The study of dislocation core structures from first principles has flourished in the decade since the first calculation which confirmed reconstruction of the single period core of the 90o partial in silicon . Supercell calculations reproduce this result  and also find a double period structure which could be degenerate or even lower in energy depending on the elastic environment [3-5]. In addition, more refined approaches have given us simulated EELS spectra and the interactions between point defects and dislocations [1,6,7]. Kinetic Monte Carlo simulations using first principles (or tight binding) energies and activation barriers have developed our understanding of kink dynamics beyond the understanding of the Hirth-Lothe model [8,9]. Modelling is thus in a position to tackle complex dislocation problems which include static and dynamic chemical interactions, which is the subject of the present contribution.
H at dislocations in silicon has been shown to enhance dislocation glide (HEDG effect), both in theory  and experiment . Carrying through the theoretical analysis from silicon to diamond, indicates a similar effect should operate. A curiosity of the HEDG experiment in silicon is the unknown mode of intermediate storage of H which appears to drive the HEDG in the absence of the activating plasma . This storage could be forms of H which occur prior to the formation of H platelets – we therefore present first principles study of clusters of Hn where 20 > n > 4 and show that shear relaxation for these defects is more important than dilation.
Finally, we turn to diamond. We note that undercoordination often occurs in dislocation cores and that carbon is expected to differ from silicon in that this undercoordination can be accommodated by rehybridisation and formation of double bonds (which are unfavourable in silicon). We describe the recently discovered ‘ethene-like’ core of the 90o partial in diamond, which is metastable, and the conditions under which it may become the locally stable state .