POSSIBLE ROLE OF NON-HYDROGEN-BONDED UNITS IN THE CHEMISTRY OF LIQUID WATER

Pure liquid water is thought to contain ca. 8 mol% of non-hydrogen-bonded (tree) OH groups and lone-pair (LP) groups. It is suggested that, as well as playing important roles in the physical and spectroscopic properties of liquid water, these groups are important in controlling solvation and in certain chemical reactions. The postulate is that if a reaction that makes a significant contribution to the overall rate involves (OH)free groups, then solutes that increase or decrease the [(OH)free] will increase or decrease the rate. Conversely, if (LP)free groups are reactants, then changes in [(LP)free] will similarly affect the reaction rate. The hydrolysis of tert-butyl iodide in binary solvent systems has been measured spectrophotometrically at 7-degrees-C, and it is shown that the results can be reasonably understood in terms of the above theory. It is clearly established that basic aprotic cosolvents lead to rapid decreases in the rates of S(N)1 reactions, and our results are in good agreement. For simple 1 : 1 electrolytes the salt effect results in rate enhancement, as expected for an 'ionisation' process. However, tetra n-butylammonium bromide causes a dramatic fall in the rate of hydrolysis. This is expected, if the key reaction involves (OH)free groups, as is indicated by the rate decreases caused by basic solutes. Thus, for simple salts, cations and anions have similar solvation numbers and so induce only small changes in the concentrations of 'free' water groups. However R4N+ ions do not form bonds to water so the anions, which solvate by hydrogen bonding to OH units, cause a large fall in [(OH)free], and hence there is a large negative contribution to the rate of hydrolysis. Taking the primary solvation number of Br- as 6, the results agree well with those for solvents such as dimethylformamide, which has a solvation number of ca. 2. Finally, it is pointed out that reactions involving attack by oxygen via an electron pair, such as the extraction of H+ from an organic reactant should behave in just the opposite manner. This is indeed the case.