2660 OILUSTONE mm POUND SOLUBILITY INFLUENCES. CCCLXIII.-Solubility Influences. Part I . The EiJect of Some xalts Sugars and Temperature on the Solubility of Ethyl Acetate in Water. By S m GLASSTONE and ALBERT POUND. ALTHOUGH much work has been done on the influence of salts on the solubility in water of various non-electrolytes (for chief refer-ences see Eyre Brit. Assoc. Rep. 1910 447 ; 1912 820; Rivett and Rosenblum Trans. Faraday Soc. 1914 9 297; Linderstrom-Lang Conapt. Rend. Trav. Lab. Curbberg 1924 15 No. 4) there seems to have been very little attempt made a t a systematic investigation. Very little work too has been done on the influ-ence of non-electrolytes such as sugars on the solubility of other non-electrolytes such as ethyl acetate ether and aniline and as far as the present author is aware there has been no systematic investigation of the effect of a mixture of substances either elec-tzolytes or non-electrolytes on the solubility of a sparingly soluble, neutxal substance.It seemed very probable that a complete examination of the so-called " salting out " effect would throw mme light on the larger problem of solution and the work described below was intended to be a contribution towards a more systematic survey of the problem than has yet been made. In the present work the solubility of ethyl acetate has been determined at 25" and a t 50" in pure water and in the presence of various sugars and of the chlorides bromides and iodides of the alkali metals and of ammonium. Some rough measurements have also been made of the solubility of various salts in ethyl acetate saturated with water (about 3%) and an interesting qualitative connexion between these values and those of the solubility of ethyl acetate in salt solutions has been established.The solubility of ethyl acetate in water has also been determined a t 0" lo" and 37". Philip (J. 1907 91 711) has shown the advantage of expressing solubility results of this kind in terms of grams or gram-moles per lo00 grams of solvent rather than per litre of solution. A further change is now made the results being expressed in terms of the number of gram-mols. of water required to dissolve one gram-mol. of ethyl acetate in the presence of various molecular quantities of added sugar or salt; in this form the results are useful for various calculations and for comparison with one another.EXPERIMENTAL. Ethyl acetate made by Roberts's method ( J . Soc. Chem. Id., 1924 43 295~) was purified from alcohol by distillation over calcium chloride and then by several fractionations over phos PART I. THE EFFECT OF WISE SALTS suams ETC. 2661 phorua pentoxide; only the final constant-boiling fraction was used in this work. The salts and sugars were the pnrest com-mercial specimens mainly supplied by the British Drng HOW, LM. and no attempt w a made to purify them; slight impurith had very little effect on the solubility of ethyl acetate and this was probably less than the experimental error. The variom solutions were made up by careful weighing of the substance and the water and were saturated with ethyl acetate a follows Since the solubility of ethyl acetate in water and in aqueous solntions decreases with rise of temperature (see below) the solution was shaken with a sllght excess of the ester at a temperature below 25" or 50" and placed in a thermostat at 25" or 50"; the e x w of ethyl acetate then separating caused the liquid to become cloudy.In the course of an hour or two the aqueous liquid waa clear again, and was a saturated solution of the ester at the temperature of the thermostat. Care was always taken that the excess of ethyl acetate present was not so lmge that the amount of water or salt dissolved by it could not be neglected. For analysis a quantity of the saturated solution (3 to 8 g., depending on the ester concentration) was traderred rapidly in a warmed pipette to a stoppered bottle and weighed; care was taken that none of the acetate layer was drawn into the pipette.In the cases of the various salts which it was desired to recover, and of the sugars the weighed solution was diluted with water, washed into a distilling flask and the ethyl acetate and some water distilled over and collected in water in such a way as to avoid loss of ester. The residue in the flask wa8 always tested to make sure that no acetic acid which might have resulted from the hydrolysis of the ester by boiling water remained behind. The ester wa8 then hydrolysed with standard sodium hydroxide and estimated in the usual way. When the original solution contained an ammonium salt distillation wa8 always n-, and in case the salt had become hydrolysed during this process and ammonia distilled over the solution after alkali hydrolysis, was boiled for a few minutes without the reflux condenser in order to expel ammonia.If distillation wtw unnecessary the weighed sohtion was diluted and hydrolysed directly. Resulk-The columns headed m and w give the number of g.-mols. of added substances and of water respectively required to make a solution which will be saturated with 1 g.-mol. of ethyl acetate at the temperature stated. The significance of the figures in the column headed R is explained onp. 2664. In the absence of any added substance the value of w (wo) is 66.15 at 25" and 80.98 at 50" 2662 QLdssTONE AXD POUND SOLWIIZl!Y z " C E S . Lithium chloride.712. to. n. 0.667 76-67 15-7 2.27 98.92 14-5 6.28 133.7 10.7 24-85 231-5 6.7 134.5 476.2 3-3 0.858 98.65 20-6 2.905 125.9 15.0 8.08 171.8 11.2 31.55 294.1 6.7 181.2 641-0 3.1 Sodium chloride. 0.219 70.68 20.7 0.477 76.65 22.0 1.540 94.34 18-3 4-86 142.0 15.6 12.03 218.8 12-7 27-10 352.1 10.6 59.30 581.4 8.7 0.273 88-12 26-1 0.589 94.80 23.4 2-02 123.5 21-0 6-43 188.0 16.6 14.40 261.8 12.6 33.42 434-8 10-6 85-00 833-3 8.9 Potassium chloride. 0-168 69.15 17-8 0.369 74-75 23.3 1.093 86.21 18-3 3-06 112.9 15-2 6.96 155-0 12.7 12-19 203.7 11.3 22.90 284.9 9.5 0.213 87.67 -0.458 92.73 25.5 1.438 113.3 22.4 4.09 162.7 17.6 8.87 197.6 13.1 14-70 245.7 11-2 31.07 386.1 9.8 Rubidium chloride. 0-578 77-52 19-7 1.454 93.10 18-5 5.22 135-9 13-3 14.20 224.7 12-5 67-8 546-4 7.1 Lithium bromide.m. w. n. 25". 0.287 68-16 7.0 0.777 71-19 6 4 1.824 77.14 6.0 4.85 91.19 5.2 13-26 96-27 2.3 15.66 40.65 -50". 0.354 84.03 8.6 0.936 85-47 4.8 2-325 98-46 7.5 6.12 115.2 5.6 18.40 133-3 2-9 19.58 83.5 -Sodium bromide. 25". 0.229 71-10 -0-630 76.76 16.8 1.560 87.72 13.9 4.91 121.2 11.2 12-10 173.3 8.8 47-40 367-6 6.4 50". 0.294 91.66 -0.794 96-64 19-6 2-69 116.3 13.1 6.43 158-7 12.1 15-48 221.7 9.1 52.60 408.2 6-2 Potassium bromide. 25'. 0-249 69.20 12.3 0.610 76.70 17.3 1.39 84-03 12.8 4-29 114.9 11.3 9-44 148.6 8.7 21-90 220.8 7.0 50". 0-370 85-09 11.0 0-737 92.59 15.7 1.774 107-1 14.7 5.46 146.0 11.9 12-65 199.2 9.3 30.68 308.6 7.4 Rubidium bromide.25". 0.421 73.53 17.5 0-993 80-65 14.5 2-65 94-69 11.2 7-66 132.5 8-6 24-00 222-7 6.5 Lithium iodide. m. w. n. 0.476 65.90 -0.906 62.60 -1.967 58.40 -3.44 38.5 -0.566 78.5 -1.112 76.9 -2.36 70.5 -5.02 56.0 -0.912 4-56' -Sodium iodide. 0.163 67.14 6.7 0-432 68-25 4.8 0-774 68-78 3-5 2.165 73.37 3.3 6-46 82.58 2.5 21.32 109-4 2.0 0-243 84.37 13.6 0.535 84-75 7.0 0.978 86.58 5.6 2-81 95.24 5.0 8-44 107.8 3.2 27-76 142.5 2.2 Pota,ssium iodide. 0.150 68.03 12.6 0.395 69-69 9.0 0-860 71.43 6-1 2.02 75.19 4.5 6-45 87.72 3.3 20-19 125-0 2-9 0.181 82-26 7.0 0.473 83-33 4.9 1.045 86.81 5.5 2-44 90.91 4-1 7.71 106.7 3.3 24-22 149.9 2.8 Rubidium iodide. 0.298 66-82 2.2 0.635 68.78 4.1 1475 69-69 2.4 2.66 73.42 2.7 5-69 82-58 2.9 12.53 107-0 3.PABT I. !PEE EFBECI! OF SOME SAL!l!S SUGAES ETO. 2663 Rubidium chloride. m. W. n. 0.742 94.5 18.2 6-62 172-4 13.8 84.4 680.3 7.1 c8?sium chloride. 0.427 74.52 19.6 1-004 86-21 20.0 3-065 113.6 15-5 7.04 163.1 12.4 26-70 294-1 8.5 132.5 671-1 4.6 0.530 92-59 21.9 3-802 141.0 16-7 29.95 330.0 8.3 197.8 1ooO-0 4-6 Ammonium chloride. 0.525 76-54 1-568 89.14 4-28 112.4 8.41 139.7 14.95 174-5 0.632 92.14 1.973 112-2 5-66 148.4 11-42 189.8 18.61 216.9 Dextrose. 0.0692 68.6 0-1423 69.3 0-392 73.6 0.871 78-6 1.486 84.1 2-262 90-6 3-315 98-5 0.0826 81.8 0*1708 83-1 0.462 86.6 1.016 91.5 1-705 96.6 2.593 103-6 3.722 110-7 194 14-7 10.8 8.7 7-2 17.6 15.8 11.9 9.4 7.8 25.4 22-0 19.0 14.3 12-1 10-8 9.7 9.7 12.3 12-1 10.3 9.1 8.7 8.0 Rubidium bromide.m. 10. n. 60'. O-SO2 87-64 13.2 3-23 119-8 12-0 40.5 375.9 7.2 Caesium bromide. 25". 0.321 70.68 14-1 0-693 74.52 12-1 1-835 86.81 11.2 3.47 97.96 9-1 8-80 135.0 7.8 50". 0.400 88.12 17.8 2-236 105.8 11.1 11.92 179.9 8-3 35-10 275.5 5.5 Ammonium bromide. 25". 0.273 70-0 0.725 74-4 1.680 81-9 4.33 94-0 8-55 109.5 16.38 122.7 50'. 0.338 86.5 0.839 90-8 2.107 102.5 5.55 120.6 11-77 160.8 21.78 177-0 L~VUloSe. 0.394 73.8 0.867 78-3 1.445 81-9 2.178 87-0 4.270 99-8 8965 131-0 25'. 50'. 0-475 89.0 1-045 94-3 1-720 97-4 2.620 101.0 4-76 111-2 8.89 134.6 13.6 11.4 9.4 6.4 5-1 3-5 16-2 11-7 10-2 7.1 5-9 4.4 19-4 14.0 10.9 9-6 7-9 7.5 16.6 12.7 9.5 7-9 6-3 6.1 Rubidium iodide.m. w. n. 0.358 80.33 -1.788 84.37 1.9 7-26 106.3 3.3 C h u m iodide. 0.237 67-79 6.9 0.638 68.87 6.1 1-241 70.78 3.7 2-70 72.30 2-3 0.290 82.88 6.5 1487 84-79 2.6 3-44 93-01 3.5 Ammonium iodide. 0.422 64.4 -0.866 63.2 -1.87 60.8 -4-74 57.4 -12.42 62-1 -0.613 80.4 -1.094 79.6 -2.418 78.7 -6.28 76.2 -16-32 88.8 0-5 sucn>ee. 0-0734 0.190 0-406 0.696 1.010 1-936 3-57 6.10 0.091 0.233 0.508 0.833 1.206 2.36 4.18 6-81 66.4 68.5 12.9 69.6 8.5 75-0 12.7 77-6 11.3 86.7 10.6 101.5 9.9 115.9 8.1 82.1 83-9 87.0 89.6 92.9 106.4 119.0 €29.2 12-1 12.4 11.8 10.3 9.9 10.4 9.1 7-Lactose.25". 50". m. W. 0-0706 69-4 0.190 72.5 0-423 77-4 0.739 83.0 1 -094 55.7 1-575 94.7 I 7 n. m. 46-0 0.0545 33-4 0,222 26-6 0-487 21-4 0-854 20-6 1.234 18-1 1.762 W . 83.3 84.7 89-0 1oc)-o 105-9 94.8 7 n. 27.1 16-7 16.4 16.2 15-6 14.1 DiscusSion. In general the effect of one substance in reducing the solubility of another has been explained along two merent lines. Euler (Z. phySiW. Chem. 1899 31 360) suggested that the addition of a mlt to water increases the internal pressure md this results in a decreztsed solvent power for a neutral solute; this theory was supported by Geffcken (ibid.1904 49 257) but watj adversely criticised by Levin (ibid. 1906 55 503). On the other hand, Rothmund (ibid. 1900 33 401) suggested that the reduction in solubility is due to the added salt becoming hydrated in solution, so that the molecules of water involved in the hydration are no longer available for the dissolution of another substance. Although Rothmund later criticised this point of view (ibid. 1909 69 523), it received support from Baur (Ahren's Samdung 1903 8 466), Lowry (Trans. Fura~?uy Soc. 1905 1 197) and Philip (J. 1907, 91 711). The last author was the fist to make use of solubility determinations in salt solutions in order to calculate the hydration values of various salts ; by assuming that the reduction in solubility of the neutral substance is entirely due to the water molecules removed in the salt-hydrate average values for the number of molecules of hydrate water per molecule of salt can be obtained.Philip's method of calculating hydration values has been applied to the results obtained in the present work and the average degree of hydration of the salts etc. at various concentrations which is equal to (w - w,,)/m is given in the column headed n. Since there is some doubt it5 to the condition of molecules and ions in concentrated solutions the hydration values for different salts are beat compared at infinite diIution; consequently all the hydration numbers obtained above have been extrapolated roughly to zero concentration. The results are given below ; as the values do not vary appreciably between 25" and 50° average results are recorded.Hydration of Salts at InJinite Dilution. LiCl ... 27 XaPl ... 24 KCI ... 22 hiH,C1 ... 21 RbCl ... 20 CsCl ... 20 LiBr ... 9 ? XaBr... 19 KBr. .. 17 NH,Br ... 16 RbBr ... 16 CsBr ... 15 LiI ... ? NaI ... 14T KI ... 129 NHJ ... ? RbI ... 5 ? CsI ... 8 ' PLBT I. THE EmEm or SOME SALTS m u m. 2665 The multa for the lithium rralte and almost all the iodides are uncertain; the solubility Sgnres in i@cs (pp. 2662-3) show that lithium bromide at the high& concentrations and lithium and ammonium iodides at almost d concentrations increase the solubility of ethyl acetate in water. The increase is probsbly connected with the formation of a compound between the salt and the ester; concenfrated solutions of iodides containing ethyl acetate have a distinct yellow colour which is not due to the presence of free iodine and can be attributed only to the presence of some complex aubgtance in solution.There was no evidence of a meta-thetical reaction between the salt and the eerfer. The remarkable fact that at 25" ethyl acetate and a 60% solution of lithium iodide in water are miscible in all proportions merib further investigation. Some experiments have been made on the solubility of the various salts in ethyl acetate saturated with water; lithium chloride, d u m potassium ammonium rubidium and durn iodides are slightly soluble whilst lithium bromide and iodide are comider-ably soluble (roughly 30% and Myo respectively at 25"). It is with the salts which are soluble in ethyl acetate therefore that anomalous results have been obtained; it follows then that whenever the added salt either combines with or is soluble in the neutral solute the hydration values calculated by the method described above are useless.Although the hydration values given above a,re in good agree-ment with those calculated by other authom from a variety of solubility measurements (Philip loc. cit.; Philip and Bramley, J. 1915 107 377; McArthur J . Physical Cbm. 1916 20 496; Thorne J. 1921 119 262; Manchot Jahrstorfer and Zepter, 2. umrg. Ch. 1924 141 G) it d m not follow thaf hydration is the main or even the subsidiary cause of salting-out. This effect may be due to some other fundamental property of each ion or molecule which is independent of the nature of the mb-etance being salted-out provided no compound formation o(x1111& In the case of the mgm some of the hydration vdues calc~&td by the method described above appear to be incredibly large and to vary considerably with temperature and so it is probable a t other factors are operative.It is seen also that molecules of a non-electrolyte have the power of reducing very considerably the solubility of ethyl metab in water. Euler and his co-workers (2. lYW-. 1917 23 192; 2. plrysid. Ch. 1924 140 113) appeatr to have tacitly assumed that only ions are responsible for salting-out but it is clear that this assumption is not justifiable. McKeown also ( J . Amer. Ciaem. SOL 1922 44 1203) in attempting to aswas the salting-out power 2666 GIJBSL'ONE AND POUND SOLUBILlTY INFLUENCES.PART I. of sodium- and chlorine-ions amume~ that the dting-out effect of undksociated molecules is very small; he finds that when ether is salted out with sodium chloride the effect is entirely due to the chlorine ions. It should be pointed out however that since the equations from which this result is obtained me admittedly approximate the conclusion is of little value; h o s t equally good agreement with most of the equations may be obtained by rising entirely Werent values from those of McKeown. In the case of ethyl acetate it is clear that both anions and Bations have definite salting-out power; the kations would be placed in the order Li>Na>K>NH,>Rb>Cs and the anions in the order c1> Br > I. Various authors have attempted to obtain equations which connect the solubility of a neutral substance in a salt or other, solution with the concentration of added salt; in general an equation of the type log 8 = a - kc has been found to be most satisfactory where 8 is the solubility of the neutral substance, c the concentration of salt and a and k are constants (for references, see Thorne h.cit. and Linderstrom-Lang loc. cit.). In the present work it has not been possible to find any one equation which will fit all the results up to the highest concentrations of added salt ; in general the logarithmic equation was found to hold good in the form log w = h / w + a where a and k are con-stants for a given salt and w and rn have the same meaning as before up to concentrations of 2-3N.For some salts-Bodium, potassium and rubidium chloridethe agreement waa very good almost up to the srtturation point. In those cases in which the d t was soluble in ethyl acetate the logarithmic equation waa not obeyed at all. In the presence of lithium chloride the solubility of ethyl acetate may be expressed by the straight-line equation w = lcma + a and a similar equation holds good for the more concentrated sugar solutions. The fact that the results will not all conform to one simple equation suggesh that the salting-out effect is due to several difEerent factors on which it is hoped, further investigations will throw light. 8olubility of Ethyl Acetate in Water. As the literature is very deficient in measurements of the solu-bility of ethyl acetate in water at different temperatures a number of determinations have been made by the method described above for salt solutions with the following results :-Grams of Ethyl Acetate dissolved by 100 gram~ of Water.10.40 at 0"; 8.96 at 10"; 7-39 at 25"; 6-65 at 37"; 6-04 at 50' The following @ma have been obtained by interpolation from a graph :-7.85 86 20" ; )I.W 8f 30" ; 6.50 8t 40". The solubility thus decreases steadily aa the temperature is raised from 0" to 50". stbmmw. (1) The solubility of ethyl acetate has been determined at 25" and 50" in solutions of the chloride bromide and iodide of the alkali metals and of ammonium and in solutions of dextrose, laevulose sucrose and lactose. (2) It is shown that hydration of the salt may be one of the factors responsible for the salting-out effect; this effect may, however be due to some other fundamental property of salt ions or molecules. Molecules aa well as ions probably have considerable salting-out power. (3) The solubility of ethyl acetate in salt solutions is best e x p d by a logarithmic equation e.g. log w = kmlw + a; the application of this equation however is limited to the more dilute salt golutions. (4) The solubility of ethyl acetate in water has been determined at 0" lo" 25" 37" and 50" ; the solubility decreases with increasing temperature. The authors are indebted to the Chemical Society for a grant from its Research Fund which defrayed part of the expense of this work. Their t h a h are also due to Mr. A. L. Stephens B.Sc., and Mr. W. R. P. Hodgson B.Sc. for valuable d t a n c e . UN'IVEaSrrP COLLEGE b T E B . [Received Aolguet 11M 1925.