NEW REACTIONS OF PROPYLENE GLYCOL ALGINATE 453 reaction either with PGA alone or mixed with PVA. In a fexv tests made with the 1:13 diamino hexane reaction such compounds were found not to have any effect. The fact that 1:4 diamino butane gives a gel with PGA while 1:4 dihydroxy butane prevents viscosity increases in PGA and PGA, PVA mixtures is further evidence that different types of reaction are involved. l{EACTION BETX¾EEN PGA AND I)IHYDROXY COMPOUNDS In view of the gelling and viscosity increases with diamines and with polymeric hydroxy compounds in contrast to the general inhibiting effects with low molecular weight hydroxy compounds, some intermediate molecular weight hydroxy compounds were examined. 1:13 dihydroxy hexane was found to have no effect either way on the concentration of PGA which would give a viscosity increase in alkaline conditions, at either 5øC or at 20øC. It is possible therefore that a longer chain dihydroxy compound could be used, as with polyvinyl alcohol or starch, to give viscosity increases at lower PGA concentrations. Higher dihydroxy hydrocarbons are not readily available, and in any case would have limited solubility in water, but a large range of polyethylene glycols is made commercially. Tests have been made with Carbowax 400, 1 000 and 6 000 and with Polyox &5. These have molecular weights ranging from about 400 to 100 000, i.e. with about 13 to 8 000 carbon atoms in the chain. In no case was any effect found at 20øC, but all the polyglycols had an effect at 5øC. 2% PGA and 10% polyglycol was used: the Carbowac 400 showed only a slight viscosity increase, but all the others showed definite gel formation. It is remarkable that they were all very similar in their behaviour despite large differences in molecular weight. NATURE OF THE REACTIONS The rheological changes in the reaction mixtures, and the formation of insoluble products on drying can most reasonably be explained by some type of cross linking which builds up a three dimensional network. There is little evidence regarding the nature of the chemical linkages involved, but a possibility is that the hydroxy compounds form esters and that the amines form amides. It is at first sight surprising that transesterification can take place in aqueous alkaline solutions but it is known that starch can

454 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS be acetylated by treatment with vinyl acetate in somewhat similar condi- tions. (7), It is suggested that the cross linking reaction is only one of several competing reactions which are taking place simultaneously. With high molecular weight polyfunctional compounds only a very small propurtion of the potentially reactive groups need enter into cross linking reactions to give marked changes in physical properties. It is understandable therefore that a transesterification or amidation reaction in which only very few of the ester groups of the PGA take part could form enough cross links to produce the physical changes that have been noted. From the examples given it will be seen that the conditions for breakdown of the links are not greatly different from those for their formation, so that to get the viscosity increases and gel formation a narrow range of conditions has to be found where the rate of formation is greater than the rate of destruction. In view of the effect of calcium and other divalent metals in forming insoluble salts with alginates, it might be suggested that, for example, the 1:6 diamino hexane forms an insoluble salt with alginic acid formed by hydrolysis of the PGA. This has been shown not to be the case by preparing the 1:6 diamino hexane salt of alginic acid which is fully soluble in water. Furthermore the addition of the hydrochloride of 1:6 diamino hexane to a sodium alginate solution produces no appreciable change in viscosity. USES Possible uses for the effects described fall into two groups. (a) Those depending on the formation of an insoluble product. (b) Those depending on high viscosity and semi gel formation, e.g. suspension of solids. The best established application is the modification of gelatine films to allow photographic processing to be carried out at higher temperatures than is possible with unmodified gelatine. Films modified in this way have the advantage in processing of being more water absorbent than those of formalin hardened gelatine. A possible use of the same gelatine modifying process is in the manufacture of hard gelatine capsules as the product, although insoluble in water, is broken down by digestive enzymes. PGA, either alone or in combination with other polymers, can be used to provide water resistant binders for pellets or granules. From the cost point of view it is an advantage to keep the proportion of PGA as low as it is technically