Search

Abstract: Carbon permeation into iron, a very important initial stage in iron-catalyzed heterogeneous reactions such as Fischer-Tropsch synthesis (FTS), is explored theoretically, to extend our thermodynamic and kinetic understanding of the process. The interaction of C atoms with five model surfaces (Fe (100), (110), (111), (211), (310)) was studied in six distinct ways. In the first, the random deposition of C atoms on the Fe surfaces was simulated by molecular dynamics, with C atoms released gradually. It shows that the early stages of carburization is a C permeation process, without much disturbance to the Fe surfaces. In the second approach, C atoms were approached to the surfaces sequentially. They bind readily (by 7-9 eV per C) to the surfaces, but to a different extent—strongest on Fe (100), and weakest on Fe (111). Addition of further C atoms proceeds with a slightly decreasing magnitude of the chemisorption energy, because of the increasing positive charges on the Fe atoms. At a certain coverage, different on each surface, C atoms prefer in calculation to go subsurface. C2 units formed on some of the surfaces. In a third approach, detailed transition paths of C permeation subsurface were calculated, with associated barriers in the order Fe (100) > (111) > (310) > (211) > (110). Differences in stacking geometries of the Fe layers in these surfaces appear to be the main cause of the variation. Comparing C permeation with surface migration on clean surfaces, the barrier of the former is smaller than that of the latter for most of the surfaces, except Fe (111). At intermediate C coverage, the (100) surface also prefers migration to permeation. In a fourth approach, we calculate that with increasing carbon chemical potential, the surface energies of iron (110), (111), and (211) surfaces decrease, while those of (100) and (310) first decrease, then increase. Based on these surface energies, a Wulff construction of nanoparticle facets is made. In a fifth approach, the position in energy of the d-band centers of the Fe surfaces upon C permeation was studied. For all the surfaces, the d-band centers move away from the Fermi level with increasing C coverage, and start to resemble those of the bulk carbide phases at high C coverage. In the last approach, we show that C permeation not only lowers the barriers of model reactions for CH4 formation and C-C chain propagation, two competing processes in FTS, but also changes the selectivity of the two competing processes. At high C coverage, chain propagation becomes preferred. A general picture emerges of C permeation on Fe surfaces as a stepwise process with opposite thermodynamic and kinetic preferences.Graphical Abstract: [ABSTRACT FROM AUTHOR]

The effect of Cu promoter on CH 4 formation from CO hydrogenation on Fe(100) has been investigated using spin-polarized density functional theory computations on periodic slab models. Two pathways of CH 4 formation are considered: (a) direct dissociation of CO followed by successive C hydrogenation and (b) successive hydrogenation of CO followed by deoxygenation of CH x O (x = 1–3). On the clean and Cu doped surfaces, direct dissociation of CO followed by successive C hydrogenation is more favorable kinetically and the overall CH 4 formation is endothermic. Compared with the clean surface, the Cu doped surface can suppress CH 4 formation by raising the barrier from the co-adsorbed CO + 4H. Methanol formation is unfavorable on both surfaces kinetically and thermodynamically. Reversely, methanol decomposition is favorable. It is also noted that Cu doping lowers the adsorption ability of the Fe(100) surface and therefore the adsorption energies of all species on the potential energy surfaces. [ABSTRACT FROM AUTHOR]

Spin-polarized density functional theory computations have been performed to investigate n CO + m H 2 co-adsorption and mutual interaction on the Fe(1 1 0) and Fe(1 1 1) surfaces. It is found that CO pre-coverage affects hydrogen adsorption strongly, while H pre-coverage hardly affects CO adsorption. It is possible to have CO + H 2 co-adsorption on Fe(1 1 0), while very strong preference of CO pre-coverage on Fe(1 1 1) has been found. On Fe(1 1 1), it is also found that each CO molecule blocks on average 2H adsorption sites at all CO pre-coverage. All these findings are consistent with available experimental results. Compared with CO/H 2 ratios in gas phase, CO/H 2 surface ratios are very different and much more sensitive to temperature than pressure. Our results provide the basis for exploring the mechanisms of iron-catalyzed conversion of synthesis gas. [ABSTRACT FROM AUTHOR]

Highlights: [•] Most, medium and least stable Fe5C2(111), (−411) and (001) surfaces are used. [•] Adsorption of H2O molecules and O atoms was calculated at high coverage. [•] Surface stability correlates with the adsorption energies of H2O and O. [•] High coverage H2O prefers hydrogen bonding instead of adsorption on iron. [•] The number of surface O atoms depends on temperature and H2O/H2 ratio. [ABSTRACT FROM AUTHOR]

Highlights: [•] H2O dissociative adsorption on the χ-Fe5C2(010) surface was investigated. [•] The χ-Fe5C2(010) surface has both iron and carbon regions. [•] Surface iron region is active for H2O dissociative adsorption and H2 formation. [•] Pre-adsorbed oxygen atoms play a significant role for further H2O adsorption. [•] Surface carbon region is inactive for H2O dissociative adsorption. [ABSTRACT FROM AUTHOR]

• Coverage dependent hydrogen adsorption on Ir(111), Ir(100) and Ir(110) surfaces. • Change of adsorption site on Ir(100) from low to high coverage. • Surface dependent repulsive interaction on Ir(111). • Surface dependent attractive interaction on Ir(100) and Ir(110). • Regular line-shaped adsorption structures were found on Ir(100) and Ir(110). Hydrogen adsorption on the perfect Ir(111) as well as the metastable and unreconstructed Ir(100) and Ir(110) surfaces up to saturation coverage has been systematically computed using periodic density functional theory and ab initio atomistic thermodynamics for understanding the interaction mechanism of hydrogen on iridium surfaces. On the Ir(111) surface including van der Waals dispersion, hydrogen adsorption prefers the threefold hollow sites at low coverage and the top sites at high coverage; in agreement with the experiments (Phys. Rev. B 60 (1999) 14016). The computed adsorption energy and desorption temperature of hydrogen agree with the experiments [−0.57 (fcc-3H) and −0.53 (hcp-3H) vs. −0.55 eV; 180 and 325 K vs. 190 and 310 K, respectively]. On the Ir (100) surface, the bridge adsorption sites are preferred in the whole coverage range, in agreement with the LEED pattern (Phys. Rev. B 73 (2006) 75430), however, adsorption energy and desorption temperature are slightly overestimated by including van der Waals dispersion (−1.50 vs. −1.02 ± 0.15 eV; 470 vs. 425 – 389 K). On the Ir(110) surface, short-bridge sites are preferred at low coverage and the top sites become dominant at high coverage, and the calculated desorption temperatures are close to experiments by including van der Waals dispersion (210 and 365 K vs. 220 and 375 K). At low coverage, the different configurations of hydrogen adsorption on the Ir(111) have the similar energies, indicating their negligible repulsive interaction, while the Ir(100) and Ir(110) surfaces prefer regular line-shape adsorption configurations due to attractive interaction, and such adsorption configurations have not been observed experimentally. Our results show that differences in adsorption configurations and energies are associated with their differences in surface structures, and in turn explain the need of different methods in computing the adsorption properties on different surfaces. Such surface-dependent properties should also be possible on other metal surfaces. Image, graphical abstract [ABSTRACT FROM AUTHOR]