Surface silicon-deuterium bond energy from gas-phase equilibration - Comment

In a recent paper [Phys. Rev. B 48, 4492 (1993)] Wampler, Myers, and Follstaedt (WMF) have reported measurements of equilibrium adsorption of deuterium on cavity walls in crystalline silicon. Their procedure provides in principle a mon reliable source of information regarding the Si-H bond energy than any previous work. Here we propose an analysis of the data that has some advantages over the analysis given by WMF and that yields a significantly higher binding energy. We first argue that the measure of binding most directly following from the observations is the free energy at 800 degrees of a deuterium atom attached to an average surface silicon tetrahedrally bonded to three other silicons, relative to an 800 degrees surface with a ''dangling bond'' at this site and a deuterium at rest far outside: this foe energy is 3.29+/-0.1 eV. We then discuss ways of extracting from this a value for the binding energy Eg at absolute zero, to compare with recent predictions from first-principles quantum-mechanical calculations. This step requires some assumptions about the effect of chemisorption on crystal vibrations; reasonable assumptions give about 3.15 eV. with a probable error modestly larger than that of the 800 degrees C foe energy; the theoretical predictions range above and below this value by one or two tenths of an eV, depending on assumptions about reconstructions.