Comment on 'Surface reconstruction on Si(100) studied by the CNDO method'

In a recent paper Ong and Chan (see ibid., vol.1, p.3931 (1989)) examined seven different asymmetric dimer reconstructions for a Si(100) surface using the CNDO (complete neglect of differential overlap) method. The p(4*1) and c(4*2) reconstructions were found to be energetically more favourable, followed by p(2*2) and (2*1). Zandvliet comments only on the results of the four members of the (2*1) family, since experimentally the other asymmetric dimer reconstructions are not observed on the Si(100) and the Ge(100) surfaces.