipette, transfer a 100-ml aliquot into a 500-ml conical beaker and add 1 ml of 4 per cent. magnesium sulphate solution and 20 ml of 30 per cent. sodium potassium tartrate solution. Then add 10 per cent. sodium hydroxide solution until the change point of phenol red is reached, followed by 15 ml in excess. Add 20 ml of 2 per cent. 8-hydroxy- quinoline solution and boil to give selective precipitation of the zinc and magnesium. Filter the precipitate on a Whatman No.40 filter-paper, dissolve in hot diluted (1 + 1) hydrochloric acid and separate the zinc from the magnesium oxinate by adjusting the solution to pH 5 and adding a quantity of an acetate buffer of this pH. Filter the zinc oxinate, dissolve it in hot 25 per cent. sulphuric acid, evaporate to fumes of this acid, and oxidise the organic matter with nitric acid, sp.gr. 1-42. Titrate the zinc with a standard potassium ferrocyanide solution by a back-titration procedure with naphthidine as the indicator. Refer to the original paper for full details.17 Store in a tightly stoppered bottle. Cool, add 10 ml of water and re-cool. Refer to the original paper for further details.l* . RESULTS AND DISCUSSION The above procedures were thoroughly tested by applying them to the analysis of 1.0-g portions of standard copper-base alloys.The three British Chemical Standard alloys, Bronzes A and C and Brass B, were chosen for this work and the analytical results were tabulated to enable an easy comparison with the accepted figures for these alloys. An examination of Table I showed that a fairly close agreement existed between the two sets of results, thereby establishing the suitability of the procedures recommended above for the inspection analysis of these types of alloys. By using automatically controlled potential electrolysis, satisfactory procedures have been made possible for the complete accurate analysis of copper-base alloys, especially those18 MILLER AND WHITTEM : CONTROLLED POTENTIAL ELECTROLYSIS [vol.77 available in limited supply. Moreover, they have assisted in the general speeding up of the analysis of these alloys by eliminating certain familiar waiting periods necessary in the chemical methods to allow complete precipitation of some alloy constituents before proceeding with the determination of the other constituents. As the control of the electrolysis is fully automatic, time is available during the first 90 minutes of any analysis for devoting t o other work, because it is only necessary to prepare, weigh and change electrodes during this period. TABLE I COMPARISON OF RESULTS BY THE RECOMMENDED PROCEDURES WITH THE ACCEPTED RESULTS Percentage composition r A > Alloy Cu+Sb Cu Mn Fe Ni A1 PbfSn Pb Sn Zn Sb B.C.S. “A.” By diff. By diff. Analyst A . .. . 85.82 85.58 - 0.06 0.03 - 11.53 1.85 9.68 1.82 - Analyst B . . 85.77 85.53 - 0.07 0.03 - 11.54 1.81 9.73 1.79 - Standard analysis - 85.5 - 0.07 trace - - 1.83 9.70” 1.86 0.24 B.C.S. “C.” Analyst A . . . . 86.90 86.86 - 0.07 0.10 - 10.24 0.38 9.86 2.50 - Analyst B . . . . 86.89 86.85 - 0.065 0.095 - 10.25 0.37 9.88 2.50 - Standard analysis - 86.85 - 0.06 0.09 - - 0.41 9.80 2-53 0.04 B.C.S. “B.” By diff. Analyst A . . . . 58.81 58.76 1.05 0.88 1.00 1.54 2.56 0.75 1.81 34.16 - Analyst B . . . . 58.82 58.77 1.06 0.89 1.01 1.54 2.57 0.77 1.80 34.11 - Standard analysis - 58.8 1.03 0.91 1.01 1.62 - 0.78 1-75 33.9 0.05 * Figure reported by Schoeller, W. R., and Holness, H., Alaalyst, 1946, 71, 217. An exception to this rule arises, however, when a preliminary spectrographic examination has shown the presence of antimony in more than trace amounts.This time is then most advantageously utilised in chemically determining the percentage of this element on a separate portion of the sample. After the completion of the second electrolysis at -0.70 volt with respect to the saturated calomel electrode, it is possible for a competent analyst to carry out the determination of the remaining elements simultaneously. For example, whilst precipitates in the aluminium and zinc determinations are digesting, progress can be made with the determination of iron, manganese and nickel, since relatively simple and rapid techniques are involved in the determination of these elements. The absorptiometric method was chosen for the determination of as many elements as possible after the electrolysis in preference to the polarographic method because nearly every laboratory has an absorptiometer available, whereas polarographs are still generally limited to the larger laboratories.Therefore, after building the unit for controlling the electrolysis, the majority of metallurgical laboratories should be able to put these procedures into operation by using normally available equipment. If a laboratory already possesses polarographic equipment, however, it should be able to use this technique to improve and simplify the determination of zinc in bronzes by incorporating the ammonia - ammonium chloride base- electrolyte recommended by Linganeg for the determination of this element. When all the necessary reagent solutions have been previously prepared, the time needed for a full analysis is from 4 to 5 hours.This is a great saving over the time required for the full analysis of scarce samples by accepted chemical procedures, but only a small saving in time when unlimited amounts of samples are available. In the former case it is impossible to proceed with the determination of the majority of the alloying constituents until lead has been completely removed as lead sulphate by allowing the solution of the alloy to stand for several hours. These circumstances do not arise in the latter case, however, since whilst the lead is precipitating it is possible to continue with the determination of iron, manganese, nickel, aluminium, etc., on other portions of sample by precipitating and filtering away interfering alloy constituents. Although the electronic controlling instrument described in this paper was primarily built to assist in the electrogravimetric determination of certain constituents of non-ferrous alloys, it should also be applicable to controlling mercury-cathode electrolyses for reducing the concentrations of interfering substances in inorganic analytical problems to reasonable pro- portions prior to determinations by the polarographic technique.Moreover, its operationalJan., 19521 I N THE ANALYSIS OF COPPER-BASE ALLOYS 19 characteristics are such that it should be applicable to the types of problems listed by Lingane,lg including the identification of the oxidation states corresponding to polarographic steps, the electrolytic preparation of organic and inorganic compounds, coulometric analysis and automatic potentiometric titrations6 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. REFERENCES Diehl, H., “Electrochemical Analysis with Graded Cathode Potential Control,” G. F. Smith! Caldwell, C. W., Parker, R. C., and Diehl, H., Ind. Eng. Chem., Anal. Ed., 1944, 16, 632. Lingane, J. J., Ibid., 1945, 17, 332. Allen, M. J., Anal. Chem., 1950, 22, 804. Lamphere, R. W., and Rogers, L. B., Ibid., 1950, 22, 463. Lingane, J. J., Ibid., 1949, 21, 497. Lingane, J. J., and Jones, S. L., Ibid., 1950, 22, 1169. Penther, C. J., and Pompeo, D. J., Ibid., 1949, 21, 178. Lingane, J. J., I n d . Eng. Chem., Anal. Ed., 1946, 18, 429. Osborn, G. H., Metalluvgia, 1949, 40, 111. Lindsey, A. J., and Sand, H. J. S., Analyst, 1934, 59, 335. Torrance, S., Ibid., 1937, 62, 719. Lingane, J. J., I n d . Eng. Chem., Anal. Ed., 1945, 17, 640. Lindsey, A. J., Analyst, 1950, 75, 104. Sand, H. J. S., “Electrochemistry and Electrochemical Analysis,” Blackie & Son Ltd., London, Milner, G. W. C., and Groom, H., Metallurgia, 1951, 44, 271. Milner, G. W. C., Anal. Chim. Acta, in the press. Milner, G. W. C., and Townend, J., Analyst, 1951, 76, 424. Lingane, J. J., Anal. Chiwz. Acta, 1948, 2, 584. Chemical Co., Columbus, Ohio, 1948. 1940, Volume 2, p. 74. BRAGG LABORATORY NAVAL ORDNANCE INSPECTION DEPARTMENT JANSON STREET, SHEFFIELD, 9

ISSN:0003-2654    

DOI:10.1039/AN9527700011

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Jan., 19521 I N THE ANALYSIS OF COPPER-BASE ALLOYS 19 Radioactive Tracer = Paper Chromatography Techniques* BY F. P. W. WINTERINGHAM, A. HARRISON AND R. G. BRIDGES (Presented at the meeting of the Physical Methods Group on Tuesday, May 22nd, 1951) The uses and potentialities of combined radiochemical and paper chromatography techniques are discussed. The principle of the methods is to associate one or more radioactive isotopes with one or more substances separated on a paper chromatogram. The labelled substances can then be located and estimated by scanning the paper radiometrically. A simple device for doing this automatically and its use in quantitative work is described. THE use of the techniques of paper chromatography as a micro-analytical tool is well established. In many instances paper chromatography is capable of separating substances in amounts below the limits of chemical detection.At other times the resolving powers of paper chromatography may be undiscovered or unexploited because of the lack of suitable chemical methods of detection. The application of radioactive tracer techniques to paper chromatography may permit the separated components not only to be located and characterised, but also to be estimated quantitatively. In the combined technique the principle is to associate one or more suitable radioactive isotopes with one or more of the components of the mixture, either before or after chromatography. The components separated on the chromatogram are then located and * An abbreviated version of this paper was read to the Isotope Techniques Conference in Oxford on July 17th, 1951.After this paper was presented to the Physical Methods Group, a paper by R. R. Williams and R. E. Smith describing a method for automatically scanning and recording radioactive paper chromatograms appeared in Proc. Soc. Exp. Biol. Med., 1951, 77, 169.20 WINTERINGHAM, HARRISON AND BRIDGES [Vol. 77 estimated by their associated radioactivity. Sometimes the radioactivity can be used for characterising or identifying a particular component when its association can be made to depend upon a specific chemical reaction. Location and estimation are carried out by systematically scanning the paper chromatogram with a Geiger - Miiller counter, for example, or by clamping the whole or part of the paper chromatogram to a photographic plate in the dark.In the photographic method the radioactive zones, if sufficiently intense, will have “auto-radiographed” themselves on the finished plate. The scanning method is capable of great sensitivity and is quantitative, but the photographic method is sometimes more convenient in qualitative work. PREPARATION OF LABELLED PAPER CHROMATOGRAMS The association of a suitable radioactive isotope with one or more bands of the resolved components of the paper chromatogram can be brought about in three ways. LABELLING OF MIXTURE BEFORE PAPER CHROMATOGRAPHY- as in tracer experiments. The mixture for chromatography may already contain one or more labelled components, For example, Benson, Bassham, Calvin, Goodale, Haas and H 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 A B C D E SteDkal have studied RF value Fig.I. Experimental radio-chromatograms Chromatogram of protein hydrolysate from wheat g r o w Strip spotted with known amino acids and exposed to Neutron activated chromatogram of C,H,Cl, isomers. Neutron activated chromatogram of a bromine analogue Neutron activated chromatogram of bromine analogues the path of radioactive carbon in the photo-swthetic incor-oration on T3. 1slI-labelled CH,I. of DDT. of DDT derivatives of {his element by respiring plant cells. Labelled intermediates weredlocated and &timated on paper chromatograms by auto-radiographic and counting techniques. The metabolism of a radioactive bromine analogue of DDT absorbed by susceptible and DDT-resistant houseflies has been ~ t u d i e d ~ , ~ by the combined techniques, in which paper chromatograms were scanned by the methods to be described.Keston, Udenfriend and Levy* have analysed an unlabelled mixture of amino-acids by treating the mixture with a labelled reagent before applying paper chromatography. The paper chromatograms were cut up into small sectionsJan., 19521 RADIOACTIVE TRACER - PAPER CHROMATOGRAPHY TECHNIQUES 21 that were mounted in turn below a Geiger - Miiller tube and the measured rate of counting plotted against distance along the strip. In this way they were able to determine quantita- tively glutamic acid, serine, glycine and alanine in a mixture of amino-acids. Another example in which the mixture already contained labelled components is illustrated in Fig.1A. A paper chromatogram was made from a protein hydrolysate prepared from some wheat grown on a solution containing radioactive sulphur-35. The paper was scanned radio- metrically and the “radiochromatogram” shown was obtained by plotting the 35S-activity against R, value. The two peaks correspond respectively to cystine and methionine biosynthesised in the plant. TREATMENT OF THE PAPER CHROMATOGRAM WITH A LABELLED REAGENT- Little attention seems to have been paid as yet to the possibilities of this method. We have made experiments on the possible location of certain amino-acid groups by methylation with l3lI-labelled methyl iodide. No attempt has been made to apply this particular reaction as a quantitative tool and its description here is merely included to illustrate the principles of the method, which, it is believed, has considerable potentialities.The important condition for the successful application of the method is that a labelled reagent must be chosen that will react selectively with one or more of the compounds separated on the paper chromatogram but not with the paper material itself. A strip of Whatman No. 1 filter-paper was spotted with solutions of different amino- acids, dried and then exposed to l3lI-labe1led methyl iodide vapour at room temperature. Methylation of the amino-acids resulted in the liberation of radioactive iodine in the amino- acid zones and particularly in the methionine zone, probably as a result of the formation of the methyl methionine sulphonium iodide. After pumping away the excess of methyl iodide, the paper was scanned radiometrically (see below) and the plotted radiochromatogram is shown in Fig.1B. The high background was apparently due to some methylation of the paper. NEUTRON ACTIVATION OF THE PAPER CHROMATOGRAM- In this method the finished chromatogram undergoes neutron irradiation in an atomic pile after evaporation of residual solvent, the pile probably being the most convenient method of radioactivation. For the success of this method, the separated components must contain or be associated with an element of suitable activation cross-section so that it can be readily assayed against any background radioactivity of the paper chromatogram itself. Fortunately the carbon, hydrogen and oxygen of cellulose did not possess significant neutron activation cross-sections in this respect, and it was found that the wide range of trace metals almost certainly present in washed Whatman No.1 filter-paper did not give rise to any serious effects under the moderate conditions of some experimental irradiations in the Harwell pile. This method has successfully been applied to the radioactivation of bromine-containing organic compounds by taking advantage of the slBr XY_ts2Br reaction. For example, an inactive bromine analogue of DDT and three derivatives were separated on a reversed-phase paper chr~matogram.~ The chromatogram was then irradiated for 3 days in the Hanvell pile at a flux of lofo neutrons per sq. cm per second. The chromatogram was scanned a few days later and the resultant radiochromatogram is shown in Fig.1E. Only the DDT analogue was present on the scanned chromatogram reproduced in Fig. 1D. The small unidentified additional peaks may have been due to activated chromatographed impurities or to con- tamination. In the method of neutron activation one important point must be borne in mind. The energy of recoil from the gamma emission in the neutron-gamma reaction is usually more than sufficient to rupture any chemical bond between the target atom and the rest of the molecule. The induced radioactivity can therefore be associated only with the component and cannot be used as a tracer in further tests with the eluted substance. Another complication may be due to the volatile nature of the recoil-freed isotope resulting in loss of activity and contamination of other parts of the chromatogram.For example, in the neutron activation of organic bromine compounds, the recoil-freed bromine is probably present in part as hydrogen bromide. It was found, however, that if the paper strip were rolled up with a second plain strip of paper, contamination was negligible. There seems to be no reason why chlorine compounds could not be estimated on a paper chromatogram by taking advantage of the 36Cl%36S reaction, and many applications could be suggested. One22 WINTERINGHAM, HARRISON AND BRIDGES : [Vol. 77 possible application is to the analysis of a mixture of the isomers of hexachlorocyclohexane, the gamma isomer of which is a well-known insecticide. A mixture containing 10 pg of the a, /3,y and 6 isomers was chromatographed under the conditions used for the DDT analogues.3 The chromatogram was then irradiated at a flux of loll neutrons per sq.crn per second for one week and scanned a fortnight later. The radiochromatogram is shown in Fig. 1C; the peaks are due to sulphur-35 arising from the chlorine of the partly separated isomers by the 35C1*+35S reaction. This particular application is being further investigated. RADIOMETRIC SCANNING When the paper chromatogram has been prepared so that the resolved components are labelled or associated with a suitable radioactive tracer, the distribution of the tracer must be determined quantitatively. Fink, Dent and Fink5 have located 1311-labelled compounds by placing the chromatogram against an X-ray plate in the dark so that the radioactive zones could be located auto-radiographically after fixing and developing the exposed plate.This method does not easily lend itself to quantitative work and there is reason to believe that it is far less sensitive than counting techniques. It has been estimated6 that, in order to obtain a reasonable contact auto-radiograph of a 82Br-labelled compound, a radioactive disintegration density of the order’ lo9 disintegrations per sq. cm (of the paper chromatogram) would be required. If this figure were obtained, say, in a 24-hour exposure, it would correspond initially to about 0.5 microcuries of bromine-82. One-hundredth of this activity, or 5 x 10-3 microcuries, could be determined to within -+_5 per cent. by Geiger counting for 1 minute with typical apparatus. A simple method of scanning a paper chromatogram is to cut it up into small equal sections, which are mounted in turn below an end-window Geiger tube. This method is very tedious and, unless each section is cut exactly and precisely mounted, errors will resixlt.AUTOMATIC SCANNING DEVICE- In Figs. 2 and 3, respectively, are shown the front elevation and plan of a device for automatically scanning 19-inch unidimensional paper chromatograms. Whatman No. 1 filter-paper is conveniently ’available in rolls of this width. Two-dimensional chromatograms are scanned by cutting the sheet into the equivalent number of suitable strips. The strip, 0, is wound around the 5-05-cm drum, H, and passes over the 5.05-cm idler roller, M, under tension from the 100-g weight, Q, which is fastened to the free end of the strip by means of a “bull-dog” clip, P.The other end of the strip is held to the drum by the spring steel wire, Y, passing through the drum slot, K. The wire is anchored to the drum at J and screwed on the opposite side at 2. The drum, H, drives an eight-bladed escapement wheel, D, by a 4 to 1 gear, F - R, so that one-eighth of one revolution of I> allows the strip, 0, to advance, under tension, almost exactly 0-5 cm. The diameter of the drum, H, is such that the changing thickness of paper on the unwinding drum has a negligible effect on the distance the paper moves with each escapement movement, and the mean value of this distance is 0.5 cm for the average chromatogram. The roller, M, is mounted exactly below a 1Q-inch by 0-5-cm slot, L, cut in the platform, N, so that a transverse 0-5-cm section of the paper strip is exposed through the slot.The effect is that, for every escapement movement (see below), consecutive 0-5-cm sections of the paper strip are exposed to the thin end-window, WI, of the Geiger-Muller tube, GM (Fig. 4), which is housed inside the lead castle, LC. The milled nuts, E, allow the platform to be removed for cleaning and inspection. The heavy frame, B, is screwed to the edge of the laboratory bench, A, through which a suitable hole is cut for the descending strip, 0. The drum and roller shafts run in phosphor-bronze bearings mounted in the bearing plate, C. So far, scanning has been limited to beta counting, the Geiger - Muller tube being relatively insensitive to the gamma-rays from bromine-82, iodine-131 and so on.The platform, N, must be sufficiently thick to eliminate all the beta particles from neighbouring but unexposed sections of the strip, an important factor in the resolving powers of the scanner. Steel plate one-eighth of an inch thick has been found to be satisfactory. The platform and frame are constructed in stainless steel, the drums in aluminium. The left prong, V, of the escapement rocker, U, normally engages a blade of the escapement wheel, D, under the tension of a small spring, T, attached to the rocker and frame. The armature of the solenoid, X, momentarily.pulls back the rocker via the connecting rod, W, against the spring, the right Each drum has a 3-mm flange on both sides.Jan., 19521 RADIOACTIVE TRACER - PAPER CHROMATOGRAPHY TECHNIQUES 23 prong engaging the opposite-but-one blade of the escapement wheel, so that when the rocker returns to the normal position, D rotates one-eighth of a complete revolution, but no more.The friction clutch, S, can be disengaged by relaxing the milled nut, I, so that the drum, H, is free of R and hence of the escapement mechanism. New strips can then be wound on by means of the handle, G. The lead castle (Fig. 4) is in two parts to facilitate handling. In assembling, the Geiger - Muller tube, GM, is placed exactly over L. The lower half of LC is then lowered into position so that the prongs, PR, engage the flange of GM. The top half of LC is then placed in position so that the contact disc, CD, engages the anode, AN EHT and earth connections are then made at AL and ET.The insulating glass tube, GL, is flared at the lower end to guide AN on to CD. The solenoid, X, is the electro-mechanical 1- IOcm.-l U Fig. 2. Timer-controlled device for scanning paper chromatograms radiometrically-front elevation register unit taken from a timing unit type 1003, which is available in many laboratories equipped for beta counting. Its operation requires 50-volt D.C. pulses, which can be drawn from the timing unit a t &-second, 1-second, Q-minute or 1-minute intervals. By inserting additional scale-of-two units, as shown in Fig. 5, the solenoid can be made to operate at 2-minute, 4-minute or %minute intervals, and so on. The probe unit amplifies pulses from the Geiger - Muller tube and feeds them to a rate-meter type 1138A. The 0 to 100-millivolt potential difference across the output potentiometer is proportional to the mean rate of counting; alternatively the rate-meter output can be obtained as a 0 to 5-milliamp.current in the same circuit. This output is fed to a pen-recording milliammeter or to a self-balancing potentiometer. In this way the rate of counting is automatically plotted against distance along the strip, which gives the required radiochromatogram. In our experience the milli- ammeter recorder was more suitable for this type of work, because the tendency of a self- balancing potentiometer to over-balance gave rise to fluctuations greater than the intrinsic statistical fluctuations of particle counting. The statistical fluctuations of the rate-meter output for a given mean rate of counting can be modified by the variable capacitance of the integrating circuit of the rate-meter.It is important that the corresponding integrating time is less than the time intervals of the scanner, otherwise the rate-meter will be unable to keep pace. On the other hand the integrating time must be sufficiently large for the statistical fluctuations in the counting rate to be small compared with the radioactivity peaks of the chromatogram. The essential features of this apparatus are : (a) the paper chromatogram is automatically scanned in a geometrically uniform manner, (b) the movement of the paper is controlled by a simple escapement mechanism operated by conventional timing equipment, (c) provision is made for plotting the radiochromatogram automatically and (a) all parts of the scanning unit are easily accessible for cleaning purposes, decontamination and so on.24 WINTERINGHAM, HARRISON AND~BRIDGES : [Vol.77 QUANTITATIVE INTERPRETATION OF RADIOCHROMATOGRAMS Ideally the counting geometry will be the same for each section. A separated component will normally be distributed over several sections of the paper chromatogram. The weight of component is invariably small compared with the weight of paper, so that self- absorption of the beta particles assayed will depend only upon the density, or weight per unit area, of the paper strip, which for practical purposes can be taken to be constant. On this basis the total weight, w, of the labelled component is proportional to the sum total of the net rates of counting (corrected for background, radioactive decay, dead-time losses and so on) of the relevant sections, i.e., where a is the corrected net rate of counting of each relevant section.The constant K depends on the specific radioactivity of the original component and on the over-all counting efficiency. In practice it has been found convenient to scan a chromatogram made with a w = K C a .. .. .. . . .. ' * (1) A Fig. 3. Timer-controlled device for scanning paper chromatograms radiometrically-plan known weight of compound of known specific activity. The position of this compound also Serves to identify the peaks of the original chromatogram. When the radiochromatogram is plotted as net rate of counting against distance along the strip, w is proportional to the area enclosed by the relevant part of the curve (see below).With the assembly described above, it has been found convenient to measure these areas as plotted on the recorder by means of a planimeter. In comparing different radio- chromatograms quantitatively it is important that the radiochromatogram is plotted on this basis, as the total radioactivity is the product of the mean rate of counting (over the relevant sections) and the distance in an absolute sense over which the radioactivity is spread. For the purposes of identification, however, the RF value is the important property, and the radiochromatogram should be then plotted as net rate of counting against RF value. Other factors of significance in quantitative work are discussed below.Jan., 19521 RADIOACTIVE TRACER - PAPER CHROMATOGRAPHY TECHNIQUES 26 Decay corrections-Usually the time taken to scan a unidimensional paper chromatogram is short compared with the half-life of the isotope being assayed, so that corrections for decay can be applied to the measured areas as a whole and based on the mean time of scanning.When the time of scanning is not relatively short, the time scale corresponding to each part of the strip must be recorded and different corrections applied according to position on the strip. Dead-time corrections-Correction for counter dead-time or quench-time losses for every section of a radiochromatogram would be tedious. For this reason it is recommended that quench and rate-meter-paralysis times be kept to a minimum.For example, if a Geiger t! M Fig. 4. Timer-controlled device for scanning paper chromatograms radio- metrically-cross section through Geiger-Muller tube and lead castle and worm's eye view of lead castle counter is effectively quenched for 100 microseconds after each pulse, this correction for a maximum rate as high as 100 counts per second will be only 1 per cent. As the rate of count necessarily varies from zero to zero through the maximum for a resolved component, the dead-time error in its estimation will be less than 1 per cent. Geometry efects-Ideally the geometrical distribution of radioactivity would be the same for every exposed section of the paper chromatogram. In practice variations occur, especially in strips cut from a two-dimensional sheet, when the radioactive zone would not be expected to be in the middle region of the section, except by chance.The beta-particle sensitivity of an end-window type of Geiger tube varies according to the position of the source below the window, becoming less as the source becomes more off-centred. For this reason Boursnel17 has used a large-area window tube for scanning. In unidimensional chromato- grams, however, the radioactive zones appear to be sufficiently similar to permit assays based on adequate controls with known amounts of radioactive material. It is possible that back-scattering from the sides of the platform slot partly compensates for any geometrical effects due to zones being off-centre. For two-dimensional chromatograms a tube of larger window area than the one depicted in Fig.4 might be desirable. However, a simpler alternative is to cut the two-dimensional sheet into strips sufficiently narrow to eliminate significant geometrical effects. Strips 1 cm wide have been found to satisfy this condition. The narrow strips are then wound round the middle of the winding drum, and the unwinding strip keeps to the middle of the idler drum without relying on the flanges for guidance.26 WINTERINGHAM, HARRISON AND BRIDGES : [Vol. 77 Self-absorption-Self-absorption of a constant source will vary according to the variations i n paper density. Corrections for this factor in quantitative comparisons along a uni- dimensional strip are tedious, but the following calculations indicate that the effect is small.For the purpose of the calculation it was assumed that a resolved component would be spread over about 5sq. cm of paper. The coefficients of variation in mean density of sections G-M scanner . & pulse Additimd scales of two input Fig. 6 . Timer-controlled assembly for radiometric scanning of paper chromatograms with automatic recording-unit to unit connections measuring 5 sq. cm cut from a sheet of Whatman No. 1 filter-paper chosen at random was 3-3.9 per cent. The mean density was 8.785 mg per sq. cm. The effect of a variation of 3-10 per cent. on the self-absorption and on the corresponding corrections in the assay of some typical isotopes was calculated by means of Libby’s self-absorption equation.* The results are summarised in Table I.TABLE I EFFECT OF PAPER DENSITY VARIATION ON SELF-ABSORPTION Proportion of beta particles unabsorbed by paper of mean density Variation of self-absorption correction factor corresponding to a variation of +_ 10% in Isotope 8.785 mg/sq. cm paper density 14c 0.366 & 7.4% 3% 0.431 & 6.5% a2Br 0.787 +_ 2.4% 36c1 0.843 1.7% S 2 P 0.957 &0-5y0 l S l I 0.828 +_ 1.9% It will be seen that, even for soft-beta emitters, this variation is unlikely seriously to impair quantitative work. Resolution-When two or more labelled components are incompletely resolved, an arbitrary decision must be made on the way to divide the area of the plotted radiochromato- gram. Alternatively, an ingenious technique developed by Keston, Udenfriend and Levy4 may be used. In this method one of the unresolved components is labelled by means of a second isotope of sufficiently different radiation characteristics to enable its boundary on the chromatogram to be determined by selective scanning.Jan., 19521 RADIOACTIVE TRACER - PAPER CHROMATOGRAPHY TECHNIQUES 27 Corrections for recorder-drum speed, for rate of scanning and for background-The radio- chromatogram is plotted by the recorder as counting rate (ordinate) against distance along the strip (abscissa).In equation (1) the activity or rate of counting, a, will be constant, if statistical fluctuations and decay are neglected, for each exposed section ; Xu is therefore proportional to the area A enclosed by each radioactivity peak of the radiochromatogram, so that w = K’A, where K‘ depends on the over-all counting efficiency (Le., observed rate of counting divided by absolute disintegration rate) and upon the units in which A ismeasured.In practice it is convenient to adjust the recorder-drum speed (provision is made for this in the majority of recorders) so that radiochromatograms plotted from RF = 0 to RF = 1 at different scanning rates will be represented by equal or comparable distances along the recorder paper. The scanning rate should be varied by altering the timer-controlled intervals between pulses fed to the scanner solenoid, according to the level of activity on the paper strip. For example, low activities will require a lower scanning rate than will high activities for the same statistical accuracy (see below). Corrections for different recorder-drum speeds and scanning rates in quantitative work are then made in the following manner.Suppose, in the calibration of the apparatus for quantitative work, a known weight, w, of a labelled substance is scanned. Let s be the rate of scanning, d the drum speed and Y the rate of counting corresponding to unit scale (ordinate) of the recorder. Let A be the area enclosed by the radioactivity peak corresponding to the labelled substance on the radio- chromatogram. This area on the recorder paper can be measured in any convenient units (to suit a particular planimeter, for example). The value of A must be corrected for decay in the usual manner, and also for background, which is simply the observed counting rate over the part of the strip that is free from labelled material.Then K’ can be derived from the expression K’ = w/A. A second strip containing an unknown weight w‘ of the same labelled substance is now scanned at a rate s‘ and recorder drum speed d’. Let A’ be the corresponding net area on the radiochromatogram and Y’ the rate of counting equivalent to unit scale of the recorder, It follows that- ,a s’rt d’s r - * .. .. .. .. w’ = K’A - In the apparatus described above, r’/r will be lo** where n is 0, 1, 2, 3 and so on, corresponding to the sensitivity ranges of the ratemeter. The fraction s’/s may be more conveniently replaced by t/t’, where t is the timer interval or time of exposure of each section and will therefore be a simple multiple of 1 second or of 1 minute. Precision in quantitative work-In radiometric assays it is rare to attempt to determine a labelled substance by calculating the absolute disintegration rate from the observed counting rate.This involves accurate data on the effects of geometry, counter efficiency and self- absorption and on the specific radioactivity of the substance itself. Assays are usually based on the rate of counting observed with a known weight of the substance under standardised conditions. In the methods described in this paper, the assays are similarly based on radiochromatograms made with known weights of labelled material. In these circumstances the mean recoveries are complete, and precision or reproducibility becomes the important point. The principal factors affecting precision, apart from manipulative errors such as those involved in transfering material by micro-pipette to the paper strip, are self-absorption, statistical errorsg inherent in all random particle counting, and geometry.The effects of self-absorption have been discussed. The statistical coefficient of variation of a net count obtained as the difference between the total count NT and the background count Ng over the same interval of time is k l 0 0 d N ~ + Ng/(N= - NB). It can be shown that the coefficient of variation of an assay by scanning is + 1 0 0 d d ( A ~ + AB)/(AT AB), where AT and Ag are the total and background areas in cm-counts per second and d is the recorder-drum speed in cm per second. The particular contribution of geometrical factors to variation, such as an asymmetrical distribution of active material over an exposed section, has not been investigated.Data are available, however, on the total variation due to all errors, including those due to manipulation, incomplete chromatographic resolution and so on, as a result of some resolution experiments with *2Br-labelled derivatives. Single labelled substances or simple mixtures were resolved on reversed-phase paper chromatograms3 and scanned. In one experiment the separated components were activated after chromato- graphy (see above). All recoveries were estimated in terms of a reference chromatogram28 WINTERINGHAM, HARRISON AND BRIDGES [Vol. 77 made with a known weight of pure substance and scanned. The results are shown in Table 11. In all experiments fairly active samples were used and weights of the order of one-hundredth or one-thousandth of those given could have been assayed with comparable accuracy.They are not intended to demonstrate the sensitivity of the methods. TABLE I1 RADIOMETRIC ASSAY OF *%-LABELLED DERIVATIVES SEPARATED ON PAPER CHROMATOGRAMS Weight applied w3 99 .. .. 1.0 99 .. .. 1.0 99 .. .. 2.0 99 .. .. 5.0 97 .. .. 5.0 + (p-BrC,H,),C: CCl, . . .. 0-5 ($-BrC,H,),CH.CCl, . . .. 50 + (p-BrC,H,),C : cc1, . . .. 50 + (p-BrC,H,),CH. COOH .. 50 Compound or mixture to chromatogram, (#-BrC,H,)&H.CCl, . . .. 0.2 (P-BrC,H,),CH.CCls . . .. 1.0 (p-BrC,H,),C : cc1, . . .. 20 +(p-BrC,H,),CO . . .. .. 20 + (p-BrC,H,),CH.COOH .. 20 Recovery by scanning, Remarks % 109 Labelled compound applied singly 104 99 99 99 99 111 99 99 99 99 105 99 99 39 99 85 99 99 99 99 92 99 99 99 9s 111 97 }Applied as a labelled mixture Applied as a labelled mixture Applied as a mixture. Activated after chromatography 107 118 100 126 62 90 Average recovery 101.2 The authors are indebted to Mr. W. K. Cordaroy for the construction of the escapement mechanism and scanning frame in the instrument shop of the Laboratory, and to the E.R.D. Engineering Company for constructing the specially designed lead castle shown in Fig. 4. This paper is published by permission of the Department of Scientific and Industrial Research. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Benson, A. A., Bassham, J. A., Calvin, M., Goodale, T. C., Haas, V. A., and Stepka, W., J . Amer. Winteringham, F. P. W., Loveday, P. M., and Harrison, A., Nature. 1951, 167, 106. Winteringham, F. P. W., Harrison, A., and Bridges, R. G., Ibid., 1960, 166, 999. Keston, A. S., Udenfriend, S., and Levy, M., J . Amer. Chem. SOC., 1947, 69, 3151. Fink, R. M., Dent, C. E., and Fink, K., Nature, 1947, 160, 801. Winteringham, F. P. W., J . Chem. SOC., 1949, S416. Boursnell, J. C., Nature, 1950, 165, 399. Libby, W. F., Anal. Chem., 1947, 19, 2 . Rainwater, L. J., and Wu, C. S., Nucleonics, 1947, 1, 60. Chem. SOC., 1950, 72, 1710. PEST INFESTATION LABORATORY LONDON ROAD SLOUGH, BUCKS.

ISSN:0003-2654    

DOI:10.1039/AN9527700019

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Jan., 19521 SMITH AND ALLISON 29 Paper Chromatography of Radioactive Penicillin BY E. LESTER SMITH AND D. ALLISON (Presented at the meeting of the Physical Methods Group on Tuesday, May 22nd, 1951) Biosynthetic 36S-penicillin of very high specific activity has been chromatographed on buffered paper strips. The distribution of the penicillins along the strips was quantitatively assessed by the Goodall and Levi bio- autographic method on agar plates inoculated with B. subtilis and by two radiometric methods ; sections containing individual penicillins were cut from the paper chromatograms, radio-autographs being used as guides, and either they were “counted” directly under a thin end-window Geiger - Muller tube or aqueous extracts were evaporated on planchettes for counting. The good agreement between the three methods confirms the validity of the Goodall and Levi method.AT least five varieties of penicillin, having the structures shown in Fig. 1, are produced by current fermentation techniques and may be present in commercial penicillin. S CH, / \ . R. CO-NH-CH-CH I I 1 /‘\CH, CO-N-CH*COOH Penicillin R -CH2*C,H, -CH2*C6H40H -CH,*CH : CH.CH,.CH, -CH2*(CHJ3*CH3 Heptyl (K) -CH2*(CH,),*CH3 Benzyl (GI A 2-Pentenyl (=I p-Hydroxybenzyl (X) Amy1 Fig. 1. Structure of penicillins Many methods are available-biological, chemical and physical-for the assay of total penici1lins.l However, it is now usually required to determine benzylpenicillin (penicillin G) specifically, and this is much more difficu1t.l The need has arisen partly because it is common manufacturing practice to add to the fermentation medium precursors containing the benzyl group and these tend to give a product consisting largely of ben~ylpenicillin.~~~ Moreover, the clinical efficacy of benzylpenicillin has been investigated more thoroughly than that of other penicillins, except heptylpenicillin (penicillin K), which is known to be less effective.%-Ethyl piperidine was originally claimed to precipitate exclusively benzylpenicillin as a crystalline salt: but it has been shown that the method is neither fully quantitative nor specific; that it seems to give results of the right order of magnitude is probably due to compensating errors5 8 Similarly it was claimed that only benzylpenicillin would crystallise in the form of the isopropyl ether complex of the free acid.’ The yield is too low for this to be adaptable as an analytical method; in any event, as will be shown later, other penicillins co-cryst allise with this isopropyl ether complex.CHROMATOGRAPHIC SEPARATION- Chromatography is the only means known for separating the penicillins completely from one another, and the only chromatographic methods suitable for routine use involve separation on buffered filter-paper strips, followed by bio-autographic delineation of the zones due to the various penicillins. This technique was originally devised by Goodall and Levis ; a number of modified methods have subsequently been de~cribed.~ 9 1 0 9 1 1 Provided careful attention is given to various points of technique, in particular to control of humidity, this micro-chromatographic method gives good separation of the penicillins.However, the zones of inhibition arising from the developed strips are usually somewhat elliptical, which makes quantitative assessment difficult. Goodall and Levi30 SMITH AND ALLISON : PAPER CHROMATOGRAPHY [Vol. 77 applied an arbitrary factor of 1.2 in calculating the results, while others have preferred to induce a similar degree of ellipticity in the standards by developing strips spotted with pure standard penicillins alongside the unknowns. Another approach-too laborious for routine use-has been to use a narrow strip cut into numerous small squares after development; the penicillin in each square was assessed separately by applying it to a seeded agar plate.12 ISOTOPIC TRACER TECHNIQUES- Radioactive penicillin of very high specific activity (up to 600 millicuries per gram) has been prepared for other purposes by fermentation of a synthetic medium containing sulphur-35 as sulphate.Concentrations up to 1 curie per litre have caused no obvious damage to the mould. It was desired to know the composition of this materia113v14 in terms TABLE I PAPER CHROMATOGRAPHIC ANALYSIS OF RADIOACTIVE PENICILLIN Crystalline sodium salt Total penicillins r 3 A By biological method By tracer method P I Average of Average of Eluates fro: L Penicillin species 3 strips, 3 strips, 1st strip, 2nd strip, 1st strip, %w/w %w/w % w/w % w/w %w/w - - - p-Hydroxybenzyl (X) .. 0.1 0.2 d2-Pentenyl (F) n-Amy1 (A) n-Heptyl (K) Benzyl (G) . . 84.7 83.8 87.5 88.2 89-1 ..2.8 3.5 2.5 2.4 2.8 .. 5-0 4.9 4.7 4-6 3.6t .. 7.4 7.6 5.3 4-8 4.5 (including XI) Non-penicillin *S (from paper chromatograms) . . . . - - 13 24 14 Non-penicillin *S (by direct - 5 t Taking to dryness inadvertently during extraction may have caused loss. ether extraction) . . . . - - of individual penicillins, and the opportunity was taken to check the bio-autographic method by tracer techniques. A preliminary experiment has already been described.l5 Accordingly, the radioactive penicillin was spotted out on buffered strips, which were developed with ether in the special apparatus described by Goodall and Levis The developed strips were left in contact with X-ray film in a cassette for several days and the film was then developed. With these radio-autographs as guides, the corresponding paper strips were cut into squares or, where necessary, rectangles, in such a way as to avoid having parts of different penicillin zones on the same square.The squares were then fitted into planchettes and measured radiometrically in a thin end-window Geiger - Miiller counter. From the total counts for each penicillin species, the proportions by weight (more strictly, the molar proportions) could readily be calculated. Alternatively, again using the radio-autographs as a guide, a strip was cut into sections each containing the whole of one penicillin species. Each section was then extracted by boiling for a few minutes with very dilute phosphate buffer. An aliquot of each extract was evaporated down on a planchette and the radioactivity measured.Corresponding strips developed with ether at the same time were plated out on agar. The results were calculated, against developed benzylpenicillin standards, in terms of pro- portions by weight of the penicillin species. The results of the biological and the two tracer techniques are recorded in Tables I and 11. In view of the errors inherent in methods that involve micro-chromatography of labile substances, the agreement between the various techniques is satisfactory. Clearly the tracer methods are not available for routine use with non-radioactive penicillin ; their value lies in providing independent evidence for the validity of the biological method, which in some quarters has been accepted only reluctantly. The radio- autographs and the radiometric assays revealed that an appreciable part of the radioactivity The tracer experiments also brought to light an interesting phenomenon.Jan., 19521 OF RADIOACTIVE PENICILLIN 31 remained in the position of the original spot, where there was no corresponding zone of biological activity.At first we supposed that this spot was due to sulphur-containing impurities or to penicillin decomposition products in the penicillin preparations. These explanations appeared doubtful when it was found that the phenomenon persisted with TABLE I1 PAPER CHROMATOGRAPHIC ANALYSIS OF RADIOACTIVE PENICILLIN Impure sodium salt Total penicillins P - Average of Average of Eluates from Penicillin species 3 strips, 3 strips, 1st strip, 2nd strip, % w/w % w/w % w/w % w/w r A -l By biological method By tracer method - p-Hydroxybenzyl (X) ..0.1 0-1 Benzyl (G) d2-Pentenyl (F) n-Amy1 (4 wHeptyl (K) .. 85.8 87-5 88.5 84.2 .. 3.2 3-0 2.5 [5*1?] .. 3.9 3.6 4.2 4.9 .. 7.0 5.8 4-8 5.8 Non-penicillin *S (from paper chromatograms) . . .. - 32 32-5 - purified crystalline penicillin and that the amount of radioactivity in the initial spot differed in replicate experiments. It then seemed more likely that this spot arose through decom- position of penicillin after application to the strip. This suggestion was confirmed by an independent assessment of the non-penicillin radioactive substances present in the crystalline penicillin. A dilute solution of this penicillin was acidified and rapidly extracted at ice temperature with three l-ml portions of ether.The total counts in the three extracts and aqueous residue are shown in Table 111. It is clear from the trend that practically all the TABLE 111 ESTIMATION OF NON-PENICILLIN SULPHUR COMPOUNDS BY ETHER EXTRACTION Radioactivity as percentage of Initial Value Experiment 1 Experiment 2 r A \ 1st Extract . . .. .. 87.0 83.0 2nd 3 ) . . .. .. 8.0 12.3 3rd 39 . . .. .. 1.5 2.0 Aqueous residue . . .. 3.5 2.7 penicillin is extracted, leaving about 3 per cent. of other sulphur-containing substances. It was confirmed that penicilloic acid, produced by inactivation of radioactive penicillin with penicillinase, was not readily extracted from acid solution with ether and remained substantially at the origin after chromatography on buffered filter-paper. Destruction of penicillin on the paper was confirmed in further experiments in which the strips were left for various periods between spotting and development.It is clear from the results in Table IV that destruction increases with time of standing and is more severe on strips kept in a dry laboratory atmosphere than on those maintained in a humid condition and kept in a refrigerator. I t had been shown in earlier experiments that penicillin is highly unstable in concentrated salt solutions, and these are the conditions that obtain when the penicillin spot dries out on the strip, since this has been previously soaked in 20 per cent. phosphate buffer. It is more difficult to explain why so much penicillin is inactivated when the minimum time elapses between spotting and development, and also why destruction does not continue during development with ether.If it did, then clearly there would be a radioactive streak right down the strip, whereas in fact very little radioactivity is detected between the penicillin zones. It seems probable that evaporation of ether from the strips in a moist atmosphere causes water to condense on them, so diluting the buffer solution and rendering it less destructive towards penicillin; the limp appearance of the strips on removal from the developing tank would support this suggestion.32 SMITH AND ALLISON PAPER CHROMATOGRAPHY [Vol. 77 The bearing of these observations on the conduct and validity of the chromatographic assays needs to be considered. It is clearly desirable to minimise destruction of penicillin by maintaining the strips in a humid condition and avoiding delay between spotting and developing.Nevertheless, destruction of penicillin will not invalidate the assay, which TABLE IV PERCENTAGE OF RADIOACTIVITY LEFT IN INITIAL SPOT Residual radioactivity, % f h 7 Time between spotting After storage in After storage in and development Experiment humid refrigerator dry laboratory 4 hour A 24, 23 B 15, 17 19, 18 24 hours A 23, 28 32, 33 5 hours B 20, 25 51, 49 23 hours A 42, 62 24 hours B 14 32 measures only the froportions of the penicillin species, unless inactivation of the different penicillins occurs a t different rates. The phenomenon does, however, throw doubt on the wisdom of using developed standards. NOTES ON CHROMATOGRAPHIC PROCEDURE- In the course of several years' experience, a number of modifications of the paper chromatography technique have been tried from time to time.The following represent the main features in which our current method differs from that described by Goodall and Levi. We use Whatman No. 1 filter-paper already cut into strips 1-8 cm by 30 cm when purchased. The strips are immersed in 20 per cent. potassium phosphate buffer at pH 6.2, pressed between blotting-paper and kept in a blotting-paper folder until used a few hours later. Numerous methods have been tried for the control of humidity, but none have been found to give absolutely dependable results. Our present practice is to spot the strips in a large refrigerator at about 4" C. The test solutions and standards are measured from Agla micrometer syringes. To prevent the developed heptylpenicillin zone passing right off the strip, we have sometimes found it necessary to cover only about the top three-quarters of the central drum with wet bandage.Quantitative assessments are made with two benzyl- penicillin standards differing in concentration by a factor of 10; these solutions are spotted on to buffered strips and developed alongside the test strips. ISOTOPE DILUTION ASSAYS FOR BENZYLPENICILLIN- . specifically. I t was hoped to utilise the 35S-penicillin in an isotope dilution assay for benzylpenicillin The method would be to add to the sample of penicillin or fermentation broth TABLE V ~SOPROPYL - ETHER COMPLEX OF PENICILLIN ACID Original penicillins A f 7 Crystalline Yellow Cyclohexylamine Penicillin species sodium salt calcium salt salt A B C p-Hydroxybenzyl (X) .. 0-2 0.3 0.7 1.0 87.6 77.1 A 3-Pentenvl .. 1.8 3.8 Benzyl ? (XI) .' n-Amy1 (Aj .. 4.8 6.9 n-Heptyl (K) - . 4.9 10.9 trace 0.3 56.6 6.9 13.8 22.4 Recrystallised isopropyl ether complexes from A B c 0.1 0.1 trace 0.5 0.9 0.5 97.2 94.7 93.0 0.7 1.6 2.4 0.9 1-1 2.3 0.6 1.6 2.8 a small known weight of 36S-penicillin G of known specific activity. From the mixture a specimen of pure penicillin G would be isolated (not quantitatively) and its specific activityJan., 19521 O F RADIOACTIVE PENICILLIN 33 measured. It was expected that the isolation could be effected through the isopropyl- ether complex. The accuracy of the method clearly depends on the ability to free the standard and the final specimen completely from other penicillins.Unfortunately, as Table V shows, the isopropyl- ether method is in fact far from adequate as a means of purification. This experience serves to re-emphasise the fact that isotope dilution methods can give misleading results unless one has avilable a dependable method for purifying (in however poor yield) the substance to be determined. An isotope dilution method for benzylpenicillin that partly avoids this difficulty has recently been published.lG The marker penicillin is labelled with the stable isotope carbon-13 in the phenylacetyl group (radioactive carbon-14 could be used alternatively). The final specimen need not then be purified rigorously, because only benzylpenicillin carries the benzyl group, and the final measurement is, in any event, made on the phenylacetic acid liberated on hydrolysis.It is still imperative, however, that either the labelled standard or a reference standard of purest benzylpenicillin should be available. Grateful acknowledgments are made to Miss Hazel Thorpe and Mr. M. Robson for assistance with the radiometric assays. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. “Symposium on Methods of Penicillin Assay: their Purpose, Scope and Validity,’’ W. Heffer & Sons Ltd., Cambridge, 1948; reprinted from Analyst, 1948, 73, 197-216; 244-257. Lester Smith, E., and Bide, A. E., Biochem. J., 1948, 42, xvii. “The Chemistry of Penicillin,” Princeton University Press, Princeton, New Jersey, 1949, p. 9. Sheehan, J . C., Mader, W.J., and Cram, D. J., J . Amer. Chem. SOC., 1946, 68, 2407. Boon, W. R., “Symposium on Methods of Penicillin Assay: their Purpose, Scope and Validity,” “Report of the Analysts’ Sub-committee of the Ministry of Health Conference on the Differential Trenner, N. R., and Buhs, R. P., J . Amer. Chem. SOL, 1948, 70, 2897. Goodall, R. R., and Levi, A. A., Analyst, 1947, 72, 277. Winsten, W. A., and Spark, A. H., Science, 1947, 106, 192. Fording, 0. B., and Breed, N. D., Fed. Proc., 1949, 8, 197. Glister, G. A., and Grainger, A., Analyst, 1950, 75, 310. Karnovsky, M. L., and Johnson, M. J., Anal. Chem., 1949, 21, 1125. Rowley, D., Miller, J., Rowlands, S., and Lester Smith, E., Nature, 1948, 161, 1009. Rowlands, S., Rowley, D., and Lester Smith, E;, J . Chem. SOL, 1949, Supplement No.2, p, S405. Lester Smith, E., “Partition Chromatography, Craig, J. T., Tindall, J. B., and Senkus, M., Anal. Chem., 1951, 23, 332. W. Heffer & Sons Ltd., Cambridge, 1948, p. 9; Analyst, 1948, 73, 202. Assay of Penicillin,” Analyst, 1949, 74, 79. Biochem. SOC. Symposia No. 3, 1950, p. 89. GLAXO LABORATORIES LIMITED GREENFORD, MIDDLESEX DISCUSSION MR. C . R. BOND asked whether there was any possibility that a radioactive sample might inactivate more quickly than an ordinary sample of penicillin when spotted out on the paper strip. DR. LESTER SMITH replied that it was improbable that the radioactivity of these penicillin samples would be intense enough to cause inactivation. MR. N. STRAFFORD asked whether the authors considered that the radiochemical method gave more accurate results than the Goodall and Levi method and whether it was being used for routine control. DR. LESTER SMITH replied that the radiochemical method was probably more accurate than the original Goodall and Levi method, but unfortunately i t was only applicable to very highly radioactive samples of penicillin. MR. R. C. CHIRNSIDE asked if the decomposition of the penicillin that had been detected in the course of the work on the radioactive assay minimised the value of the Goodall and Levi method. He had under- stood Dr. Lester Smith to say that such decomposition could be appreciable in the course of an hour or two, whereas in the Goodall and Levi method the tests were allowed to stand overnight. Might the amount of decomposition therefore be sufficient to affect the results adversely ? DR. LESTER SMITH replied that the decomposition of penicillin on the paper strips was certainly a disturbing feature. However, it would only invalidate the Goodall and Levi method if the various penicillin species were decomposed at different rates. The inactivation did not appear to continue during development with ether, and could therefore be minimised by transferring the strips to the developing tank as soon as possible after the test spots had dried.

ISSN:0003-2654    

DOI:10.1039/AN9527700029

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

34 COOK : THE FRACTIONATION OF URINARY 17-KETOSTEROIDS [Vol. 77 A Rapid Clinical Method for the Estimation and Fractionation of Urinary 17-Ketosteroids BY E. R. COOK Total 17-ketosteroids in urine can be rapidly estimated, but further partitioning into a- and fl-17-ketosteroids requires a considerable amount of work, This paper describes attempts to reduce this work to a minimum by separating smaller quantities of material and by a direct estimation of the p-ketosteroid component as its digitonide. Considerable time is saved without loss of accuracy. Results obtained with pure 17-ketosteroids and with pure 17-ketosteroids added to urine extracts indicate recoveries of above 93 per cent. ALTHOUGH many methods have been proposed for the determination and fractionation of urinary 17-ketosteroids, they are all essentially similar, involving acid hydrolysis of the water-soluble ketosteroid conjugates and extraction of the crude total 17-ketosteroids with a lipoid solvent, removal of non-specific chromogenic material by a chemical separation into non-ketonic and ketonic fractions, with a further separation of the ketonic fraction into 3(a)-hydroxy- and 3(/3)-hydroxy-17-ketosteroids. The 17-ketosteroids in the various extracts are estimated by colorimetric or polarographic methods.For clinical purposes, crude total 17-ketosteroids can be rapidly estimated in 10ml of urine by the method of Drekter, Pearson, Bartczak and McGavack,l but few non-specialist clinical laboratories could possibly carry out the lengthy and detailed procedures necessary for further fractionation.I t is thought, however, that the increasing significance of steroid metabolism, especially with the prospect of increasing availability for clinical use of adreno- corticotrophic hormone, requires the development of rapid, simple fractionation procedures. A method is described, based on the routine procedure of Reiss, Hemphill, Gordon and Cook,2 by which the crude extract from 250 ml of urine may be rapidly purified and partitioned into 3(a)-hydroxy- and 3(/3)-hydroxy-17-ketosteroids, the results being available within 30 hours of receiving the urine specimen. METHOD REAGENTS- Acetic acid-Heat AnalaR glacial acetic acid under reflux for 1 hour with 2 g of chromic acid per litre and then fractionally distil (see Orton and Bradfield3).Benzene-AnalaR benzene has proved satisfactory, and the benzene distilled from extracts may be used again if washed with water, dried over anhydrous sodium sulphate and fractionally distilled. AZcohoZic m-dinitrobenxene-Make an accurate 2 per cent. w/v solution with purified alcohol and m-dinitrobenzene purified by the method of Callow, Callow and ern men^.^ The solution should be kept in the dark and discarded after 4 to 5 days. Ethanol-Dry commercial absolute alcohol by heating under reflux with calcium oxide for 1 to 2 hours, then fractionally distil through a 6-pear column. Heat the distillate under reflux for 1 hour with 10 g of semicarbazide acetate per litre and again fractionally distil. Alcoholic 2.5 N potassium hydroxide-Shake the appropriate weight of potassium hydroxide pellets with pure alcohol and centrifuge to remove potassium carbonate. Titrate an aliquot of the clear supernatant liquid against 0.25 N hydrochloric acid and adjust the normality to 2.5 This reagent must be kept in a refrigerator and discarded after 4 days.Girard’s reagent T5-On account of its hygroscopic nature, the solid reagent should be kept over silica gel in a desiccator. 0.01 N by addition of alcohol. Make the solution in acetic acid just before use. THE HYDROLYSIS AND EXTRACTION OF URINARY 17-KETOSTEROIDS- An examination of the many procedures for the hydrolysis and extraction of 17-keto- steroids reveals that most investigators prefer either independent acid hydrolysis followedJan., 19521 COOK : THE FRACTIONATION O F URINARY 17-KETOSTEROIDS 35 by extraction with ether or benzene, or simultaneous hydrolysis and extraction with benzene.Bitman and Cohen6 have shown that destruction of 3(~)-hydroxy-17-ketosteroids may occur during the usual heating under reflux for 10 minutes with 15 per cent. v/v of concentrated hydrochloric acid, and we have accordingly made use of the simultaneous hydrolysis and extraction procedure of Hamburger,' modified to deal with 250 ml of urine. Procedure-Acidify 250 ml of urine with 25 ml of 40 per cent. v/v sulphuric acid, add 50 ml of benzene and heat the mixture under reflux on an electric hot-plate for 30 minutes. Cool rapidly and remove the benzene extract in a separating funnel. Repeat the extraction under reflux with 50ml and then with 40ml of benzene.Wash the combined extracts with 30ml of water, followed by four 30-ml portions of 2 N sodium hydroxide and three 30-rnl portions of water, and remove the final water wash as completely as possible. Adjust the volume of benzene to 150 ml and evaporate two 5-ml aliquots to dryness in an oil-bath at 100" to 105" C ; remove the final traces of benzene in a vacuum desiccator. The binary azeotrope between water (91 per cent.) and benzene (9 per cent.) boils at 69.3" C and eliminates the need for chemical drying of the wet extracts. Estimate the total 17-ketosteroids in the dried residue by the Zimmermann reaction. THE COLORIMETRIC DETERMINATION OF 17-KETOSTEROIDS- The original method of Zimmermann* as modified by Callow, Callow and Emmens4 has proved satisfactory for routine purposes.However, traces of water in the reacting solution may depress the final colour density considerably (see Fig. l), and care must be taken to store the reagents under anhydrous conditions. Inhibition of colorimetric reaction, yo Curve c 25 50 Fig. 1. The influence of water on the Fig. 2. Standard curves prepared with crystal- colour produced in the Zimmermann line 17-ketosteroids. Curve (a) androsterone, reaction reaction mixture diluted with 10 ml of alcohol B.P.; curve (b) dehydroisoandrosterone, reaction mixture diluted with 10 ml of 85 per cent. alcohol: curve ( c ) dehydroisoandrosterone, reaction mixture diluted with 5 ml of 85 per cent. alcohol Procedwe-Add 0.2 ml of pure. alcohol to the dried 17-ketosteroid residue, followed by 0.2 ml of alcoholic 2 per cent.m-dinitrobenzene solution and 0.2 ml of alcoholic 2-5 N potassium hydroxide. Use pipettes Determine the reagent blank on 0.2 ml of pure alcohol.36 COOK : THE FRACTIONATION OF URINARY 17-KETOSTEROIDS [Vol. 77 of 1 ml capacity, graduated in 0.01 ml, with the tip drawn out to a fine jet inclined sideways; this device ensures that the liquid that invariably creeps around the tip is taken off by the wall of the tube. Shake the tubes thoroughly, firmly stopper them and incubate them in the dark at 25" 0.1" C for 1 hour. Add 10 ml of absolute alcohol and measure the colour density against the reagent blank in a photo-electric absorptiometer, using Ilford green filters with a maximum transmission at 5 2 0 0 ~ . Perform duplicate determinations, which must be repeated if the absorptiometer readings are not within 2 per cent.Read the 17-ketosteroid content from a standard graph (see Fig. 2) prepared with androsterone and checked at frequent intervals. No correction is made for the non-specific chromogenic material present in the crude benzene extract,g as the true 17-ketosteroid content is obtained after the Girard-T separation. SEPARATION OF KETONIC MATERIAL- The removal of the ketonic fraction from the crude extract by the use of Girard's reagent T follows along conventional lines. By using smaller amounts of crude steroid, and conse- quently smaller volumes of reagents, it is possible to carry out the separation very quickly, a further factor introduced in the present investigation being the increased rate of hydrolysis of the Girard T - 17-ketosteroid complex on raising the temperature of hydrolysis from 50" to 100" C.Experiments with pure 17-ketosteroids and urine extracts containing added known quantities of 17-ketosteroids indicate recoveries between 95 and 105 per cent, (Table I). TABLE 1 RECOVERY OF 17-KETOSTEROIDS AFTER THE MICRO GIRARD-T SEPARATION Experiment Dehydroisoandrosterone .. Androsterone . . .. .. Urine extract A . . .. .. Urine extract A . . .. .. Urine extract B . . * . .. Urine extract B . . .. .. Urine extract C . . .. .. Urine extract C . . .. .. Crude steroid, mg 1.8 1.0 1.8 1.0 1.8 1.0 Pure steroid added, mg 0.98 1-40 1.00 1.00 0-74 0.74 - - - 0-74 Total pure steroid found, "g 0-92 1.38 0.97 1.00 1.54 1.60 1-18 1.41 1.09 1.33 Added steroid recovered, Recovery, mg % 93-9 98.6 97.0 100.0 0.74 100.0 0.75 101-4 0.72 97.3 Procedure-While the estimation of crude 17-ketosteroids is proceeding, distil 135 ml of the wet benzene extract down to 2 to 3 ml, quantitatively transfer the distillate to a test tube with pure benzene and evaporate to dryness; remove the last traces by suction.Add 9 ml of pure benzene and dissolve the residue by gentle warming. Take an aliquot of this solution containing 1-8 mg of crude 17-ketosteroids, calculated from the result of the Zimmer- mann reaction, and evaporate to dryness by heating in an oil-bath, followed by suction. With a pipette, add 0-1 ml of glacial acetic acid containing 5 mg of Girard's reagent T, lightly stopper the tube and incubate in a bath of boiling water for 4 minutes.Cool rapidly, add 4 ml of chilled 2.1 per cent. sodium carbonate solution and gently shake the tube to remove as much carbon dioxide as possible. Pour the aqueous contents into a 20-ml separating funnel and wash the tube with 4 ml of ether; put this into the funnel, which should then be shaken vigorously. Complete the removal of the non-ketonic fraction by two further extractions with 4 ml of ether. Wash the combined ether extractions with 2ml of water and add the washing to the aqueous extract containing the Girard T - ketosteroid complex. With a pipette, add 0.5 ml of con- centrated hydrochloric acid to the aqueous solution, and then pour this back into the separating funnel and shake the funnel to remove the ether - water emulsion adhering to the walls; finally transfer the solution back to the test tube.Rinse the funnel with 2 ml of water and drain the washings into the aqueous extract, which should now be about 8ml in volume. Lightly stopper the tube, preferably by means of a glass 'bulb with a sealed tail, and place it in a small water-bath at 50" to 60" C for 3 to 5 minutes to boil off most of the dissolved ether ; finally incubate the tube in a bath of boiling water for 5 minutes. Girard's reagent T doesJan., 19521 COOK : THE FRACTIONATION OF URINARY 17-KETOSTEROIDS 37 not decompose under these conditions. Cool and pour the aqueous extract into the funnel, and remove the free 17-ketosteroids with 4ml of ether followed by three 3-ml portions of ether, washing the test tube with the ether each time.Wash the combined ethereal extracts with 2 in1 of water followed by 2 ml of 2.1 per cent. sodium carbonate and three 2-ml portions of water, and make the final separation as sharp as possible. Wash the funnel with 2 ml of ether, and add this washing to the extract; then add 2 ml of benzene and a small chip of porcelain. Evaporate the wet extract to dryness in an oil-bath; the ether boils off first, followed by the water - benzene azeotrope and finally by the remaining benzene; remove the last traces by suction. Add 2 ml of pure alcohol, gently warm the tube to promote solution of the 17-ketosteroid, then cool and take 0.2-ml aliquots for the Zimmermann reaction. PRECIPITATION OF THE 3(/3)-HYDROXY-17-KETOSTEROID DIGITONIDE- Under the appropriate conditions, digitonin combines with /3-17-ketosteroids to form a complex insoluble in ether. Earlier methods involving this reaction to separate urinary 17-ketosteroids into a- and /3-fractions have been criticised.The determination proposed by Talbot, Butler and MacLachlanlO suffers from the obvious defect that the /3-17-ketosteroid is not directly estimated, while the method of Baumann and Metzgerll gives variable results, depending upon the initial concentration of total 17-ketosteroids. The procedure of Frame12 requires a minimum of 15 mg of total steroid, although modifications permit this separation to be applied to micro quantities of total 17-keto~teroid.~J~ The determination is com- paratively lengthy and necessitates decomposition of the p-17-ketosteroid digitonide by pyridine, extraction with ether, removal of pyridine from the ether extract and subsequent estimation of the 17-ketosteroid by colorimetric or polarographicf3 methods.Preliminary investigation showed that an excess of digitonin has only a slight inhibitory action upon the Zimmermann reaction, and a method was developed whereby, after quanti- tative precipitation, a modified Zimmermann colour reaction could be carried out directly on the /3-ketosteroid digitonide. By this process 30 to 450 pg of /3-17-ketosteroid can be recovered in yields generally above 95 per cent. from 800 to 1000 pg of total 17-ketosteroid, as shown in Table 11. Below 30 pg, equivalent to 3 per cent., yields are occasionally low. This is not a disadvantage for clinical purposes, as the normal values lie between 0 and 18 per cent.TABLE I1 RECOVERY OF KETOSTEROID BY THE PROPOSED PROCEDURE Total ketosteroid t% Experiment present, Solutions of dehydroisoandrosterone Solutions of androsterone and dehydroisoandrosterone 1000 1000 1000 900 850 800 800 400 400 400 Ketonic extracts of patients' urines 1000 with added dehydroisoandrosterone 1000 1000 1000 900 900 Dehydroisoandrosterone idded, P8 400 300 200 100 50 40 30 128 86 43 427 342 256 171 171 128 86 nil 40 nil 80 nil 40 1 found, Pg 382 279 197 100 49 38 29 129 93 44 400 338 251 169 166 122 81 32 73 40 118 90 132 Recovery, 95.5 93.0 98.5 100.0 98.0 95-0 96.7 100.8 108-1 102.3 93.6 98.8 98.1 98.8 97.1 95.3 94.2 % 102.5 97.5 105-038 COOK : THE FRACTIONATION OF URINARY 17-KETOSTEROIDS [Vol.77 Procedzzre-Take an aliquot of the alcoholic solution from the Girard-T separation containing 800 to 1OOOpg of ketosteroid and evaporate it to dryness in a 15-ml tapered centrifuge tube. Add 0-2 ml of pure alcohol and 0-2 ml of a 2 per cent. solution of digitonin in 80 per cent. alcohol, and place the lightly stoppered tube in a water-bath at 70" to 75" C for 4 minutes. Cool to room temperature, add 0.1 ml of ether, firmly stopper the tube and leave it in a refrigerator overnight. Precipitation of the P-ketosteroid - digitonin complex occurs within an hour, but is not quantitative unless allowed to stand at 0" C for at least 12 hours. Centrifuge the flocculent precipitate at 2500 r.p.m. (radius P4cm) for 3 minutes and pour off the supernatant liquid carefully; then invert the tube and allow it to drain for 2 minutes.Wipe the lip with filter-paper, add 1 ml of ether and break up the precipitate with a fine glass rod. Use a further 1 ml of ether to wash the glass rod and the sides of the tube and centrifuge it at 2500 r.p.m. for 2 minutes. Pour off the supernatant ether, but do not allow the tube to drain, as the precipitate dries very quickly and easily flakes off the tube. Wash the residue with a further 2 ml of ether, centrifuge and pour off the supernatant ether as before. Add 0.5 ml of pure alcohol to the partly dried digitonide complex, together with a small porcelain chip, and evaporate the mixture to dryness in a boiling water-bath and follow by suction. This alcohol stage, introduced to break up and distribute the precipitate around the bottom of the centrifuge tube, has improved the recovery of /3-17-ketssteroids by several per cent.COLORIMETRIC ESTIMATION OF THE fl-17-KETOSTEROIDS- The presence of digitonin has little effect on the Zimmermann reaction, but an opalescent precipitate forms in a few moments if absolute alcohol is added to the reaction mixture. This may be prevented by dilution with 85 per cent. alcohol, although a slight quenching of the colour is observed. Procedure-Add 0-2 ml of pure alcohol and 0.2 ml of alcoholic 2 per cent. m-dinitro- benzene solution to the dried fl-17-ketosteroid - digitonin precipitate in the tube, mix well, bring rapidly to incipient boiling over a bunsen burner and then plunge the tube into a bath of cold water.Little, if any, alcohol is lost by this seemingly drastic treatment, which is necessary to bring all the residue into solution. Then add 0-2 ml of alcoholic 2.5 N potassium hydroxide, shake the tube well, firmly stopper it and incubate at 25" f- 0.1" C for 1 hour. Similarly incubate a reagent blank consisting of 0.2 ml of pure alcohol, 0.2 ml of 2 per cent. m-dinitrobenzene solution and 0-2 ml of alcoholic 2-5 N potassium hydroxide. Add 10 ml of 85 per cent. alcohol and measure the colour developed as before. However, owing to the slight quenching effect of the aqueous alcohol used as the diluting medium, it is necessary to draw a standard curve for these conditions (see Fig. 2, curve B). We have found it convenient to construct a further standard curve for quantities of steroid less than lOOpg, in which the absorptiometer readings are considerably increased by the addition of only half the usual amount of 85 per cent.alcohol (see Fig. 2, curve C). SUMMARY A method has been described whereby the estimation of total 17-ketosteroids in 250 ml of urine, and the determination of the separated /%fraction can be completed in 30 hours. A new feature of the method is the direct determination of the fl-fraction in the form of its digitonide; this avoids the usual decomposition stage. Standard quantities of pure 17-ketosteroids added to urine gave recoveries of at least 93 per cent. The procedure is recommended as a rapid and satisfactory routine determination for use when the estimation of fractional 17-ketosteroids may be of diagnostic value. The author wishes to express his thanks to Dr. Max Reiss, Dr. J. J. Gordon, Miss M. Rooks and Miss J. Pelly for their advice and assistance in this work. Thanks are also due to Dr. C. L. Hewett of Organon Laboratories Ltd., Glasgow, for supplies of pure dehydroiso- andros t erone.Jan., 19521 WHALLEY AND WYATT 39 REFERENCES 1. Drekter, I. J., Pearson, S. J., Bartczak, E., andMcGavack, T. H., J . Clin. Endocrinol., 1947,7, 795. 2. Reiss, M., Hemphill, R. E., Gordon, J. J., and Cook, E. R.. Biochem. J., 1949, 44, 632. 3. Orton, K. J. P., and Bradfield, A. E., J . Chem. Soc., 1927, 983. 4. Callow, N. H., Callow, R. K., and Emmens, C. W., Biochem. J., 1938, 32, 1312. 5. Girard, A., and Sandulesco, G., Helv. Chtm. A d a , 1936, 19, 1095. 6. Bitman, J., and Cohen, S. L., J . Biol. Chem., 1949, 179, 455. 7. Hamburger, C., Acta Endocrinol., 1948, 1, 375. 8. Zimmermann, W., Hoppe-Seyl. Z., 1936, 245, 47. 9. Gibson, J. G., and Evelyn, K. A,, J . Clin. Invest., 1938, 17, 153. 10. Talbot, N. B., Butler, A. M., and MacLachlan, E., J . Biol. Chem., 1940, 132, 595. 11. ,Baumann, E. J., and Metzger, N., Endocrinology, 1940, 27, 664. 12. Frame, E., Ibid., 1944, 34, 175. 13. Butt, W. R., Henley, A. A., and Morris, C. J. 0. R., Biochem. J., 1948, 42, 447. THE BIOCHEMICAL AND ENDOCRINOLOGICAL RESEARCH DEPARTMENT THE BRISTOL MENTAL HOSPITALS BARROW GURNEY, BRISTOL November, 1950

ISSN:0003-2654    

DOI:10.1039/AN9527700034

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Jan., 19521 WHALLEY AND WYATT 39 Porcelain Microchemical Apparatus for Gravimetric Analysis BY C. WHALLEY AND G. H. WYATT An investigation into the behaviour of porcelain micro-apparatus has been made by members of a British Standards Institution Sub-committee. The behaviour both on heating and on acid treatment was investigated in order t o provide standard methods of testing and to suggest limits to the changes in weight that could be allowed. The results showed that on heating to temperatures in excess of 95OOC softening and discoloration of the glaze occurs and is accompanied by increase in weight. Acid treatment of ordinary micro-crucibles showed very little attack, but with filter crucibles incor- porating an unglazed porcelain filtering base the losses were large and inconsistent.In view of the results of these tests, the authors would welcome information from other sources as to the behaviour of porcelain apparatus used in gravimetric micro-analysis and the tests applied in deciding whether the apparatus is suitable for this purpose. THE British Standards Institution is producing a series of standards for microchemicaI apparatus (B.S. 1428), and a sub-committee, of which one of us is the present chairman, was established in January, 1948, to consider accessory apparatus connected with gravimetric micro-analysis. Included in this committee’s considerations were the following : combustion boats, crucibles, filter crucibles and filtersticks. These items are available in several materials, including porcelain. It was considered desirable that the standard should control the quality of the materials used, but the only available data for porcelain appeared to be those in the draft revision of B.S.914:1940, “Tests for Laboratory Porcelain,” the title of which has been changed to “Quality of Laboratory Porcelain.” This draft describes the methods of testing glazed porcelain recommended for laboratory apparatus (i.e., essentially macro- apparatus) and specifies the performance of such items when so tested, but it does not include filtration apparatus incorporating porous porcelain. The committee therefore undertook an examination of porcelain micro combustion boats, crucibles and filter crucibles. The following members of the British Standards Institution Sub-committee were responsible for the experimental investigations: Miss I.H. Hadfield, Prof. H. B. Nisbet, Messrs. A. Bennett and G. Ingram and the authors. It was thought that it would be useful to publish the results, most of which were obtained on experimental batches. The main object of the committee was to devise a form of treatment that would bring the articles to a constant weight when used for further experiments, but with the treatment described below it was usually not found possible to achieve stability. The members of the committee do not feel that they can make any recommendations as t o the use of porcelain for gravimetric micro-analysis without further work and would welcome any information or suggestions about the use or testing of porcelain microchemical apparatus40 WHALLEY AND W'YATT PORCELAIN MICROCHEMICAL [Vol.77 during the last ten years. of The Analyst. Comments should be addressed to the authors, care of the Editor COMBUSTION BOATS AND CRUCIBLES It was understood that these two items are customarily manufactured from similar materials. Therefore the boats were tested for constancy of weight on ignition and the crucibles for resistance to acid attack; both methods were derived from B.S. 914 (draft revision) and are described below. (a) "Method of test for constancy of weight o n ignition (carried out on complete articles or broken pieces). Wash the articles or pieces in cold 10 per cent. hydrochloric acid followed by distilled water, dry and ignite them at a dull red heat. Allow to cool and weigh, and repeat the ignition until a constant weight is obtained.Then heat for 2 hours in a muffle furnace at a temperature of 950" to 1000" C, allow to cool and weigh again. " It was found that a second ignition of the combustion boats for 2 hours at 650" to 800" C resulted in changes of weight ranging from a loss of 11 pg to a gain of 30 pg; 60 per cent. of the boats gained in weight (mean loss = 6 pg, mean gain = 11 pg). Further ignition, however, rapidly led to constancy of weight to within 2 pg. When the boats were ignited at the higher temperature (950" to 1000" C) there was a large gain in weight of the majority of those tested and the glaze became appreciably discoloured, as if by oxidation of ferrous to ferric oxide; Of 16 boats tested only one lost weight (2 pg); the remainder gained weight by amounts ranging from 30 pg to 1100 pg (mean 387 pg).Subsequent ignition at 950" to 1000" C resulted in reduced gains in weight, but the increases remained highly significant and there was no indication that a constant weight could be expected after a reasonably small number of ignitions. The initial weights of the boats were about 0.7 g, so that the mean gain was 5-5 mg per 10 g of total weight; B.S. 914 (draft revision) specifies a maximum of 0.1 mg per 10 g and the committee had considered that a limit of 0.01 mg per 10 g should be attained for microchemical apparatus. The above tests were also applied to micro crucibles of 3-ml capacity and it was found that at 950" to 1000" C the glaze softened. (b) "Method of test for resistance of glaze to acid (carried out on complete dishes).Wash the dish in cold 10 per cent. hydrochloric acid followed by distilled water, dry to constant weight at 120" C and, when cold, tare it against a similar dish. Fill it to three- quarters of its total capacity with constant boiling-point hydrochloric acid, cover with a clock glass and heat on a steam-bath for four hours. Then wash with distilled water and dry to constant weight at 120" C taring against the same dish as before." Micro-crucibles of 3-ml capacity subjected to this test changed in weight by amounts ranging from a loss of 20 pg to a gain of 6 pg. Calculation from measurements of the crucibles showed that the average loss in weight was about 0.05 mg per square decimetre of surface. It should be noted, however, that the distribution of the weight changes showed that they were probably controlled by non-uniformity of surface condition during weighing rather than by loss of substance.It is to be concluded that the glaze is resistant to hydrochloric acid attack and that the limiting loss of 0.1 mg per square decimetre of surface considered necessary by the committee could be attained by the manufacturers (B.S. 914, draft revision, specifies a maximum loss of 1 mg per square decimetre). FILTER CRUCIBLES Filtersticks and filter crucibles are manufactured from similar materials and the corn- mittee's attention has been confined to the latter in the belief that the findings would also be applicable to filt erst icks. Resistance to hydrochloric acid-A series of tests was devised by the committee as follows- (i) Immerse the filter crucibles in concentrated hydrochloric acid, sp.gr.1-16, on a steam-bath for 6 hours; suck through the filter one crucible-full of acid and wash with distilled water; dry, ignite at 650" C, cool and weigh. (ii), (iii) Repeat treatment (i) twice, but each time reduce the treatment with con- centrated acid to 1 hour. (iv), ( v ) Repeat the treatment twice, but each time use 2 N hydrochloric acid for 1 hour.Jan., 19521 APPARATUS FOR GRAVIMETRIC ANALYSIS 41 A set of 18 porcelain filter crucibles, prepared for experimental purposes only, was supplied in three separate groups, and samples from each group were tested by three independent observers. The behaviour of all the groups was found to be similar by all observers; the changes of weight are summarised in Table I (crucibles only were tested, i e ., without lids). TABLE I LOSS OF WEIGHT OF SET NO. 1 OF PORCELAIN FILTER CRUCIBLES AFTER TREATMENT WITH HYDROCHLORIC ACID 18 crucibles examined Loss in weight, pg A f \ Treatment Range Mean (i) . . .. .. .. . . 400 to 2700* 1500* (ii) . . .. .. .. .. 12 to 192 77 (iii) again repeated3 . . . . .. 5 to 12 9 (iii) . . .. .. .. .. 42 to 420 162 (iii) repeatedt . . .. .. .. 87 to 627 299 (iv) . . .. .. . . .. 61 to 533 214 ( 4 3 * .. .. .. .. 109 to 242 146 * Excluding one abnormal loss. of 4200 p g . j- Test by two investigators only. $ Test by one investigator only. . Further treatments of various kinds failed to effect any improvement in the performance of these crucibles.Another set of porcelain crucibles from another experimental batch was tested in a similar manner. The results were more regular than those shown in Table I and the losses of weight were less, but they were still too great to permit use of the filter crucibles for gravimetric micro-analysis. It was also evident from the results of these tests that the attack of 2 N hydrochloric acid was greater than that of the concentrated acid. A third experimental set of 12 porcelain filter crucibles embodying a modified filter base was then tested by two observers in the manner described above, but including additional treatments with concentrated nitric acid and with distilled water. The new reagents were included lest hydrochloric acid should exert an unexpected specific action.The results are TABLE I1 LOSS OF WEIGHT IN pg OF SET NO. 3 OF PORCELAIN FILTER CRUCIBLES AFTER TREATMENT WITH ACIDS AND WITH DISTILLED WATER Duration and nature of treatment r A \ Crucible No. 1 2 3 4 5 6 7 8 9 10 11 12 1 hour, conc. HC1 793 815 1 hour, 2 N HC1 1644 1067 1 hour, conc. HNO, 407 448 46 hours, 2 N HC1* 376 367 934 312 329 339 1 hour, conc. HCl 62 60 1 hour, 2 N HC1 72 45 1 hour, conc. HNO, 30 24 1 hour, 2 N HCl 100 336 699 253 261 265 1 hour, conc. HCl 51 40 1 hour, 2 N HC1 54 25 1 hour, conc. HNO, 10 17 1 hour, 2 N HCl 274 167 187 378 138 232 1 hour, conc. HCl 13 0 1 hour, 2 N HCI 56 63 1 hour, conc. HNO, 25 5 3 hours, 2 N HCl 1029 914 1263 1207 922 973 1 hour, conc. HCl 69 1 hour, 2 N HCl 68 60 1 hour, conc. HNO, 21 15 1 hour, water 71 35 53 120 51 121 1 hour, conc.HCI 69 1 hour, 2 N HCl 77 76 1 hour, conc. HNO, 5 7 2 hours, water 32 38 76 58 120 63 * This acid was at room temperature; all other treatments were on a steam-bath.42 WHALLEY AND WYATT [Vol. 77 shown in Table 11. It is evident that hydrochloric acid was not specific in its attack and these results confirmed that the dilute acid removed more material than did the concentrated acid. The losses of weight after heating with water are surprising, especially when it is noted that this treatment followed repeated extractions with dilute acid. Finally three crucibles, Nos. 7, 8 and 9, were placed in a Pyrex beaker containing distilled water that was alternately heated nearly to boiling and cooled. The water was hot for a total of €4 hours and at room temperature for a total of 32 hours.Another similar beaker containing distilled water only was treated in the same manner to serve as a control. The crucibles lost in weight 92, 92 and 115 pg respectively, and the water was found to have extracted from the three crucibles about 130 pg silica and a trace of alkali. The results of the tests on ordinary micro-crucibles as described above show that glazed porcelain supplied by the same manufacturer withstands treatment with hydrochloric acid ; it must, therefore, be concluded that with the filter crucibles it is primarily the porous- porcelain filter base that is attacked. Committee members also examined the behaviour of transparent silica filter crucibles and the results are given here for comparison with those obtained with porcelain filter crucibles of approximately the same size and weight. During the initial 6-hour treatment with con- centrated acid, the 17 crucibles and lids tested lost weight, the amounts varying from 13 t o 600 pg (mean 131 pg).These losses did,not appear to be important, because the crucibles were used in the condition in which they were received from the makers and it was to be expected that drastic treatment would remove foreign matter and loose particles that might be present. After treatments (ii) and (ziz) the three investigators obtained rather different results, which are, therefore, shown in full in Table 111. TABLE I11 CHANGE I N WEIGHT I N O F SILICA FILTER CRUCIBLES AFTER AN INITIAL TREATMENT WITH CONCENTRATED HYDROCHLORIC ACID Treatment r A I (ii) (iii) (ii) (iii) ( i v ) (4 repeated repeated Observer A - Crucible No. 1 2 3 4 5 Crucible No. 1 2 3 4 5 6 Crucible No. 1 2 3 4 5 6 Observer B- Observer C- .. -11 - 16 - 17 - 15 . I 0 - 37 + 7 - 13 .. -31 - 27 - 26 - 25 .. - 5 - 36 + 7 - 19 .. $29 - 66 + 4 - 15 .. - 55 -51 - 128 .. - 2 - 90 - 79 * . - 72 - 70 - 30 .. -128 - 18 - 54 .. - 47 - 232 - 38 .. - 9 - 43 - 37 .. - 36 - 6 + 2 .. -11 - 7 + 1 .. - 25 - 7 - 1 .. - 35 - 10 + 12 .. - 15 - 11 - 6 .. -11 - 2 + 4 -t 2 -- 5 -+ 7 -- 8 -t 6 -- 5 - 6 -6 + 2 +4 -8 $7 are the NOTES-obSerVer A treated crucibles and lids separately; the algebraic sums of the changes in weight shown. Observer B repeated the 6-hour and two l-hour treatments with concentrated hydrochloric acid on same filter crucibles. The test was severe, but after sufficient treatment with acid the silica filter crucibles would attain reasonably constant weights. Some of the irregularity in the results may have been due to difficulty in bringing the surface of the silica to the same condition before each weighing , c/o THE EDITOR, The Analyst 7-8, IDOL LANE LONDON, E.C.3 October, 1951

ISSN:0003-2654    

DOI:10.1039/AN9527700039

出版商:RSC    

年代:1952

数据来源: RSC