ISSN:0003-2654    

DOI:10.1039/AN95378FX051

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BX053

出版商:RSC    

年代:1953

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95378BP141

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

NOVEMBER 953 Vol. 78, No. 932 THE ANALYST EDITORIAL ANALYTICAL ABSTRACTS THE Society of Public Analysts and Other Analytical Chemists has provided abstracts of the literature on analytical chemistry for readers of The Analyst ever since the journal was first published in 1876. For the last four years these abstracts were prepared by the Bureau of Abstracts and supplied to members of the Society and to subscribers to The Analyst under the title British Abstracts C. The Bureau of Abstracts will cease to exist at the end of 1953, and British Abstracts will not appear in their present form thereafter. In these circumstances, the Council of the Society has decided to resume publication of abstracts of analytical literature. British Abstracts C will be replaced in the new year by a new publication to be called Analytical Abstracts. This will be edited by Norman Evers, B.Sc., Ph.D., F.K.I.C., and published by W. Heffer Clt Sons Ltd. for the Society, and it will cover the whole range of analytical literature in the same way as did Abstracts C. Tt is expected that it will be possible to extend its usefulness to readers as time passes. Analytical Abstracts will appear each month bound separately from ?'he Analyst and in The two journals, The Analyst and Analytical Abstracts, will be available in 1954 to members of the Society and to subscribers on the same terms as obtained for The Anal-yst and Abstracts r in 1953. format similar to that of the abstracts included in The Analyst prior to 1950. 629

ISSN:0003-2654    

DOI:10.1039/AN9537800629

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

630 ERSKINE, STROUTS, WALLEY AND LAZARUS A SIMPLE VOLUMETRIC [VOl. 78 PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS A Simple Volumetric Method for the Routine Determination of Glycerol BY J. W. B. ERSKINE, C. R. N. STROUTS, G. WALLEY AND W. LAZARUS Glycerol is oxidised a t room temperature by a solution of sodium meta- periodate to give formaldehyde and formic acid. The formic acid, after the excess of periodate has been destroyed with ethylene glycol, is titrated with standard alkali, with phenol red as indicator, to give a measure of the glycerol present. Carbon dioxide must be excluded, and the oxidation must be carried out in the dark to minimise side reactions between sodium meta- periodate and the formaldehyde and formic acid. A number of other polyhydric alcohols and sugars interfere, but these are not likely to be present in the glycerol-containing products of soap-making and fat-splitting, such as crude glycerine, lye and soap.The most likely organic impurities in these products are trimethylene glycol, which does not react, and polyglycerols, which do not yield formic acid; these substances, therefore, do not interfere as they do in the acetin and dichromate methods. Glycols with adjacent hydroxyl groups, like ethylene glycol, react, but produce only formaldehyde. The method has been checked against distilled glycerine, and against soap-lye crude glycerine, soap lyes and soaps containing known added amounts of glycerol, and also on commercial crudes, soap lyes and soaps by comparison with the acetin and dichromate methods.DURING the last forty years two standard methods* have been available for the determination of glycerol in crude glycerine, namely, the dichromate and the acetin methods. The former measures the total oxidisable matter and is, therefore, unsuitable when the glycerine contains other oxidisable matter not easily separated from the glycerol. The acetin method is more specific and measures the hydroxyl content of the product. The most likely impurities in commercial glycerine that would be returned as glycerol by this method are polyglycerols, for which a correction can be made, and trimethylene glycol,? for which no correction is usually made, although a method1 exists for its determination. In recent years new methods have been suggested for the determination of glycerol, and of these the most important is that based on oxidation by periodate.Glycerol is oxidised by periodates to give formaldehyde and formic acid; under certain conditions the amount of formic acid produced gives a measure of the glycerol present. Other hydroxy bodies that may be present in glycerine include glycols and polyglycerols, but it has been shown that these compounds do not react with periodates to give formic acid. Polyhydric alcohols with three or more adjacent hydroxyl groups, e.g., sorbitol and sugars, yield formic acid, which would interfere with the determination. They are not likely to be present in crude glycerine, soap lyes or soaps. For practical purposes the periodate oxidation is specific for glycerol in these products. Methods based on oxidation with sodium metaperiodate have been developed for glycerine by Bradford, Pohle, Gunther and Mehlenbacher2 and the American Oil Chemists’ Society (A.0.C.S.).3*4 Although the latest A.O.C.S.method (Ea. 6-51) represents a notable advance, it has certain disadvantages, the most important being the need to titrate to two different end-points, the dependence of the result on the weight of sample used in the determination and (for some laboratories) the necessity of using a pH meter. These disadvantages are absent from the method now described, which, based on extensive work carried out in the laboratories of Imperial Chemical Industries Ltd. (Nobel Division) and of Unilever Limited, consists in a simple acidimetric titration.* Analysis of Crude Glycerine, International Standard Methods, 191 1, and Supplement No. 1 (I.S.M., One per cent. of trimethylene glycol is returned as 0-8 per cent. of glycerol by the acetin method. 191 1) British Executive Committee, 1914. Referred to in this paper as “standard methods.”Nov., 19531 METHOD FOR THE ROUTINE DETERMINATION OF GLYCEROL EXPERIMENTAL INVESTIGATION OF THE CONDITIONS AFFECTING THE DETERMINATION OF GLYCEROL BY SODIUM The e$ect of light an,d tem$eruhre--In early experiments with the A.O.C.S. method, erratic results were sometimes obtained, especially with distilled glycerine ; this was attributed to the various degrees of exposure to light between different experiments. I t was also found in the course of experiments on a modification of the Newburger and Bruening methodJ5 which was ultimately developed into the method described later, that, in daylight, the result was affected by the temperature at which the reaction was carried out.A series of experiments was then carried out in which a distilled glycerine* was oxidised with sodium metaperiodate according to the procedure described on p. 635, under various conditions of temperature and illumination. The results are shown in Table I and indicate that (2) the glycerol content was accurately determined when the oxidation was carried out in the dark either at 0" C or at 25" C, (ii) that the amount of exposure to light had an appreciable effect on the amount of glycerol found, a result over 5 per cent. too high being recorded in bright sunlight at 25" C, and (iii) that at 0" C the effect of light, although not so marked, was still appreciable.63 1 METAPERIODATE OXIDATION- TABLE I ?'HE EFFECT OF LIGHT ON THE DETERMINATION OF GLYCEROL BY SODIUM METAPERIODATE Separate weighings of distilled glycerine (90.25 per cent. of glycerol) were used in each determination by the method described on p. 635 Time of oxidation = 1 hour Glycerol found , I In the dark at : In daylight a t : -7 % Yo 0" c, 25" c, 0" c, 25" C, Daylight conditions 90.30 90.30 90.24 90.40 90.31 90.74 Dull, indirect 90.32 90.26 90-94 95.68 Sunny, direct 90.20 90.46 90- 16 90.52 Dull, indirect 90.12 90.36 90.32 90.16 Mean 90.25 90.32 90.86 91.76 Sunny, indirect % % - - - - - - Experiments were also carried out to study the action of sodium metaperiodate on the products of the glycerol oxidation, formic acid and formaldehyde, as it was probable that these side reactions caused the anomalous results found in daylight.Portions of solutions containing known amounts of formic acid and formaldehyde were treated with sodium metaperiodate under various conditions of light and temperature, and the formic acid present at the end of the reactions was measured. The amounts of formic acid and formaldehyde taken were roughly equivalent to those that would be present at the end of a determination in which the amounts of sample and reagents recommended in the method on p. 635 were used. The amount of sodium periodate was approximately equal to the excess of oxidant remaining at the end of such a determination.The results are shown in Table I1 and indicate that (i) sodium metaperiodate reacts with both formic acid and formaldehyde, (ii) the reactions proceed slowly in the dark but are greatly accelerated by daylight, and (iii) between 0" and 20" C, temperature has no appreciable effect when the oxidation takes place in the dark. Head and HughesJ6 who also carried out experiments on the action of sodium periodate on formic acid and formaldehyde, report similar findings. Experiments 17 to 21 show that the presence of sodium iodate reduces the oxidation of formic acid and formaldehyde. As iodate is formed in the oxidation of glycerol, the main reaction, the influence of these side reactions is in practice less than would be expected from the earlier results shown in this table.* A multiple-distilled glycerine, containing no impurity but water, whose glycerol content had been accurately determined from specific gravity and moisture determinations.632 ERSKINE, STROUTS, WALLEY AND LAZARUS: A SIMPLE VOLUMETRIC TABLE I1 IN DAYLIGHT AND DARKNESS Ar 0" AND 20" C Mixture titrated with 0.1 N sodium hydroxide after destruction of excess of periodate with ethylene glycol Time of oxidation == 1 hour [Vol. 78 THE OXIDATION OF FORMIC ACID AND FORMALDEHYDE BY SODIUM METAPERIODATE Expt. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 A Lighting Temp- con- HCHO erature, ditions added, "C g Nil 0.6 0 In the Nil 0.6 20 Nil 0.6 %O 20 0.ti 0.6 i:} Dduailight Nil 20 0.6 Nil 0.6 20 0.6 Nil ;:\ Bright 0.6 20 daylight 0.6 20 Nil 20 J 0.6 0.6 0.6 dark I3 G HCOOH added, g 0.4257 Nil 0.4257 0.4257 0.4257 Nil 0.4257 Nil 0-4257 0.4257 Nil 0.4174 0.4174 Nil 0.4174 0.4174 Nil 0.4174 0.4174 0.4256 0.4256 NalO, added, R Nil Nil 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1-5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 U NalO, added, g Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil 4.0 4.0 4.0 4.0 4.0 li HCOOH found, g 0.4257 0*0003 0.4264 0.4259 0.4258 0.0002 0.4263 0.0127 0.4093 0.4354 Nil 0.4187 0.4194 0.0360 0.3667 0-4407 0*0090 0.4140 0.4214 0.4255 0.4254 1; HCOOH found, as glycerol, g 0.8514 0*0006 0.8524 0.8518 0,8516 0.0004 0.8526 0.0254 0.8186 0.8708 Nil 0.8374 0.8388 0.0720 0.7334 0.8814 0.0180 0.8258 0.8428 0.8510 0.8508 Difference E-B, g + 0.0003 -1 0.0005 -4- 0.0002 j 0~0001 to*ooo2 - t- 0~000G -t 0.0 127 -0.0164 - 10.0097 Nil +0*0013 -t 0*0020 + 0.0360 - 0.0507 10.0233 + 0.0090 - 0.0025 + 0.0040 - 0*0001 - 0~0002 Glycerol error, % + 0.07 -t0*12 + 0.05 + 0.02 -1- 0.Ob $0.14 + 2.98 - 3.85 + 2.28 Nil - $- 0.3 1 4- 0.48 + 8.47 - 12.15 + 5-58 3-2-12 - 0.60 + 0.96 - 0.02 - 0.04 n7oTE-h experiments 2, 6, 8, 14 and 17 the glycerol error has been calculated on the assumption that a sample weight of 0.85 g of glycerol was taken.TABLE I11 TIME OF OXIDATION AND EFFECT OF SODIUM CHLORIDE Separate weights (about 0-8 g) of glycerine taken for each test and determination carried out by procedure given on p. 635 Time of oxidation, minutes 60 gravity measurement 60 30 60 60 30 30 60 30 Soap lye crude 1 . . .. .. .. 60 30 Distilled glycerol, 90.26 per cent. by specific Soap lye crude 2 .. .. .. .. 60 30 Sodium chloride added, Nil Nil Nil 1 1 1 1 2 2 Nil Nil Nil Nil Glycerol found, 90.20 90.36 90.35 90-24 90.24 90.38 90.40 90.28 00.25 84.27 84.30 82.47 82.44 Yo THE TIME OF OXIDATION AND THE EFFECT OF SODIUM CHLORIDE- Weighed quantities of distilled glycerine were treated with sodium metaperiodate according to the method described on p. 635, the oxidation being allowed to proceed for 30 minutes or 1 hour. Bradford and his co-workers2 have stated that sodium chloride retards the oxidation, so, A study was made of the time required for the oxidation to go to completion.Nov., 19531 METHOD FOR THE ROUTINE DETERMINATION OF GLYCEROL 633 as soap-lye crudes contain about 10 per cent. of sodium chloride, some tests were also carried out in which large quantities of sodium chloride (relative to the amount of the glycerol) were added.The results, shown in Table 111, indicate that under the conditions used, an oxidation time of 30 minutes is sufficient, even in the presence of much sodium chloride. Some results for crude glycerines are also shown. THE EFFECT OF TRIMETHYLENE GLYCOL- the presence of trimethylene glycol. glycol added, but it was found that it had no effect under the test conditions. According to Bradford et ~ 1 . ~ the periodate oxidation gives low figures for glycerol in Several determinations were made with trimethylene THE EFFECT OF CARBON DIOXIDE- It was found that the titration was affected by atmospheric carbon dioxide (or other acidic gases) and that the sharpness of the end-point was considerably increased by passing a slow stream of nitrogen (freed from carbon dioxide) through the solution being titrated.I t was also found that the passage of nitrogen through the solution caused no volatilisation of formic acid, as shown by the figures in Table IV. TABLE Iv NON-VOLATILITY OF FORMIC ACID UNDER THE CONDITIONS OF THE TEST Formic acid titrated with 0.1 N alkali. Volume of solution at the start: 200 ml A R C Total time Loss of of passage of Formic acid Formic acid formic acid Loss of minutes g g g % nitrogen, present, found, ( A - B), formic acid, 0 0.4632 0-4633 Nil Nil 10 0-4632 0.4632 Nil Nil 30 0.4632 0.463 1 0~0001 0.02 0 0.4404 0.4404 Nil Nil 10 0.4404 0.4404 Nil Nil 30 0.4404 0.4402 0~0002 0.05 ,. I H E DETERMINATION OF GLYCEROL I N PURE GLYCERINES, CRUDES, SOAP LYES AND SOAP BY THE PERIODATE PROCEDURE- Pure gZycerines-Tests that were carried out on a pure multiple-distilled glycerine (see Table I, columns 1 and 2) show excellent agreement between the figures for glycerol by the periodate method and those from specific gravity and moisture determinations. Similar results were attained with other samples of pure distilled glycerine. Experiments were also carried out on glyceryl sesquicarbonate, a crystalline derivative that was prepared in a state of high purity; the results for glycerol, after hydrolysis, were in excellent agreement with elementary analysis. Mixtures containing known amounts of glycerine-In order to check the accuracy of the periodate method for glycerol in soap-lye crudes and to show whether the method could be used for soap lye and soap, it was tested on made-up soap-lye crudes, soap lyes and soaps containing known amounts of glycerol.The glycerol was determined by the procedure described on p. 635. As the glycerol content of soap is low, some tests were carried out in which the amount of glycerol to be determined was reduced to one-tenth, with a corresponding reduction in the reagents and volumes; a 10-ml micro-burette was used for the titration. The satisfactory recoveries of glycerol shown in Table V indicate (2) that the glycerol in crudes is accurately determined with no interference from trimethylene glycol and polyglycerol, and (ii) that the method can be applied to soap lyes and soaps (after acidifying, cooling and filtering to remove fatty acids). When the amounts taken for the determination are reduced to one-tenth, the precision is reduced but the results are still satisfactory.CommerciaZ firoducts-Table VI shows results by the periodate method for glycerol in industrial soap-lye crudes, saponification crudes, soap lyes and soap, together with com- parison results by the standard acetin and dichromate methods. The figures indicate that there is good agreement between the periodate method and the standard methods.634 TABLE V REPLICATE ANALYSES OF MADE-UP SOAP-LYE CRUDES, BY THE PERIODATE METHOD SOAP LYES AND SOAPS Soap lye Glycerol Type of product taken, g Soap-lye crude.. . . 0.9210 0.9238 0.8413 0.8999 0.8904 0.9276 . . . . 0.8694 0.8642 0.9174 0.8575 0.8956 0.8320 0.8928 0.9333 0.8979 0.9490 0.9286 0.9327 0.0968 0.0968 0.0888 0.0888 0.0888 ERSKINE, STROUTS, WALLEY AND LAZARUS: A SIMPLE VOLUMETRIC [Vol.78 ANALYSES OF COMMERCIAL SOAP-LYE CRUDES, SAPONIFICATION CRUDES, SOAP Soap (fatty acids removed before determination) Poly- glycerol taken, g 0.005 0.005 0.006 0-005 0.005 0.005 Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Tri- methylene glycol taken, g 0.013 0.013 0.013 0.013 0-013 0.013 Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Sodium chloride taken, g 0.1 0.1 0.1 0.1 0-1 0.1 1.79 1-79 1.64 1-64 1-64 1-64 Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Nil Soap taken, g Nil Nil Nil Nil Nil Nil 0.14 0.14 0-13 0.13 0.13 0.13 160-0 160.0 160.0 160.0 160.0 160.0 10.0 10.0 10.0 10.0 10.0 Glycerol found, g 0-9195 0.9245 0.8406 0.8999 0.8899 0.9249 0.8673 0.8650 0.9144 0.8542 0-8910 0.8320 0.8775 0.9284 0.8877 0.9310 0.9140 0.9220 0.0962 0.0960 0.0891 0.0904 0.0893 Recovery, % 99.84 100.08 99.92 100~00 99-94 99-71 99.76 100.10 99.70 99-62 99-50 100*00 98.3 99.5 98.9 98.1 98.4 98-9 99.4 99.2 100-3 101-8 100.6 TABLE VI LYES AND SOAPS BY THE Type of product 79 2 .. .. 3 . . .. 77 4 . . . . 97 5 . . .. 6 . . . . Saponification crude* 1 . . 97 2 .. 7 9 3 . . 77 1 .. ¶¶ 5 .. 97 6 .. 99 2 . . .. 77 3 .. .. 99 4 .. .. 7) 5 .. .. 9 ) 6 .. .. 9) 2 .. .. .. 99 3 .. .. .. Soap-lye crude* 1 . . . . 7, 39 Soap lye 1 . . .. Soapt 1 . . .. .. * Corrected .. .. .. . . .. .. .. . . . . .. .. .. . . .. . . .. * . .. .. * . PERIODATE AND STANDARD METHODS Glycerol, % Acetin Dichromate Periodate method method method 84.29 - 84.27 82.44 - 82-53 83.83 - 83-83 82.36 - 82-47 82.16 - 81.92 83.23 - 83.09 91.94 - 91.90 91.80 - 92.09 86.62 - 86-53 93-03 - 93.06 89.93 - 90.00 89.17 - 89.03 6-30 6.40 6-35 6-18 - 6.19 6-10 6.37 6.29 - 9.50 9.48 - 8.99 8.84 - 9.09 8-92 - 0.33 0.28 - 0.46 0.41 - 0.42 0.40 A r -7 for total acetvl value of residue.t Glycerol expressed on total fatty matter. NOTE-The figures for soap-lye crudes and saponification crudes have been analysed statistically, and the results show that there is no significant difference between results by the two methods.Nov., 19531 METHOD FOR THE ROUTINE DETERMINATION OF GLYCEROL 636 METHOD PREPARATION AND STANDARDISATION OF REAGENTS- Sodium metaperiodate*-Dissolve 228 g of pure periodic acid , which may conveniently be prepared electrolytically,7 in 1 litre of water contained in a 3-litre beaker.Add slowly with constant stirring 1 litre of a M solution of sodium hydroxide. Transfer the contents of the beaker to a 3-litre round-bottomed flask, place it in a water-bath, and distil off the water under reduced pressure until the volume of liquid that remains is about 250 ml. Allow the solution to cool, filter off the sodium metaperiodate on a sintered-glass funnel, porosity grade 3, and dry the salt overnight at 100" to 110" C. The yield is 80 per cent., calculated on the periodic acid, and its solubility in water is 10 g per 100 ml at 20" C. Dissolve 0.36 g of the salt in 100 ml of water contained in a flask with a ground-glass stopper.Add 7-5 g of analytical reagent grade sodium bicarbonate followed by 3 g of analytical reagent grade potassium iodide. Titrate the solution with 0.1 N sodium arsenite solution that has been standardised against 0.1 N iodine immediately before use. Determine the composition of the material as follows. Stopper the flask and set the mixture aside for exactly 15 minutes. 1 ml of 0.1 N sodium arsenite = 0*010696 g of NaIO,. Material made in the manner described will have a purity of 99.8 per cent. or more. Sodium metaperiodate solution, 3.33 per cent. w/v-Measure 150 ml of water into a 1-litre conical flask and boil to expel carbon dioxide. Close the flask with a rubber bung carrying a soda-lime guard tube, swirl the contents of the flask to dissolve the salt, and cool the flask in running water.Larger quantities of the reagent may be made up pro rata, but the solution is not stable and must be freshly prepared each day. Sodium hydroxide solution, 0.1 N-In order to avoid indicator errors it is essential that the sodium hydroxide used to titrate the formic acid shall be standardised against 0.1 N hydrochloric acid itself standardised against pure anhydrous sodium carbonate with phenol red as indicator. Full details of the standardisation of the hydrochloric acid are given in the literature.* In this titration it is necessary to add an excess of acid to the alkali, then to boil to expel carbon dioxide and finally to titrate the residual acid. From a 105-ml bulb-burette, measure into a hard-glass flask, containing 50 ml of water, 85 ml of 0.1 N hydrochloric acid standardised against pure anhydrous sodium carbonate with phenol red indicator.Boil the solution for a few minutes to expel carbon dioxide. Close the flask with a rubber stopper carrying a soda-lime guard tube and cool to room temperature. Wash down the sides of the flask with a little water free from carbon dioxide, add 0.2 ml of phenol red indicator and titrate the acid with the 0.1 N sodium hydroxide solution to a pale pink end-point while passing a slow stream of nitrogen free from carbon dioxide through the solution. The flow of gas should be regulated to avoid loss of solution as spray. From the corrected volumes of acid and alkali, calculate the factor for the 0.1 N sodium hydroxide. Determine the glycerol equivalent of the reagent at intervals (or for each batch of reagent) by means of standard glycerol, the concentration of which is known from its specific gravity9 and its moisture content.Add 5 g of sodium metaperiodate. DETERMINATION OF GLYCEROL- Weigh into a dry 500-ml conical flask, a sample portion containing 0.7 to 0.8 g of glycerol.? Wash down the sides of the flask with water and then add 50 ml of water and 1 ml of M hydro- chloric acid. Boil the solution for 3 minutes to expel carbon dioxide, close the flask with a rubber stopper carrying a soda-lime guard tube and cool in running water to room temperature. Add 0-2 ml of phenol red indicator and neutralise the solution with 0.1 N sodium hydroxide solution standardised to the same indicator.: Add 150 ml of a 3.33 per cent. w/v solution of sodium metaperiodate, swirl, stopper the flask again and allow it to stand in the dark for * We are indebted to Messrs.K. Sporek and J. Templeton for the preparation and analysis of pure At the time this work was carried out the material available commercially was t For the determination of glycerol in soaps (after removal of fatty acids), a sample containing one- Use a 6 or $ The colour at the approach to the end-point is a pale reddish-brown: the titration must therefore sodium metaperiodate. not sufficiently pure, but supplies at the requisite standard of purity can now be obtained. tenth of this amount may be taken, with a corresponding reduction in reagents and volumes. 10-ml micro-burette for the titration. be continued until the definite pink colour of the indicator is seen.636 MORRIES AND STUCKEY: THE CHROMATOGRAPHIC DETERMINATION OF [VOl.78 30 minutes. Wash down the sides of the flask with carbon dioxide-free water, add 5 ml of redistilled ethylene glycol and set the mixture aside in the dark for 20 minutes. Titrate the liberated formic acid from a 90-ml bulbed burette with 0.1 N sodium hydroxide standardised as described above; pass a stream of nitrogen free from carbon dioxide through the solution at such a rate that no liquid is lost by spraying. The end-point is indicated by a definite pink colour.* A determination can be completed in an hour and a half. Since this work was submitted for publication a paper by L. Hartman has appeared (1. A$@. Chem., 1953, 3,303) describing a “Rapid Determination of Glycerol by the Potassium Yeriodnt e Met hod.” REFERENCES Carry out a blank titration at the same time. 1 ml of 0.1 N NaOH = 9.209 mg of C3H,0,. 1. 2. 3. 4. 5. ti. 7. 8. 9. Cocks, L. V., and Salway, A. H., J . SOC. Chem. Ind., 1922, 41, 1 7 ~ . Bradford, P., Pohle, W. D., Gunther, J. K., and Mehlenbacher, V. C., Oil G. Soup, 1942, 19, 189. “Report of the Glycerin Analysis Committee,” ,S)ctober, 1946, J . Amer. Oil Chenz. SOC., 1947,24, 18. “Report of the Glycerin Analysis Committee, Newburger, S. H., and Bruening, C. F., J . Ass. 08. Agric. Chern., 1947, 30, 651. Head, F. S. H., and Hughes, G., J . Chem. SOC., 1952, 2046. Willard, H. H., and Ralston, R. R., Trans. Electrochem. SOC., 1932, 62, 239. Analytical Chemists Committee of Imperial Chemical Industries Ltd., Analyst, 1950, 75, 577. Bosart, L. W., and Snoddy, A. O., I n d . Eng. Chenz., 1927, 19, 506. September, 1950, Ibid., 1950, 27, 412. IMPERIAL CHEMICAL INDUSTRIES LIMITED NOBEL DIVISION STEVENSTON, AYRSHIRE RESEARCH DEPARTMENT UNILEVER LIMITED PORT SUNLIGHT, CHESHIRE May 21st, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800630

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

636 MORRIES AND STUCKEY: THE CHROMATOGRAPHIC DETERMINATION OF [VOl. 78 The Chromatographic Determination of Glutamic Acid in Wheat Gluten and Gluten Hydrolysates BY P. MORRIES AND R. E. STUCKEY A method has been devised for the determination of glutamic acid in gluten and in gluten hydrolysates. Total dicarboxylic acids (glutamic acid and aspartic acid) are determined by adsorption on Amberlite IR-4B resin a t a pH value of 3 to 4 and subsequent elution with N hydrochloric acid; the amount of aspartic acid present is then found by means of descending paper- strip chromatography with a phenol - water system. The paper chromato- graphy of gluten hydrolysates has been investigated with phenol, collidine and isobutyric acid as the mobile phase; the precision of the estimation of a single spot on a water - isobutyric acid chromatogram was found to be -& 12 per cent.(P = 0-95). MANY of the available data relating to the glutamic acid content of proteins are based on the work of Jones and Moeller,l which involves the precipitation of the calcium or barium salts of glutamic acid from strong aqueous alcohol. This procedure is somewhat lengthy, however, and cannot readily be adapted for routine use; in addition, Chibnall, Rees, Bailey and Williams,2 who subjected the method to an extensive critical examination, stressed the necessity for solubility corrections. It was thought that chromatographic procedures for the estimation of glutamic acid were worth investigating in the hope of evolving a method more convenient for routine evaluation of gluten hydrolysates.The methods available for the chromatography of the dicarboxylic amino-acids fall into two groups: first, the traditional chromatographic techniques on alumina or, more recently, on ion-exchange materials, and, secondly, partition chromatography on columns of starch or silica gel or on paper strips. Both methods were investigated for their suitability as quantitative analytical procedures and to find a satisfactory combination of chromato- graphic methods. * The colour at the approach to the end-point is a pale reddish-brown; the titration must therefore be continued until the definite pink colour of the indicator is seen.NOV., 19531 GLUTAMIC ACID IN WHEAT GLUTEN AND GLUTEN HYDROLYSATES QUANTITATIVE PAPER CHROMATOGRAPHY 637 Many recent methods for the analysis of mixtures of amino-acids have involved the paper chromatographic techniques developed by Consden, Gordon and Martin.3 We have used their method for some time for the routine separation and examination of amino-acids and consider it possible to determine glutamic acid in wheat gluten hydrolysates by means of a similar technique.The method used in the work reported here is based on that suggested by Atkinson, Stuart and Stuckey.4 Developing solvents were removed from the papers by washing with ether rather than by oven-drying, as Fowden and Penny5 have shown that drying at elevated temperatures causes serious losses of amino-acids. The hydrolysates to be examined were produced by hydrochloric acid digestion of wheat gluten; the excess of acid was removed by treatment with silver oxide.The amounts of gluten hydrolysates used were adjusted to contain 20 to 40pg of glutamic acid for each chromatogram. PROCEDURE- Weigh about 1 g of wheat gluten and digest it with 10 ml of concentrated hydrochloric acid on a bath of boiling water for 4 hours under a reflux condenser. Cool, dilute to 25 ml and neutralise the excess of acid by adding dry silver oxide a little a t a time, mixing thoroughly after each addition; when neutralised to a pH value of approximately 7, as determined by indicator papers, allow the precipitate to settle and place 2p1 of the supernatant liquid on a strip of Whatman No. 1 filter-paper by means of an “Agla” micrometer syringe. Place the filter-paper strip in a chromatograph chamber containing an atmosphere saturated with water and the solvent used (usually isobutyric acid or phenol), develop the chromatogram (descending) overnight and remove the developing solvent by washing with ether and drying a t room temperature.Spray the chromatogram with a 1 per cent. solution of ninhydrin in n-butanol containing 10 per cent. of pyridine and heat in an oven at 80” C for 30 minutes. Cut out the appropriate spots and elute with 5 ml of a 20 per cent. solution of pyridine in water. Gentle swirling in a test tube is usually sufficient to extract the colour within 5 minutes; if necessary, remove fibres by centrifugation. Measure the absorption a t 570 mp on a suitable spectrophotometer and compare the results with those obtained from extracts prepared similarly from standard solutions of glutamic acid run side by side on thesame sheet of paper.RESULTS- In the first series of quantitative experiments a mixture of isobutyric acid and water was used to attain separation. The spot corresponding to the RF value of glutamic acid (0.38 in isobutyric acid by comparison) was assumed to contain glutamic acid only, in order to determine the precision of the estimation. The results shown in Table I were obtained on a single gluten hydrolysate by comparison with a solution of glutamic acid as standard. TABLE I THE PRECISION OF GLUTAMIC ACID ESTIMATION BY PAPER CHROMATOGRAPHY Number of Standard Material results Maximum, Minimum, Average, deviation Error, % % % Yo Gluten hydrolysates (single Solution of amino-acids spot) .. .. .. 16 14.5 11-4 12.5 0.737 (P = 0.95) 12 (double spots) . . . . 1% 1.34 1.23 1-25 0.036 (P = 0.98) 7 (P = 0.96) 6 A study of the RF values indicated that phenol was the most likely solvent to effect separation of glutamic acid from other amino-acids. But the determination of glutamic acid by the above method in a synthetic mixture of amino-acids approximating in composition to a wheat gluten hydrolysate gave results about 5 per cent. too high. This interference was thought to be caused chiefly by serine; this was confirmed with a two-dimensional chromato- gram and phenol and collidine - lutidine solutions as the solvents. I t is possible that this638 MORRIES AND STUCKEY: THE CHROMATOGRAPHIC DETERMINATION OF [Vol. 78 interference could be eliminated in a one-dimensional chromatogram with a phenol - ammonium hydroxide solvent, but with this solvent high paper-blanks were obtained for the ninhydrin reaction.Although the two-dimensional paper chromatogram was reasonably satisfactory, it was hoped that a simpler method would be found for routine use. Fowden6 published a method that avoids this paper blank, but these difficulties and the doubtful accuracy of the procedure for quantitative work encouraged a search for other methods. ION-EXCHANGE RESIN TECHNIQUES Glutamic acid is one of the few dicarboxylic amino-acids present in protein hydrolysates, the only other of importance being aspartic acid. A study of the chromatogram with phenol and collidine - lutidine showed that aspartic acid gave a well-defined spot not contaminated by other amino-acids when phenol was used as solvent in a one-dimensional chromatogram.Ion-exchange resins have been used by a number of workers for the separation of the di- carboxylic amino-acids. Therefore, if a procedure were available for the determination of total dibasic acids (glutamic acid plus aspartic acid), the aspartic acid could then be deter- mined by paper chromatography and the glutamic acid found by difference. Several schemes for the separation of amino-acids by ion-exchange materials have been published; those for the separation of the dicarboxylic amino-acids depend on the use of a weakly basic anion-exchanger. Cannan7 determined the dicarboxylic acids in several proteins by treating them repeatedly with Amberlite I R 4 resin in a flask.Tiselius, Drake and HagdahP used a column technique and estimated the amino-acids in the eluates by means of a Kjeldahl nitrogen determination. Partridge and Brimleyg separated glutamic acid from mixtures of other amino-acids with Ile-Acidite B, their yield of pure glutamic acid being about 82 per cent. of the theoretical. Drakelo effected the separation of glutamic acid and aspartic acid on columns of Amberlite IR-4 resin, using a Kjeldahl nitrogen estimation to determine the glutamic acid in the eluates. Consden, Gordon and Martin,ll also using I R 4 resin, described a method of determining glutamic acid and aspartic acids in proteins and their subsequent separation ; they determined amino-nitrogen in the eluates by the copper method of Pope and Stevens,12 but failed to obtain quantitative recoveries of amino-acids from synthetic mixtures, and attributed this to the presence of formaldehyde from the resin in the eluates.In general, either recoveries were not quantitative or the methods could not easily be adapted to routine analytical estimations. In the separation of the dicarboxylic amino-acids by means of a weakly basic anion- exchange resin, the hydrolysate is run down a column of the chloride form of the resin; the dicarboxylic amino-acids are retained on the column and the mono-amino mono- carboxylic acids and the more basic amino-acids pass through. The analytical grade of Amberlite IR-4B was used in this work (the nearest equivalent in the Zeocarb series is De-Acidite E), the principle of the method being that of Consden, Gordon and Martin,ll who used IR-4.The excess of hydrochloric acid in the gluten hydrolysate was removed by treatment with silver oxide and filtration instead of repeated evaporation in vacuo. No difficulties from formaldehyde liberated from the resin were encountered, but in order to attain quantitative recovery it was essential to have the particle size of the resin sufficiently small for the concentrations used, otherwise the rate of flow through the column would need to be inconveniently slow. PROCEDURE- Reduce Amberlite I R 4 B resin, which is usually supplied as the base with a B.S.S. mesh size of 16 to 50, to 40 to 60 mesh by grinding gently under water. With it still under water, add dilute hydrochloric acid until the pH of the liquid is less than 2 and wash with water by decantation until the pH of the supernatant liquid is greater than 3.Transfer the resin as a slurry to a glass column of 1 cm internal diameter to a height of 15 cm and wash it with sufficient water to ensure that the pH of the washings is between 3 and 4. Do not allow the column to become dry at any stage during the procedure or channels will form. Dilute a suitable amount of the gluten hydrolysate to a measured volume so that it contains about 1 per cent. of the dicarboxylic amino-acids. Neutralise the solution to a pH of 3 to 4 with silver oxide and filter through a dry filter-paper. Transfer 5ml of the filtrate by means of a pipette to the top of the resin column and allow it to run through at a rate of 1 to 2 ml per minute. Wash the column with water until a portion of the washings no longer gives any reaction for amino-acids with ninhydrin.This usually occurs after aboutKov., 19531 GLUTAMIC ACID IN WHEAT GLUTEN AND GLUTEN HYDROLYSATES 630 40 ml of washings have been collected. Elute the column with about 100 ml of N hydro- chloric acid and evaporate the eluate to small bulk in uacz~o. Determine the amino-nitrogen by the copper method or the total nitrogen by a normal Kjeldahl estimation. When working with the liquors derived from gluten hydrolysates after the extraction of glutamic acid-they usually contain considerable quantities of ammonium chloride-it was found necessary to use a smaller quantity of sample, otherwise separation of amino-acids on the column was not clean.A paper chromatogram was run concurrently with the resin separation, and the aspartic acid was estimated by the method referred to above (p. 637). The amino-nitrogen from the aspartic acid is subtracted from the total dicarboxylic amino-nitrogen and the result calculated to glutamic acid. RESULTS- The method was first applied to the analysis of known amino-acid mixtures approximating in composition to gluten hydrolysates and containing glutamic acid, aspartic acid, cystine, alanine, serine, glycine, threonine, lysine, arginine, histidine, methionine, tryptophan, tyrosine, valine, leucine and phenyl alanine. Table I1 shows the results obtained. TABLE I1 RECOVERY OF TOTAL DICARBOXYLIC ACIDS FROM SYNTHETIC AMINO-ACID MIXTURES Glutamic acid and aspartic acid taken (calculated as glu tamic acid), mg 53.6 54.6 53.4 50.8 Glutamic acid and aspartic acid recovered (calculated as glutamic acid) A f 3 By amino- Total nitrogen method, nitrogen, Recovery, mg mg Y O 54.7 - 102.2 54.0 - 98.8 { 1;::: - 51.1 100.6 { :;:: - Recoveries of the combined glutamic acid and aspartic acid were good both by amino- A blank determination of total nitrogen nitrogen and by total nitrogen determinations.was negligible. TABLE I11 DETERMINATION OF TOTAL DICARBOXYLIC ACIDS IN GLUTEN HYDROLYSATES Glutamic acid Material added, % Gluten hydrolysate (1) . . - Gluten hydrolysate (2) . . - Glutamic acid liquors . . - 3.94 8-06 2-17 6-15 6-12 Pope and Stevens amino-nitrogen method Glutamic acid and aspartic Recovery of acid calculated added glutamic -- as glutamic acid, acid, % % 7-85 - 11-85 101.6 15.59 100.5 9-05 - 11.15 97.0 15.1 98.4 3.3 - 9.35 98.5 Glutamic acid and aspartic acid calculated as glutamic acid from total nitrogen, 9-7 % - - 9.9 - - 6.9 - Table I11 shows results obtained with gluten hydrolysates and with glutamic acid liquors.Recoveries of added glutamic acid were good, the results in each instance being for total dibasic acids calculated as glutamic acid. The total nitrogen retained on I R 4 B resin always exceeded the amino-nitrogen with actual hydrolysates and liquors, indicating the value of a method for determining glutamic acid by an cc-amino-nitrogen estimation in the eluates, rather than by determining total nitrogen, as some workers do. In this connection the comments of Fromageot and Cola+ are of interest; they point out that the de-amination640 MORRIES AND STUCKEY [Val.78 of certain amino-acids, particularly tryptophan , results in the formation of acidic nitrogenous substances that would be retained on I R 4 B resin. TABLE IV GLUTAMIC ACID AND ASPARTIC ACID CONTENT O F GLUTENS Total dicarboxylic amino-acids (calculated as glutamic Aspartic acid Glutamic acid Protein Glutamic acid Material acid) , in gluten, in gluten, (N x 5.7), in protein, YO % % Yo Y O Maize gluten . . 21.7 1.5 Wheat gluten (1) . . 29.2 2.5 Wheat gluten (2) . . 32.4 1.9 Wheat gluten (3) . . 32-4 2.2 Wheat gluten (4) . . 25.5 2.3 Wheat gluten (5) . . 29.5 2.3 20.0 70.5 28.3 26.4 68-8 38-4 30.2 79.6 38.0 30-4 71.8 42.3 23.2 64.2 36.1 27.2 65.3 41.6 Table IV shows some results for glutens.That the estimation of aspartic acid by paper chromatography should suffer from greater inaccuracies than the estimation of total dicarboxylic amino-acids is not a serious disadvantage in dealing with proteins like those in gluten, in which the amount of aspartic acid is fairly small compared with the amount of glutamic acid. For comparison with previous figures the results of Jones and Moeller may be quoted; these workers found 25 per cent. of glutamic acid in glutenin gnd 44 per cent. in gliadin, these being the chief proteins of gluten. More recently, Miller, Leiffe, Shellenberger and Miller,14 Rice and Ramstead15 and Pence, Mecham, Edder, Lewis, Snell and Alcott,16 using microbiological methods, have reported glutamic acid contents of wheat gluten from 30.4 to 35.5 per cent., calculated to a theoretical gluten containing 17.5 per cent.of nitrogen (corresponding to a nitrogen to protein conversion factor of 5.7). The results quoted in Table IV are somewhat higher than these, although the samples examined were reputed to have a high gliadin content. The authors wish to thank the Directors of The British Drug Houses Ltd. for permission to publish these results. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Drake, B., Nature, 1947, 160, 602. 11. 12. 13. 14. 15. 16. Jones, D. B., and Moeller, O., J . Biol. Chem., 1928, 79, 429. Chibnall, A. C., Bailey, K., Rees, M. W., and Williams, E. F., Biochem. J., 1943, 37, 360. Consden, R., Gordon, A. H., and Martin, A. J. P., Ibid., 1944, 38, 224. Atkinson, R. O., Stuart, R. G., and Stuckey, R. E., Analyst, 1950, 75, 447. Fowden, L., and Penny, J. R., Nature, 1950, 165, 846. Fowden, L., Biochem. J., 1951, 48, 327. Cannan, R. K., J . Biol. Chem., 1944, 152, 401. Tiselius, A., Drake, B., and Hagdahl, L., Experimentia, 1947, 3, 15. Partridge, S. M., and Rrimley, R. C., Biochem. J., 1948, 42, 443. Consden, R., Gordon, A. H., and Martin, A. J. P., Biochem. J., 1948, 42, 443. Pope, C. G., and Stevens, M. F., Ibid., 1939, 33, 1070. Fromageot, C., and Colas, R., Biochem. Biophys. Acta, 1949, 3, 417. Miller, B. S., Leiffe, J . Y., Shellenberger, J. A., and Miller, G. D., Cereal Chem., 1950, 27, 96. Rice, A. C., and Ramstead, D. E., Ibid., 1950, 27, 238. Pence, J. W., Mecham, D. K., Edder, G. H., Lewis, J. L., Snell, N. S., and Alcott, H. S., Ibid., 1950, 27, 335. THE BRITISH DRUG HOUSES LTD. GRAHAM STREET, LONDON, N.1 April 9th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800636

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Nov., 19533 DUNCAK AND POHTEOUS 641 The Identification and Determination of the Lower Straight-chain Fatty Acids by Paper Partition Chromatography BY R. E. B. DUNCAN AND J. W. PORTEOUS* An improved method of detecting and determining the lower (C,-C,) straight-chain fatty acids is described ; it involves running paper chromato- grams in n-butanol - ammonia mixtures and spraying with a mixture of methyl red - bromothymol blue indicators in formalin. The critical factors for obtaining reproducible chromatograms applicable to the range 6 to 80 pg of each acid are described. A variation has been noted in the relationship between spot area and the acid concentration with the indicator used. The method has been applied to two problems, volatile fatty acid production in grass fermentations, and the metabolism of A ctinowzyces israeli.THE difficulties underlying the chromatographic separation and identification of the lower volatile fatty acids have been summarised by E1sden.l Since that date several paper partition chromatographic separations of the volatile fatty acids have been devised by Brown,2 Brown and Hall,3 Hiscox and Berridge,4 Kennedy and Barker,5 Miettinen and Virtanens and Reid and Lederer.' All these methods are similar in principle: 2 to 10-p1 spots containing 5 to 1OOpg of each of the fatty acids as their sodium or ammonium salts are applied to the chromatographic paper ; the solvent system is n-butanol, ethanol or both, equilibrated with ammonium hydroxide ; the spots are subsequently revealed by spraying the dried chromato- gram with an acid - base indicator solution; bromothymol blue, bromophenol blue, bromo- cresol green and bromocresol purple solutions adjusted to various pH values have been used.The fatty acid hydroxamates have also been used (Fink and Fink,8 and Thompsong). Of the methods available we have found that of Reid and Lederer' to be the most satisfactory because of its simplicity and quantitative nature. We could not, however, obtain consistent results with less than 20 pg of the individual acids, so this paper is concerned with the systematic elaboration of the critical factors involved in a more reliable technique. METHOD REAGENTS- fraction boiling between 116" and 118°C. n-Butanol-Redistil as required through a 12-inch fractionating column, collecting the Ethanol-Reagent grade.Ammonium hydroxide solutions-(a) 1.5 N ammonium hydroxide. Distil reagent grade ammonium hydroxide, sp.gr. 0.880, into distilled water at 2" C to give an approximately 10 N solution. (b) 3 per cent. ammonium hydroxide. Prepare by diluting reagent grade ammonium hydroxide, sp.gr. 0.880, with tap water. Formaldehyde solution, 40 per cent.-Made from formalin, B.P. Indicator spray-Dissolve 200 mg each of methyl red and bromothymol blue in a mixture of 100 ml of formalin and 400 ml of ethanol. Add 0.1 N sodium hydroxide until a reading of 5.2 is obtained on a glass-electrode pH meter. Fatty acids-Acetic, propionic, n-butyric, n-valeric, n-hexanoic and DL-laCtiC acids. Acetic and DL-lactic acids were of analytical reagent quality and the others of reagent quality in the work described here.Dilute this as required to 1.5 N with freshly distilled water. PROCEDURE- Standard acid solzdions-Prepare separate aqueous solutions of each member of the homologous series, acetic to hexanoic acid, and titrate against the pH meter to a pH value * Present address : The Rowett Research Institute, Bucksburn, Aberdeenshire.642 DUNCAN AND PORTEOUS : THE IDENTIFICATION AND DETERMINATION [Vol. 78 of 8.0 with freshly prepared 1.6 N ammonium hydroxide. Mix these stock solutions and dilute them as required to give solutions in which the individual acid concentrations range from 5 to 80pg per 5 4 . Apply 5-p1 spots of the desired mixed acid solution to the chromatographic paper as described below. The solvent system-Equilibrate n-butanol and 1.5 N ammonium hydroxide by shaking equal volumes of them together.Use 15-ml aliquots of the upper (butanol) layer as the mobile phase for each chromatogram; place an equal volume of the lower (aqueous) layer separately in the tank to maintain equilibrium conditions. Preparation of the chromatographic paper-Wash 11.25-inch by 9-25-inch sheets of Whatman No. 54 chromatographic paper by serial transfer in pairs through three shallow photographic dishes each containing 400ml of the butanol layer of a 1 + 1 butanol and 1.5 N ammonium hydroxide system to which 5 per cent. v/v of ethanol has been added after separation of the phases. Hang the papers individually in a fume cupboard, keeping them out of contact with any metal, until dry, and then weigh them to the nearest 100mg.Running the chromatogram-Arrange the starting points for the chromatograms at intervals of 2-5 cm along a line 6 cm from the bottom of the sheet, with the outer points not less than 5 cm from the sides. Apply spots of standard and unknown solutions from a 5-pl pipette. Transfer the sheets to the mobile (butanol) phase in Petri dishes in the tank, and run the chromatograms by upward displacement for 15 hours at 18” to 20” C, during which time the solvent reaches the top of the paper. Dvying and spraying the pa$ws-Remove individual papers from the tank at 15-minute intervals, weigh them immediately, and hang them to dry at room temperature in still air in a fume cupboard. Reweigh at 10-minute intervals until the papers return to their original dry weight; 10 minutes later spray the paper with the solution of methyl red and bromothymol blue indicator. Deliver the indicator as evenly as possible from an all-glass “atomiser,” with a rotary pump supplying the air blast.A t the first even tinting of the paper faint yellow spots usually show on a pink background. Continue to spray until these spots are obliterated and the paper is evenly damped through to the back. About 30 ml of indicator solution, delivered in approximately 90 seconds, is required for each sheet. Colour development on the chromatogram-Acid spots are shown on the sprayed chromato- gram, which has an intense red colour, by dipping the paper (in cylindrical form) immediately into ammonia vapour contained in a tall cylinder 6 inches in diameter at the bottom of which is a layer of 3 per cent.v/v ammonium hydroxide. It is necessary to expose the paper to the ammonia vapour as evenly as possible, and to develop the alkaline (dark green) colour of the indicator rapidly on the paper. After the paper has been exposed for 2 to 3 seconds to the ammonia vapour, remove the paper, open it out and hold it away from the ammonia for about 1 minute, during which time the acid (red) colour of the indicator returns evenly to the whole paper. Coil the paper in the reverse direction, invert it and again hold it in the ammonia vapour for a few seconds until the alkaline colour reappears as an even background to the acid spots, after which remove the paper and allow the acid colour to return evenly to the whole paper.Repeat this procedure several times, gradually increasing the length of exposure to the ammonia vapour to 10 to 15 seconds as the time taken for the alkaline colour to develop in the ammonia vapour increases. After four or five exposures acid spots fail to appear at all, but reappear with increased intensity in later exposures to the ammonia, and a t the same time the green background becomes more and more stable between exposures. Finally, intense red spots persist on a green background. The chromatograms can be marked at any time within the next 20 minutes. Thereafter the background returns slowly to an even pale brick-red colour; the acid spots remain visible as dark red areas for some time, but over periods ranging from 1 to 24 hours, depending on the concentration of the acid in the spot, the acid spots merge completely into the background.Quantitative analysis-Trace the acid spots on graph paper and measure the areas. The standard acid solutions chromatographed along with the unknown solutions are those containing 20 and 50pg per 6 p l of each acid. Estimations are made from the mean of four results from quadruplicate chromatograms. A linear relationship exists between spot area and the acid content of the spot; the spot areas of standard acids are plotted against the concentration of each acid to give regression lines from which the concentrations of acids in experimental solutions are estimated. Allow 10 minutes for each wash. The acids show as orange or red spots.Nov., 19531 OF THE LOWER STKAIGHT-CHAIN FATTY ACIDS 643 EXPERIMENTAL The method described above gives the optimum conditions for the chromatographic identification and determination of the normal C, to C, fatty acids; it was elaborated only after exhaustive investigation of each of the possible variable factors involved.Reid and Lederer7 used bromocresol purple as an indicator spray; acid spots on the chromatograms showed as yellow spots on a purple background. Using this method we found that acid spots often failed to appear in several of eight replicate chromatograms treated identically. When acid spots did appear they were often indefinite and the back- ground, instead of being entirely purple, presented a purple and yellow mottled appearance. The identification of “unknown” spots was then difficult.With standard fatty acid solutions, acid spots could be identified fairly consistently at concentrations of 20 pg per 5 p1; at 10 pg per 5 pl identification was possible in about 75 per cent., and at 5 pg per 5 p1 in about 50 per cent. of the chromatograms. To increase sensitivity and reproducibility it was necessary to eliminate this mottling effect and to improve spot definition and colour contrast. Optimum conditions for each variable factor in the present technique were found. These conditions are described and discussed for each successive step in the method. PRE-TREATMENT OF CHROMATOGRAPHIC PAPER- Whatman Nos. 1, 4 and 54 filter-papers have been used, all three giving satisfactory results after suitable pre-treatment. Untreated papers and those saturated with vapour from concentrated ammonium hydroxide, sp.gr.0.880 (Reid and Lederer’) showed marked mottling of the final chromatogram, as did papers washed with 1.5 N ammonium hydroxide. Washing with redistilled 92-butanol prevented some of the mottling in the final chromatogram. Washing with the butanol phase of a system of equal volumes of butanol and 1.5 N ammonium hydroxide almost completely prevented mottling. During the washing procedure an aqueous phase sometimes separated out and on these occasions results were poor, presumably owing to uneven washing of the paper. Ethanol (5 per cent. v/v) added to the butanol phase after separation from the equilibrated system prevented the subsequent formation of an aqueous phase. RF values of the acids varied with the paper and the pre-treatment, but were constant for any one set of conditions (see Table I).Lactic acid and acetic acid have the same KF value and are indistinguishable at low concentrations. When each of these two acids was present to the extent of 30 pg per 5 p1 or more, it was found that lactic acid gave a darker and more permanent spot within the acetic acid spot, so that the two could be observed simultaneously. TABLE 1 RI: VALUES OF THE NORMAL c, TO c, FATTY ACIDS ON VARIOUS CHROMATOGRAPHIC PAPERS AFTER UPWARD DEVELOPMENT WITH n-BUTANOL SATURATED WITH 1.5 N AMMONIUM HYDROXIDE Rp values Acetic Propionic Butyric Valeric Hexanoic Paper* acid acid acid acid acid 0.13 0-2 1 0.3 1 0.4 1 0.52 0.13 0.22 0.35 0-49 0.63 0.10 0.19 0.30 0.42 0.54 0.11 0.20 0.33 0-45 0.60 0.09 0.16 0.25 0.33 0.44 0.09 0.15 0.27 0.40 0.52 A r 3 1 (W) 1 (U) 4 (W) 4 (U) 54 (W) 54 (U) NoTEs-(W), washed with butanol saturated with 1-5 N ammonium hydroxide (see text) ; (U), untreated * Numbers refer to Whatman chromatographic papers. papers.APPLICATION OF FATTY ACIDS TO THE PAPER- The volatility of the lower fatty acids and of their ammonium salts has been the cause of one of the difficulties in the elaboration of a chromatographic technique (Elsdenl and Reid and Lederer7). In the present work, however, it has been possible to obtain satisfactoryw DUNCAN AND PORTEOUS : THE IDENTIFICATION AND DETERMINATION [Vol. 78 qualitative results over the range 5 to 50 pg per 5 p1 by spotting the papers with the sodium, barium or ammonium salts of the acids 30 minutes before transferring the papers to the chromatographic tanks.The sodium or the barium spots did not overlap the acetic acid spot, provided the total cation on the paper did not exceed approximately 0-5 micro-equivalents. At higher concentrations, the cation spots tended to obscure the acetic and even the propionic acid spots. Consequently the ammonium salts have been used in routine work (cf. Reid and Lederer'). In qualitative experiments up to 40 pl of the ammonium salts have been applied to each starting point, 5 pl at a time, by the usual intermittent drying technique, without appreciably affecting the final result. For quantitative work, however, it is preferable to make only one 5-pl application to each starting point and to transfer the paper immediately to thetank.Room temperature did not exceed 20°C. PUK~FICATION OF SOLVENTS- In initial experiments a broad acidic band extending 3 cm up from the foot of the paper was apparent in the final chromatogram. This band was eventually traced to acidic material absorbed from the atmosphere by the stock ammonium hydroxide solutions. The use of 1.5 N ammonium hydroxide made by dilution of freshly distilled ammonia eliminated this acidic band and decreased mottling in the final chromatogram. Redistilled n-butanol was used throughout the investigations. DRYING THE CHROMATOGRAM- It is undesirable to spray damp chromatograms, as movement and spreading of spots can take place, and in the present instance residual ammonia in the paper tends to swamp the indicator.Papers dried over long periods at room temperature (18" to 20" C) in still air also gave poor results, presumably because the ammonium salts of the acids volatilised. Results were best on spraying the papers 10 minutes after they had returned to their original dry weight in still air at room temperature. Con- siderable variation in drying times is apparent for identically treated papers. Table I1 shows typical drying times. TABLE 11 THE LOSS OF MOBILE PHASE (a-BUTANOL SATURATED WITH 1.5N AMMONIUM HYDROXIDE) FROM VARIOUS CHROMATOGRAPHIC PAPERS, 11.25-INCH BY 9.25-INCH, WHEN DRIED IN STILL AIR AT 18" TO 20" C Mobile phase on the paper at various times after removing the paper from the chromatographic tank, g 0 mins. 10 mins. 20 mins. 30 mins. 40 mins. 50 mins.60 mins. 6.8 4.3 2.6 0.9 0.0 0.0 0.0 7.3 4-8 3.3 1.8 0.6 0.0 0.0 7.1 4.6 3-0 1.5 0.5 0.0 0.0 7.2 5.5 3.8 2.0 1.0 0.3 0.1 5.8 4.0 2.8 1.4 0.6 0.1 0.0 5-2 2.8 1.4 0.3 0.2 0.0 0.0 4.3 1.5 0-1 0.0 0.0 0.0 0.0 6.1 3.9 1.9 0-3 0.1 0.0 0.0 5.8 2.5 1.0 0.3 0.0 0.0 0.0 6.3 4-2 2.5 1-1 0.5 0.1 0.0 A c \ NoTEs-(U), Untreated; (W), washed with n-butanol saturated with 1.5 N ammonium hydroxide. * Numbers refer to Whatman chromatographic papers. INDICATOR SPRAY- Theoretically an indicator with a pH range of between 4 and 9 should be used to reveal the fatty acids. Accordingly, a number of indicators within this range and just outside it were selected for trial. These indicators, with their pH ranges, were: bromophenol blue (3.1 to 4.4), bromocresol green (4.0 to 5.6), methyl red (4.2 to 6-3), bromocresol purple (5.2 to 64), bromothymol blue (6.0 to 7.6), phenol red (6.8 to &a), cresol red (7.2 to 8 9 , thymol blue (8.0 to 9-6), phenolphthalein (8.3 to 10.5) and thymolphthalein (9.3 to 10.5).There was no colour differentiation between background and acid spots with bromo- phenol blue, thymol blue, phenolphthalein or thymolphthalein. The remainder of theNov., 19531 OF THE LOWER STRAIGHT-CHAIN FATTY ACIDS 645 indicators were taken singly (0.04 per cent. w/v in a mixture of ethanol and formalin) and in pairs (0-02 per cent. w/v of each) to give 21 indicators, all of which were put through comparative tests with unwashed papers. Of these 21 indicators, those containing bromo- cresol green were poorest owing to lack of good colour contrast, those containing cresol red or bromocresol purple gave good colour contrast, whilst those containing bromothymol blue, phenol red or methyl red gave good contrast and improved spot definition.In this last group, one mixture, bromothymol blue and methyl red, was markedly superior to the rest and was adopted for the method used; the concentration of each component indicator was increased to 0.04 per cent. w/v to intensify the colours. COLOUR DEVELOPMENT IN THE FINAL CHROMATOGRAM- Successful colour development depended on intermittent, but even, exposure of the sprayed chromatogram to the optimum concentration of ammonia vapour in the manner described. Attempts to develop the colour in an “ammonia tank” of large cross-sectional area with either continuous or intermittent exposure of the paper to the ammonia vapour were not successful, presumably because of uneven loss of vapour when the tank was opened to insert the paper.In the present work a tall cylinder, at least 12 inches deep and of just sufficient diameter (6 inches) to accommodate the roll of paper, has been used; whenever the paper was removed the cylinder was closed by a glass plate to maintain a uniform concentration of ammonia. In the present technique there was a tendency for a strip round the edge of the paper to develop more rapidly than the rest of the paper. For this reason, particularly in quantitative work, spots were not applied within 5 cm of the edge of the paper. QUANTITATIVE ANALYSIS- Using bromocresol purple as the spray reagent and Whatman No.1 paper, Reid and Lederer’ obtained a linear relationship between spot area and the logarithm of the acid concentration. With Whatman No. 4 paper and the technique described above, but with Reid and Lederer’s spray reagent, this relationship has been confirmed. When, however, the mixed indicator spray of methyl red and bromothymol blue was used, a linear relationship between spot area and concentration of acid was observed over the range 16 to 80 pg per 5 p1 for each acid. With standards solutions containing 16 and 80pg per 5 p l of each acid, regression lines from the means of spot areas in quadruplicate runs have been drawn and the spot areas for other concentrations of acids predicted. The deviations of the predicted from the actual concentrations are shown in Table 111.TABLE I11 DEVIATIONS OF PREDICTED ACID CONCENTRATIONS FROM ACTUAL ACID CONCENTRATIONS WHEN SPOT AREA IS PROPORTIONAL TO ACID CONCENTRATION FOR THE NORMAL c, TO c, FATTY ACIDS Range of concentra- Acid tion, Pg Per 6 PI Acetic . . . . 8 to 80 ,> 7, Propionic . . . . Butyric . . . . Valeric . . . . 9 1 Hexanoic . . . . 7 7 Number of concentra- tions tested 7, >I Mean deviation, 3-16 3-16 3-0 3-71 2.88 PLg Per 5 Pl Standard Bias deviation, Pg Per 5 Pl +0.278 3.75 + 1.58 3.55 +0*417 3.60 + 0.958 4.32 + 1.04 3.48 Error a t mid-point of range, 7.8 7.4 7.6 9.0 7.25 % It will be noted that the bias, i.e., the “mean deviations taking acount of sign,” is persistently positive (Table 111). With the logarithmic relationship, Reid and Lederer obtained negative figures and suggested these might be partly caused by personal errors in over estimating spot areas of the acids a t lower concentrations.The perimeter of the coloured spot observed in the sprayed and developed chromatogram marks the minimum concentration contour at which the indicator in the spray reagent will show the acid in the spot. The lower the detectable concentration contour, i.e., the more sensitive the reagent, the larger the spot and the less well-defined is the spot boundary.646 DUNCAN AND PORTEOUS [Vol. 78 Visual appreciation of the coloured spot boundary, i.e., the precision with which the spot area can be estimated, will depend on the colour contrast between the spot and the back- ground. Both of these possible sources of variation in the estimated size of the visible spot are inherent in the method and could vary with the indicator used.I t is not surprising, therefore, that the use of two different indicators should give two different relationships between visible spot area and acid concentration, nor is it surprising that the experimental deviations from these relationships are of opposite sign in the two methods. Personal errors in tracing spot boundaries and estimating spot areas add a further complication. It is clear that experimental conditions should be rigidly standardised if the method is to be used quantitatively. With the optimum conditions described above, mottling was completely eliminated from the chromatograms; colour contrast and spot definition were excellent. The normal C, to c6 fatty acids at concentrations between 5 and 80pg per 5p1 (0.1 and 0.8 per cent.w/v) were readily handled simultaneously in one application to give reproducible chromatograms without tailing or overlapping of spots. Identification over the range stated and determina- tion over the range 16 to 8Opg per 5p1 were consistent for each acid. Two assistants, untrained in any form of chromatographic technique, were immediately able to attain results conforming with the authors’. Fig. 1 shows a typical chromatogram. APPLICATIONS One of us (R. E. B. D.) has successfully applied the technique described to the identifica- tion and determination of the C, to C, fatty acids in laboratory grass - water fermentation mixtures and in field silages. In both instances it was necessary to have available a method for detecting and estimating the C, and C, acids at small concentration in the presence of relatively large concentrations (7 to 10 times as much) of the C, to C, acids as a preliminary to fuller investigation by gas-phase partition chromatography (James and Martinlo).One of us (J. W. P.) has used the method to discover whether or not the slow-growing anaerobic pathogen Actinomyces israeli produced any of the C, to c6 fatty acids. I t has been shown by Erikson and Porteousll that the organism produced lactic acid when grown in a glucose broth medium containing lactic acid and sometimes also acetic, propionic and butyric acids; a particularly sensitive method applicable to low concentrations of fatty acids was therefore essential. We wish to acknowledge the receipt by one of 11s (K. E. H. D.) of a D.S.I.R. grant and by the other (J. W. P.) of a M.R.C. personal grant. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCE s Elsden, S. I<., in “Biochemical Society Symposia, No. 3,” Cambridge Unix-ersity Press, 1949, p. 74. Brown, F., Biochem. J., 1960, 47, 598. Brown, F., and Hall, L. l’., Nature, 1950, 166, 66. Hiscox, E. R., and Berridge, N. J., Ibid., 1950, 166, 522. Kennedy, E. P., and Barker, H. A., Anal. Chenz., 1951, 23, 1033. Miettenin, J. K., and Virtanen, A. I., Ann. Acad. Sci. Fenn., 1951, -411, KO. 41, 1. Reid, R. L., and Lederer, M., Biochem. J., 1951, 50, 60. Fink, K., and Fink, R. H., Proc. SOC. Exp. BioE. Med., 1951, 78, 135. Thompson, A. R., Australian J . Sci. Research, B, 1951, 4, 180. James, A. T., and Martin, A. J. P., Biochem. J., 1952, 50, 679. Erikson, I)., and Porteous, J . W., .I. Gen. 2Wicwbio/., 1953, 8, 464. DEPARTMENT OF BIOLOGICAL CHEMISTRY UNIVERSITY OF ABERDEEN Muy 13th, 1953Fig. 1. Separation of the C, to C, volatile fatty acids. The same chromatogram photographed by (a) reflected In and ( b ) transmitted light. ( b ) the grain of the chromatographic paper is enhanced when photographed by transmitted light Concentrations of each acid (pg per 5 p1) from left to right: 5, 10, 20, 30, 50, 80.

ISSN:0003-2654    

DOI:10.1039/AN9537800641

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Nov., 19533 PRIMAVESI 647 The Determination of isoButyraldehyde in n-Butyraldehyde BY G. R. PRIMAVESI isoButyraldehyde and n-butyraldehyde are reduced to the corresponding alcohols by sodium borohydride. The isobutyl alcohol is then estimated absorptiometrically by means of a modified Komarowsky reaction for higher alcohols, of improved reproducibility, carried out under controlled conditions. IN Koinarowsky’s method1 for the determination of higher alcohols by means of the colour produced on heating with concentrated sulphuric acid and an alcoholic solution of salicyl- aldehyde, isobutyl alcohol gives a colour from 100 to 150 times deeper than that produced by an equal amount of n-butyl alcohol. This difference in the behaviour of the two alcohols forms, after a reduction by sodium borohydride, the basis of a proposed method for the absorptiometric determination of iso- but yraldehy de in n-bu t yraldehyde .Many modifications have been made to Komarowsky’s method; these include the use of substituted benzaldehydes other than salicylaldehyde or furfural, variations in the methods of cooling or heating the reaction mixture, and variations in the method and order of adding the reagents and test sample~.2,~,4,51617,*~~ Most of these modifications deal with considerably larger amounts of higher alcohols than the method now described. Some authors have had trouble with high blanks. It has been found that the following variables affect the degree of colour developed: the proportions of sulphuric acid, ethanol, water, salicylaldehyde and n-butyl alcohol present in the reaction mixture; the technique of adding the reagents and mixing the reaction solution; the temperature at which the reaction is carried out; the duration of heating and the time of standing after cooling the mixture.Of these variables the addition of the reagents and mixing the reaction mixture need the closest attention; the others are important, but are merely a matter of ordinary control. The total amount of ethyl alcohol plus water is more critical than slight variations in the amount of either. The blank gave no difficulty with the ethyl alcohol used; this was P.I. rectified spirit (96.1 per cent.) supplied by The Distillers Co. Ltd. This spirit is free from aldehydes and contains less than 1 p.p.m. of higher alcohols as isobutyl alcohol, which figure represents the whole of the yellow colour produced by the reagents and sulphuric acid.The amount of isobutyl alcohol that gives a reading on the Spekker photo-electric absorptiometer of 0.1 higher than the blank under the conditions described is 6 pg; the corresponding figure for n-butyl alcohol is 830 pg. The object of the technique now described is to prevent, as far as possible, any heating of the reaction mixture during the mixing process and subsequently to control accurately the duration and temperature of heating. If this is done carefully the standard error of a single determination is equal to a Spekker drum reading of about 0.010 anywhere on the scale. As the object was to estimate isobutyraldehyde in n-butyraldehyde, all series of experi- ments were made with a standard concentration of total butanol and various ratios of iso- to n-butyl alcohol.A blank on ethyl alcohol was included in each series. The conditions described were determined by means of many statistically planned experiments and were chosen to give the maximum difference between iso- and n-butyl alcohol, combined with reasonable insensitivity to slight variations in timing and dilution. METHOD REAGENTS- EthyZ alcohol-This must be free from aldehydes and higher alcohols. P.I. rectified alcohol (96-1 per cent.) supplied by The Distillers Co. Ltd. is satisfactory. Sdicylaldehyde reagent-Dissolve 0.1 67 g of laboratory grade salicylaldehyde in 100 ml of ethyl alcohol. The concentration is arbitrary, but for use with a standard graph it must be the same as that used in the preparation of the graph.Accordingly, the aldehyde content648 PRIMAVESI THE DETERMINATION OF [Vol. 78 should be determined by the hydroxylaminelO or other suitable method and the strength of reagent should be adjusted if necessary. Concentrated suulphuric acid-This should be of recognised analytical purity. PREPARATION OF THE SAMPLE- Prepare an ethanolic solution of the sample, by successive dilution if necessary, to give the required total butyl alcohol content. Of the calibration series given below, one was done with a total ethyl .alcohol volume of 6-5 ml by using a 3-ml sample of a 0-045 per cent. v/v solution of total butyl alcohol in ethyl alcohol, the isobutyl alcohol ranging from 0 to 4 per cent.v/v of the total butyl alcohol. The other series was done with a total ethyl alcohol volume of 6 ml by using a 3-ml sample of a solution of 0.02 g of total butyl alcohol in 100 ml of ethyl alcohol, the isobutyl alcohol ranging from 0 to 7-84 per cent. w/w of the total butyl alcohol. PROCEDURE- Measure from a burette 10ml of concentrated sulphuric acid into a 7 x 1-inch test tube and add from a pipette 3 ml of the salicylaldehyde reagent; the tube should be held vertically, resting on the bench, and the solution run carefully down the side to form a layer; without moving the tube, add, similarly, from a pipette 3 ml of the sample solution. Immediately immerse the end of the tube in solid carbon dioxide - ethyl alcohol mixture (temperature about -78" C) and swirl vigorously for 15 seconds.Remove the tube and mix the contents thoroughly. This is best done by first twirling the tube held almost horizontally, so as to wet the entire inner surface, and then swirling the tube held vertically. Repeat the mixing at least once more. It is most important to get the two liquids of widely differing densities, viscosities and volatilities completely homogeneous without undue heating; the mixture should be at or slightly below room temperature when mixed, and, in order to attain this end, the time of cooling in the carbon dioxide bath should be varied to suit the mode of action of the operator. When its contents are homogeneous, place the tube in a thermostat bath at 80" C for exactly 30 minutes; remove it and cool it in tap water.Exactly 10 minutes after removal from the bath, read the optical absorption on the Spekker absorptiometer, fitted with a 2-cm cell and Ilford No. 605 filters, against distilled water in a 2-cm cell. For reproducible results the greatest care must be taken to ensure that the addition and mixing of the sample and reagents are done in a reproducible manner. PRELIMINARY TREATMENT AND REDUCTION OF SAMPLE- Take 1 ml of sample in 50 ml of water and flash distil into an ice-cooled 100-ml standard flask until no "tears" are visible in the condenser and for 3 or 4 minutes further. The volume taken need not be exact as the final aliquot is adjusted later, but should be within 5 1 0 per cent. of 1 ml. Make the distillate up to 100 ml with water and immediately take two 10-ml aliquots.Analyse one aliquot for total aldehyde by the hydroxylamine method; to the other in a 50-ml calibrated flask add 10ml of a freshly prepared solution of 0.15 g of sodium borohydride in 100 ml of water; the amount is not critical but it must be an excess to ensure complete reduction of all the butyraldehyde (not merely the isobutyraldehyde) . The reagent is unstable, but keeps for at least a day in aqueous solution. Allow the solution to stand in the stoppered flask for at least 10 minutes and then make it up to 50 ml with water. The time of standing can, if necessary, be prolonged indefinitely. From the titre given by the hydroxylamine method calculate the volume of reduced aldehyde solution required to give exactly 15mg of total butyl alcohol; take this volume and make it up to 50 ml with pure ethyl alcohol. This total butyl alcohol concentration of 0-030 g per 100 ml in the final solution was chosen for estimating about 4 per cent.of isobutyraldehyde in rt-butyraldehyde. If higher or lower concentrations are expected, the total butyl alcohol content of the final solution and the standard graph should be modified accordingly. The reduction of the aldehyde by sodium borohydride was based on work by Chaikin and Brown.ll They state the reaction to be- 4RCHO + NaBH, + 2NaOH + H,O +4RCH20H + Na3B03 But results of better reproducibility were attained by omitting the alkali, the reaction presumably being- THE DETERMINATION OF iSOBUTYKALDEHYDE I N W-BUTYRALDEHYDE 4RCHO + NaBH, + 2H,O + 4RCH,OH + NaBO,Nov., 1953: ~SOBUTYRALDEHYDE IN n-BUTYRALDEHYDE 649 The reagent solution seems to be stable for at least a day, provided that the solution is not acid.NaBH, + HC1+ 3H,O -+ NaCl + H3BO3 + 4H2. Mole for mole, n-butyr- aldehyde gives about one-fifteenth and isobutyraldehyde about one-fifth as much colour as isobutyl alcohol. It is therefore essential to ensure that no aldehyde or potential aldehyde is present at the end of the reduction stage. Polymers of the aldol type would probably give a colour even if reduced to the corresponding hydroxy-alcohols, and paraldehyde-t ype polymers are probably not reduced by the neutral or slightly alkaline reagent. I t is in order to remove interfering substances of this kind that the sample is flash distilled immediately before analysis. The distillation is carried out in a large excess of water to save manipulation and unnecessary oxidation of the distillate.It is assumed that any oxidation that does take place affects the two aldehydes proportionately. Neither iso- nor n-butyric acid give a colour in the test. ABSORPTIOMETRIC ESTIMATION OF THE ~SOBUTANOL- solution in ethyl alcohol as sample. standard graph prepared as described below. PREPARATION OF THE STANDARD GRAPH- Synthetic mixtures were made by weighing the two aldehydes direct into a small stoppered weighing flask. Each aldehyde was flash distilled immediately before making up any mixture and each mixture was analysed immediately after making up. These precautions were taken to ensure that a minimum quantity of butyric acid was present in the standard samples.The two aldehydes used as samples were obtained from a careful fractionation of the ordinary laboratory stock of each aldehyde. The isobutyraldehyde fraction used had b.p. 64.6” C at 760 mm of mercury and nzo of 1.3728 and the n-butyraldehyde fraction b.p. 74.9” C at 760 mm of mercury and n”,” of 1.3795. RE su LTS Acid causes immediate evolution of hydrogen- Aldehydes react in the Komarowsky test to give a colour. Proceed with the Komarowsky test exactly as described above, using 3 ml of the final Read the proportion of iso- to n-butyraldehyde from a Four replicate determinations were made on 1 ml of each mixture. The final colori- metric estimation was made in duplicate in each determination. Tables I and I1 show results for standard series of solutions.In Table I11 the volume of the aliquot used in making the final dilution is given to show the kind of variation that can occur without greatly affecting the result. STANDARD SERIES- TABLE I RESULTS FOR STANDARD SERIES: 0 TO 4 PER CENT. v/v OF ZSOBUTYL Reaction mixture: 10 ml of concentrated sulphuric acid, 3.5 ml of 0.143 per cent. v/v solution of salicylaldehyde in ethyl alcohol, 3 ml of 0.045 per cent. v/v solution of total butyl alcohol in ethyl alcohol Blank on A \ ALCOHOL IN n-BUTYL ALCOHOL Iso- in n-butyl alcohol, per cent. v/v ethyl alcohol 0 1 2 3 4 Spekker readings . . . . 0.193 0.297 0.174 0.298 0.183 0.287 0- 184 0.278 0.306 0.293 0.299 0.309 0.296 Mean .. .. . . 0.184 Colour due to isobutyl alco- hol, i.e., mean - “0” - figure . .. . . . Standard error . . . . 0.008 Ditto as a percentage of “iso” colour . . .. - - 0.010 0.539 0.738 0.938 1.095 0-535 0-738 0.931 1.110 0-541 0.748 0.939 1.094 0.525 0.748 0.922 1.076 0.732 0.735 0.732 0.739 0.739 0.535 0.933 1.094 0.239 0.007 3 0-443 0.006 1.4 0.637 0.008 1.2 0.798 0.014 1.8650 PRIMAVESI THE DETERMINATION OF pr01. 78 TABLE I1 RESULTS FOR STANDARD SERIES: 0 TO 8 PER CENT. V/V OF iSOBUTYL ALCOHOL IN ?Z-BUTYL ALCOHOL Reaction mixture: 10 ml of concentrated sulphuric acid, 3 ml of a solution of 0-167 g of salicylaldehyde in 100 ml of ethyl alcohol, 3 ml of a solution of 0.020 g of total butyl alcohol per 100 ml of 10 per cent. v/v water in ethyl alcohol Blank on ethanol Spekker readings . . 0.1 11 0.114 0.1 13 Mean . . . . . . 0.113 Colour due to isobutyl alcohol, i .e . , mean - “0” figure . . Standard error.. . . Ditto as a percentage of “iso” colour . . . . - - - 7 0 0.190 0.186 0.194 0.179 0.178 0.182 0.185 __ 0.006 isoButanol in n-butanol, per cent. w/w 0.98 1-96 0.308 0.403 0.293 0-405 0.294 0.409 0-386 0.393 0-384 0.298 0-397 2-94 3.92 0.528 0.618 0.512 0.631 0.514 0.633 0.595 0.610 0.612 0.518 0-617 4-90 5-88 0.718 0.815 0-719 0.826 0-732 0.829 0.827 0.817 0-812 0.723 0.821 0.113 0.212 0.333 0-432 0.538 0-636 0.008 0.010 0.009 0.014 0.008 0.007 7 5 3 2 1 3 6.86 7:84 0.939 1.012 0.909 1.018 0.934 0-997 1.000 0.998 0.992 0.927 1-003 0.742 0.818 0.016 0.010 b) d 1 TABLE rrr APPLICATION OF PROCEDURE TO KNOWN MIXTURES OF BUTYR-4LDEHYDE isoButyraldehyde in n-butyraldehyde, per cent. w/w 0 0.94 2.33 3.47 4.38 1” 2T Col.1 Col. 2 Col. 1 Col. 2 Col. 1 Col. 2 Col. 1 Col. 2 10.78 0-319 10.39 0.496 10.81 0.700 11.15 0.882 10.72 1.000 0.329 0.488 0-684 0.863 0.999 9.84 0.337 10.92 0.468 11.32 0.679 10.34 0.891 10.52 1.035 0.322 0.495 0.676 0.936 1.006 9.97 0.343 10.59 0.514 12-45 0.663 10.76 0.898 10.95 0-956 0.353 0.479 0.681 0.880 1.010 10.45 0.332 10.45 0.483 10-39 0.714 11.26 0.881 11.85 1.020 0.319 0.476 0.693 0.841 1.009 A r 7 Mean absorptiometer Colour due to isobutyr- aldehyde, i.e., mean - Standard error.. . . Ditto as a percentage of “iso” colour . . . . readings . . . . “0” figure . . . . 0.332 0.487 0.686 0.884 1 -007 - 0.155 0-354 0.552 0.675 0.012 0-015 0.016 0.028 0.023 10 5 5 3 - NOTE-Subsequent work has shown that the n-butyraldehyde used as a standard contained 0.9 per cent.of isobutyraldehyde, so all Spekker absorptiometer readings in Table I11 should be reduced by 0.15 unit. * Columns headed 1 show the volume of reduced aldehyde solution used in the final dilution to 50 ml. t Columns headed 2 show the corresponding absorptiometer readings. The author thanks Mr. D. R. Read for help in planning and interpreting statistical experiments, and the Directors of The Distillers Company Limited and of British Petroleum Chemicals Limited for permission to publish this paper. REFERENCES 1. 2. 3. 4. 5. Komarowsky, A., Chcrn. Ztg., 1903, 27, 807 and 1086. “Report of the Royal Commission on Whisky and other Potable Spirits,” H.M. Stationery Office, Von Fellenberg, T., Chenz. Ztg., 1910, 34, 791. Penniman, W. B. D., Smith, D. C., and Lawske, E. I., Ind. Eng. Chem., Anal. Ed., 1937, 9, 91. Coles, H. W., and Tournay, W. E., Ibid., 1942, 14, 20. London, 1909.Nov., 19531 ~SOBUTYRALDEHYDE IN WBUTYRALDEHYDE 6. 7. 8. 9. 10. 11. Federico, L., and Cioffi, R. M., Chim. e Industr., 1947, 29, 298. Custance, H. M., and Higgins, M., Analyst, 1949, 74, 310. Gierer, S., and Hoffman-Osterhof, O., Mikrochemie, 1951, 38, 232. Osborn, G. H., and Mott, 0. E., Analyst, 1952, 77, 260. Maltby, J. G., and Primavesi, G. R., Ibid., 1949, 74, 498. Chaikin, S. W., and Brown, W. G., J. Amer. Chem. SOC., 1949, 71, 122. THE DISTILLERS Co. LTD. RESEARCH AND DEVELOPMENT DEPARTMENT GREAT BURGH, Epsow, SURREY 65 1 April 24th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800647

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

Nov., 19531 ~SOBUTYRALDEHYDE IN WBUTYRALDEHYDE 65 I The Quantitative Separation of Copper, Lead and Tin by Cathodic Deposition BY G. H. AYLWARD AND A. BRYSON A method is described for the quantitative separation of copper and lead from tin by cathodic deposition from phosphoric acid solution. The tin forms an anionic complex that is not reduced a t the cathode and the copper is separated from the lead by deposition a t a controlled potential. The lead is deposited in a metallic form and the tin is determined volumetrically. The difficulties associated with the quantitative cathodic deposition of lead have been investigated and optimum working conditions have been established to overcome these difficulties. IN recent years increasing attention has been given to the problem of separating copper, lead and tin by controlled-potential deposition.Several of these methods1 y 2 7374 depend on the deposition of copper by controlled cathode-potential followed by simultaneous cathodic deposition of tin and lead. The tin and lead deposit is then dissolved and the metals are determined separately. It is possible, however, to increase the difference between the deposition potentials of these two metals by forming stable anionic stannic complexes. Lassieur5 used this method to separate copper and lead dioxide simultaneously from tin in a solution containing hydrofluoric and nitric acids. Boric acid, sodium oxalate, hydroxyl- amine and sodium hydroxide were then added and the tin was deposited cathodically from the hot solution. His results reported to the nearest milligram show errors in each estimation of up to 1 mg.Lingane and Jones6 made a complex of tin with tartrate and separated copper from lead in slightly acid solution in the presence of hydrazine dihydrochloride. After deposition of lead, hydrochloric acid was added and the tin was deposited. The lead figures show losses of from 0.3 to 2.1 mg, whilst five tin figures have an average deviation of 0-7 mg. Norwitz' used phosphoric acid to keep tin in solution in a nitric acid electrolyte and thus avoided the tedious separation of metastannic acid before electrolytically depositing copper. The stability of tin phosphate complexes is shown by the distillation of antimony chloride a t 160" C without interference from stannic chloride if phosphoric acid is present.* Polarograms were taken for copper, lead and tin in phosphoric acid and revealed little change in the half-wave potentials of copper and lead, whereas no step could be found from the tin in acid or ammoniacal solutions.These observations prompted the present investiga- tion, which gives a method for the successive cathodic deposition of copper and lead from phosphoric acid, followed by tin if desired. We believe that the volumetric method for tin is more suitable and therefore have used this method throughout our work. The cathodic deposition of lead presents several difficulties, which have been well recognised in the past and which have led many to believe that quantitative electro-deposition of the metal is impossible. Among these difficulties are the unsatisfactory nature of the deposit and the tendency to re-solution during the washing process.During the progress of this research, these and other difficulties were met and systematically investigated. As a result it has been possible to formulate a method that reduces these errors and allows a quantitative determination of the element. In the procedure adopted, the alloy containing copper, lead and tin is dissolved in hydro- chloric and nitric acids, phosphoric acid is added and the solution acids are expelled by652 A4YLWARD AND BRYSON: THE QUANTITATIVE SEPARATION OF [Vol. 78 fuming. The solution is diluted to 120 ml and copper is deposited at -0.35 volt with respect to the saturated calomel electrode (S.C.E.) by controlled-potential electrolysis.After weighing, the copper-covered cathode is replaced in the solution and the lead is deposited, without control, as the tin phosphate complex is not decomposed in this electrolyte. When lead deposition is complete, the solution is neutralised with ammonium hydroxide. The electrolysis beaker is withdrawn and the plated electrode is washed with three successive solutions of ammonium sulphate and then with distilled water, alcohol and ether. The cathode is dried and weighed. Hydrochloric acid and nickel shot are added to the electrolyte and, after boiling, the reduced tin is titrated with iodine. EXPERIMENTAL APPARATUS- The apparatus comprised a Griffin and Tatlock electrolysis unit adapted for manual potential control by wiring a 240-volt “Variac” transformer into the electrode input circuit. The e.m.f.to the cell could be controlled by varying the alternating voltage applied to the transformer - rectifier circuit of the instrument. With this circuit the cathode potential could be held to within 20 mV of the required value. The smaller platinum gauze electrode, 75 sq. cm in area, was generally used as the cathode, as investigation showed that the lead deposition area should be as small as possible. Stirring was automatic and it was found that the design of the stirrer had some effect on the nature of the lead deposit. The most satisfactory stirrer was one that forced the flow of the electrolyte evenly through the mesh of the platinum gauze. Under these conditions the deposit is more uniform and finer in texture.ELECTRO-DEPOSITION OF COI’PER- If the initial current-density does not exceed 1.5 amperes per sq. dm, copper will be deposited from phosphoric acid electrolyte as a bright salmon-coloured plate. If the cathode potential is not allowed to become more negative than -0.35 volt with respect to the S.C.E., separation of the copper from the lead and tin is quantitative. Four hundred milligrams of copper are deposited completely in 35 minutes on a platinum cathode 125 sq. cm in surface area. ELECTRO-DEPOSITION OF LEAD- Lead is deposited quantitatively from this electrolyte, but gives a dark powdery non- adherent deposit. This trouble was encountered in the early days of lead electroplating and was overcome by B e t t ~ , ~ who added gelatin to the plating bath. By the addition of 1 ml of a 0-5 per cent.gelatin solution to the electrolyte a bright grey metallic deposit is obtained. RE-SOLUTION OF LEAD DURING WASHING- Schoch and Brown,l in their method for depositing lead from hydrochloric acid, washed the lead by substituting for the electrolyte successive beakers of distilled water, alcohol and ether. The greatest error in ten results was 0-7 mg with an average deviation of 0.3 mg. They recommended this method in preference to the anodic deposition of lead dioxide. Sandlo preferrcd to deposit the dioxide, owing to the oxidisability of the lead deposit. Lingane and Jones,6 who deposited metallic lead from tartrate solutions, recognised the loss of lead during washing and commented: “This loss averages 1-5 rf: 0.6 mg and is sufficiently repro- ducible so that it can be added as a correction to obtain lead results which generally will be correct to well within 1 mg.” The technique of washing described by Schoch and Brown gave wash solutions that invariably contained lead, although none remained in the electrolyte. This finding was reported by Kny-Jones, Lindsey and Penney,ll in connection with the re-solution of tin deposits from hydrochloric acid.These workers recognised that the loss was due to solution of the deposit in the film of electrolyte left in contact with the electrode on removing the electrolysis beaker. The procedure suggested by these authors of neutralising the electrolyte with ammonium hydroxide before washing was adopted for the present problem. But polaro- graphic investigations showed that copper, equivalent to several tenths of a milligram, was dissolved from those parts of the copper-plated electrode exposed to the slightly ammoniacalNov., 19531 COPPER, LEAD AND TIN BY CATHODIC DEPOSITION 653 solution during the washing.This difficulty was overcome by ensuring that the lead deposit completely covered the copper. By this washing technique, loss by solution caused by the electrolytic action between the lead deposit and the platinum electrode is negligible. ADSORPTION OF PHOSPHATE ON THE LEAD PLArE- The weights of lead found by using this washing technique were consistently high. It was considered that the increase of weight was not caused by oxidation during drying of the deposit because a quick low-temperature method was used.It was proved, after numerous tests, that the high figures were caused by adsorbed phosphate ions. The modified washing technique described below was adopted; it removed the adsorbed phosphate film and simul- taneously reduced the amount of re-solution of the lead deposit. THE WASHING TECHNIQUE- Experiments were carried out on lead-plated electrodes after deposition on copper-plated platinum from solutions of the analytical reagent grade metal. Distilled water, ammonium nitrate, ammonium sulphate, ammonium chloride and hydroxylamine wash solutions were tried. A fresh lead-plated electrode was immersed three times in 120 ml of each solution by moving the beaker up and down with the stirrer running. Each solution was tested polarographically for lead. The only solution to offer any advantages over distilled water was 0.1 per cent. ammonium sulphate.Less lead was present in this solution and the deposit did not tarnish during drying. In a second set of experiments, separate plated electrodes were treated with various numbers of ammonium sulphate washings and the lead deposit was weighed each time. The deposits were dissolved in nitric acid and the weight of adsorbed phosphate determined by the phosphomolybdate method. Each solution was tested for dissolved lead and its phosphate content was determined. The results are shown in Tables I and 11. TABLE I THE REMOVAL OF PHOSPHORIC ACID FROM LEAD DEPOSITS BY INCREASING THE NUMBER OF WASH SOLUTIONS Amount of Wash method lead taken, lead found, Difference, lead plate, Amount of Amount of phosphate on mg mg mg mg (a) 1 beaker of 0.1% (NH,),SO, .. 50.0 51.0 1.0 0.80 (b) 2 beakers of 0.1% (NH,),SO, . . 50.0 50.3 0.3 0.18 (c) 2 beakers of O.lyo (NH,),SO, and 1 beaker of distilled water . . 50.0 49.9 0.1 0-03 (d) 3 beakers of 0.1% (KH,),SO, and 1 beaker of distilled water . . 50.0 49.9 0.1 negligible The results in Table I indicate that it is necessary to wash the lead deposit with a t least three separate volumes of wash solution to reduce the weight of adsorbed phosphate to less than 0.05 mg. TABLE I1 PHOSPHATE AND LEAD TESTS ON WASH SOLUTIONS OF (d) IN TABLE I Wash beaker Phosphate content, for lead, Polarographic test mg mg First 171.6 0.12 Second 6.4 less than 0.05 Third 0.46 negligible Fourth 0-12 negligible From the results in Table I1 it is evident that ammonium sulphate effectively prevents A small amount of lead is present Results are good for as little as 10 mg re-solution of the lead during the washing procedure.in the first wash beaker but not in the other three.654 AYLWARD AND BRYSON: THE QUANTITATIVE SEPARATION OF [Vol. 78 by plating on a smaller surface area and using the four-beaker wash and finally an alcohol and then an ether rinse. METHOD FOR DETERMINING THE TIN The tin can now be quantitatively deposited if hydrochloric acid and hydrazine dihydro- chloride are added, but we have used the iodimetric method after reducing the tin with nickel shot. PROCEDURE- To the weighed sample, containing not more than 400 mg of copper, add 5 ml of 10 N hydrochloric acid and then 5 ml of 15 N nitric acid.When solution is complete, add 10 ml of phosphoric acid, sp.gr. 1.75, and evaporate the phosphoric acid to a syrupy consistency to remove the solution acids. Cool, add 120 ml of distilled water and electrolyse, using platinum electrodes, with the smaller gauze electrode as the cathode. Control the cathode potential a t -0.35 volt with respect to the S.C.E. For the best copper deposit the initial current density should not exceed 1.5 amperes per sq. dm. When the current becomes constant at the control potential, after approximately 35 minutes, switch off the stirrer and lower the electrolysis beaker, washing the electrodes, calomel electrode and stirrer with distilled water. Rinse it in alcohol, dry at 105" C, cool and weigh. Add 1 ml of 0.5 per cent.gelatin solution to the electrolyte and adjust the electrolyte l e ~ l so that all the copper is covered by the solution (approximately 200 ml). Electrolyse at a current density that causes no hydrogen gas to be liberated during the deposition of the lead. This is approxi- mately 0-4 amperes per sq. dm, or at a cathode potential not lower than -0.8 volt with respect to the S.C.E. When electrolysis is complete, after 75 to 90 minutes, neutralise the electrolyte with 15 N ammonium hydroxide. Switch off the stirrer and quickly replace the electrolysis beaker by a 150-ml beaker containing 120 ml of 0.1 per cent. ammonium sulphate solution. With the stirrer running, immerse the electrodes three times by quickly moving the wash beaker up and down. Repeat this procedure with two fresh sulphate solutions and then wash the beaker successively with distilled water, alcohol and ether.Switch off the current and remove the cathode. After the removal of the copper and the lead, transfer the solution to a 500-ml conical flask and add to this the first lead wash solution. Add 75 ml of 10 N hydrochloric acid, reduce the tin with nickel shot and determine the tin iodimetrically using starch solution to indicate the end-point. Switch off the current and remove the cathode. Replace the weighed copper-plated electrode in the circuit. Quickly dry the deposit at 100" C, cool and weigh. RE s u LTS Analyses of mixtures of copper, lead and tin are shown in Table 111. Solutions for all the separations were made up from analytical reagent grade metals.TABLE I11 RSULTS OF COPPER, LEAD AND TIN DETERMINATIONS FOR SYNTHETIC ALLOYS Amount of copper I taken, found, mg mg 385.0 383.2 386.2 386.0 400.9 401-1 358-2 358.2 275-0 275.0 245-2 245.1 75.0 75.0 45.2 45.0 Higher range for lead . . Amount of lead taken, found, mi3 mg 97.5 97.5 76.0 76-2 49.8 49.6 25.0 24.9 10.0 9.8 10.0 9.9 - f - - 150.0 149.8 200.6 200.3 350.2 350-0 400-3 400.5 Amount of tin ------h-----7 taken, found, mg mg 15.6 15.4 38-9 38.8 51.0 50.9 103.8 103-6 - - - - 198.2 198.0 195.5 195.4Nov., 19531 COPPER, LEAD AND TIN BY CATHODIC DEPOSITION 655 APPLICATIONS The separation can be applied, with equally good results, to the determination of the three metals in copper-based alloys and to the determination of copper and tin in white metals.In 10 ml of phosphoric acid no more than 450 mg of copper and 60 mg of bismuth will remain in solution. Nevertheless, more than 500 mg of each of the following metals, tin, antimony, zinc, nickel, iron, lead, cadmium, aluminium, chromium and manganese, can be present in the solution without the formation of insoluble phosphates. Copper is separated from antimony, which is partly reduced to the metal at approximately -0.76 volt with respect to the S.C.E. The antimony, however, is deposited simultaneously, but not quantitatively, with the lead. The problem of removing arsenic and antimony before electrolysing the copper is at present under investigation. On neutralising the electrolyte with ammonium hydroxide to wash the lead deposit, the phosphates present prevent the precipitation of iron and aluminium. The cathode potential must not be more negative than the deposition potential of nickel, iron or zinc, for they will, if present, be deposited from slightly ammoniacal phosphate solution.Nickel is deposited first a t -1.1 volts with respect to the S.C.E., if 200mg of the metal are present in 200 ml of solution at room temperature. If the initial conditions of applied voltage are adhered to for the lead deposition, on neutralising the electrolyte the cathode potential will be more positive than this value and nickel will not be deposited. This method enables the determination of copper, lead and tin in an alloy to be completed within 34 hours. It avoids the tedious separation of the tin as metastannic acid and the separation of large amounts of lead by fuming with sulphuric acid. The method has two further advantages: it avoids filtering and, if the solution of the alloy and the electrolytic procedures are carried out in the same beaker, only one transfer of the analysis solution from vessel to vessel is necessary in the course of the determination. If bismuth is present it will be deposited before all the copper is removed. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. REFERENCES Schoch, E. P., and Brown, D. J.. J . Amer. Chem. SOL, 1916, 38, 1660. Lindsey, A. J., and Sand, H. J., Analyst, 1934, 59, 335. Torrance, S., Ibid., 1938, 63, 488. Milner, G. W. C., and Whittem, K. N., Ibid., 1952, 77, 11. Lassieur, A., Ann. Chim., 1925, 3, 269. Lingane, J . J., and Jones, S. L., Anal. Chem., 1951, 23, 1798. Norwitz, G., Ibid., 1949, 21, 523. Scott, W. W., and Furman, N. H., “Standard Methods of Chemical Analysis,” Fifth Edition, D. Van Nostrand Co. Inc., New York, and The Technical Press Ltd., London, 1939, Volume I, p. 69. Betts, A. G., U S . Patent, 1902, No. 713,277 and No. 713,278. Sand, H. J., “Electrochemistry and Electrochemical Analysis,” Blackie and Son Ltd., London, 1940, Volume 11, p. 73. Kny-Jones, F. G., Lindsey, A. J., and Penney, A. C., Analyst, 1940, 65, 498. DEPARTMENT OF ANALYTICAL CHEMISTRY SCHOOL OF APPLIED CHEMISTRY N.S.W. UNIVERSITY OF TECHNOLOGY SYDNEY, AUSTRALIA April 27th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800651

出版商:RSC    

年代:1953

数据来源: RSC

摘要:

666 WYATT DCETHYLAMMONIUM DIETHYLDITHIOCARBAMATE FOR THE [VOl. 78 Diethylammonium Diethyldithiocarbamate for the Separation and Determination of Small Amounts of Metals Part I. The Successive Determination of Small Amounts of Copper, Manganese and Iron in Organic Compounds BY P. F. WYATT A scheme for the successive separation and absorptiometric determina- tion of copper, manganese and iron is described, whereby iron is isolated as cupferrate, and copper and manganese separately as diethyldithiocarbamates. Methods are described for the absorptiometric determination of all three elements as their coloured diethyldithiocarbamate complexes in chloroform, and alternative absorptiometric procedures are indicated. Tests on random mixtures of the three metals over the ranges 20 to 1000 pg of iron, 5 to 70 pg of copper and 10 to 200 pg of manganese show good recoveries.THE determination of small amounts of copper and manganese in dyestuffs and other organic compounds used in the rubber industry and in rubber-proofed fabrics is often required, because these metals accelerate ageing of the rubber. In the method described here, iron, copper and manganese are successively separated and determined absorptiometrically. Separation of iron is necessary to avoid interference with the copper and manganese deter- minations, and provision is made for its determination if required. The method depends on- (i) the quantitative separation of ferric iron from copper and manganese by extraction of its cupferron complex with chloroform, provided the reaction is carried out at sufficiently high acidity, (ii) the complete separation of copper from manganese by extraction of its diethyl- dithiocarbamate complex from mineral acid solution, and (iii) the formation of a highly coloured (purplish-brown) complex of manganese with diethylammonium diethyldithiocarbamate in acetate-buffered solution, which is extractable with chloroform.METHOD REAGENTS- All reagents must be of recognised analytical quality. Aqueous cup ferron soltdion, 5 per cent .-Prepare freshly from Hopkin and Williams' AnalaR reagent and filter. Diethylammonium diethyldithiocarbamate solution-Dissolve 2 g of Hopkin and Williams' AnalaR reagent in 200 ml of redistilled chloroform and preserve the reagent in the dark in an amber-coloured bottle. Reject the solution when it begins to show any yellow discoloration.Standard ferric iron solution-Dissolve 0.7023 g of ferrous ammonium sulphate, (NH,),S04.FeS04.6H,0, in about 100 ml of water containing 5 ml of concentrated sulphuric acid, add a slight excess of bromine water, boil off the excess of bromine, cool, and dilute to 1 litre a t 20" C with water in a graduated flask. One millilitre is equivalent to 100 pg of iron. Standard copper solution-Dissolve 0.3928 g of copper sulphate, CuSO,.SH,O, in water and dilute to 1 litre at 20" C with water in a graduated flask. One millilitre is equivalent to 1OOpg of copper. Dilute 10.0 ml of this solution to 100 ml a t 20" C with water in a graduated flask as required. Standard manganese solution-Measure 45-5 ml of 0.1 N potassium permanganate into a 250-ml beaker.Alternatively dissolve 0.144 g of potassium permanganate in about 50 ml One millilitre of this diluted solution contains 1Opg of copper.YOV., 19531 SEPARATION AND DETERMINATION OF SMALL AMOUNTS OF METALS 657 of water. To either of these solutions add 5 ml of diluted sulphuric acid (1 + 1) and then a saturated aqueous solution of sulphur dioxide, about 1 ml at a time, until the solution is just decolorised. Dilute to about 150ml with water and boil gently for 15 minutes. Cool, transfer to a 500-ml graduated flask and dilute to the mark a t 20" C with water. One millilitre contains 100 pg of manganese. Dilute 10.0 ml of this solution to 100 ml at 20" C with water to give a solution containing 10 pg of manganese per ml, as required.PREPARATION OF SOLUTION- Weigh accurately 2 to 5 g of prepared sample into a 100-ml Kjeldahl flask and decompose the sample with sulphuric and nitric acids, using 4 to 6 ml of concentrated sulphuric acid, depending on the weight of sample taken. Finally clean up the solution with a little perchloric acid.1 Evaporate twicc to fuming with 5 to 10-ml portions of water to remove all nitric acid, and allow to cool. Add 10 ml of water and 10 ml of diluted hydrochloric acid (1 + l ) , and boil gently for 3 to 5 minutes. Cool, filter through a small filter if necessary, washing with the minimum amount of water, transfer to a 50-ml separating funnel and dilute to 30 to 35 ml with water. Weigh 2 to 5 g of it in a 25 to 30-ml platinum crucible, support the crucible in a hole of suitable diameter cut in asbestos board, and reduce the contents to ash at as low a temperature as possible, finally igniting in a muffle furnace controlled at 500" & 50" C, until all carbon is removed.Fuse the ash with 2 g of potassium pyrosulphate, cool, heat with 10ml of diluted hydrochloric acid (1 + 1) and 10 ml of water until the melt is dissolved and add 5 ml of diluted sulphuric acid (1 + 1). Transfer the solution to a small beaker, add 2 ml of bromine water to ensure oxidation of all the iron, and boil off the excess of bromine. Cool, filter unless clear, transfer to a 50-ml separating funnel and dilute to 30 to 35 ml with water. With some materials it is preferable to reduce the sample to ash. Carry out a blank test on the reagents simultaneously with the test.SEPARATION AND DETERMINATION OF IRON- Add 1 ml of cupferron solution, mix well, add 5 ml of chloroform and shake the mixture for 30 seconds. Run the chloroform layer into a dry 25-ml graduated flask, wash with a few drops of chloroform, without mixing, to remove any drops of highly coloured extract, and transfer the wash to the flask. Add a further 0.5 ml of cupferron solution to the contents of the separating funnel and mix. If all the iron is removed the turbidity will be almost white. Extract with two or three further 5-ml portions of chloroform, adding the extract to the flask and taking care to introduce no water into the flask. Dilute to the mark with chloroform, mix, and immediately place in the dark. Transfer the aqueous layer to a 100-ml conical flask, add 1 ml of 60 per cent.perchloric acid and evaporate to fuming. Continue to heat, adding a few drops of nitric acid if necessary, until the residue is colourless, add 5 ml of water and evaporate to fuming. Measure the optical density of the chloroform extract in a 1-cm cell by means of a Spekker absorptiometer, using Ilford No. 601 violet filters with H503 heat filters and the solution obtained from the blank test as reference solution. Calculate by means of a calibration graph, prepared as described below, the weight of iron corresponding to the optical density. Calculate the amount of iron in parts per million in the sample. Preparation of calibration graj5h-Add known amounts of standard ferric iron solution containing 100 pg of iron per ml, equivalent to, say, 0, 100, 200, 300, 400 and 500 pg of iron, to a mixture of 5 ml of diluted sulphuric acid (1 + l ) , 10 ml of water and 10 ml of diluted hydrochloric acid (1 + 1) and cool.Transfer each standard in turn to a 50-ml separating funnel, dilute to 30 to 35ml with water, add 1 ml of cupferron solution and continue as described for the test solution. Determine the optical density of the solutions to which standard iron solution has been added, using the solution to which no standard iron solution has been added as reference solution. Plot a graph relating optical density to micrograms of iron. Nom-Measurement of the colour of the cupferron complex has the disadvantage that the excess of cupferron is apt to cause a green colour to develop, which increases rapidly in intensity, particularly if the extract is exposed to light, and this leads to incorrect results.For accurate work it is therefore safer to transfer the cupferron extract to a 100-ml conical Reserve for the determination of copper and manganese.658 WYATT : DIETHYLAMMONIUM DIETHYLDITHIOC.4KBAMATE FOR THE [Vol. 78 flask, add 2 ml of diluted sulphuric acid (1 + l ) , evaporate the chloroform, and decompose the residue with small amounts of nitric acid, followed by 0.5-ml of perchloric acid. Then complete the determination of iron by one of the well-known methods, e.g., by the thio- glycollate or o-phenanthroline method. Alternative determination of iron-Absorptiometric determination of the iron can be effected by means of diethylammonium diethyldithiocarbamate, as follows.After decomposing the chloroform extract of the cupferron complex, add 5 ml of water and evaporate to fuming. Add 15 ml of water, boil for 1 minute, cool and dilute the solution to 25.0ml in a measuring flask. Transfer the whole or a suitable aliquot portion of this solution, containing up to 250pg of iron, to a 50 or 100-ml graduated separating funnel, and add 20 ml of 4 M sodium acetate solution. Immediately add 10 ml of diethylammonium diethyldithiocarbamate reagent, shake vigorously for 1 minute, allow to separate, and carefully transfer the chloroform layer to a dry 25-ml graduated flask. Wash with 1 or 2 ml of chloro- form, without mixing, and run the washes into the flask. Extract with two further 5-ml portions of carbamate reagent, shaking for 30 seconds each time, transfer the extracts to the flask, and dilute to the 25-ml mark with chloroform.Mix and set aside in the dark for 15 minutes, and measure the optical density within the next hour, in a 1-cm cell, using Ilford No. 601 violet filters with H503 heat filters. If the solution shows the slightest turbidity, run it through a dry 9-cm Whatman No. 1 filter-paper before measuring the optical density. Preparation of calibration graph-Add known amounts of standard ferric iron solution containing 100 pg of iron per ml, equivalent to 0,50,100, 150,200, 250 pg of iron, to 2 ml of dilnted sulphuric acid (1 + 1) contained in a 50 or 100-ml graduated separating funnel, dilute to 25m1, add 20ml of 4 M sodium acetate solution and 10ml of diethylammonium diethyldithiocarbamate reagent, and extract the iron exactly as described for the test solution , diluting the combined extracts to 25.0 ml with chloroform in a graduated flask.Measure the optical densities of the solutions to which standard iron solution has been added against the solution containing no added iron as reference solution. Plot a graph relating optical density to amount of iron. SEPARATION AND DETERMINATION OF COPPER- Add 15ml of water to the reserved solution (p. 657), boil for 1 minute, cool and add 1 ml of 5 per cent. sodium metabisulphite solution. Transfer to a 50-ml separating funnel and dilute to 25 ml with the water used for rinsing. Add from a burette 10.0 ml of diethyl- ammonium diethyldithiocarbamate solution and shake vigorously for 40 seconds.Allow to separate, dry the stem of the funnel with filter-paper, and run sufficient of the chloroform layer into a 1-cm cell, filtering through a dry filter-paper unless the extract is perfectly clear. Cover the cell with a small sheet of glass and set it aside in the dark for 15 minutes. Reject the remainder of the chloroform extract and wash the aqueous layer with small amounts of chloroform with brief shaking until the chloroform washes are colourless. Transfer the aqueous layer to the original 100-ml flask, washing with a little water, and heat until occluded chloroform is evaporated. Continue to evaporate until the volume is reduced to 25 to 30ml. Cool and reserve the solution for the determination of manganese. Measure the optical density of the chloroform extract of the copper in a 1-cm cell by means of a Spekker absorptiometer, using Ilford No.601 violet filters with H503 heat filters and the solution obtained from the blank test as reference solution. Calculate from a calibra- tion graph, prepared as described below, the weight of copper corresponding to the optical density, and hence find the amount of copper in parts per million in the test solution. Preparation of calibration graph-Add 0, 1.0, 2.0, 3.0, 4.0 and 5.0 ml of standard copper solution containing 10 pg of copper per ml to 5.0 ml of diluted sulphuric acid (1 + 1) contained in 50-ml conical flasks, dilute to about 15ml with water and cool. Transfer each solution in turn to a 50-ml separating funnel, dilute to 25 ml with water and extract with 10.0 ml of diethylammonium diethyldithiocarbamate solution as described for the test.Measure the optical density of the solutions to which standard copper solution has been added using the solution to which no standard copper solution has been added as reference solution. Plot a graph relating optical density to amount of copper. SEPARATION AND DETERMINATION OF MANGANESE- Add a small piece of litmus paper to the reserved solution (above), make just alkaline with ammonium hydroxide, then make just acid with diluted hydrochloric acid (1 + l ) , andNOV., 19531 SEPARATION AND DETERMINATION OF SMALL AMOUNTS OF METALS 659 add 0.5 ml in excess. Add 0-5 ml of 5 per cent. sodium inetabisulphite solution and cool. The solution at this stage should be colourless.Transfer it to a 100-ml separating funnel, add 10 ml of 4 M sodium acetate solution and dilute to 50 to 60 ml with water. Extract by shaking for 30 seconds with 10 ml and then 5 ml of diethylammonium diethyldithiocarbamate solution, washing with a little chloroform, without mixing, between each extraction. Transfer the extracts and washings to a dry 25-ml measuring flask and dilute to the mark with chloro- form. Filter through a dry 9-cm Whatman No. 1 filter-paper into a dry 50-ml conical flask, close the flask with a cork or glass bulb, and set it aside in the dark for 5 minutes. Measure the optical density within the next half hour in a 4-cm cell by means of a Spekker absorptio- meter, using Ilford No. 601 violet filters with H503 heat filters and the solution obtained from the blank test as reference solution.Find from a calibration graph, prepared as described below, the weight of manganese corresponding to the optical density and calculate the amount of manganese in parts per million in the test sample. Preparation of cazibration gra$h-Add 0, 2.0, 4.0, 6.0, 8-0 and 10.0 ml of standard manganese solution containing 1Opg of manganese per ml to 5 0 m l of diluted sulphuric acid (1 + 1 ) contained in a 100-ml conical flask, dilute to 25 to 30 ml with water and continue as described for the test solution. Determine the optical density of the solutions to which standard manganese solution has been added, using the solution to which no standard man- ganese solution has been added as reference solution.Plot a curve relating optical density to weight of manganese in micrograms. NoTE-If desired, the diethyldithiocarbamate extraction may be used merely to separate the manganese, and the determination can be completed by the periodate method. This further step is always necessary if nickel and cobalt are present, as these remain with the manganese and form coloured complexes with the reagent. Transfer the combined chloroform extracts to a 50-ml conical flask, add 2 ml of diluted sulphuric acid (1 + l ) , evaporate the chloroform and continue to evaporate to fuming. Decompose the diethyldithiocarbamate by heating with small amounts of nitric acid, until a clear colourless digest is obtained. Add 10 ml of water and again evaporate to fuming. Dilute with 20 ml of water, add 2 ml of diluted phosphoric acid (1 + 1) and about 0.2 g of potassium periodate and boil gently for 1 minute.Set aside at just below boiling tempera- ture until the permanganate colour is fully developed, cool well and dilute to 25ml in a graduated flask. Measure the optical density in a 4-cm cell, using Ilford No. 604 green filters with H503 heat filters. Establish the calibration graph by adding known amounts of standard manganese solution (over the range 0 to 200pg of manganese) to 2ml of diluted sulphuric acid in 50-ml conical flasks and proceeding as described for the test solution. Measure the optical densities and plot a graph relating optical density to amount of manganese. Proceed as follows. DISCUSSION OF METHOD IRON- If the amount of iron present is abnormal, say, above 500,ug, it will be necessary to extract with further 0 6 m l portions of cupferron solution and 5-ml portions’of chloroform, until the last extract is free from iron colour, but a large excess of cupferron should be avoided.The measurement of the optical density will then be carried out at a suitable higher dilution. Provided pure reagent is used, the extracts are not exposed to bright light and the measurement of the optical density is not unduly delayed, good results are obtainable by direct measurement of the colour of the ferric cupferrate complex. If a number of tests are done, it is best to measure the optical densities of the blank value of the reagents and each test as soon as they are extracted, and to correct for the reagent blank reading.If it is decided to adopt the safer course and decompose the cupferron extract, final absorptiometric determination of the iron by means of thioglycollic acid or o-phenanthroline is quite satisfactory. Determination by extraction with diethylammonium diethyldithio- carbamate in chloroform from sodium acetate buffered solution as the dark brown ferric complex provides a simple, sensitive and accurate method of determination, as the cupferron separation will have isolated the iron from other interfering elements. COPPER- provided the acidity is sufficiently great. The cupferron extraction affords a most complete separation of iron.2 Although copper forms a stable complex with cupferron, it remains in the acid solution Before proceeding with the separation of the copper,660 WYATT : DIETHYLAMMONIUM DIETHYLDITHIOCARBAMATE FOR THE [Vol.78 it is essential to destroy the copper cupferrate and any free cupferron in the solution. Although it is customary to extract copper as its diethyldithiocarbamate complex from ammoniacal citrate solution, extraction from acid solution is equally complete, provided that a chloroform- soluble diethyldithiocarbamate is used, so that an excess of the reagent is maintained in the chloroform extract.l The procedure has the advantage of making the separation more selective, bismuth being the only element to interfere under the conditions described. If bismuth is present, the copper can be separated and re-extracted as follows. Shake the combined chloroform extracts with 2ml of 10 per cent.potassium cyanide solution for 15 seconds. Add 8 ml of water and transfer the chloroform layer to another separating funnel. Again shake it with 2ml of potassium cyanide solution, add 8ml of water and reject the chloroform layer, washing the potassium cyanide extracts with a little chloroform. Combine the potassium cyanide extracts in a 100-ml conical flask, add 2 ml of diluted sulphuric acid (1 + 1) under a hood, and boil down to half its volume to destroy the potassium cyanide. Cool, dilute to about 25 ml and extract directly with 10.0 ml of diethylammonium diethyl- dithiocarbamate reagent, measuring the optical density as before. MANGANESE- Manganese forms a purplish-brown compound with diethyldithiocarbamates, but this does not appear to be as stable as those formed by other metals.Although it can be extracted from slightly ammoniacal solution containing a little citrate or tartrate, it is better to extract manganese from acetate-buff ered solution under such conditions that a reasonably high concentration of diethyldithiocarbamate is maintained in the chloroform. By this means full recovery of manganese is obtained and the colour remains reasonably stable. SENSITIVITY- The sensitivity of the methods can be gauged from the observed optical densities shown in Table I, obtained with a Spekker photo-electric absorptiometer, Ilford filters and a tungsten-filamen t lamp. TABLE I SENSITIVITY OF THE METHODS Amount, Pg 10 20 30 40 50 60 80 100 150 200 250 300 400 500 Optical densities with Ilford colour filters A 1 > Iron Copper, Manganese Thioglicollic Diethyldi? Diethyldi- Fy I Cupferron acid thiocarbamate thiocarbamate thiocarbamate Periodate 0.23 0.45 0.67 0.88 1.11 - - - 0.105 0.215 0.32 0.43 0.535 0.80 1.06 - - -- I 0.092 0.185 0.38 0.565 0.77 0-95 - - - 0.595 0.89 1.19 Dilution, ml- 25 50 25 10 25 25 Cell size, cm- 1 4 1 1 4 4 Filter No.- 601 604 601 601 601 604 Green Filter c o l o u e - Violet Green Violet Violet Violet RECOVERIES- shown in Table 11. Recoveries obtained from random mixtures of known amounts of the three metals areNOV., 19531 SEPARATION -4ND DETERMINATION O F SMALL AMOUNTS OF METALS 661 TABLE I1 RECOVERY EXPERIMENTS Amount of metal added Amount of metal recovered and reagent used A h f- > I 7 Fe, cu, Mn, Fe, cu, Mn, Pg Pg PQ Pg Pg Pi? 500 20 20 490.0 (u) 20.0 (c) 18.5 (c) 100 50 50 102.0 (a) 49.5 (c) 48.0 (G) 300 70 100 295.0 (u) 68.0 (G) 100.0 (G) 400 50 80 400.0 (b) 49.9 (c) 78.0 (c) 200 20 20 203.0 (b) 20.6 (G) 21.0 (G) 50 70 50 52.0 (b) 69.0 (c) 47.0 (G) 300 10 10 291.0 (b) 10.6 (c) 10.3 (G) 1000 50 50 980.0 (b) 51.0 (c) 51.5 (c) 20 5 200 19.5 (b) 4-8 (c) 198.0 ( d ) 250 10 150 247.0 (c) 9.9 (c) 148.0 ( d ) 30.0 (c) 48.0 (d) 150 30 50 150.0 (c) Reugents-(a) Cupferron, (b) thioglycollic acid, ( c ) diethyldithiocarbamate, ( d ) periodate. REFERENCES Strafford, N., Wyatt, P. F., and Kershaw, F. G., Analyst, 1945, 70, 232. Strafford, N., and Wyatt, P. F., Ibid., 1947, 72, 54. 1. 2. DYESTUFFS DIVISION IMPERIAL CHEMICAL INDUSTRIES LTD. HEXAGON HOUSE, BLACKLEY MANCHESTER, 9 April 20th, 1953

ISSN:0003-2654    

DOI:10.1039/AN9537800656

出版商:RSC    

年代:1953

数据来源: RSC