It would seem therefore that the real necessity for the use of sulphate of alumina for precipitating the rosin is not so much the necessity for forming acid aluminum resinates, but the fact that the rosin precipitated in this way has a more colloidal character, than that thrown out of solution by other coagulants, and consequently will have greater covering power and efficiency as a water repellant. This explanation is opposed to the theory that rosin acids in the form of emulsion or suspensions are efficient sizing agents, for it is obvious that visible floating rosin has lost its colloidal character and its covering power.
The fact is that the so called free rosin emulsions, when properly made, contain but a very small amount of rosin acid in the emulsified form, practically all of it remaining in solution in the dilute soap. The art of preparing good rosin size emulsions (using the term as generally understood) is therefore the ability to dilute a solution of rosin acids, without the actual liberation of rosin in the emulsified form. The difficulty in doing this will explain the erratic results obtained by Remington and other investigators when endeavoring to determine the effect of rosin size containing dissolved rosin.
The laboratory difficulties involved are shown by an article by Otto Kress and R. T. Struthers, published in Paper, April 1913. Their results show that from a rosin saponified with 15% of sodium carbonate, over 98% was obtained by them in hot dilute aqueous solution, and that from a rosin saponified with 10% of sodium carbonate, only 50.6% was brought into actual solution in hot water. It is quite possible however, to dilute such a rosin soap holding in solution about 45% of rosin acids, to an aqueous solution of 2% solids, without having any of the rosin become insoluble. In this condition all of the rosin acids can enter into chemical reactions with other solutions and will precipitate from solution in a very bulky colloidal mass.
Between the extremes of physical condition just described, there are a great number of intermediate stages. The rosin acids may be partly liberated by dilution in the form of small visible floating particles and coarse granular masses and a part may be in a state of colloidal solution. It is safe to say that all rosin particles which are sufficiently coarse to be classed as suspensions, have lost the greater part of their sizing value. That portion of the rosin acids which is in colloidal solution is still effective for sizing purposes because it has the property of becoming fixed upon the fibres by absorption. This action can only take place, however, when the paper stock is free from such electrolytes as may discharge the colloid before it reaches the fibre. Dilute acid resinate solutions may contain variable proportions of dissolved acid resinates, colloidal rosin, and rosin suspensions, and the relative proportion of these is what determines the basis of its waterproofing possibilities. Assuming that the sizing value of these solutions varies directly with the amount of rosin acids that are in true and colloidal solution, we have a measure of efficiency which checks very closely with actual mill results.
