Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
155 The Influence o f Hydrogen-ion Concentration upon the Adsorption of Weak Electrolytes hy Pure Charcoal.— Part II. By H akold J ohn P helps , B.A., B.Sc., Department of Biochemistry, Oxford, Ramsay Memorial Research Fellow. (Communicated by C. N. Hinshelwood, F.R.S.—Received April 30, 1931.) In a recent communication results were described which indicate th a t propionic, hexoic, and succinic acids are adsorbed by purified charcoal only as unionised molecules.* There was no evidence th at anions of these acids were adsorbed to any measurable extent. Thus the amount of these acids adsorbed from mixed solutions of any one acid and its sodium salt is propor tional to the amount of unionised acid present as calculated from the known ionisation constants and the hydrogen-ion concentration of the solutions.*}* Further workj has led to the conclusion th at the presence of an unionised carboxyl group is essential for adsorption to take place. On the other hand, the adsorption curve for the bases w-propylamine and w-butylamine at different hydrogen-ion concentrations did not follow the ionisation curves very closely. Strong preferential adsorption of the unionised molecule was observed but there was quite considerable adsorption from solutions of such acidity that no unionised amine could exist in them. The adsorption of both bases was found to fall off gradually and continuously with increasing acidity from p H 11 to Pn 3. In view of the fact that the charcoal used in this work was Norit charcoal purified by treatment with strong halogen acids, it was thought possible th at very small traces of these acids remained after the washing to which the char coal was subjected, and that these traces of acid caused the adsorption of basic ions by direct chemical combination. A stream of air in which a sample of the acid-treated charcoal was ignited was free from hydrogen chloride, but the retention of small quantities of hydrogen fluoride was not impossible. It seemed desirable, therefore, to obtain a pure charcoal by some method other than acid extraction. It was found that a fairly active adsorbent char coal could be prepared from “ ashless ” filter paper. * Phelps and Peters, ‘ Proc. Roy. Soc.,’ A, vol. 124, p. 584 (1929). t Tt being assumed that an equilibrium exists between adsorbed molecules and those in solutions. t Phelps, ‘ J. Chem. Soc.,’ 1929, p. 1724.
Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
156
H. J. Phelps. Preparation of Purified Charcoal.
Whatman’s No. 40 “ ashless ” filter circles were damped and compacted as far as possible and then carbonised by gradual heating in platinum vessels. Residual organic matter was then removed and the charcoal activated by heating at 1200° C. in a reducing atmosphere in a gas furnace for half-an-hour.* The adsorbent power of this charcoal, for the substances studied, is slightly less than half that of the acid-purified Norit charcoal previously used. Experimental. In order to make this investigation comparable with my previous work, the electrolytes investigated were: propionic acid, n-propylamme and w-butylamine. All were purified by distillation. The general technique of the adsorption experiments also exactly followed that previously employed. Various hydrogenion concentrations were obtained by partial neutralisation of the electrolytes with the requisite quantities of strong acids or bases. As has previously been pointed out, the resulting mixtures of the weak acids or bases and their salts are to some extent self-buffering and consequently there are no large changes of hydrogen-ion concentration during adsorption. For each experiment a total volume of 22 c.c. of solution (approximately 0*2 per cent, in each case) was shaken with 200 mg. of charcoal for about 12 hours at 20° C. At the end of each experiment the mixture was centrifuged and the clear liquid decanted and analysed by the appropriate method, as described in Part I ( loc.cit.) and in the c Journal of the Chemical Society ’ ( cit.). equilibrium values of the hydrogen-ion concentration were determined by the bubbling hydrogen electrode. Experimental Results. The effect of hydrogen-ion concentration upon the adsorption of w-propylamine and n-butylamine is shown in figs. 1 and 2 respectively. It will be seen that in each case the adsorption is very exactly related to the degree of ionisation of the bases at different hydrogen-ion concentrations. The adsorption * The charcoal was heated in a silica crucible 8 cm. in diameter and 11 cm. high. Above the charcoal was placed a layer of small circles of filter paper to provide a reducing atmo sphere in the early stages of heating. The crucible was provided with a lid. The supply of gas and air to the furnace was adjusted for nearly complete combustion. There was probably a small amount of oxidation during the early stages of heating but this was almost certainly confined to the protective layer of paper and the charcoal immediately adjacent to it. Both the ash from the paper and the surface layers of the charcoal were rejected.
T
Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
Hydrogen-ion Concentration and Adsorption.
157
6
4
2
Fig. 1.—The adsorption of ra-propylamine plotted against hydrogen-ion concentration (continuous curve). The adsorption is plotted as milligrams of base adsorbed by 200 mg. of charcoal. The dotted curve represents the approximate ionisation curve of %-propylamine plotted to such a scale that the maximum adsorption of the base is equivalent to zero ionisation.
F ig. 2. —The adsorption of w-butylamine plotted against hydrogen-ion concentration.
The adsorption is plotted as milligrams of base adsorbed by 200 mg. of charcoal.
Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
158
H. J. Phelps.
rises from nil at p s 7-8 to a sustained maximum in solutions more alkaline than ps 11. The value of the ionisation constant of w-propylamine given by Ostwald* is 4*7 X 1CT4. These results indicate that charcoal prepared from pure cellulose adsorbs the molecules of the normal amines, but that it will not adsorb their ions to any measurable degree. The influence of hydrogen-ion concentration upon the adsorption of propionic acid is shown in fig. 3. It will be seen from the broken curve that the adsorp tion does not show the rapid fall with increasing alkalinity th at is shown in the
F ig. 3.—The adsorption of propionic acid by filter-paper charcoal (broken curve) and by
the same charcoal after acid-treatment (continuous curve), plotted against hydrogenion concentration. The adsorption is plotted as milligrams of acid adsorbed by 200 mg. of charcoal. The dotted curve represents the approximate ionisation curve of pro pionic acid, plotted to such a scale that the maximum adsorption of the acid by acidtreated charcoal is equivalent to zero ionisation.
adsorption of this acid by purified Norit charcoal. It falls off gradually from pH1 •5 to p H6. This can hardly be due to the adsorption of propionate ions as there is no adsorption in solutions more alkaline than p B 6, at which hydrogenion concentration, the concentration of propionic acid molecules, may be considered to have fallen to zero. Ostwald gives K for propionic acid as 1*34 X 10~5.f It seems probable that the adsorption of propionic acid is influenced by changes in hydrogen-ion concentration in solutions more acid than p H3, on account of some ionising impurity in the charcoal itself. The results * ‘ Z. Phys. Chem.,’ vol. 33, p. 361 (.1898). t ‘ Z. Phys. Chem.,’ vol. 3, p. 170 (1889).
Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
Hydrogen-ion Concentration and Adsorption.
159
a t least indicate a disturbance of the equilibrium between the solution and the surface. It is well known th at the most highly purified filter papers contain a certain amount of basic impurity. The nature of these impurities has been discussed by Fresenius.* Langef has found th at it is almost impossible to remove the last traces of silica, calcium and iron from celluloses of vegetable origin. It seemed reasonable to attribute the anomalous adsorption of propionic acid by filter paper charcoal to the presence of these substances. A sample of the charcoal was, therefore, digested with a mixture of hydro chloric and hydrofluoric acids and evaporated slowly to dryness. The charcoal was then repeatedly washed with distilled water until the washings were chloride-free, and subsequently heated under reduced pressure. In this way the treatment to which the purified Norit charcoal had been subjected was exactly imitated. After this treatment the filter-paper charcoal was found to be almost as selective an adsorbent of molecular propionic acid as the acid-purified Norit charcoal. The curve relating adsorption with hydrogen-ion concentration is shown in fig. 3 (continuous curve). In the case of acid-treated filter-paper charcoal, adsorption falls from a maximum value to nil between p H 3-2 and p a 6. In the case of acid-purified Norit charcoal the values are 3-8 and p n 6 respectively. The results agree as closely as can be expected with the view th a t only the unionised acid molecule is adsorbed. By analogy with the results obtained with acid-purified Norit it might have been expected that the acid-treated filter-paper charcoal would have acquired the property of adsorbing propylammonium ions. This was not the case. A series of experiments with this charcoal showed th a t the adsorption of n-propvlamine by it was influenced by hydrogen-ion concentration in exactly the same way as was the adsorption by untreated filter-paper charcoal. The actual amount of base adsorbed was also, within the limits of experimental error, identical in the two cases. Discussion of Results. The hypothesis that the adsorption of basic ions by purified Norit charcoal was due to direct chemical combination between these ions and traces of acid impurity in the charcoal, seems to be justified by the results here described. Such impurities as there are in charcoal prepared from filter paper are certainly * ‘ Z. Analyt. Chem.,’ 1883, p. 241. f * Ber. Deut. Chem. Ges.,’ 1878, p. 823.
Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
160
H. J. Phelps.
basic. This charcoal adsorbs only unionised amine molecules. The presence of basic impurities, however, in the filter-paper charcoal does not give it the property of adsorbing acid ions. In view of the fact that propionic acid is not appreciably ionised in solutions more acid than p H, one would expect that the adsorption of the acid would also remain constant with increasing acidity beyond p n 3. This does, of course, happen with acid-purified Norit charcoal and also with acid-treated filterpaper charcoal. The steady rise in the adsorption of propionic acid upon untreated filter-paper charcoal with increasing acidity up to 1-5 at least must, therefore, be a property of the impure charcoal itself. It has been suggested by Rideal* that the molecular adsorption of weak acids observed with purified Norit charcoal might be apparent only. If one imagines that hydroxyl ions are expelled from the charcoal and replaced by acid anions the observed adsorption would be indistinguishable from that produced by true molecular adsorption. This argument is in essence the same as that advanced by Miller,f except in so far as Rideal would attribute the origin of the hydroxyl ions to gases adsorbed by the charcoal, while Miller postulates a layer of water molecules on the charcoal surface. If the adsorp tion of weak acids be regarded as proceeding by the replacement of hydroxylions, it must be assumed, either that the number of available hydroxyl-ions decreases with increasing alkalinity, which seems improbable, or it must be assumed that the hydroxyl-ions are more easily displaced from the surface as the hydrogen-ion concentration in the liquid phase rises. If one imagines that the expulsion of hydroxyl-ions from the charcoal is determined by the con centration of hydrogen-ions in the liquid phase, one would expect that with any one charcoal the adsorption of all weak acids would be influenced by hydrogen-ion concentration in the same way. The results which have been described for maleic and fumaric acids, for example, show that this is not the case. If the adsorption of the simple amines is to be explained on the same lines, it must be assumed to proceed by the displacement of hydrogen-ions from the charcoal surface by ionised base. Further it must be assumed that such hydrogen-ions are replaced progressively more readily as the hydrogen-ion concentration in the liquid phase falls from to 11. Apart altogether from the complex assumptions that any ionic representation of the adsorption of weak electrolytes may demand, it is hard to imagine any ionic mechanism of adsorption, as a result of which the adsorption rises progressively as the * “ Surface Chemistry,” 2nd ed., p. 277. f ‘ J. Phys. Chem.,’ vol. 28, p. 992 (1924).
Downloaded from http://rspa.royalsocietypublishing.org/ on March 27, 2018
Hydrogen-ion Concentration and Adsorption.
161
concentration of adsorbable ions falls, and which would account for the fact th at the adsorption is maintained at a maximum value in solutions of such “ reaction ” that the concentrations of adsorbable ions in them must be very small indeed. On the other hand, the assumption that the adsorption of the weak electrolytes studied proceeds very largely, or exclusively, through the medium of unionised molecules, adequately explains all the results so far obtained. There is no reason to suppose that the adsorption of strong electrolytes, in which case there is a considerable concentration of ions under practically all conditions, proceeds in a similar way. Much evidence of the ionic adsorp tion of strong electrolytes exists. It has been shown, also, by Schilov and his co-workers* that the adsorption of strong electrolytes by charcoal depends on the presence of adsorbed or combined gases on the charcoal surface. Summary. It has been found that charcoal prepared from W hatman’s No. 40 filter circles adsorbs w-propylamine and w-butylamine only as unionised molecules. The amount of base adsorbed at any given hydrogen-ion concentration is, therefore, proportional to the concentration of unionised molecules existing under the conditions. Propionic acid is adsorbed only as unionised molecules, but some impurity in the charcoal prevents true equilibrium between the molecules in solution and those at the surface being established. Acid-treatment of the charcoal removes this impurity and there is good relation between the concentration of propionic acid molecules in solution and the amount of adsorption that takes place. The acid-treatment of the charcoal has no influence on the adsorption of w-propylamine or n-butylamine. These results justify the assumption th at a sufficiently purified charcoal will present a perfectly neutral surface to a water solution of a weak electrolyte. Such a surface will adsorb weak acids and bases only as unionised molecules. Acid-treated filter-paper seems to approximate very closely to an ideal neutral surface. My sincere thanks are due to Professor R. A. Peters for his advice and encouragement and to the Ramsay Memorial Trustees for a Fellowship which enabled me to undertake this work. I must also express my indebtedness to the Deputy Master and Assayers of the Royal Mint in whose laboratories the filter-paper charcoal was prepared. * Schilov and Tschmutov, ‘ Z. Phys. Chem.,’ vol. 148, p. 233 (1930); Schilov, Tschmutov and Schatunovskaja, ‘ Z. Phys. Chem.,’ vol. 149, p. 211 (1930); and also by Frumkin, ‘ Koll. Z.,’ vol. 51, p. 123 (1930). VOL. CXXX1II.— A.
M
