(472g) First-Principles Theoretical Analysis Of Carbon Nanostructures And New Carbon Allotropes

Singh, T., University of Massachusetts - Amherst
Behr, M. J., University of Minnesota
Aydil, E. S., University of Minnesota
Valipa, M. S., University of Massachusetts
Maroudas, D., University of Massachusetts, Amherst

Diamond is a technologically important material due to its unique properties such as extreme hardness, high electrical resistivity, high thermal conductivity, and chemical inertness. It is used widely in the fabrication of machine tools, optical coatings, X-ray windows, and light-emitting optoelectronic devices. The synthesis of diamond or diamond-like thin films at room temperature by chemically-driven allotropic transformations has evinced keen interest recently; one such method to synthesize such films involves the treatment of carbon nanotubes (CNTs) and carbon nanofibers (CNFs) with H atoms from an H2 plasma. However, the development of systematic strategies for depositing thin films of diamond with desirable properties requires a fundamental understanding of the interactions between H atoms and the growth surface, which remains unclear. In this context, we aim at analyzing the crystalline C phases that are observed upon exposure of CNTs to an H2 plasma and to investigate the role of H in determining the structure of the transformed C films.

This presentation focuses on a detailed analysis of the various crystalline phases of C observed upon exposing CNTs to an H2 plasma, as well as the potential role of H in determining the structures of the resulting C nanocrystals. In our studies, we have combined experimental characterization tools based on high-resolution transmission electron microscopy (HRTEM) and electron diffraction with first-principles density functional theory (DFT) calculations. Our DFT calculations are carried out within the generalized gradient approximation and employ plane-wave basis sets, ultra-soft pseudopotentials, and slab supercells. Our experiments reveal that, upon exposure of CNTs/CNFs to an H2 plasma, a transformation is induced of the graphitic crystalline network to an amorphous carbon matrix, along with the formation of a web of crystallites (~ 5-10 nm in diameter) that are found embedded in the amorphous matrix. Subsequent HRTEM and electron diffraction studies have revealed the presence of two distinct crystalline phases embedded within the amorphous matrix. The observed electron diffractions and lattice spacing from these two phases are consistent with a face-centered cubic (fcc) lattice with lattice parameter, a, of 4.25 Å and a body-centered cubic (bcc) lattice with a = 3.0 Å; these structures are distinct from pure fcc carbon and bcc carbon phases, respectively.

Using DFT calculations, we have analyzed the structure of several new allotropes of pure carbon, which are candidate structures to explain the experimentally observed carbon phases. Our results indicate that, in the absence of H in the crystal structure, none of the allotropes can account for the lattice parameter of the observed carbon phases. Hence, we consider the incorporation of H in the interstitial sites of the cubic phases of carbon, and investigate the effect of varying the composition, CxHy (x,y < 3), of the carbon phases. Our calculations show that the bond-center (BC) configuration is the most stable interstitial H defect configuration in diamond carbon; in fact, the carbon lattice expands in order to accommodate the extra H atoms located at BC sites. However, there is a critical H concentration within the diamond lattice that onsets the stability of the CxHy lattice structure. From a comparison of the formation energies of the C2H and CH structures, we infer that there exists an intermediate configuration, with a lattice parameter of ~ 4.27 Å, which renders the CxHy lattice structure stable; this stable structure also can provide an explanation for the experimentally observed fcc phase. Similar analysis performed for the bcc phase reveals that there exists an intermediate configuration, with composition that varies between that of the CH (a ~ 2.90 Å) and CH2 structures (a ~ 3.10 Å), whose lattice parameter can account for that of the observed bcc phase. Hence, based on our theoretical analysis of C structures and allotropes, we conclude that the experimentally observed carbon nanostructures can be explained by accounting for the presence and concentration of H in the C lattice.