M. I. Heggie & C. D. Latham / Stability of the hypothetical superhard carbon metal H-6
We have taken the 4 atom unit cell with chains at right angles to each other (Hoffman's structure [5], figure 1), relaxed it to zero pressure and obtained a formation energy per atom of 0.96 eV with respect to graphite at 0 K. Then we built a larger tetragonal unit cell comprising 16 atoms built from the 4 primitive unit cells. This allowed the unhybridised p orbitals to relinquish their role in pi bonds and form sigma bonds with each other. When relaxed the structure contained 5-fold and 7-fold rings (as opposed to 6-fold rings in diamond) with a formation energy of 0.86 eV per atom. The resulting structure will be insulating because the valence electrons are all tied up in covalent bonds.
Taking the 6 atom unit cell of H-6 (chains at 60° to each other) and forcing the unhybridised p orbitals to remain in pi bonding yielded a formation energy per atom of 1.28 eV per atom for H-6, higher than the previously calculated tight binding value of 0.43 eV with respect to graphite [1] but closer to the ab initio result [6] of 0.92 eV. According to the tight binding calculations there is a barrier of 0.55 eV per atom for the transformation of this phase into diamond, but our results give no such barrier in agreement with the ab initio calculations [6]. This is an important difference since our results and ab initio ones indicate H-6 is not even meta-stable with respect to diamond. In connection with the failure of tight binding we should note here that the proximity cell potential gives the correct graphite to diamond transition energy (0.33 eV [2]) while the tight binding method gives 0.71 eV [1].
[Abstract] [Introduction] [Computer Models] [Results] [Acknowledgement] [References] [Figure 1]
| Christopher D. Latham | HTML 3.2: [W3C][WDG] |