Two models from the recent literature, proposed to describe the appare
nt effect of Ge coverage on silicon monohydride desorption kinetics, a
re evaluated based on their ability to fit temperature-programmed deso
rption data, an evaluation of the physical consistency of the estimate
d kinetic constants, and a comparison with the effects of other atomic
impurities (B, P, C) on hydrogen desorption from Si(100). The first m
odel attributes the decrease in the peak temperature for silicon monoh
ydride (beta(1)) desorption with increasing Ge coverage to a long-rang
e electronic effect that reduces the activation barrier uniformly over
the entire surface. It is shown that this model fails to fit the high
Ge coverage data unless the preexponential factor also decreases by a
pproximately nine orders of magnitude, which is physically implausible
. The second model considers the possibility of an alternate pathway t
o depopulate the silicon monohydride phase, in which hydrogen diffuses
from the silicon to the germanium surface phase, and desorbs rapidly
from a short-lived GeH intermediate. The estimated activation barrier
for the surface migration step of 25+/-1 kcal mol(-1) is thermodynamic
ally consistent with the energetics of hydrogen desorption from Si and
Ge, and its magnitude is intermediate between the estimated activatio
n barriers for hydrogen surface diffusion on Si and Ge. Although Ge ma
y modify the energetics of the interaction of hydrogen and silicon, it
is concluded that such an effect alone is insufficient to describe th
e shift and broadening of the beta(1) feature, while the GeH intermedi
ate model succeeds, even in the absence of any such effect.