1072 DAWSON AND MCCRAE : METAL-AMMONIA COMPOUNDS IN CXI1.-Metal-Ammonia Compounds in Aqueous Solu- tion. Part IV. The influence of Temperature on the Dissociation of Copper-Ammonaa Sulphate. By H. M. DAWSON and J. MCCRAE. IT has already been shown (this vol., p. 496) that the distribution of ammonia between water and chloroform at 20° is not quite independent of the concentration, and the variation is exhibited by the distribution coefficient curve on p. 497. In our first communication (Trans., 1900, "7, 1239), we have shown that a complex compound, probably Cu4NH,*S04, is formed when excess of ammonia is added to a solution of copper sulphate ; this complex compound is dissociable, and only when there is a large excess of ammoriia in the solution does the number of fixed ammonia molecules per atom of copper approximate to four.We have been able to follow the extent of this dissociation a t constant temperature with decreasing total ammonia concentration, and now we have endeavoured to ascertain the influence of temperature on the dissociation. On account of the method which we employ, it was impossible to use a temperature much above the ordinary, and we have not deemed it advisable to work above 30°, which is our upper limit. For lower temperatures, we are only limited by the freezing point of the solution, but the lowest temperature at which we found it convenient to work was 109 This gives us a range of ZOO, and it was expected that the extent of dissociation at the lower and higher temperatures would be sufficiently different to indicate the temperature influence.AQUEOUS SOLUTION.PART IV. 1073 Diatribution of Ammonia between Water and Chloroform at varying Temperatures. The method of experimentation was exactly as has been described in our previous communications (Zoc. cit.). I n order to ascertain the amount of ammonia fixed by the salt in solution from the distribution coefficient determined with this solution, i t is necessary to know the dis- tribution ratio between pure water and chloroform, and since this varies very considerably with the temperature, determinations of the latter had to be carried out at different temperatures. The temperatures chosen were 10' and 30°, and the following results were obtained : Concentration of NH, in aqueous part. Grams per litre. Temperature 10" : 5'211 6'974 8.123 8-701 12'835 17'370 17'428 Temperature 30" : * 6 '092 6.658 8.497 10 -203 11 '660 13.601 16.660 Concentration of NH, in the chloroform. Gram per litre.0-1686 0.2261 0'2633 0'2847 0.4253 0,5864 0.5862 0.2250 0.2949 0,3761 0'4558 0'5232 0'6133 0.7575 Distribution coefficient, - C1. c1 k. 30.91 30'84 30 734 30'57 30 *I 8 29'67 29.73 22'63 22.57 22.65 22-39 22 29 22.18 22-00 * In a previous publication (Trans., 1900, 77, 1243), we have given the co- efficient at 30" as 23'2, but we prefer to take the numbers given above for the reason pointed out on p. 496. These figures show exactly the same relationships as those previously obtained for the distribution coefficient at 20° (this vol., p. 496), namely, for solutions less than 0*5 normal with respect to ammonia the distri- bution coefficient remains practically constant at constant temperature, but with more concentrated solutions this ratio diminishes with increas- ing concentration of tbe ammonia.The variation of the distribution coefficient within the same limits of ammonia concentration is very nearly the same at the three temperatures, consequently the curve1074 DAWSON AND McCRAE : METAL-AMMONIA COMPOCNDS IN given on p. 497 may be used to express the results obtained at 10' and 30' if we add to the ordinate 4.54, or subtract from it 3.66. I n other words, the three curves representing the dependence of the dis- tribution coefficient on the ammonia concentration at the different temperatures are almost parallel. It is evident that the distribution coefficient at constant ammonia concentration is not a linear function of the temperature, for the difference in the coefficient €or loo (from 20° to 30°) is in one case 3.66 units, whilst in the other case, for loo (from 10' to 20°), it is 4.54. The results obtained have been plotted as curves (not reproduced here), and from these the values at any concentration of ammonia can be read off.Experiments at 10' and 30° with Copper Sulphate Solutions. A t 10Oand 30°, experiments were made with 0.1 normal copper sul- phate solutions, and at 10' with 0.05 normal solutions. The results so obtained are recorded below, and they may be compared with those which we have previously found at 20°. In the table on p. 1075 we give the various data, the molecules of ammonia fixed per molecule of salt (or per atom of copper) being cal- c , - h 2 culated from the formula, 17,07? 2 in which k is the distribution coefficient for pure water at the same concentration of ammonia in the chloroform, c1 and c2 the observed ammonia concentrations in the aqueous and chloroform layers, and n the normality of the salt solu- tion.We are unable to take account of the physical action of the dissolved material, but, as already stated (p. 511), we believe this to be, in the case of coppermdts, very small. When the distribution coefficient attains a high value, correspond- ing to a small amount of free ammonia in the salt solution, the error in the determination of the fixed ammonia may be relatively large on account of the small quantity of acid used in the titra- tion of the chloroform.Further, if the coefficient is not much greater than that for pure water, the accuracy is not great, since the calculation involves the difference between these two values. The most accurate values are those obtained when the concentration of ammonia in the aqueous phase is between 6 and 9 grams per litre. As no extreme accuracy can be claimed for the individual figures, we have drawn smoothed curves representing the molecular amount of ammonia fixed per atom of copper in 0.1 N solution at loo, 20", and 30" with varying ammonia concentration. The numbers for 20' have been recalculated in accordance with the later determina-AQUEOUS SOLUTION. PART IV. Concentration of NH, in aqueous layer. Grams per litre.C1' 1075 Concentration of NH, in CHCI,. Gram per litre. c2' Temperature 10". 0.1 N copper sulphate : 5-427 6.688 7'171 8-931 10.675 13.130 8-322 0.0765 0.1153 0.1317 0.1679 0.1872 0.2446 0.3273 0-05 N copper sulphnte : 3.580 5.347 7.095 9'800 0.0666 0.1224 0*1786 0,2679 Temperature 30: 0.1 N copper sulphate : 5.347 7.062 8'650 10.310 11.960 13-720 0.1072 0.1790 0.2441 0.3170 0 '391 5 0'4699 Distribution coefficient. k'. 70.92 58.01 E4.45 49 *56 47.71 43.65 40.12 53.76 43'68 39.73 36-58 49 *86 39-45 35.44 32.55 30.55 29-21 Coefficient corresponding to c2 for water. k. 30'94 30.91 30.89 30'88 30.87 30.82 30.57 30.94 30.90 30.87 30.75 22'W 22.64 22.62 22-60 22.52 22-39 Molecular ratio fixed NH,. 17-01 n of cuso,: e, -kc, 2,- 1 : 3.52 3 '66 3'64 3'69 3.67 3 '66 3'68 3 '56 3 '66 3 '70 3.66 3 '42 3'53 3'66 3-68 3.68 3-76 tions of the distribution coefficient between pure water and chloro- form at this temperature.Instead of reproducing these curves, me have constructed the table given on p. 1076 indicating points on the curves. As would be expected, the figures indicate that, at constant ammonia concentration, the extent of dissociation increases as the temperature rises, and it may be stated that if a comparison be made between the numbers obtained with 0.05 Ncopper sulphate solutions at loo and 20°, the same concluaion is evident. The differences between the numbers obtained at the different temperatures are, however, too small to admit of any quantitative statement as to the influence of1076 MELDOLA AND EYRE: 6 grams per litre ............ 7 , , , , ............ a , , , , ............ 9 , , , , ............ Table showing the number of molecules of ammonia Fxed per atom of copper in 0.1 N solution. 3.55 3'63 3'67 3.70 Concentration of NH, in the aqueous phase. 10". 20". 3 '46 3.55 3 '61 3.67 1 30". -I 3'42 3 *51 3 '58 3-62 temperature on the dissociation of the complex copper ammonia sulphate. THE YORKSHIRE COLLEGE, LEEDS.