Diffusion of hydrogen, deuterium, and tritium on the (110) plane of tungsten

Publisher Website
Google Scholar
Abstract
Diffusion coefficients for thermally activated and tunneling diffusion as well as mean square fluctuation data are presented for 1H, 2H, and 3H on W(110), as functions of coverage θ and temperature. If D is written as D0 e−E/kT for thermally activated diffusion, there are only small isotope effects, consistent with zero point energy differences for E, which increases slightly with coverage. There are dramatic isotope effects in D0, which can be approximated by ke−cM−1/2 at low θ, k and c being constants. It is suggested that this behavior results from the large vibron spacings relative to phonon energies, and that adsorbate induced changes in the surface phonon spectrum of W(110) are responsible for the changes in D0 with θ which are also very substantial. For tunneling diffusion isotope effects are remarkably small. This is rationalized in terms of additive mass corrections resulting from adsorbate–substrate interactions, which make the effective masses of the three isotopes very nearly the same. The variation in Dtunneling with θ is explained in terms of increasing barrier heights and slight decreases in effective mass as θ increases. Differences in the behavior of 1H and 3H on the one hand and 2H on the other in the tunneling regime at θ>0.6 are tentatively explained in terms of the statistics of the isotopes. The effect of oxygen on tunneling diffusion is also presented and discussed. Mean square fluctuations, proportional to the two‐dimensional compressibility of the adlayer show isotope effects, decreasing monotonically with increasing mass. Mixtures were also examined. In the activated regime D0 varies very strongly with composition at total constant θ=0.9, going through minima when the mole fraction of the lighter isotope in each binary mixture is ∼75%. Evidence for phase segregations in the mixtures is also presented: Diffusion and fluctuation results for mixtures are most easily interpreted by assuming complete miscibility of any two isotopes above 133 and below 90 K but segregation into two phases between these temperatures, the two phases coming closest to being pure at 115–118 K.