|dc.description.abstract||Methane cracking on transition metal surfaces is a catalytically important reaction. It is a key step to produce hydrogen and carbonaceous nanomaterials, such as carbon nanotubes (CNT) or carbon nanofibers (CNF), which display unique mechanical and electrical properties, and have been widely used as electronic components, polymer additives, gas storage, and catalyst support materials. Although the catalytic methane cracking and CNT/CNF growth have drawn lots of attentions, the understanding of the catalytic methane cracking properties and CNT/CNF growth mechanism is still limited.
To develop a better understanding of the catalytic methane cracking and the CNT/CNF growth process, the activation of the C−H bond of methane and the creation C−C bonds on transition metal catalysts, especially Ni, have been studied at atomic level using Density Functional Theory (DFT). Ni is of particular interest because, among the different metals commonly used in the methane cracking and CNT/CNF production, Ni-based catalysts show very good catalytic activity at relatively moderate temperatures. In this research, factors that affect the methane dissociation properties, e.g. effects of the catalyst structure, carbon deposition, oxide support and alloying, were analyzed using DFT calculations. The study of the Ni catalyst surface topology effect on methane dehydrogenation properties was conducted on various Ni catalyst surfaces, i.e., Ni (100), Ni (111) and Ni (553). The transition states for methane sequential dehydrogenations on the three surfaces were identified. The results show that the adsorption of CHx (x=1-3) and H species is favoured on less packed surfaces, e.g., Ni (100) and Ni (553). Moreover, it was found that the Ni (553) and Ni (100) promote the dissociation of CHx species by lowering the activation barriers when compared to Ni (111).The above study was conducted on clean Ni catalyst surfaces. During the reactions, however, there will be carbon atoms deposited on the Ni surface. To provide a more realistic modeling of the reaction, the study of Ni catalytic methane cracking is then further extended by taking into account the effects of carbon atoms depositions. Methane dissociation on clean, surface-carbon, and subsurface-carbon-covered Ni (111) surfaces were investigated. The results show that the existence of surface and subsurface C atoms destabilized the adsorption of the surface hydrocarbon species when compared to the clean Ni (111) surface. Moreover, it was found that the presence of carbon atoms increase the CHx (x=4-1) species activation barriers especially on the surface–carbon-covered (1/4 ML) Ni (111) surface, where CHx (x=4-1) species encounter the highest energy barriers for dissociation due to the electronic deactivation induced by C−Ni bonding. The calculations also show that CHx dissociation barriers are not affected by neighboring C atoms at low surface carbon coverage (1/9 ML).
The DFT study of Ni catalytic methane dissociation, so far, only focuses on Ni catalyst surface. However, in the actual process, the Ni catalyst is usually deposited on oxide support; little is known about the effect of the support, especially the metal-support interface, on the dissociation properties of methane. Therefore, the dissociations of methane and hydrogen on Ni cluster supported on γ-Al2O3 support were investigated using DFT calculations. Two systems: Ni4 cluster supported on the spinel model of γ-Al2O3 (100) surface, S(Ni4), and on the non-spinel model of γ-Al2O3 (100) surface, NS(Ni4), have been used to model Ni4/γ-Al2O3. For both models, it was found that CH4 and H2 dissociations are kinetically preferred at the particular Ni atoms located at the nickel-alumina interface when compared with the top of the Ni cluster. Also, the study of CH3 and H adsorption on different sites of the S(Ni4) and NS(Ni4) show that CH3 and H bonded with the Ni atom at Ni4/γ-Al2O3 interface are more stable than at the top site adsorption. Hirshfeld charge analysis showed that the surface Al atom works primarily as a charge donation partner when CH3 and H are bonded with the Ni atom at the interface. This also resulted in an up shift of the d-orbital around the Fermi energy, which finally stabilized the interface adsorption by this Al (donor)–Ni–adsorbates (acceptor) effect. The results obtained in the present analysis indicate that the metal-oxide interface plays an essential role in the dissociation of methane and hydrogen.
During the methane cracking process, carbon is deposited on the catalyst. Part of these carbon atoms will exists in the form of CNT, and some of them is deposited as encapsulating carbon in the form of graphene, which causes catalyst deactivation. To understand the role of metal elements in the growth of CNT or graphene, some crucial processes occurs on the (111) surface of different transition metals, i.e., Fe, Co, Ni, and Cu was analyzed using DFT. These processes consist of methane cracking to produce C, C atoms surface diffusion and C nucleation reactions. This study showed that Ni-based catalyst is a suitable substrate for growing CNT: it has a moderate reactivity towards C−H bond activation; low energy barrier for C atom surface diffusion, and a relatively high nucleation barrier for the surface C atoms. Meanwhile, this study also showed that Cu may be a suitable catalyst for synthesis of graphene due to the low diffusion and nucleation barriers of C adatoms on Cu. One key limitation of Cu is the low reactivity of this metal towards methane dissociation, which dominates the growth rate and reaction conditions of the process. Since Fe and Ni were found more reactive towards C−H bond breaking reactions, the results from this study indicate that Cu based alloys, e.g. Cu8Ni, may be a suitable catalyst for the mass production of graphene.
To further extend the understanding regarding the behavior of the carbon atoms during the Ni catalyst CNT growth, the structure, nucleation energetics, and mobility of carbon intermediates up to 6 atoms on the Ni (111) surface were investigated. This study showed that carbon clusters were more thermodynamically stable than adsorbed atomic carbon, with linear carbon structures being more stable than branched and ring structures. The results also showed that carbon chains have higher mobility than branched configurations. The transition states and energybarriers for the formation of different carbon clusters were also studied. The results suggest that the formation of the branched carbon configurations is kinetically favored as it presents lower energy barriers than those obtained for carbon chains.
Furthermore, based on the above DFT calculations results, a Ni catalytic CNT growth mechanism based on carbon species surface diffusion was developed. A multi-scale modeling approach that integrates DFT calculations and kinetic Monte Carlo (KMC) simulation was developed, in which the energetic results obtained from DFT calculations were used to set-up the kinetic database for the KMC simulation. The KMC simulations explicitly follow the elementary steps involved in the CNT growth that include CH4 dissociation, C surface and bulk diffusion, C nucleation, C3 trimer diffusion and C and C3 incorporation into CNT wall. The KMC simulations show that CNT growth is dominated by the C surface diffusion. Moreover, it was found that the surface diffusion of the small C cluster, e.g., C3 trimer, is also a critical step in the growth mechanism of the CNT. It prevents fast nucleation of the C atoms on the catalyst surface, and therefore inhibits the deactivation of the catalyst. The CNT growth rates predicted by KMC simulations fit well with the experimental data, verifying the proposed CNT growth mechanism. This study will therefore provide insight regarding the mechanism and kinetic properties of Ni catalytic methane cracking and CNT growth process.
In summary, a systematic theoretical investigation of the catalytic methane cracking and CNT growth process was performed in this study. It was found the catalyst structure, carbon deposition, and the γ-Al2O3 support has significant effect on the CHx dissociation properties. Moreover, DFT analysis also shows that the reactivity of the catalyst towards C−H bond activation and CNT or graphene growth varies with different transition metals. Finally, based on the DFT study of the carbon cluster nucleation, a CNT growth model that accounts for carbon cluster diffusion and nucleation was proposed. Using the kinetic parameters that obtained by the DFT calculations, a KMC simulation was developed. By comparing the CNT growth rate predicted by the KMC simulations with the experimental data, the proposed CNT growth mechanism is validated.||en