\3Zt L 94 3 THE EFFECTS OF NON-ELECTROLYTES ON THE SOLVENT POWER OF WATER BY CLARENCE B. LOVELL THESIS FOR THE Degree of BACHELOR OF SCIENCE IN CHEMICAL ENGINEERING COLLEGE OF LIBERAL ARTS AND SCIENCE UNIVERSITY OF ILLINOIS 1921 Digitized by the Internet Archive in 2015 https://archive.org/details/effectsofnonelecOOIove I wish to express my sincere thanks and appre- ciation to Dr. J. H. Keedy for the interest that he has shown in thi b investigation, and for his cooperation in interpreting the experi- mental data. - ‘ . - - ' fable of Gontenta I . II . Ill . IV. V. VI . Introduction ....... Historical Experimental Materials ....... Procedure ...... Glucose Solutions Glycerol Solutions .... Acetone Solutions . Urea Solutions ..... Di scussion Solvent Action of Organic Solutes Group Influences ..... Hydration Effects Salt Formation in Urea Solutions . Restrictive Action . Page 1 1 2 3 6 11 1 ? 30 31 94 34 37 35 Summary 31 Bibliography Table of Plate3. Page I . Benzoic Acid in Glucose Solutions 4 II . Salicylic Acid in Glucose Solutions 5 III. Specific Gravity Curve of Glycerol Solutions 8 IV. Benzoic Acid in Glycerol Solutions 9 V. Salicylic Acid in Glycerol Solutions 10 VI. Specific Gravity Curve of Acetone Solutions 13 VII . Benzoic Acid in Acetone Solutions 14 VIII . Salicylic Acid in Acetone Solutions 15 IX. Specific Gravity Curve of Urea Solutions 17 X. Benzoic Acid in Urea Solutions 18 XI. Salicylic Acid in Urea Solutions 19 XII. Composition Diagram: Benzoic and Salicylic Acids in Glucose Solutions 92 XIII. Composition Diagram: Benzoic and Salicylic Acids in Glycerol Solutions 93 XIV. Composition Diagram: Benzoic and Salicylic Acids in Acetone Solutions 25 XV. Composition Diagram: Benzoic and Salicylic Acids in Urea Solutions 96 > ) . The Effects of h on - Electrolytes on the Solvent Power of Water . I . Introduction * The effects of non-electrolytes on the solvent power of water have not been developed very thoroughly. That is to say, that while the solute iray repress the solvent power of water, the solute itself may have a specific solvent action, and in this way the total solubility may be increased rather than diminished. It is to be expected that two media which are solvents in the absence of each other should retain their solvent power when mixed. This specific solvent power of each, however, may be considerably modified by the presence of the other. Furthermore, solids, which in the dry condition are not considered solvents, may when dissolved function as sol- vents. The accumulation of some data on this point has been the purpose of this investigation. II . Historical . The influence of dissolved substances on the solvent action of water has been extensively investigated, both from the theoretical and experimental standpoints. Presuming the complete analogy between liquid and gaseous solutions, Nernst developed the doctrine that "The concentra- tion of the molecules of the solute remains unchanged by the (1) - . ' - * r . p presence of other substances ^ The same principle was as- sumed by A. A. Noyes^ in his work on the influence on solu- bility of the presence of dissociated substances. Arrhenius,' 3 ’ however, using the data of Hojes on the solubility of thal- lium chloride, showed that this was not even approximately true; that the solvent power of a given medium was diminished both by the presence of dissolved molecules and by the pres- ence of ions. The influence of ions in reducing solubility is well illustrated in the familiar "salting out" effect. The theoretical aspects of this repression of solubility have never been satisfactorily developed. The attraction of the electric charges of the ions doubtless bring about a contrac- tion of the solvent water ( M ©lectrostrietion" 4 ) which would cause a diminution of its solvent power. The withdrawal of the solvent water to form hydrates of the molecules and ions as suggested by Jones® and others, would result in a decrease in the active mass of the water, and a corresponding reduction in its solvent action. For the most part, however, chemists agree that a sufficient explanation of this effect has not been given. Perhaps all that can be done at the present is to accept the evasive, but helpful expression of Washburn® that the "thermodynamic environment" of water is modified by the presence of dissolved substances. Ill . Experimental . Material s --For this work, weak organic acids are espec- ially suitable because of their rather low solubility in water, their almost negligible ionization, and the ease by which - -7 • ' 3 their solubility can be accurately determined by titration. Benzoic and salicylic acids were chosen as the solutes for this work. Both of these were "chemically pure" products of the Merck brand, and their purity was checked by the fol- lowing tests: The benzoic acid melted at 121° -122 ° , and its solubility at 25° was found to be 0.336 grams per 100 grams of water. The salicylic acid melted at 156°, and its solu- bility at 25° was found to be 0.222 grams per 100 grams of water. These results agree well with the accepted values found in the literature. As the non-electrolytes for the solvent medium, glycerol, glucose, urea and acetone were used. These were of good grade, although when dissolved all of them except acetone showed a slight acidity. Consequently blank titrations were run in all determinations in these solu- tions, and corrections were made for the acidity indicated. Solutions of these were made up as needed in strengths vary- ing from 1 J to 25^o by weight in 100 parts of solution by volume. Upon studying the curves obtained, it was found that when the results were converted to a percent-by-vreight basis -- that is, grams per 100 grams of solution -- an apparently simpler relation was brought out. The specific gravities of these solutions were either obtained from the "Landolt-Born- stein Tabellen" or by actual determination, using a Westphal balance. Procedure --These solutions were prepared in a closely regulated thermostat. An excess of the solid acid was added to 200 cc. of the aqueous solution in a 500 cc . Erlenmeyer flask. The flask was stoppered with a rubber stopper through - I ' : e k epnb fega 4>1 * ■ i; if ' : : :?; ■*. ■ ■ - i i 4 ;PI O 3"i CS ;-M L W u r-WT * Z5 Gim Per IOOGcrr\ (*lijceiro\ So\u.tion GirnDissoWed-By H*0 Present in /oo&m GiUjcerot Solution 20 JS 10 Gtrn YerlQQ&m Solittipa; -J #_1 I 11 especially in the case of benzoic acid. Another interesting, fact is that benzoic is more soluble than salicylic acid. This was the case in all determinations and is in accord with Bourgoin’s® work. He found that below 40° salicylic acid is less soluble than benzoic, while above 40° it is more soluble. Glycerol Solutions --The results of solubility determi- nations in glycerol solutions are shown in Table II , and the corresponding solubility curves in Plates IV and V. These data, if plotted on the grams-per-lOG cc. basis, give a curved Table II . Solubilities in Glycerol Solutions . Benzoic Acid * Salicylic Acid Glycerol by Weight Solubility in 100 g. Calculated Solubility in Solubility in 100 g. Calculated Solubility in Solution Water present Solution Water present at /° G G G G 0.0 .3360 .3360 .2220 .2220 .998 .3365 .3330 . 2225 .2195 4.95 . 3450 .3200 .2260 .2120 9.765 .3520 .3030 .2340 .2000 14.5 .36 20 .2875 .2410 .1898 19.1 .3730 .27 20 .2495 .1795 S3. 5 .3839 .2570 .8565 .1698 line. On the other hand, the graph obtained on a basis of grams per ICO grams of solution is a straight line. The sol- ubility cf benzoic acid and salicylic acids in glycerol was determined in the usual way, and values of 1.93 grams and 1.59 grams, respectively, per 100 grams of solvent. From these values the solubilities of these acids may be calculated for the glycerol fraction of the solution, just as it was calcti- lated for the water fraction, on the aseiimption that there is . ' r.- 9 . 1? # no restrictive influence involved. For example, in 100 grams of a 23.5 fc solution of glycerol, the solubility of benzoic acid in the water fraction (76.5 grams) would be .2570 grams; in the glycerol fraction (23.5 grams) .4536 grains. The total solubility calculated on the assumption of no restrictive in- f fluence would then be .7365 grains. The value actually found is .3839 gram. The difference, .3267 grams, represents the total repression of solubility in this solution. In the same way the restrictive influence for salicylic acid in 100 grams of 23.5$ glycerol solution was calculated to be .2870 grams. This conclusively shows that solubility is not an additive quantity; that while both components of the solution exert a solvent action, there is seme restrictive influence in opera- tion. The solubility curves in glycerol and glucose solutions are very similar, which strongly suggests that analogous rela- tionships exist in both cases. The striking difference is, glycerol solutions are the better solvents. Acetone Solutions --The results of the solubility deter- minations in acetone solutions are given in Table III, and the corresponding solubility graphs in Plates VII and VIII. The solubilities increase very rapidly for the higher concen- trations of acetone. As in the case of glycerol solutions the total solubility is not the sum of the solubilities of the water and acetone fractions, considered separately. Upon plotting the solubilities on a ” composition” diagram (See Plate XIV, page 25), curves resulted which are similar - , * % 13 ior\$ j±i±i 14 Grm Per /oo Gem Solution. 15 PlateMT Solubility of Saiicijlic Acid, in Different Percent Solutions of Acetone vO »n (*m Per ioo(krr\ of Solution 17*k- . GrnVzr 100 Urea So tut von |SolabiUtij of Benzoic Acid inDif jerervll Percent Solutions of llfea. frrH— « \gr t*> C* I 'oajiyi }uao3^ ^1 $L m Gm Ver/oo(km of Solution. (* mper lOO&m Urea Solution 19 piatcxi Solubilituof SaUcuhc Acid, in Different Percent Solutions of Llrea. £ d (n Q) ft c o 2 •< 4-* m 3 CO T>3yrn *o &m Per looGm of Solution .. ; Table III. Solubilities in Acetone Solutions. 20 Benzoic Acid Salicylic Acid Acetone by Weight Solubi li ty Calculated Solubility Calculated in 100 g. Solubility in in 106 g. Solubility in Solution Water present Solution Water present * G G G G 0.0 .3360 .3360 .2820 .8220 1 .0025 .3610 .3329 .2320 .2200 5.04 .5150 .3190 .3320 .2110 10 . 18 .8 850 .3020 .5510 .1995 15.35 1.3900 .2846 .9220 .1880 80.65 2.4650 .2663 1.6100 .1760 86.00 5.1200 .2489 3.4600 .1644 to those obtained by Seidell 9 for ethyl alcohol solutions. He, however, offered no explanation why these graphs assumed such a shape. It is also to be noticed that the curves of both acids when drawn on a basis of grains of acid per IOC grams of solution have practically the same slope. Urea Solutions — The results of solubility determinations in urea solutions are shewn in Table IV, and the solubility graphs in Plates X and XI. The solubility of both acids in- creases rather rapidly with concentration of urea. Both Table IV. Solubilities in Urea Solutions . Benzoic Acid Salicylic Acid Solubility Calculated Solubility Calculated urea in 100 g. Solubility in in 100 g. Solubility in by Weight Solution Water present Solution Water present % G G G G 0.0 .3360 .3360 .2220 .2220 .997 .3455 .3330 . 8290 .2200 4.93 .4040 .3195 .3043 .2110 9.75 .4790 .3030 .3860 .2005 14.45 .57 25 .2880 .4710 .1900 19.00 .6700 .2720 .4775 .1800 23.48 .7650 . 257 2 ♦ 48 25 .1700 * • * . • . . V • * . . - ■* . ‘ or' . ' . . . curves bend slightly up to a urea concentration of about 14.45??, and then an inflection takes place, not very marked in the ben- zoic acid curve, but sharp in the salicylic acid curve. The peculiarities of these graphs will be discussed later. I V . Discussj on . Solvent A ction of Organic Solute s --From the solubilities as determined above, it is seen that the solubilities of ben- zoic and salicylic acids in a definite amount of solution is decreased by the presence of glucose, but increased by the pres- ence of glycerol, urea and acetone. The concentration of the water in the glucose solutions, however, is greatly reduced. For example, 100 cc. of a 25% glucose solution contains only 84.46 grains of water. From this it may be seen that the sol- ubility in glucose solutions is too great to be accounted for by the solvent action of the water actually present. It fol- lows, therefore, that either the presence of the glucose in- creases the solvent power of water, or that glucose itself has a solvent action on the acids. The first assumption is rejected since it is at variance with the work of Noyes° and Arrhenius^, who found that the presence of dissolved substances, especially of ions, tend to reduce rather than increase the solvent power of water. On the other hand, it is not sur- prising that an organic substance like glucose should be a solvent for benzoic and salicylic acids. This viewpoint is supported by the following experiment: Some solid glucose and benzoic acid were carefully heated together in a test . : , •- . - - » i ■ ■ HtO ioo °fo Percent Composition Glucose m% 10 X'O *\Q j\Q fo L\0 1 i jo . ' 60 . j ft ? .. i . . 1 too Hxo too*/* Percent Composition G\\\cero\ iqqJo °4 tube, giving a perfectly homogeneous mixture. There was a slight darkening during the process due to the formation of caramel and a little steam was evolved. Upon cooling a layer of benzoic acid separated on top of the melt. Group Inf luences - -Certain inferences may be drawn as to the effect of certain groups on the solvent power of an or- ganic solvent for benzoic and salicylic acids. First of all, it should be noticed that the presence of OH-groups reduces the solubility of these acids, as evidenced by their low sol- ubility in glucose and glycerol solutions. The fact that glucose is a poorer solvent than glycerol is doubtless due to the larger number of OH-groups in the former. The increased solubility in acetone solutions might be anticipated from the two CH^-groups, since both acids are fairly soluble in hydro- carbons. The CO-group is probably hydrated in solution, be- coming C(GH)p, and has a repressing influence on solubility. As 100 % acetone is approached, however, (see Plate XIV, page ° 5 ) , the solubility becomes greater, indicating that anhydrous acetone is a much better solvent than the hydrate. This leads to the conclusion that the CO-group has a greater solvent ac- tion on these acids than the C(OK)s> -group. From the same point of view, the high solubility in urea solutions is to be attributed to the NH^-groups. As confirmatory evidence of this may be cited the fact that aniline, CgHg.NH^, is a better solvent for these acids than benzene, CgHg. Hydration Effects - -Jones and his collaborators have in- J . . L. ' - Composition Gurve ofB enzoic and. Sal icy Uc Actd-* in Acetone Solu tion Composition Curve of Benzoic and Salicijic Acids in lire a Solutions H*0 /ooy 0 Percent Composition Urea my 0 sis ted strongly that the deviation of solutions from what might be called their "ideal" behavior is due to hydration effects. According to this "solvation” hypothesis, molecules of the solvent combine with solute molecules and ions to form loose molecular compounds called "solvates”. This theory gives the most immediate explanation of the fact that the solvent power of solutions does not follow the law of mixtures. The water is partly consumed in forming, hydrates of the solute, thereby reducing the number of effective molecules. In many cases it is probable that a series of hydrates is formed, each having a specific solvent action of its own. To illustrate: Sei- dell 9 found that camphoric acid in water-ethyl alcohol mixtures had a maximum solubility in about 87$ alcohol. This indicates that alcohol molecules of a certain degree of hydration have a higher solvent action than the pure alcohol. For the other acids whose solubility he measured in water-alcohol mixtures, the maximum solubility was for 100$ alcohol. This change in hydration makes solubility effects all the more complex, and since probably other influences are involved, it may be suf- ficient to attribute this deviation from ideal behavior to a modified thermodynamic environment of the solution. Salt Formation in Urea Solutions — The possibility of salt formation in the case of urea solutions is very strong, and it is altogether likely that the increasing solubility of the acids with increase in the urea concentration is largely due to this effect. Such action would affect the solubility of °8 the acids in two opposing ways, viz.: The withdrawal of urea from the solution to form salts would reduce the solvent power of the solution, since urea seems to have a high specific sol- ubility for these substances; on the contrary, the withdrawal of the acid by salt formation would increase its solubility. Attempts to prepare urea benzoate and urea salicylate failed. h either of them has been reported in chemical literature, and so their existence, it must be granted, is problemmatical . If they do exist in solution they are probably rather soluble and considerably hydrolyzed by the water. Upon attempting * to re- move the water by boiling, urea only would remain since both of the acids are readily volatile in steam; if the water were removed by evaporation at low temperatures, the free acids, by reason of their low solubilities, would tend to separate first . The rather abrupt break in the solubility curve of sal- icylic acid in urea solutions (see Plate XV, page °6) may be easily interpreted by means of the Fhase Rule in the light of the assumption of salt formation. In a urea solution contain- ing an excess of salicylic acid, there are (counting the water) three components. The phases are three — ■ solid salicylic acid, solution, vapor. Therefore, substituting in the ex- pression for the Phase Rule, F = G + 2 - P, we obtain F = 2. That is, there are two degrees of freedom. From this it would follow that for a definite temperature, 25°, there should be a definite solubility of salicylic acid for every concentra- tion of urea. But, by reference to the figure, it will be i . * _ — - seen that above 14.45^urea concentration the solubility of the acid becomes practically constant. This means that another phase has been added to the system, presumably solid urea salicylate, so that now F = 1. There is to be mentioned another complication in dealing with urea solutions. The decomposition of urea into ammonia and carbon dioxide is catalyzed by acids, the latter uniting with the ammonia to form ammonium salts. That this does occur was shown qualitatively by adding an excess of salicylic acid to 100 cc. of 25 % urea solution. Upon standing for several days at temperatures slightly above that of the room the excess of acid gradually went into solution with an evolution of carbon dioxide. While ammonium salicylate is almost neu- tral and would affect the titration results only slightly, it may considerably influence the solubility of the molecular salicylic acid. Restri cti ve Action - -The simplest mode of behavior when a substance is dissolved in a mixture of two mutual solvents is that each of the solvents should retain its specific sol- vent power. In such a case the solubility of the substance should be additive and would follow the law of mixtures. The solubility graph in such a case would be represented by a straight line, as shown diagrammatically in Figure 1 (see next page), where A and B are the two solvents and the solubility of the solute C is represented by the ordinates. The straight line DE would represent this ideal behavior — that is, when . - . ■ - - » 30 too?, ft /oo%B the solvent power of one solute is not changed by the presence of the other. More generally the solubility graph is a curve of the general form DFE, as will be discussed later. Ben- zoic acid and salicylic acid in glu- cose solutions (see Plate XII, page p P) give graphs that appear straight lines up to a concentration of 50^ glucose, which is somewhat above the limiting concentration for satisfactory work. This may be due to the fact that the curvature is inappreciable, and too much significance should not be given to the solubility values for pure glucose as in- dicated by the projection of these graphs. More probably the true values would be considerably greater. On account of glu- cose being a solid at 25 ° no attempt was made to determine these values directly. Figure 1 In the case of binary solvents, the general condition is that solubility is not an additive quantity, and does not fol- low the law of mixtures. Each solvent seems to restrain the solvent action of the other, so that the combined solubility is less than if each were operating separately. The graph is not a straight line but a curve, of the general shape DFE. An expression for the solubility of a substance in a bi- nary system of solvents may be developed as follows: Let A and B be the two solvents, where a and b represent the number . j ■ of grans (or moles) of each present, and 3 the solubility. The expression for the total solubility is then S = aS^ + bS-g - abk. The last term represents the repression of solubility and has not justification beyond the assumption that it is dependent on the concentrations of both A and B. This last term becomes nil whenever a or b is zero — that i 3, whenever the solute is a single substance. The graph of the above equation is of the general shape of the curve DPS in Figure 1. By using the data for a given curve values for k may be derived which are roughly constant. Table V shows the values calculated from the foregoing solubility data. Table V. Values for k in £> = aS A + bSp - abk . Benzoic Acid Salicylic Acid entration in in in in Solution 1 Glycerol Acetone Glycerol Acetone 1% k = .00243 k = .00273 k = .00202 k = .00 281 5 jo .00253 .00276 .00208 .00301 10 fc .00264 .00279 .00217 .00309 155$ .00 274 .00279 .00226 .00314 20 $ ♦ 00 286 .00250 .00236 .00308 .00300 .00157 .00247 .00254 Average , .00 270 .00252 .00226 .00 295 V . Summary . 1. The solubilities of benzoic and salicylic acids have been measured in solutions of glucose, glycerol, acetone and urea at 25° for the range of ifo to 23.5^6 by weight. 2. These organic solutes exert a solvent action on ben- * - - ' 32 zoic and salicylic acids. 3. The influence of certain groups on solubility has been considered. 4. The total solubility in all cases was found to be less than the combined solubilities of the components operating separately . 5. This repression of solvent action i3 probably largely due to hydbation effects. 6. The formation of urea salicylate in urea solutions, as interpreted by the Phase liule, was indicated. 7. An empirical expression for solubility in a binary medium has been developed. ' ' . 33 VI . Bibliograp hy . 1. Hernst: Theoretical Chemistry, p. 357 (1911). 2. A. A. Hoyes: Z. physik. Chem., 6 , 243 (1890). 3. Arrhenius: Z. physik. Chem., 51 , 824 (1899). 4. Drude und Herns t: Z. physik. Chem., P5, 79 (1894). 5. H. C. Jones: Z. physik. Chem., 15 , 419 (1894). 6. Washburn: Principles of Physical Chemistry, p. 179 (1915). 7. Hoffman und Langbeck: Z. physik. Chem., 51, 385-434 (1905)* 8. Bourgoin: Compt. rend., 87, 62-64 (1903). 9. Seidell: Trans. Am. ■Electroch. Soc., 13, 319-331 (1908).