Classical models of metal desorption from semiconductor surfaces

7 Classical models of metal desorption from semiconductor surfaces 

The most plausible models to have been developed for metal—semicon-dutor systems are all based on the original concepts of Arthur [331], with Kern and his co-workers [279, 282, 332] being principally responsible for more recent extensions and elaborations. 

Arthur assumed the metal film to be in the form of two-dimensional 

At steady state, dn/dt = 0, i.e. at constant temperature the precursor state population is constant, and this simplification allows the desorption rate, — dn/dt, to be written as 

There are, however, three very important implicit assumptions in this model, apart from those of an ideal interface. Firstly, since desorption is only allowed to occur from a constant precursor state population (dn/dt = 0), it is effectively always a zeroth-order process. If a different order is observed, desorption is not the rate-limiting step. The second point is that this treatment is only appropriate for cases where the metal-metal bond energy (around the peripheries of the islands) is less than that for the metal- semiconductor, since for the opposite case the weaker adsorbate—surface bond will not prevent an atom desorbing once it has acquired sufficient energy to break the (stronger) metal—metal bond. Thirdly, no provision is made for possible diffusion of the adsorbate into the substrate during desorption. 

---