Communication: Origin of the difference between carbon nanotube armchair and zigzag ends.

In this work, we have found that the difference between armchair and zigzag ends of carbon nanotubes (CNTs) does not pertain at close study for individual bonds and thus alternative strategies need to be developed to reach the ultimate goals in selective growth. Based on first-principles simulations, the difference between binding strengths for CNTs of different chirality was investigated using hydrogen dissociation energies at their passivated ends. When all H atoms are removed collectively we find the well-known difference: that armchair bonds are much weaker than zigzag ones, which is typically seen for both CNT ends and graphene edges. However, when individual H atoms are removed we find almost no difference in hydrogen dissociation energies, small difference in bond lengths, which by association means small difference in C-C and M-C binding energies. We show convincingly that the difference in binding energy between armchair and zigzag ends is due to a fragment stabilization effect that is only manifested when all (or several neighbouring) bonds are broken. This is because at armchair ends/edges neighbouring dangling bonds can pair-up to form C≡C triple bonds that constitute a considerable stabilization effect compared to the isolated dangling bonds at zigzag ends/edges. Consequently, in many processes, e.g., catalytic growth where bonds are normally created/broken sequentially, not collectively, the difference between armchair and zigzag ends/edges cannot be used to discriminate growth of one type over the other to achieve chiral selective growth. Strategies are discussed to realize chirality selective growth in the light of the results presented, including addition of C2-fragments to favor armchair tubes.