No abstract

Confirmation of the existence of aluminum trilaurate has been obtained by using infrared absorption to follow the course of the reactions of trimethylaluminum and aluminum isopropoxide with lauric acid. The trisoap is shown to exist as micelles of high molecular weight in solution, the size decreasing on dilution or elevation of temperature. On hydrolysis to the disoap, hydrogen bonding does not appear to play a significant role in the formation of polymer structures. Coordination of carboxylate groups with aluminum is shown to be a major feature of the structure of both tri- and di-soaps. Spectral assignments in the 6.0 to 6.4 μ region have been made for various types of inter- and intra-molecular coordination. The effects of dilution, hydrolysis, and hydrochlorination of aluminum trilaurate in solution have been assessed by viscosity, light-scattering, and infrared measurements.

A light-scattering study of aluminum distearate in dilute benzene solution suggests that it is a linear molecule of random coil shape. This is further substantiated by intrinsic viscositymolecular weight data, the constancy of the ratio of the molecular weight to the sclu;ire of the radius of gyration (Rg), the value of in Flory's intrinsic viscosity -nlolecular weight relationship, and the close agreement betwcen the theoretical and experimental results found for the unit length of the polyn~er chain. This average unit length has been calculated to be 4.46 A. INTRODUCTIONThe fundamental structure of aluminum soaps remains rather obscure, although most authors (1, 2, 3, 4,5,6,7,8) agree that these substances dispersed in organic solvents are high polymers, the monomeric unit being the basic aluminum disoap, e.g. aluminum hydroxydistearate. Opinions differ, however, as to the nature of the chemical bonding between the monomeric units. McGee (6) and Gray and Alexander ( 5 ) have proposed an alternation of covalent and co-ordination bonds between the aluminum atom and oxygen atom of the hydroxyl radical to form the polymer chain. McBain (2) favors an association of the molecules into micelles, while McRoberts and Schulman (3) attribute the weak type of bonding exhibited by these molecules to hydrogen bonding. Recently the present authors (8) and Bauer et al. (7) gave evidence in favor of a structure similar to that proposed by Gray and Alexander, but with the carboxylate group also acting as a bridge between two aluminum atoms. These co-ordination bridge structures could easily be broken and could account for the changes in physical properties of solutions or gels on dilution, peptization, or'on aging.I t has always been recognized that aluminum soap solutions are relatively unstable and deteriorate with time, with a gradual lo\vering of the molecular weight of the polymer. This instability has not as yet been fully explained but the breakdown is accelerated by moisture and in general by traces of acid or other polar compounds. The rate of change depends 011 the concentration, the more concentrated solutions being relatively more stable. A complete explanation of the physicochemical properties of aluminum soap solutions is impossible without a detailed linomledge of the structure of the soap molecule. I t was considered important to obtain more clata 011 molecular weight and viscosity characteristics of aluminum stearate solutions and deduce from these, if possible, the average size and shape of the molecules. I t was recognized that interpretation of the results would be hampered by the instability of the polymer, so the experimental techniques employecl were designed to diminish as much as possible its effect on the experimental data. Ailother limitation of the interpretation of the data lies in our lack of knowledge of molecular weight distribution characteristics and of any method of fractionation for this polymer.Aluminum distearate swells in cold benzene and complete solution can be effected only by ...

In the c o u~s e of investigations of the properties of polymers of high ~nolecular weigl~t by the light-scattering method, anomalous results were obtained, lvhich were found to be due to reflection of light a t a glass-air interface. The reflection that was found to be important was that of the primary beam back into the scattering vol~inle. In this work the Light-scattering Apparatus "B" described by Hadow, Shetier, and Hyde ( 2 ) was used. A substantially parallel beam of

No abstract