Abstract We have developed the reverse Monte Carlo (RMC) modeling method to generate two physically acceptable models for tetrahedral amorphous carbon (ta-C). The models have been produced by fitting to experimental neutron diffraction data and by applying various constraints consistent with chemical and physical knowledge of the material. In particular, three different carbon atom types (tetrahedral, planar and linear) have been defined and a new approach to applying coordination number constraints has been developed. The models were generated from an initial random distribution of 3000 atoms and from these we have determined ring number and coordination number distributions that are statistically averaged over the whole structure. We are able to look in detail at the distribution of different types of carbon bonding sites within the box and at the effects of incorporating 5 at.% H into the structure. The new models show that atoms with sp2 bonds tend to cluster and form polymer-like chains, which interlink the regions of sp3 or diamond-like bonding. Also, the relative flexibility of these polymeric chains results in more porous, less dense regions of the network. A comparison with other models for ta-C is also made.
Read full abstract