next up previous contents
Next: Annealing of the damage Up: General description and justification Previous: Comparison with ab initio

   
Generation of a damaged region in diamond

In a previous publication [112], we searched for the minimum energy (the "displacement energy" Ed) required to permanently remove a carbon atom from its initial diamond lattice site for various displacement directions. We found that for a "random" direction, defined as an ``off-axis'' direction, 7o off the <100> axis in the xy plane and 5o perpendicular to it [113], Ed =52 eV, in agreement with results of previous calculations and with the average of the displacement energies experimentally found for different directions in the diamond crystal[112,114].

To create a heavily damaged region in diamond we have energetically displaced bulk atoms with a momentum which corresponds to an initial kinetic energy of 416 eV (i.e 8 times Ed). These calculations required the use of the relatively large sample $(10\times 8\times 8$unit cells), so that the effects due to the damage were contained in less than 1/3 of the total volume of the sample. Although the calculations were carried out at 0K, the introduction of kinetic energy to some atoms has led to the local heating of the sample, as will be described in the next section. The Berendsen's energy dissipation process [115] described above was thus applied, to rapidly quench the configuration around the defects and return the crystal back to $\sim$ 0 K within about 2.5 ps after one bombarding event. Up to 12 carbon atoms (each with an energy of 416 eV) were knocked, one at a time, into the same volume from slightly different directions along ``off-axis'' orientations, so that the damage regions created by all recoiling atoms overlapped, and the largest component of the initial velocity of the 12 atoms displaced was x.

To check the generality of our results, we have displaced in a similar manner 12 different atoms at different ``off-axis'' directions, also at T= 0 K. The similarity of the structural configurations of the damage obtained for the two cases, and their behavior under post-bombardment annealing, as deduced from statistical analysis, proved that the general nature of the heavily damaged region is independent of the details of the damaging process. We selected an initial kinetic energy of 416 eV imparted to 12 atoms because under these conditions we obtain a heavily damaged region inside an intact diamond matrix. Higher energies would require larger samples to minimize boundary effects, at the cost of computation time. Much lower energies would produce a too small damaged region to yield statistically meaningful results.

We also created sample with a much lower density of defects. This was obtained by the energetic displacement of just one atom, with an initial kinetic energy of 416 eV. Exactly the same boundary conditions an energy dissipationas those above, were applied in this case. By comparing the results obtained with the heavily damaged and the slightly damaged samples, we could study the role play by the defect density on the post annealing structure.


next up previous contents
Next: Annealing of the damage Up: General description and justification Previous: Comparison with ab initio
David Saada
2000-06-22