Reports of the Academy of Sciences of the USSR

1961, Volume 139, No. 4

CHEMISTRY

E. A. Levitskii, V. N. Maksimov, and I. Yu. Marchenko

ON THE POLYMERIC NATURE OF \(5/6\) BASIC ALUMINUM CHLORIDE AND THE POSSIBILITY OF THE EXISTENCE OF ALUMINUM OXYCHLORIDES OF HIGHER BASICITY

(Presented by Academician V. A. Kargin on 18 III 1961)

A number of studies \((^{1-3})\) have shown that soluble basic salts of aluminum are prone to the formation of complex compounds with high molecular weight. It may be assumed that the tendency toward the formation of such products is most clearly manifested in systems of high basicity. From this point of view, it is of interest to investigate a series of physical properties of \(5/6\) basic aluminum chloride \(\mathrm{Al}_2(\mathrm{OH})_5\mathrm{Cl}\) \((^4)\).

In the present work, the task was set of determining the possibility that \(5/6\) basic aluminum chloride forms products of polymeric structure. Polymeric substances are characterized by a number of specific properties, the most important of which include: 1) increased viscosity of solutions; 2) swelling in solvents and the formation of systems intermediate between a solid and a liquid; 3) the ability to form fibers and films (anisotropy of properties).

Obviously, if \(5/6\) basic aluminum chloride is capable of forming polymers, then these products must possess the above-mentioned complex of physical properties.

The preparations used in the investigation were prepared by dissolving metallic aluminum or pure aluminum hydroxide in hydrochloric acid. As is known, both of these methods make it possible to obtain products with a high ionic weight \((^2)\).

With an increase in the concentration of \(5/6\) basic chloride in the system \(\mathrm{Al}_2(\mathrm{OH})_5\mathrm{Cl}—\mathrm{H}_2\mathrm{O}\), the appearance of the system changes as follows:

The transition along the series solution—powder is completely reversible. When small amounts of water are added to the powder, it swells with formation of a gel. With increasing dilution, the oxychloride passes into a true solution. Let us note that swelling upon addition of a solvent is typical of polymers. The gelatinous product has noticeable thixotropy, which is characteristic of polymers whose macromolecules are somewhat branched.

Previously \((^5)\), certain properties of solutions of \(5/6\) basic aluminum chloride were investigated. It was shown that, in the concentration range studied (0–200 g/l \(\mathrm{Al}_2\mathrm{O}_3\) at 20° and 0–300 g/l \(\mathrm{Al}_2\mathrm{O}_3\) at 75°), the density of the solutions changes strictly linearly. As is known \((^6)\), \(\mathrm{Al}_2(\mathrm{OH})_5\mathrm{Cl}\) is a substance with practically unlimited solubility. In the concentration ranges studied, the density changed in a manner entirely analogous to that observed in aqueous solutions of electrolytes, where, with increasing concentration, no qualitative changes occur (for example, formation of hydrates) \((^7)\).

However, the character of the change in viscosity of an \(\mathrm{Al_2(OH)_5Cl}\) solution at 20 and 75° differed sharply from the curves for the change in viscosity of aqueous electrolyte solutions with change in concentration \((^{5,7})\). These curves proved to be very similar to graphs constructed for such a typical polymeric substance as polyisobutylene in various solvents (heptane, hexane, toluene). This is confirmed by Fig. 1, where the data obtained on the change in viscosity with change in concentration \((^{5})\) are interpreted in terms of reduced viscosity. As can be seen from the figure, the characteristic viscosity of a solution of \(5/6\) basic chloride at 20° is noticeably different from zero and is 0.06. Thus, in a number of properties the solutions of \(5/6\) basic aluminum chloride studied are identical with polymer solutions.

Fig. 1. Reduced viscosity of solutions of \(5/6\) basic aluminum chloride at 20° C

Fig. 2. Thermomechanical curve of a glassy specimen of \(5/6\) basic aluminum chloride

Of considerable interest was the study of the dependence of the mechanical properties of condensed (glassy) \(5/6\) basic chloride on temperature. Tests of a specimen obtained in the form of a tablet 8 mm in diameter and 4 mm high were carried out on dynamometric balances under a compressive load of \(1.59\ \mathrm{kg/cm^2}\), with a load duration of 10 sec. The results are given in Fig. 2, from which it is seen that, as the temperature is raised, \(5/6\) basic chloride passes through three physical states: at 25° it changes from the glassy state to the highly elastic state (characteristic only of polymeric substances), and at 90° to the viscous-flow state.

In the course of the experiments it was also repeatedly noted that \(5/6\) basic aluminum chloride is inclined to form films. The anisotropy of the properties of the substance has been used in a number of works. Thus, recently there have appeared reports on the use of \(5/6\) basic aluminum chloride as a binder in refractory cements \((^{8})\) (concentrated solutions of \(\mathrm{Al_2(OH)_5Cl}\) possess good adhesive ability) and on the preparation, from a solution of the basic chloride, of fibrous boehmite, for which various applications have been found \((^{9})\).

Thus, from the totality of its properties, \(5/6\) basic aluminum chloride may be regarded as a substance capable of forming products of polymeric structure.

The existence in solution of any basic compound is proved by the presence of the corresponding inflection on the potentiometric titration curve. However, polymerization of less highly basic compounds can distort the course of the potentiometric titration curve. Klenert and Denk \((^{12})\) observed a complex character in the change in the amount of alkali that must be added to a solution of the basic chloride in order to reach the potential jump (in titration to \(\mathrm{Al(OH)_3}\)) during aging of the solution (see Fig. 3).

In seeking to explain this phenomenon, the authors assumed that, in parallel with hydrolysis, other reactions occur, for example, for a solution of composition \(\mathrm{Al(OH)_2Cl}\).

The concept of the polymeric nature of \(5/6\) basic aluminum chloride makes it possible to explain the known experimental data associated with

the question of the possibility of the existence of higher basic chlorides (i.e., intermediate in composition between \(\mathrm{Al_2(OH)_5Cl}\) and \(\mathrm{Al(OH)_3}\)) as individual substances. Recently a number of reports have been made on the discovery of higher aluminum oxychlorides. Denk et al. \((^{11,12})\) and Tanabe \((^{13})\) found the \(8/9\)-basic chloride \(\mathrm{Al_3(OH)_8Cl}\) and the \(11/12\)-basic chloride \(\mathrm{Al_4(OH)_{11}Cl}\).

\[ \mathrm{Al(OH)_2Cl + Al(OH)_3 = Al_2(OH)_5Cl,} \tag{1} \]

\[ \mathrm{3Al(OH)_2Cl = 2Al(OH)_3 + AlCl_3,} \tag{2} \]

whereby the first reaction promotes a decrease in the consumption of alkali, and the second an increase; raising the temperature promotes the occurrence both of reaction (2) and of the hydrolysis reaction

\[ \mathrm{Al(OH)_2Cl + HOH = Al(OH)_3 + HCl.} \tag{3} \]

Fig. 3. Change in the amount of \(0.1\,N\) NaOH solution added until the “potential jump” is reached, during aging of solutions of basic aluminum chloride of different basicity (according to the data of \((^{12})\)) at \(t=20^\circ\) (a) and \(60^\circ\) (b).

The unconvincing nature of such an explanation is obvious. First, even if all the reactions given do take place, this is in no way reflected in the balance of hydroxide ions in the system. Secondly, it is known that the \(5/6\)-basic chloride is titrated with alkali to \(\mathrm{Al(OH)_3}\) quite readily \((^6)\). Consequently, if the matter were only in the processes considered by the authors, the amount of alkali solution added up to the “potential jump” on the titration curve would remain unchanged during aging.

The essence of the observed phenomenon is that part of the aluminum-containing basic cations is “masked” during aging and therefore is not titrated with alkali. It is quite possible that this is directly connected with the formation, during the aging of the \(5/6\)-basic chloride, of high-molecular compounds (polymers), in which (at the given rate of titration with alkali) only part of the aluminum can be neutralized. Other changes in the system (aging of aluminum hydroxide, hydrolysis of the \(5/6\)-basic chloride, and of the high-molecular product itself) will lead to opposite results (an increase in the alkali consumption before the “potential jump” is reached). Heating evidently promotes hydrolysis to a greater extent than polymerization, which also causes a qualitative change in the character of the phenomenon with increasing temperature.

It is possible that this same “masking” of part of the cations of the basic salt within the macromolecule also explains the appearance of inflections on the potentiometric titration curves in the basicity range \(\mathrm{OH/Al}=2.5\div3\). Indeed, titration of part of the aluminum-containing cations enclosed within the macromolecule should proceed at a higher concentration of hydroxide ions than titration of these same cations freely present in solution. Therefore, when the “masked” cations are titrated at

additional inflections shifted toward higher pH of the medium and higher basicity of the solution should appear on the potentiometric curve. Still greater difficulties are associated with titrating polymerized \(5/6\) basic aluminum chloride with sulfate ion, with formation of a precipitate of \(5/6\) basic aluminum sulfate, for example

\[ 2\mathrm{Al}_2(\mathrm{OH})_5\mathrm{Cl}+\mathrm{Na}_2\mathrm{SO}_4 = (\mathrm{Al}_2(\mathrm{OH})_5)_2\mathrm{SO}_4+2\mathrm{NaCl}. \]

Denk \((^{10})\) showed that precipitate formation begins at enormous excesses of sulfate ion in solution compared with the stoichiometric amount. Although this reaction is undoubtedly ionic, even at large excesses of sulfate ion precipitate formation begins only after several hours (or days). Our experiments have shown that, when titrating at ordinary rates, turbidity in a solution of \(5/6\) basic chloride begins only after the amount of added sulfate ion exceeds by \(4\cdot 10^7\) times the amount required to attain the solubility product of \(5/6\) basic aluminum sulfate.

It is obvious that the difficulties arising in the titration of polymeric basic aluminum chloride with \(\mathrm{OH}'\) ions must be aggravated in the case of titration with \(\mathrm{SO}_4''\) ions, owing to their lower mobility and to a change in the order of the reaction.

Thus, highly basic aluminum chlorides (with basicity \(\mathrm{OH}/\mathrm{Al} > 2.5\)) may be regarded not as individual compounds, but as polymerized varieties of \(5/6\) basic chloride, \(\mathrm{Al}_2(\mathrm{OH})_5\mathrm{Cl}\).

Other known experimental data also support the hypothesis put forward here. In particular, when an excess of metallic aluminum is dissolved in hydrochloric acid, the solution tends, in the limit, toward the composition of \(5/6\) basic chloride; solutions with the ratio \(\mathrm{Al}/\mathrm{Cl} > 2\) are not formed in this case \((^{10})\). When the amount of dissolved aluminum exceeds this ratio, the character of the system changes: the solution turns from a true solution into a colloidal one \((^{10,14})\).

Received
15 III 1961

CITED LITERATURE

\(^{1}\) H. W. Kohlschütter, P. Hantelmann, Zs. anorg. u. allgem. Chem., 248, 319 (1941).
\(^{2}\) G. Denk, J. Alt, Zs. anorg. u. allgem. Chem., 269, 244 (1952).
\(^{3}\) Л. К. Лепинь, А. Я. Вайваде, ЖФХ, 27, No. 2, 217 (1953).
\(^{4}\) В. А. Каргин, Н. А. Оганджанова, ЖФХ, 10, No. 4–5, 782 (1937).
\(^{5}\) Э. А. Левицкий, Б. М. Шелгачев, Изв. высш. уч. зав., Цветная металлургия, No. 2 (1961).
\(^{6}\) М. Е. Шишняшвили, В. А. Каргин, А. Л. Бацанадзе, ЖФХ, 21, No. 3, 391 (1947).
\(^{7}\) Справочник физико-химических и технологических величин, 10, 1931, p. 93.
\(^{8}\) A. Wickers, J. Maloney, U. S. 2818345, 1957; Chem. Abstr., 52, 6747e (1958).
\(^{9}\) J. Bugosh, U. S., 2915475, 1959; Chem. Abstr., 54, (F) 9156, 1960.
\(^{10}\) G. Denk, L. Bayer, Zs. anorg. u. allgem. Chem., 267, 89 (1951).
\(^{11}\) G. Denk, Angew. Chem., 62, 340 (1950).
\(^{12}\) G. Klenert, G. Denk, Zs. anorg. u. allgem. Chem., 301, 3–4, 171 (1959).
\(^{13}\) H. Tanabe, J. Pharmac. Soc. Japan., 74, 866 (1954); Chem. Abstr., 49, 11482 c. (1955).
\(^{14}\) G. Bayley, J. Soc. Chem. Ind. (Trans.), 39, 9, 118 (1920).