A corresponding-states correlation of low-density binary-and self-diffusion coefficients is presented. The equations are simple to use, are sdciently accurate for most calculations, and correlate those data used in their derivation somewhat better than calculations based on the Lennard-Jones potential if potential parameters have to be estimated from the critical properties. The Enskog kinetic theory of dense gases is used in modified form to obtain an expression for the high-density diffusion coefficient for isotopic mixtures in terms of the viscosity and compressibility of the gas. Generalized viscosity and compressibility charts are then used to construct a graph for predicting a reduced self-diffusion coefficient as a function of reduced temperature T. = T/T. and reduced pressure PI = p/pc. The effect of the pressure on the Schmidt number, Sc = p / p D , is also discussed.Finally the extension of this chart to nonisotopic mixtures is considered.In the design and operation of gasphase processes, one must have methods available for estimating various physical properties ahen experimental data are unavailable. The equation of state and thermodynamic properties of dilute and moderately dense gases can be calculated satisfactorily from intermolecular forces by means of statistical mechanics (24, 3 ) .
At higher densities the generalized chartsof Hougen and Watson (25) and the revised tables of Lydenon, Grecnkorn, and Hougen (35, 18) are more useful. The viscosity, thermal conductivity, and diffusion coefficients for dilute gases can be estimated from intermolecular forces by kinetic theory calculations (24, 3 ) . Several charts have been prepared from experimental data for the prediction of the viscosity coefficient of dense gases (8, 12, 13, 48, 252, the most recent being that of Carr, Parent, and Peck (9). The thermal conductivity coefficient of dense gases can be estimated from the chart of Lenoir, Junk, and Comings (33, I S ) , which is a generalized representation of experimental data for several gases. The earlier charts of Comings and Nathan (11) and of Gamson (16) for the prediction of the thermal conductivity coefficient of dense gases were constructed from experimental viscosity and equationof-state data by means of the Enskog dense-gas kinetic theory (15), inasmuch as there were not enough high-density thermal conductivity data available for the preparation of a dimensionless graphical representation. This paper presents an empirical method for the calculation of the diffusion coefficients of dilute gases; it also describes the construction of a chart for the prediction of the selfdiffusion coefficient of dense gases analogous to those prepared for thermal conductivity by Comings and Nathan (11) and by Gamson (16).In a system composed of two substances, A and B, the binary diffusion coefficient, DAB, is defined by either of two equivalent relations (in the absence of thermal diffusion, pressure diffusion, and forced diffusion effects (4): In practice there are three types of binary diffusion systems i...