A Model for Chemical Reactions in Solution

Some indication of the importance of the solvent interactions is afforded by the observation that of the thousands of reactions which have been studied in solution, less than 20 have been capable of comparative study in the gas phase. The study of ionic reactions has been almost completely restricted to solutions for reasons which are quite understandable ionic processes are virtually nil in the gas phase at temperatures below 1000 K. However, this accounts for most of the solution reactions studied, since, as we shall see, most reactions between polar molecules involve ionic species as intermediates. Thus such common reactions as the hydrolysis of alkyl halides or of esters do not proceed at measurable rates in the gas phase (at least not at temperatures at which other, competing reactions are not dominant). The only large class of reactions which proceed conveniently in both gas and liquid states is the free radical class, and undoubtedly as 

The molecular description of rate processes in solutions is possible to the extent that a molecular description of the behavior of solutions is available. While no rigorous, molecular theory of solutions has been developed, there are a number of semiempirical models which will provide a framework for discussing the kinetics of solutions. For the purposes of our discussion it is convenient to classify solutions as those in which the average interactions between neighboring molecules are of the order of kT or less (weak interactions) and those in which they are much greater than kT (strong interactions). As we will see later, this corresponds in a crude fashion to the distinction between noiiionic and ionic solutions. In the following sections, we shall consider first the behavior of nonionic systems. 

The simplest molecular model capable of describing a liquid is the hard sphere molecule with an attractive potential well (Sec. VII. 1). There are 

Because of the close packing in liquids, each molecule in such a system will be surrounded by a number Z of nearest neighbors which will tend to 

The problem of specifying a collision between two such molecules is complicated by the range of the attractive forces. If we think about it in detail, we will find, as in the case of gases, that for each physical process of interest (i.e., pressure, viscosity, diffusion, etc.), there will be a different definition of a collision. From the point of view of a chemical reaction, no generalization is possible, but we may say for purposes of convenience, as a zero-order approximation, that a pair of molecules will be considered in a state of collision so long as their potential energy of interaction is of the order of magnitude of fcT. 

---