We have revealed a mechanism for low-temperature water
adsorption on a graphite (0001) surface that is governed by a
delicate balance between substrate-adsorbate geometryconstraints, the strength of the graphite-water bond, and the
strength of the intermolecular hydrogen bond. We hypothesise
that the water molecules that initially strike the surface adsorb
and diffuse until they bind to defect sites. On a partially covered
surface, if the molecule strikes a patch of a clean surface, it will
eventually diffuse to the nucleation centre (defect site) and form
hydrogen-bonded aggregates. If, however, a molecule impinges
on an already formed cluster, it will adsorb and stay on top of
the cluster. The first layer therefore spreads out parallel to the
surface faster than the second and higher layers develop. The
effect is particularly pronounced on a graphite surface with its
low defect concentration* that makes the average size of water
clusters relatively large. Thus the growth mode of water films
on graphite resembles that of Stranski-Krastanov, instead of
Volmer-Weber’4.
Simultaneously, the structure of the water adlayer is temperature-
dependent, which is a direct consequence of the small
energy difference between bonding in 2D/3D nets, or amorphous/
crystalline structures. These different structures have different
desorption kinetics. The apparent zero-order kinetics (independent
of the coverage) result from the relatively slow transition
between different phases in the course of the TPD run.