Estimation of Diffusion Coefficients in Gas Mixtures

A more rigorous kinetic theory than that in Chapter 2 not only supplies us with the proper form of the constitutive relations for multicomponent diffusion, it also provides an explicit relation for the binary diffusion coefficient. A slightly simplified version of the kinetic theory result is 

The parameter the diffusion collision integral, is a function of k T/e, where is the Boltzmann constant and e is a molecular energy parameter. Values of tabulated as a function of k T/e, have been published (Hirschfelder et al., 1964 Bird et al., 1960). Neufeld et al., (1972) correlated using a simple eight parameter equation that is suitable for computer calculations (see, also, Danner and Daubert, 1983 Reid et al., 1987). Values of a and e/k (which has units of kelvin) can be found in the literature—for only a few species—or estimated from critical properties (Reid et al., 1987 Danner and Daubert, 1983). The mixture a is calculated as the arithmetic average of the pure component values. The mixture e is taken to be the geometric average of the pure component values. 

A number of empirical or semiempirical correlations for estimating gaseous diffusion coefficients have also been developed. These include the method of Wilke and Lee (1955), which is based on Eq. 4.1.3 with C = 0.02199 - 0.00507 (Mi +  

TABLE 4.1 Diffusion Volumes in Fuller-Schettler-Giddings Correlation Parameters 

With T in kelvin (K), P in pascals (Pa), and M2 in grams per mole (g/mol) and C = 1.013 X 10 D will be in square meters per second (m /s). The terms and V2 are molecular diffusion volumes and are calculated by summing the atomic contributions in Table 4.1. 

---