1526 DOUCJAL: EFFECT O F HEAT C - E f e c t of Heat on Aqueous Solutions of Clz~onze Alum. By MARGARET DOUIE DOUGAL. IT has long been known that when an aqueous solution of chrome alum is heated, its colour changes from violet or purple t’o green ; the alteration, which begins at about 70°, is apparently complete in a few minutes, even in nioderately strong solutions, a t 100’. On evapo- ration, the green solution yields a green, nou-crystalline mass. If the liquid is allowed to stand, the green colour persists for some time, especially if the solution is dilute, but it gradually reverts t o the original purple, and ultimately the solution deposits the dark purple octahedra of ordinary chrome alurn. Various explanations have been given of this phenomenon. According t o Schrotter (Aqiqa.Plzys. Chem., 1841, 53, 513), Lowel (J. Phuym., [3], 7, 32l), and, more recently, I h r d (Compt. Tend., 84, l089), the alteration of colonr is simply the result of the change of bydration ; Berzelius (Ann. Phys. Chem., 61, 1) and Premy (Compt. rend., 47, 883) considered it was due to the formation of a basic salt, whilst Fischer (Kustney’s arc hi^, 14,164) and Jacquelain (Conzpt. rend., 24, 439) surmised that it was the direct effect of the separa- tion of the chromic sulphate from the alkali sulphate. None of these explanations is altogether sufficient. I t might be inferred from Graham’s work on liquid diffusion that, chrome alum would undergo dissociation in aqueous solution even in the cold, and therefore tlhe mere separation of the chromic sulphate from the potassium sulphate would not alone account for the change.That the solution experiences a profound change on heating is obvious f r o a the circumstance, pointed out by Favre and Valson (Cowpt. rend., 74, 1023), that only a, portion of the sulpliuric acid present in the green solutioii is pre- cipitable in the cold by barium chloride, and that prolonged boiling with barium chloride is needed to effect its entire separation as barium sulphate ; whereas the violet solution at ordinary temperatures at once yields up its sulphuric acid to barium chloride. The greenON AQUEOUS SOLUTIONS OF CHROME ALUM. 1527 solution further differs from the violet in containing, as first shown hy Kriiger ( A m . Phys. Chem., 61, 218), free sulphuric acid. At Professor Thorpe's suggestion, and under his direction, I have been led to investigate this phenomenon, in the expectation that the true explanation might be found to depend upon the formation of Recours's chromosulFhuric acid, which is known to behave towards barium chloride in the remarkable manner above mentioned.The inquiry was begun in the early part of 1894, in the labora- tory of t,he Royal College of Science, South Kensington, but, owing to other engagements, its completion has been delayed. I n view of the recent work of Monti ( N . Cirn., 1896, 43, 212-216) and of W. R. Whitney (Zeits. physikal. Chern., 1896, 20, 40), I am zow induced to put together the results I have obtained. A variety of physical methods may be employed to throw light upon the nature of the change attending the alteration in colour of the chrome alum solution.Sprung has studied the variation in the viscosity or internal friction of the violet and green solutions, and Lecoq d e Boisbaudran, by dilatometxic observations, has noticed the change in density. I find that the alteration in density of even a dilute solution of chrome alum, after boiling, may readily be detected by the specific gravity bottle. A weighed amount of chrome alum was dissolved in a known volume of water, and portions of the solution were heated to about 100' under such conditions that no evaporation was possible. The weight of the green liquid required to fill a specific gravity bottle was then determined, a precisely similar experiment, under identical conditions of temperature, &c., being made with the violet solution.The several weights of the two solutions obtained in three experiments were as follows : Violet solution. Green solution. Grams. Grams. I. ......... 100.4028 100- 376 5 IT. ......... 101-2537 101.1712 111. ......... 2025096 102.3567 The first experiment was made on a 1 per cent,. solution, the second on a 2$ per cent., and the third on a 5 per cent'. solution. The actual amount of the change depends on the duration of the heating and on the lengt'h of time which has elapsed since the green solution had been prepared. I t will be observed, however, that in all cases the green solution was less dense than the violet solution; hence the transforniation is accompanied by expansion, the result probably, of the production of the free sulphuric acid. I have, further, made a series of observations on the relative rates of the diffusion of the green and violet solutions.The method 5 L 21528 DOWAL: EFFECT OF HEAT adopted was that of Graham. Two pt-ecisely similar flat-bottomed glass crystallising dishes, of equal diameter and depth, and of about 100 C.C. capacity, mere placed each at the bottom of a flat-bottomed glass dish of about 1000 C.C. capacity, and of equal depth and width. These were placed on felt. Each of the smallel* dishes was then filled up to a definite point with the solution to be diffused. The dif- fusion experiments were carried on in an underground room of fairly uniform temperature, into which the direct sunlight never penetrated. The solutions to be converted into the green variety were boiled for several hoursin a flask connected with a reflux condenser.They were cooled as rapidly as possible by means of ice, and were allowed to stand in the cellar where the experiments were made, together with the violet solution, until both had acquired the temperature of the air. Every known precaution was taken to make each pair of experiments as strictly comparable as possible. The required amount of water, which, meanwhile, had been standing along with the flasks containing the chrome-alum solution and the necessary apparatus in the diffusing room, was run in fyom a dropping funnel, the end of which was furnished with a narrow, glass tube, drawn out till merely a fine jet of liquid passed, and bent at right angles, so that the water as i t flowed impinged on the side of the outer basin, and, in falling, made as little disturbance as possible in the water already in the dish.As the water reached the top of the dish containing the chrome alum solution, it was allowed to impinge on the side in such manner that the overflow of the water into the inner vessel should be as gentle as possible. The depth of the pure water above the level of the chrome alum solution was about the 8ame as that of the solution itself. The outer vessel was covered with a glass plate, a bell-jar placed over the whole, which was then covered with brown paper. In this way it was hoped that the temperature-change would be as little as possible. At the end of the period of diffusion a small glass plate, which was gently introduced into the liquid by means of a stout platinum wire, arranged so that it held the circular plate by three hooks and was readily detachable, was adjusted so as to cover the small dish. The diffusate was then syphoned off into a beaker, the few drops remaining in the dish and on the surface of the plate being removed by a finely pointed pipette. The chromium iu each diffusate was determined by precipitation with ammonia in the usual way, and the sulphuric acid in the filtrate estimated by boiling with hydrochloric acid and barium chloride, Iu all, seven independent pairs of experiments were made with two different solutions, each containing about 1 per cent.of chrome alum. The results may be tabulated as follows :ON AQUEOL-S SOLUTIONS OF CHROME ALUM.1529 Cr. ---. 0 ‘0446 0 -0345 0 9490 0 -0652 0 -0656 0 -0670 0 -04’76 Volumo of solution of chrome nluin. 0’1’742 0*1.158 0 ‘2035 0 -2697 0.2802 0-2859 0.1’7’78 90 *o C.C. 2 , 2 1 1&0 :: 150.5 ,, 143.0 ,, 64.5 ,, 0.0393 0.0314 0 -0465 0 -0641 0.0629 1 0.0643 i 0.0469 Volume of water into which diffu- sate p s d . Violet solution. Green solution. SOJ. - -~ 0 -1818 0 -1 531 0 *2184 0 293 1 0 *2929 0 -3101 0 *1788 All the experiments made are given. In considering the results it will be understood that the several pairs are, strictly speaking, only comparable with one another, as the conditions in the case of one pair and another were not necessarily the same ; time, temperature, atrength of diff usate, amount of init’ial change, amount of retrogres- sion, &c., d l affecting the results.I n the outset there was nothing to determine the best strength on which to make the observations. In dilute solutions the retrogression from green to violet is slowest ; but, on the other hand, the actual mass of the salt diffused depends on the initial strengt-h of the chrome solution. In every case it is noticed that the green diffusate contains less chromium and more sulphuric acid than the violet solution ; this is compatible with the assumption that the green solution contains a colloidal, and therefore slowly diffusing chromosulphuric acid. So far as they go, then, these results are in harmony wit’h the coii- clusion already drawn by Monti and, by inference, by Dr. Whitney, t h a t when a solution of chrome alum is heated, it is resolved into a mixture of potassium sulphate, chromylsulphuric acid, and free sulphuric acid.As Whitney has shown that only one-sixth of the sulphuric acid originally present in the chromic sulphnte is changed into free sulphuric acid, whereas Favre and Valson found that the green solution gives up one-third of the sulphuric acid present in the original chromic sulphate to barium chloride, in the cold, we may formulate the entire change as follows : The general result, however, is unmistakable. ~[C~.I,(SO,)~,K~,SO,] + H,O = [C1~~O(SO,)~]SO~ + 2K2S04 + HgSO,. The curious fact that the green solution of chrome alum, on heating, transforms its contained snlpharic acid into a state in which it, is not precipitnble in the cold by barium chloride, has its analogy in the remarkable behaviour of certain cobaltamines. Thus, as Jiirgensen bas shown, chloropurpureo-sulphnte does not yield hydro-1530 POPE : THE REFRACTION CONSTANTS cbloric acid on heating with concentrated sulphuric acid, nor does its solution give a precipitate with silver nitrate. We must assume, therefore, as in the case of the chromosulphnric acid, that one portion of the acid radicle has not the same relation towads the rest of the compound as the other. It is hardly to be expected that the fact of the elimination of the acid radicle, as free sulphuric acid, in the course of the change of the violet chromic sulphate to the green compound will be found to be singular. The observations of Gerlach would seem to show that iron alum behaves in a, somewhat similar manner. The curious behaviour of aluminium acetate, discovered many years ago by Crum, appears to be a, phenomenon of the same order. Here, too, we observe the gradual elimination of acetic acid, dissipated because i t is volatile, with the simultaneous formation of a colloidal substance. 1 desire, in conclusion, to express my thanks to Professor Thorpe for his advice and assistance in the prosecution of these experiments.