444 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS IOOO Figure 1 Io 06 I I I 15 3.'5 6 -.. d 12 30 60 Speed of rotation, rev min-1 Brookfield viscometer measurements on 3.5 • PGA solution a Alkali treated, spindle 3 b Alkali treated, spindle 4 c Untreated, spindle 2 d Untreated, spindle 3 dependent on the rate of rotation of the spindle than are those of the un- treated solution. The results in Table I show similar differences between shear sensitivity before and after treatment in those cases where the alkali has produced a viscosity increase. In some cases where there was a much greater rise in viscosity only low speeds of rotation could be used to obtain readings on the scale. Although the curves in Fig. 1 show that this procedure will give much greater differences than if readings could be made at higher rates of shear, it is considered that the low shear readings are good indication of the change in rheology of the mixture. Prolonged action of alkali leads to a drop in viscosity after the initial rise, but the high viscosity and gel structure is maintained if the mixture is acidified after a short time. In most of the experiments described acetic acid was added 15 min after addition of the alkali. A fairly general effect of the alkali treatment is to render the product insoluble after drying, as was mentioned in the case of the original patent for the treatment of gelatine films. Any gel formation can be considered as an insolubilization of a colloid, although in some cases the effect is reversible

NEW REACTIONS OF PROPYLENE GLYCOL ALGINATE 445 on heating, and in all cases the insoluble material is highly hydrated. With the mixtures under consideration, if any mixture which has shown notice- able gel formation is dried after reacidification the solid will not dissolve in water. In most cases it will swell but will not take up as much water as •vas in the gel before drying. On the other hand the water insoluble solid can be dissolved in alkalis with a pH of above 10 and the resulting solution is of low viscosity. Some examples of the changes produced The PGA was taken from material produced in commercial quantities. Several samples were used. Sample A which was used in one series of experi- ments had about 70ø//0 of the carboxyl groups esterified while the others (B, C and D) were about 80% esterified and were all very similar in behav- iour. As will be seen in Table I, a higher concentration of sample A than of sample D, the more highly esterified product, is required to produce the effect. For this reason all further experiments were made with PGA having about 80% of the carboxyl groups esterified: this is the highest proportion generally available in commercial products. (i) PGA alone The figures in Table I show that the effect of making a PGA mildly alkaline depends on the concentration of the solution and the temperature. As the concentration is raised the effect of the alkali changes from bring- ing about a drop in viscosity to causing a rise which, in some cases, is very marked. Reducing the temperature at which the alkali is added results in Table I Treatment of PGA solutions with alkali 0.5 mmole Na•CO• added g-' PGA. Reacidified with 1 mmole acetic acid g-' PGA after 15 min. PGA sample Concentration g 100g -• 4.5 4.5 5.5 5.5 5.5 2.5 2.5 3.5 4.0 Tempera- Viscosity (P) ture øC After treatment Before treatment 10 135(2) 78(20) 20 75(2) 49(20) 10 490(2) 235(20) 20 186(2) 103(20) 40 140(2) 69(20) 3 -- 7(60) 20 -- 2(60) 20 16(6) 13(60) 20 48(6) 30(60) 2 100(2) 460(20) 65(2) 47(20) 11 000(2) 3 600(10) 3•300(2) 800(20) 28(2) 10(20) 352(3) 55(60) -- 0.85(60) 150(6) 35(60) 3 600(6) 3 250(12) Note: The figure in brackets is the speed of rotation of the Brookfield viscometer spindle.