Interaction of Group V elements with GaAs surfaces

4 INTERACTION OF GROUP V ELEMENTS WITH GaAs SURFACES 

The impetus for this topic has been provided by the development of molecular beam epitaxy (MBE) as a viable thin film deposition process [ 111]. As a result, the approach has concentrated more on investigations of reaction kinetics than on electronic effects, since kinetic parameters are directly available from modulated molecular beam measurements (see Sect. 2.4.1). We will summarize here only the results for beams of As4 and As2 interacting with 100 GaAs surfaces, but closely similar behaviour is observed for other Group V elements and other Group III—V compound surfaces. The choice of tetramer and dimer beams is dictated by the evaporation behaviour of Group V elements in that elemental sources produce tetramers and Group III—V compound sources produce dimers. Monomeric species are not readily available. 

If we consider first the case of an incident As4 beam, the essential features of the model proposed by Foxon and Joyce [343] are illustrated in Fig. 43. As4 is adsorbed into a mobile precursor state and in the absence of a surface Ga population has a zero sticking coefficient but a 

These results can be explained by a process of dissociative chemisorption with a pairwise interaction of As4 molecules adsorbed on adjacent Ga lattice sites. When the As4 surface population is small compared with the number of Ga sites, the rate-limiting step is the encounter/reaction probability between As4 molecules, leading to second-order kinetics. As the As4 surface population is increased, there is an increasing probability that an arriving molecule will find adjacent sites occupied and the desorption rate becomes proportional to the number of molecules being supplied, i.e. a first-order process. 

This model has been criticized by Jewsbury and Holloway [344], who suggested on theoretical grounds that the second-order behaviour arose not from a pairwise interaction of As4 molecules, but simply because an As4 molecule could only desorb from an As site on the surface. The problem with this interpretation is that the sticking coefficient of As4 should become unity for a Ga-stable 100 surface, whereas experimentally its maximum value is 0.5 and then only for a Ga-saturated surface. 

---