338 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS ture is shown by the number of mix- ing and dispersing devices used. In addition to the various types of ordinary mixing devices, we have colloid mills, homogenizers, roller mills, microatomizers, etc. Each is best fitted for certain types of work. Rather than dwell upon the me- chanical features of these devices, let us consider the physico-chemical factors which control dispersion, so that we can see what can and what cannot be accomplished by me- chanical dispersing devices. Most of this discussion will be about disper- sion of liquids ia liquids, or emulsifi- cation: Interfacial tension between oil and water hinders the dispersion of oil into small droplets when an attempt is made to form an emulsion by agitation. A definite shearing force is required to break a droplet into smaller droplets, and the interfacial tension resists that force. Such shearing forces are set up during agitation by differential currents of flow within the liquid. Bearing in mind the fundamental principles of viscous flow, it follows that the force pulling apart two points in a dif- ferential flow field will be propor- tional to the distance between those points. If those two points are conceived to be on the surface of a droplet, the force tending to tea• that droplet apart will be seen to de- pend on the size of the droplet. Consequently with a given degree of agitation, droplets below a certain size will never experience forces strong enough to tear them, and hence a given force of agitation, no matter how long continued, will not produce an emulsion below a cer- tain droplet size, as long as inter- facial tension is constant. During the earlier stages, the average droplet size will be de- creasing and, for that reason, the rate of dispersion will be decreasing. Also 'the number of droplets and hence the rate of collisions between droplets, will be increasing. There- fore the rate of coalescence, which depends on the rate of collisions, will be increasing. Eventually a state of dynamic equilibrium is reached, where rate of dispersion equals rate of coalescence. • If a greater rate of agitation were used, a greater degree of dispersion would be achieved at the equilibrium state. Now let us suppose the agitatio.n stopped. Since the dispersing forces no longer exist, collisions caused by brownian motion, thermal currents, etc., will cause gradual coalescence until eventually we have two sep- arate layers. Ordinarily the oil will be lighter than the water. When agitation has ceased, the oil droplets will all tend to rise toward the top of the liquid mixture. There they will be forced into contact with each other by buoyant forces, and will hence coalesce much more rapidly than if they had remained suspended throughout the liquid. Another kinetic factor which can aff'ect emulsification is that of energy barriers. There is a close analogy with ordinary chemical re- actions. Probably no emulsion is entirely stable. That would occur only if interfacial tension were zero

SURFACE-ACTIVE AGENTS IN COSMETIC INDUSTRY 339 but this is the condition required for solution, and two such liquids would therefore be miscible. However, an emulsion may be stable enough for practical purposes. A chemical re- action may involve a decrease in free energy and thus be thermodynam- ically possible. However a mole- cule, before reacting, may have to acquire a large amount of energy in order to form an activated complex. This energy is surrendered when the reaction occurs and hence does not figure in the net energy of the reac- tion. Nevertheless if this "barrier" energy is high, the rate of reaction will be very small. Similarly two droplets, to coalesce, may have-to overcome an energy barrier imposed by the droplets having similar elec- trical charges. The charges may not lower the free energy of coalescence substantially, and thus not alter the thermodynamic tendency toward coalescence. However, they may have a great effect on rate of coales- cence by forming a high energy barrier. So an emulsion can be stabilized in the sense of greatly re- ducing the rate of coalescence, with- out affecting fundamental insta- bility as determined by interfacial tension. Ordinarily when a fresh interface is generated, the final equilibrium interfacial tension is reached ap- proximately in a small fraction of a second. Further, electrical • barrier forces are often not great. Con- sequently, interfacial tension is a rough guide to ease of emulsification and rate of breaking of the emulsion formed, but with numerous excep- tions. All emulsions are inherently unstable, since coalescence of emul- sified phase will, by reducing area of interface, reduce interfacial free energy. ROLE OF EMULSIFYING AGEST Emulsions of practical interest have at least three components: Two liquidsmone of which is nearly always water or an aqueous solu- tionmand an emulsifying agent. The emulsifying agent must con- centrate at the oil-water interface. There it forms films which stabilize the emulsion. Like all surface- active compounds, emulsifying agents combine polar and non-polar groups in the molecule in proper balance to give the desired effect. Various mechanisms are possible: 1. The emulsifying agent, if a soluble surface-active agent, forins an oriented molecular layer at the interface, and thus reduces inter- facial tension. This facilitates rup- ture of droplets during agitation and permits formation of a finer emul- sion by agitation forces. The very fact that an emulsion is fine reduces its rate of breaking. According to Stokes' Law, small droplets will rise to the top of the liquid more slowly than large droplets. As was noted earlier, forcing together of droplets at the surface of the mix- ture is an important factor in caus- ing separation. Further, low interfacial tension decreases the rate of separation by decreasing the proportion of colli- sions which are effective in causing