We tried seperating the dimer to give a single Si-Si bond along 110 between the Oi atoms. It was suggested that this might be the source of the 1105 cm-1 vibrational mode often seen at the same time as the dimer modes. The resultant structure is shown in Figure 6.8, with vibrational modes in Table 6.4. As can be seen, the modes are not in good agreement with the 1105 cm-1 LVM, excluding this as a model. Instead of dropping, the assymetric Oi stretch modes increase in frequency due to the proximity of the atoms. In addition the calculations gave this structure as over 1.5 eV less stable than the puckered dimer, which although probably an overestimate, suggests the defect is not stable in this configuration. The defect is not positioned symmetrically in the cluster, and this manifests itself in the assymetric atomic shifts from their ideal lattice positions, and the split top mode; in the fully symmetric structure this would be degenerate with D2h symmetry.
Earlier calculations showed that the energy of this structure could be lowered by placing the Oi atoms in perpendicular 110 planes, thus lowering the conflict between their stress fields. This roughly corresponds to cis- and trans- isomers. The separation of the O atoms is now a/8 (2 3 3). However the energy differences between these structures is quite dependant on the basis set used and number of bond centres, and so although this earlier work predicted a binding energy of 0.7 eV for the dimer, we might expect an error on this of 50% or more.