FURTHER ASPECTS OF THE PHYSICAL CHEMISTRY OF SOME NON-IONIC DETERGENTS.

The most interesting properties of detergents in solution are their surface and micellar behaviour and the factors affecting this. Aqueous solutions of non-ionic detergents are colloidal, thus the techniques applied in their study have been similar to those generally used in colloid science. It has now been accepted that molecules of non-ionic detergents having a polyoxyethylene chain sufficiently large to produce water solubility of the hydrophobic moiety, orientate themselves in micelles with the hydrophobic moiety inside and the glycol chains outside. The glycol chain confers water solubility by trapping water molecules in some way (Goto, Sugano & Koizumi, 1954; Ferguson, 1955). The exact amount of water trapped and the means by which this is effected is conjectural as, until recently, no independent method of measuring the aqueous covolume of the micelle had been reported. Hydroxonium ions, hydrogen bonding and various arrangements of the water molecules around the ether oxygens or within the glycol structure have been suggested (Chwala & Martin, 1937, 1947; Wurzchmitt, 1950; Trinchieri, 1952; Hsaio, Dunning & Lorenz, 1956; Kehren & Rosch, 1956; Rosch, 1956; Bailey & Callard, 1959 ; Schick, 1963b). From viscosity and micellar studies, Kushner & Hubbard (1954) estimated that there were 43 molecules of water per polyoxyethylene chain in a micelle of Triton X 100 (nl0). Of this number, they suggested 20 molecules were held by hydrogen bonding to the ether oxygens, the rest being physically trapped by the chain. Nakagawa & Inoue (1958) showed the number of hydrating water molecules per oxygen atom of the polyoxyethylene chain increased with chain length. Other workers (Karabinos, Hazdra & Ballun, 1955; Karabinos & Metziger, 1955; Kehren & Rosch, 1956; Reich, 1956; Rosch, 1956; Boehmke & Heusch, 1960), using data from viscosity, polarimetry, and heat of hydration, have given 1, 2, 3 or 4 water molecules per ether oxygen, depending on the chain length and the workers concerned. We have recently described a method, based on vapour pressure measurements, of estimating the micellar hydration (Elworthy & Macfarlane, 1964). The vapour pressures over gels and concentrated solutions of detergents were measured as a function of detergent concentration and, by a suitable extrapolation procedure, the concentration determined at which the solution had (within experimental error), the same apparent