A quantitative study has been made of the reaction of ethyl radicals with molecular hydrogen in the gas phase in the temperature range 240 to 320 "C. The mercury (63P1) photosensitized decomposition of hydrogen in the presence of ethylene was used to generate ethyl radicals. Extinction coefficients for the absorption of 2537 &, by mercury vapor were measured and Beer's law was shown to be obeyed under the experimental conditions used. T h e corrections required to allow for the nonuniformity of radical concentrations in the cell were small. After delineating the experimental conditions necessary to minimize secondary reactions, the rate constant (cn13 mole-' s-') for the reaction CzHS + Hz = CzH6 + H was found to be given by log,, k = 12.57 -13.718. Experiments in the presence of added carbon dioxide showed the absence of hot radical effects at the working pressure of 92 Torr of hydrogen.Canadian Jorlrnal of Chemistry, 46, 3275 (1968) Relatively few atom transfer reactions of the I type R + X, = R X + X, where X = H or D, have been studied (1-3), and in the even fewer cases where the reaction of a given radical with both H, and D, has been investigated there is some doubt about the magnitude of the isotope effect on the activation energy. For n-C,F, radicals, values of 14.0 (4) and 13.8 (5) kcal mole-' have been found for the reaction with D,, 1.9 (4) and 1.5 (5) kcal mole-' greater than for the reaction with Hz. For the reaction of C2F5 with D, and Hz (4) the activation energies were almost the same: 12.6 and 12.4 kcal mole-', respectively. An intermediate isotope effect was found in the case of CF,; values of 9.5 (6) and 10.2 (7) kcal mole-' were found for the reaction with D,, 0.9 (6) and 0.7 (7) kcal mole-' greater than for the reaction with Hz, although the probable errors quoted by the authors in these cases were such that a negligible isotope effect would also be compatible with their results. In the case of C2H5 radicals the value found for the reaction with D, was 13.3 kcal mole-' (8) while the value for the reaction with H, (9), as subsequently corrected (lo), was only 11.3 kcal mole-'.The apparent isotope effect for C2H5 radicals (8, lo), 2.0 kcal mole-', is very close to the difference between the zero point energies of Hz and D,. However, there is no theoretical justification for equivalence with zero point energy differences, and recent measurements in this laboratory (1 1) show no such correlation in the 'Present address: Research Department, Imperial Oil Enterprises Ltd., Sarnia, Ontario.case of the three-center exchange reactions of atomic and molecular hydrogen and deuterium.If we accept a value of 6.8 kcal mole-' (12) or greater (13) for the activation energy of the reaction H + CzH6 = CzHs + Hz and a value of about 6.2 kcal mole-' (14, 15) for -AH, then the activation energy for the reaction CzHs + Hz = CzH6 + H should be > 13.0 kcal mole-'. In view of the uncertainty in this value we felt it necessary to reinvestigate the reaction.Aside from possible errors in the interpretation of the ea...

The interdiffusion coefficient of thallium in thallium amalgams was measured by the capillary reservoir technique. At 25" C the values of 106D in cm2 sec-I a t 0. 00365, 0.0536, 0.0895, 0.148, 0.188, 0.283, 0.323, and 0.404 mole fraction thallium were 1. 08, 1.37, 1.50, 1.37, 1.14,0.83, 0.95, and 1.11 respectively with standard deviations in 10-S(cm2 sec-I) of 0. 03, 0.06, 0.09, 0.10, 0.09, 0.08, 0.08, and 0.11 respectively. The minimum in the value of D a t 0.288 mole fraction thallium shows that the compound T12HgS plays a significant role in the diffusion process.Values of a In a/a In N for the thallium and for the compound were used to explain the dependence of D on the concentration.This study of diffusion in thallium amalgams was undertaken because this unusual system is liquid a t room temperature from 0 t o 42 wt% thallium and because accurate thermodynamic data are available from the classic work of Richards and Daniels (1) on thallium amalgam concentration cells. For the concentration range 0.137-0.160 wt% thallium, von Wogau (2) reports that a t 11.0-12.0' C the diffusion coefficient of thallium is 1.03 )( cm2 sec-I; and, for the concentration range 0.00216 to 0.00287 wt% thallium, Furnam and Cooper (3), from polarographic measurements a t 25' C, found that the diffusion coefficient was 0.99 X cm2 sec-l. Stromberg (4), from polarographic measurements -on a 1 0 4 wtyo amalgam, deduced a value of I.60X lov5 cm2 sec-l.The experimental method used was based on that employed by Anderson and Saddington (5) with the modification that a chemical rather than a radiochemical method of analysis was used. Fick's second law of diffusion for a one-dimensional system states that where c is the concentration of the diffusing species, t is the time, and D is the diffusion coefficient with dimensions of area per unit of time. Consider a capillary of length I, with one end closed, having an initial concentration Ci uniform throughout. Let the capillary be immersed in so large a volume of solution of composition C, that a t infinite time when diffusion is complete and the composition within the capillary and outside the capillary is the same the composition is still C,. The solution of equation [I] for time t under these circumstances is where C,, is the average composition of the solution in the capillary a t time t. The series on the right side of equation [2] is rapidly convergent.In the derivation of equation [2] from equation [I] it was assumed that the diffusion coefficient is independent of the concentration. When the difference in concentrations of the two solutions is small, it may be assumed that the diffusion coefficient is constant. In this present work a chemical method of analysis was used t o evaluate the terms on the left-hand side of equation [2] with the result that in both the numerator and the denomi-