EXCHANGE 8061 '12 WT ' The Quantitative Determination of Chromium THESIS Presented to the Faculty of the Department of Philosophy of the University of Pennsylvania, in Partial Fulfil- ment of the Requirements for the Degree of Doctor of Philosophy BY THOMAS RUSH ALEXANDER, JR. PHILADELPHIA, PA. 1915 The Quantitative Determination of Chromium THESIS Presented to the Faculty of the Department of Philosophy of the University of Pennsylvania, in Partial Fulfil- ment of the Requirements for the Degree of Doctor of Philosophy BY THOMAS RUSH ALEXANDER, JR. PHILADELPHIA, PA. 1915 WASHINGTON, PA. OBSERVER JOB ROOMS 1917 The writer takes this opportunity to express his indebtedness to the suggestions, advice and unfail- ing kindness of Dr. Edgar F. Smith for any merit which this investigation, may possess. INTRODUCTION Careful analysts know that in the gravimetric determination of chromium in the form of sesquioxide the latter is almost invariably contaminated with alkali chromate.- To correct this error it has been customary after the ignition of the chromium sesquioxide to digest it with water. In the presence of alkali chromate the water is colored yellow and it has been customary to treat the sesquioxide with water until it ceases to impart a color. The colored extracts or filtrates may be reduced to a small volume in a glass vessel and mercurous nitrate introduced. The resulting mercurous chromate, after filtration, is ignited in a weighed porcelaine crucible. There remains only chromium sesquioxide. The alkali chromate so often present in ignited sesquioxide of chromium is never negligible. The error intro- duced by it varies in magnitude and should always be taken into consideration. The purpose of the present study has been to determine under what conditions, if any, chromium sesquioxide may be obtained free from alkali. 444342 EXPERIMENTS A solution of chrome alum of known chromium content was prepared. Fifty c.c. were treated with ammonia gas and after the resulting chromium hydroxide had partially dissolved in the ammonium hydroxide, heat was applied until complete precipi- tation took place. The resulting precipitate was washed by decantation with water, approximately fifty c.c. of water being used for each washing; after which the washing was continued on the filter with the aid of suction. On evaporating the wash water chromium was not found in the residue nor was it possible to get a greater purity of the oxide upon using more than five hundred c.c. of water for washing purposes. The precipitation of chromium hydroxide took place first in beakers of Jena glass ; later dishes of fused silica and finally platinum dishes were used. The chromium hydroxide was always ignited in platinum cruci- bles. Upon washing it after ignition, with water, the latter in- variably took on a yellow color. Some of the results obtained were as follows : CHROME ALUM SOLUTION Chromic Oxide Jena Glass Quartz Platinum taken found found found 0.2349 gms. 0.2375 gms. 0.2362 gms. 0.2350 gms. 0.2349 0.2384 0.2358 0.2358 0.2349 0.2372 0.2348 0.2353 0.2349 0.2375 0.2356 0.2340 0.2349 0.2376 0.2362 0.2367 0.2349 0.2370 0.2359 0.2365 0.2349 0.2369 0.2363 0.2360 0.2349 0.2372 0.2360 0.2369 Average 0.2374 " 0.2359 0.2358 " % alkali 1.06 " 0.43 0.38 A second series of experiments w r as instituted with chromium sulphate. From appended results it will be seen that chromium determinations were more accurate, but such conditions do not ordinarily obtain in the course of an analysis. Further, the determinations indicate that the presence of even small quantities of alkali salts in the solution from which the chromium is pre- cipitated render the method unreliable. Results : CHROMIUM SULPHATE SOLUTION Chromic Oxide Jena Glass Quartz Platinum taken found found found 0.2392 gms. 0.2422 gms. 0.2400 gms. 0.2400 gms. 0.2392 0.2420 0.2380 0.2398 0.2392 0.2400 ' 0.2426 ' 0.2413 0.2392 0.2401 0.2413 ' 0.2392 0.2392 0.2419 0.2405 0.2408 0.2392 0.2417 0.2401 ' 0.2395 0.2392 0.2406 0.2396 ' 0.2395 0.2392 0.2400 0.2397 ' 0.2399 Average 0.2411 " 0.2402 ' 0.2400 % alkali 0.84 " 0.41 0.33 The method suggested long ago by Berzelius for the gravi- metric estimation of chromium is familiar to all analysts. Modi- fications of it have been suggested from time to time, but these too are so familiar that a detailed description of them is not necessary. On dissolving chronium hydroxide in an excess of fixed alkali and oxidizing the solution with hydrogen peroxide subsequently precipitating mercurous chromate and igniting it to chromium sesquioxide, satisfactory and concordant results were not obtained because it was impossible to completely elim- inate the alkali salts from the mercurous chromate even by pro- longed washing with water by decantation. The sodium or pot- assium salts in the solution were responsible for the error, hence an effort to eliminate them was made along the following lines. Into a solution of chrome alum of known chromium content was brought ammonia gas until a portion of the precipitated chrom- ium hydroxide had redissolved. To this solution hydrogen peroxide was added and at the end of half an hour the oxidation of chromium oxide to chromic acid was apparently complete. It was discovered later on that if sufficient time was allowed for the hydrogen peroxide to act in the cold, before applying any heat, the oxidation of the chromium to chromic acid was always com- plete. When heat was applied too soon a portion of the chrom- ium hydroxide almost invariably remained unoxidized. To a solution such as has just been described mercurous nitrate was added. The resulting mercurous chromate was filter- ed out and washed with a dilute solution of mercurous nitrate The chromium sesquioxide remaining after the ingnition of the mercurous chromate in a platinum crucible contained chromate, so that the results were always high. It seemed possible that more complete washing of the mercurous chromate might reduce the error, hence after executing the oxidation as outlined in the preceding lines, the liquid was diluted to about four hundred c.c. and heated to boiling ; mercurous nitrate was introduced and the resulting mercurous chromate was allowed to settle, the super- natant liquid being decanted through a filter. The residue of mercurous chromate in the breaker was boiled for a minute with fifty c.c. of water, allowed to cool, and a few drops of the mercu- rous nitrate solution introduced. After the precipitate had set tied the liquid was decanted and the washing repeated. Naturally small quantities of mercurous chromate were transferred to the filter in each decantation and the operation was continued until nearly the whole of the precipitate had been so transferred. The last traces of the precipitate were assembled from the beaker and the washing continued on the filter with dilute mer- curous nitrate, suction being applied after each washing. The results obtained by this procedure are as follows : Chromic Oxide taken 0.2948 gins. 0.2948 0.2948 0.2948 0.2948 0.2948 0.2948 0.2948 0.2948 0.2948 Chromic Oxide found 0.2931 gins. 0.2939 0.2939 0.2950 0.2947 0.2948 0.2945 0.2949 0.2939 0.2951 Average 0.2943 Average Error 0.16% The ease with which a cromium salt may be converted into a chromate by the method just outlined suggested that possibly it might also prove of value in the separation of chromium from other metals. SEPARATION OF CHROMIUM FROM IRON To portions of a solution of chrome alum slightly acidulated with sulphuric acid, various amounts of ferrous ammonium sul- phate were added and the treatment with ammonia and hydrogen 6 peroxide carried out as above. After the excess of ammonia and hydrogen peroxide had been removed the precipitated ferric hydroxide was filtered and washed, then transferred to a breaker and boiling water added. The supernatant liquid was yellow, but after filtering and washing again it was colorless. Examination showed that the ferric hydroxide still contained small amounts of chromium so the experiment was repeated with this modification : after the first treatment with ammonia and hydrogen peroxide the precipitate was allowed to settle and the supernatant liquid decanted, care being taken to keep as much of the precipitate as possible in the beaker. To the precipitate water was added and the treatment with ammonia and hydrogen peroxide repeated. The ferric hydroxide so produced was found to be free from chromium except in cases where relatively large amounts of iron were present. Under these conditions it is best to repeat the addition of ammonia and hydrogen peroxide several times. The combined filtrates and wash water were boiled to expel the excess of ammonia and hydrogen peroxide. During this operation con- siderable ferric hydroxide separted and was removed by filtra- tion. The chromium was determined by precipitation with mer- curous nitrate as already outlined. Tests failed to show the pres- ence of iron in the filtrate. The results were as follows : Mohr's Salt Chromic Oxide Taken Taken Found 0.0291 gms. 0.1873 gms. 0.1868 gms. 0.1054 0.1007 0.2901 0.8063 0.3112 8.0747 0.1873 " 0.1867 0.1873 0.1873 0.1873 0.1873 0.1873 0.1870 0.1869 0.1877 0.1872 0.1868 Average Error 0.16% SEPARATION OF CHROMIUM FROM MANGANESE It seems probable that the separation of chromium from manganese could be accomplished in the same way as the separa- tion of chromium from iron. A standard solution of potassium permanaganate was made, and portions of twenty-five c.c. were reduced with hydrogen peroxide after accidulation with sul- phuric acid. A standard solution of potassium dichromate was reduced in the same way, and after boiling each solution for a 7 few minutes to remove the excess of peroxide they were poured into the same beaker and treated with ammonia followed by the addition of hydrogen peroxide. It was found necessary to treat the precipitate so produced with ammonia and hydrogen peroxide several times in order to oxidize the last traces of chromium. In all probability there is a small quantity of man- ganous chromate formed which is only slowly decomposed by ammoniacal hydrogen peroxide. Complete solution of the chrom- ium was obtained in every case after three treatments with am- monia and hydrogen peroxide, the liquid being well boiled after each addition of hydrogen peroxide and the volume kept at about one hundred c.c. The separation of the chromium and the man- ganese was complete, as is shown by the following results. The chromium was precipitated as mercurous chromate and weighed as chromium sesquioxide : Potassium Permanganate Chromic Oxide Taken Taken Found 0.0183 gms. 0.2007 gins. 0.1999 gins. 0.0183 " 0.2007 " 0.2007 " 0.0183 " 0.2007 " 0.2009 " 0.0183 " 0.2007 " 0.2011 " 0.0183 " 0.2007 " 0.2004 " Average Error 0.04% The presence of a slight yellow color in the nitrates obtained after extracting the ignited chromium sesquioxide with water is not always indicative of the presence of soluble chromate, but may be due to the presence of very finely divided sesquioxide in suspension. During the course of this investigation it was fre- quently noticed that although filtrates so obtained had a yellow color they did not respond to the tests for chromates. By shaking the liquid with finely divided asbestos and filtering through a thick layer of asbestos in a Cooch filter the suspended sesquioxide may be completely removed. VOLUMETRIC DETERMINATION OF CHROMIUM Burrion and Seneschal* have recommended the following method for the volumetric estimation of chromium in chromic salts : ^Compt. rend., 157, 1528. The chromic salt solution after the addition of an excess of alkali is treated with an excess of hydrogen peroxide to convert the chromite to chromate. After boiling to destroy the excess of hydrogen peroxide, the solution is acidulated with sulphuric acid and the chromate reduced with a measured excess of ferrous ammonium sulphate and the excess of ferrous salt titrated with a standard potassium permanganate solution. Some determina- tions carried out as recommended did not give entirely satisfac- tory results. The values obtained were low and the reason seemed to be that the boiling did not entirely destroy the excess of hydro- gen peroxide in the alkaline solution. Alkaline chromate solu- tions showed the presence of hydrogen peroxide even after long continued boiling, for after acidulation under a layer of ether the blue perchromate color could be detected. In an attempt to complete the decomposition of the excess of hydrogen peroxide a piece of platinum covered with platinum black was introduced into the solution, but this did not serve to remove the last traces of peroxide oxygen and the results were correspondingly low. Hydrogen peroxide in contact with silver oxide produces metallic silver and oxygen is evolved, so a few drops of silver sulphate solution were added to the strongly alkaline chromate solution and the chromium determination carried out as recom- mended by Burrion.* The results obtained were still a trifle low ; this may be due to the presence of small amounts of metallic silver suspended in the solution, producing gradual reduction of the potassium permanganate. This explanation seems to be all the more probable in view of the fact that the end point of the titration with potassium permanganate was rather indefinite. The purple color was slowly discharged, and if an amount of permanganate sufficient to produce a color which remained for one minute was used, the quantity of chromium indicated was low. The same titration, carried out with a potassium dichromate solution to which no hydrogen peroxide had been added, gave good values for chromium and a sharp end point. It is possible, after becoming accustomed to the use of this method to select an end point which will give quite satisfactory results. *Loc. cit. The substitution of ammonia for fixed alkaline hydroxides proved to be a great benefit, owing to the ease with which the last traces of active oxygen could be expelled from the solution. The plan adopted was the same as the one used for convert- ing the chromic salt to chromate preparatory to precipitation with mercurous nitrate. After saturation with ammonia and addition of hydrogen peroxide the reaction was allowed to con- tinue in the cold until complete. After boiling, the chromate solution was acidified and a weighed coiantity of ferrous am- monium sulphate added. The solution was diluted to a volume of nearly one liter and the excess of ferrous salt titrated with potassium permanganate. The end point of the titration is quite distinct, the only percaution necessary being that a large volume of water be added, as recommended by Burrion. The results given below were obtained by using a solution of potassium dichromate which had been standardized against iron wire and ferrous ammonium sulphate in the usual way. Portions of twenty-five c.c. were taken from a burette and reduced after acid- ulation, with hydrogen peroxide ; the excess of oxygen being removed by boiling and the chromium redetermined as indicated above. CHROMIC OXIDE Takne 0.2007 gms. 0.2007 0.2007 0.2007 0.2007 0.2007 0.2007 0.2007 0.2007 Found 0.2009 gms. 0.2013 0.2010 0.2002 0.2010 0.2006 0.2010 0.2008 0.2006 Average Error +0.04% Finally portions of twenty-five c.c. of the standard potassium dichromate solution were reduced with ferrous ammonium sul- phate and the excess of ferrous iron determined by titration with potassium permanganate. The solution was then subjected to the action of ammonia and hydrogen peroxide, the precipitated iron and manganese removed and the chromium redetermined in the filtrate by precipitation with mercurous nitrate and ignition to chromium trioxide. 10 At least three treatments with ammoniacal hydrogen per- oxide were required to completely remove the chromium. The results obtained were : CHROMIC OXIDE Taken Found by Titration Found by Precipitation 0.2007 gms. 0.2008 gms. 0.2005 gms. 0.2007 " 0.2010 " 0.2004 " 0.2007 " 0.2011 " 0.2009 " Average Error +0.14% 0.04% SUMMARY 1. In order to convert chromic salts qualitatively into chromic oxide, they should first be converted into mercurous chromate and then ignited. The quantity of alkali salts present should be small. To insure this condition ammoniacal hydrogen peroxide can be employed as an oxidizing agent. The precipi- tation of mercurous chromate should take place from a dilute solution. 2. Chromium may be readily separated from iron and man- ganese by means of ammoniacal hydrogen peroxide. 3. Ammoniacal hydrogen peroxide may be used to convert chromium salts to chromates preliminary to the volumetric deter- mination of chromium with a ferrous salt and potassium perman- ganate. 11 UNIVERSITY OF CALIFORNIA LIBRARY, BERKELEY - THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW Books not returned on time are subject to a fine of 50c per volume after the third day overdue, increasing to $1.00 per volume after the sixth day. Books not in demand may be renewed if application is made before expiration of loan period. *-. c ' - '5Jan'55DPZ 195501 REC'D Lw> NOV 22 1957 9 REC'D DEC 17 1957 olul'l mi 8 15m-4 ,'24 < UNIVERSITY OF CALIFORNIA LIBRARY