Search

In heterogeneous catalysis, supported single-atom, single-cluster, and nanocatalysts have attracted wide attention. The interaction of single atoms and single clusters with support determines the stability and activity of the catalysts. To understand the type and strength of metal–support interaction, the adsorption structures and energies of transition metals (Mn= Fe, Ru, Os, Co, Ni, and Cu, n= 1–6) on the fully dehydrated silica (001) surface have been systematically computed based on spin-polarized density functional theory (GGA-rPBE-D3). It is found that aggregated structures are more stable than singly dispersed structures, indicating stronger metal–metal interaction. The adsorption energy increases from Fe, Ru to Os, while decreases from Fe, Co, Ni to Cu. Electron transfer from the metal to support has been found for Fe, Co, Ni, and Cu, while that from the metal to support and within the metal cluster has been found for Ru and Os, indicating the additional polarization and electrostatic interaction.

The adsorption and dissociation of O2on Ir(100) has been studied with X-ray photoelectron spectroscopy (XPS), low energy electron diffraction (LEED) and scanning tunneling microscopy together with density functional theory calculations. At low coverage, XPS shows a single O 1s peak at 530.6 eV, associated with bridged O atoms in an Ir(100)–(3 × 1)–O structure. At saturation coverage, an additional peak at 529.5 eV appears, accompanied by an Ir(100)–(2 × 1)–O structure. The O 1s peak at 530.6 eV is assigned to oxygen atoms adsorbed on bridge sites of Ir atoms, and the peak at 529.5 eV to oxygen atoms on fourfold hollow sites of Ir. Based on our detailed experimental and computational study, oxygen diffusion into the subsurface region can be ruled out. The present study provides new findings on an additional oxygen state of the O/Ir(100) system. It also provides evidence for the absence of subsurface oxygen in Ir, implying that some discussions on subsurface oxygen in the literature may be misleading.

To understand the mutual interaction and synergistic effect of transition metals and supports in heterogeneous catalysis, the less coordinated and more active surface MoAatoms were substituted with Fe, Co, Ni, Cu, Pd, and Pt doping atoms for investigating the adsorption of CO, H2, H2O, and CO2as well as OH, H, and O. Metal loading affects the surface electronic structure. On these surfaces, Fe, Co, Ni, Cu, and Pd doping atoms are positively charged, indicating electron transfer from the metal to the surface, while Pt doping atoms are slightly negatively charged, revealing electron transfer from the surface to the metal. On the pure surfaces and surfaces doped with four metal atoms (4M, 25%), surface MoAatoms are most preferred adsorption sites. By replacing all surface MoAatoms with eight doping atoms (8M, 50%), the more coordinated and less active surface MoBatoms become active. Not only surface metal atoms but also surface carbon atoms are active for the adsorption of surface species. Depending on doping metals, the adsorption of surface species can become slightly more or less exothermic. Exploring the dissociative adsorption of H2O and CO2reveals metal- and loading-dependent potential energy surfaces. Full H2O dissociative adsorption is favored thermodynamically on the 4M-doped surfaces and more exothermic than on the pure surface while doping metal-dependent on the 8M-doped surfaces. CO2dissociative adsorption is thermodynamically favored on the 4M-doped surfaces, while it becomes endothermic on the 8M-doped surfaces, which prefer either molecular adsorption or equilibrium between molecular and dissociative adsorption. Comparing the adsorption of CO, OH, O, and H on the pure and doped Mo2C(101) surfaces as well as the corresponding metallic low-index M(hkl) surfaces reveals their similarity and difference. These results provide a basis for studying the mechanisms of reactions involving these surface species.

The mechanisms of iridium-catalyzed nonoxidative conversion of light hydrocarbons have been computed on flat Ir(111) and stepped Ir(211) surfaces under given conditions. On Ir(111), CH4dissociation is endergonic; the first C–H dissociation is rate-determining, and surface CH* is the most stable species. On Ir(211), CH4dissociation is exergonic; the first C–H dissociation is rate-determining as well as surface CH* and C* are the most stable species in equilibrium. On Ir(111), ethane dissociation is exergonic and has the minimum energy path of CH3CH3→ CH3CH2*+ H* → CH2CH2* + 2H* → CH2CH* + 3H* → CH2C* + 4H* → CH3C* + 3H* at 180 K, in agreement with the experiment of ethylene dissociative adsorption and the co-adsorption of acetylene and hydrogen. At 450 K, CH3C* further dissociates in the order of CH3C* → CH2C* → CHC* → C* + CH*; however, coverage CHC* prefers coupling rather than dissociation, in agreement with the experimental observation. On the stepped Ir(211), ethane dissociation is exergonic and can have surface ethylene and acetylene as intermediates, and all reactions end up with surface CHC*, which can further dissociate into surface CC* depending on hydrogen pressure. The adsorption energy and configuration of CH* and CHCH* at high coverage on Ir(111) at 450 K and CC* on Ir(211) at 510 K from the dissociative adsorption of methane, ethane, and ethylene have been discussed.

Periodic density functional theory calculations were employed to study the mechanisms of H2O dissociative adsorption on the Ni(211) surface at different coverages. Due to stronger H-bonding, H2O clustering is more favored than direct Ni–O interaction. The first dissociation step has a lower barrier and is much more exothermic than the second dissociation step (0.91/–0.89 vs 1.45/–0.01 eV). Surface OH represents the most stable and important intermediate on the clean and O-precovered surfaces. Taking H2O as oxidant [H2O = OH + 1/2H2(g); H2O = O + H2(g)], the saturation coverage of OH and O is 0.67 and 0.33 ML, respectively. Comparing the results of H2O dissociative adsorption on the Ni(111), Ni(100), Ni(110), and Ni(211) surfaces, which can represent a nickel nanocluster or nanoparticle having different facets, shows that there is no general correlation of adsorption energy (OH, H, and O) and saturation coverage (OH and O) with surface stability (energy). There is also no direct Brønsted–Evans–Polanyi relation between reaction barriers and reaction energies (H2O = HO + H, 2OH = O + H2O), but they correlate well for OH + H = H2O. This diversity offers the mechanistic insight of the H2O dissociative adsorption of nickel nanoparticles and provides the basis for investigating water-involved reactions catalyzed by nickel having different facets.

Experimentally, surface carbon hydrogenation on metallic iron forms CH4, ethylene/ethane, and propene/propane, while previous density functional theory studies showed that these reactions are highly endothermic. This disagreement can be attributed to the shortcoming of low coverage of surface carbon atoms and the underestimated role of surface hydrogen. On a p(4 × 4) Fe(110) surface with 0.25 ML carbon coverage (4 C atoms) and other free sites filled with H atoms, we studied surface C hydrogenation with spectators and found that the formation of CH4, ethylene/ethane, and propene/propane becomes exothermic. Coupling of CH + CH and CH3C + CH to acetylene and propylene is favored thermodynamically. Next, CHCH and CH3CCH can be hydrogenated into CH2CH and CH3CHCH, and the subsequent hydrogenation of CH2CH and CH3CHCH determines the formation and selectivity of alkenes and alkanes. The most important surface species is carbide HC for chain initiation or CH3C (RC for higher homolog) for chain propagation. This model of high coverage with spectators may help understand the mechanism of heterogeneous catalysis under real conditions.

Dissociation of CO often occurs on Ru surfaces with dense chemisorbed CO (CO*) adlayers, ubiquitous in the practice of catalysis, where the strong bond in CO tends to weaken via bimolecular reactions with H2before cleavage. Nevertheless, H2adsorption on CO*-crowded Ru surfaces is a rare event because of the CO*-induced barrier and the competitive adsorption between CO and H2. Here, we thereby performed a comprehensive ab initio study of the coadsorption of CO and H2as well as the following initiation reaction on Ru(0001). At 100–550 K (3 MPa, H2/CO = 2), Ru(0001) is always saturated, covered by CO with a “hexagonal” pattern, where H2adsorption is neither kinetically nor thermodynamically favored. On the basis of a systematic analysis of the electronic structure, we find that two adjacent coadsorbed CO* can serve as a promoter for the scission of the H–H bond in H2(g), leading to the simultaneous formation of two COH* molecules in one step (Ea= 1.64 eV). This so-called CO self-promoting hydrogenation route offers new insights into the fundamental mechanism for CO activation on CO*-crowded low-index Ru surfaces.

CO adsorption on the perfect Ir(111) as well as the unreconstructed Ir(100) and Ir(110) surfaces at different coverages has been computed at the level of periodic density functional theory. General agreement between theory and experiment is found in the top adsorption configuration at low coverage (vibrational frequencies) and adsorption patterns at a given coverage (low-energy electron diffraction) as well as saturation coverage (thermal desorption). At the lowest coverage, the computed CO adsorption energy on Ir(111) agrees very well with the experiment [−1.95 vs −1.91 ± 0.09 eV], whereas large differences are found on Ir(100) [−2.22 vs −2.04 ± 0.05 and −1.47 eV] and Ir(110) [−2.36 vs −1.61 eV]. The same trend is also found in the CO desorption temperature at the lowest coverage, that is, best agreement on Ir(111) [590 vs 560 K] and large differences on Ir(100) [620 vs 572 K] and Ir(110) [700 vs 620 K], while good agreement at high coverage on all surfaces. At the lowest coverage, CO dissociation has a very high effective barrier and is endothermic, in agreement with the experiment. Our results provide the basis for exploring CO activation and reaction on the Ir surfaces under real conditions.

To understand the mechanisms of Fe-based Fischer–Tropsch synthesis in a retrosynthetic way, successive dissociation of CH4, C2H6, and CH3CHO on the Fe(110) p(5 × 5) surface has been systematically computed using density functional theory. The successive dissociation of CH4[CH4→ H + CH3→ 2H + CH2→ 3H + CH → 4H + C] and C2H6[C2H6→ H + CH3CH2→ 2H + C2H4→ 3H + CHCH2→ 4H + CHCH → 4H + CH + CH] has close apparent barrier (0.69 and 0.65 eV, respectively) and is highly exothermic (−2.47 and −3.17 eV). The CH3CHO successive dissociation is barrierless and favors the minimum energy path of CH3CHO → H + CH3CO → 2H + CH2CO → 3H + CHCO → 4H + CH + CO. All of these successive dissociations result in the formation of surface species H, C, and CO, which are related with the initial steps of the adsorption and activation of H2and CO. It is particularly noted that the computed results that ethylene dissociates easily into acetylene and acetylene dissociates reversibly into CH + CH agree with that experimentally observed. Under the consideration of gaseous H2and surface carbonaceous species, the formation of CH4and C2H6has very low apparent barriers and is favored thermodynamically, in agreement with the experiment from carbon deposited on Fe(110) surface and H2gas. This indicates that surface coverage plays a very important and decisive role in determining reaction mechanisms.

To understand the initial steps of Fe-based Fischer–Tropsch synthesis, systematic periodic density functional theory computations have been performed on the single-atom stepped Fe(710) surface, composed by p(3 × 3) Fe(100)-like terrace and p(3 × 1) Fe(110)-like step. It is found that CO direct dissociation into surface C and O is more favored kinetically and thermodynamically than the H-assisted activation via HCO and COH formation. Accordingly, surface O removal by hydrogen via H2O formation is the only way. On the basis of surface CHxhydrogenation (x= 0, 1, 2, 3), surface CHx+ CHxcoupling and CO + CHxinsertion resulting in CHxCO formation followed by C–O dissociation, surface C hydrogenation toward CH3formation is more favored kinetically than the formation of CHx-CHxand CHxCO, as well as thermodynamically. Starting from CH3, the formation of CH4and CH3CO has similar barriers and endothermic reaction energies, while CH3CO dissociation into CH3C + O has low barrier and is highly exothermic. Therefore, turning the H2/CO ratio should change the selectivity toward C–C formation and propagation.

Surface O removal by H and CO on Co(0001) has been studied using periodic density functional method (revised Perdew–Burke–Ernzerhof ; RPBE) and ab initio atomistic thermodynamics. On the basis of the quantitative agreement in the H2O formation barrier between experiment (1.34 ± 0.07 eV) and theory (1.32 eV), H2O formation undergoes a consecutive hydrogenation process [O + 2H → OH + H → H2O], while the barrier of H2O formation from OH disproportionation [2OH → H2O + O] is much lower (0.72 eV). The computed desorption temperatures of H2and H2O under ultrahigh vacuum conditions agree perfectly with the experiment. Surface O removal by CO has a high barrier (1.41 eV) and is strongly endothermic (0.94 eV). Precovered O and OH species do not significantly affect the barriers of H2O and CO2formation. All of these results indicate that the present RPBE method and the larger surface model are more suitable for studying cobalt systems.

On the basis of density functional theory including dispersion correction [ωB97XD/6-311+G(2df,2p)//B3LYP/6-311G(d,p)], the thermodynamics and kinetics of the reactions of CH3OH and CH3OCH3over H-ZSM-5 have been systematically computed. For the reaction of the methylated surface (CH3OZ) with CH3OH, CH3OCH3formation is kinetically controlled and the competitive formation of CH2O + CH4is thermodynamically controlled, in agreement with the observed desorption temperatures of CH3OH, CH3OCH3, and CH2O under experimental conditions. For the reaction between ZOCH3and CH3OCH3, the formation of the framework stabilized (CH3)3O+is kinetically controlled, consistent with the NMR observation at low temperature, and the competitive formation of surface CH3OCH2OZ + CH4is thermodynamically controlled. On the basis of the thermodynamically more favored CH2O and CH3OCH2OZ, there are two parallel routes for the first C–C bond formation, from the coupling of CH3OCH2OZ with CH3OH and CH3OCH3as well as from the coupling of CH2O with CH3OH and CH3OCH3. The most important species is the methylated surface (CH3OZ), which can react with CH3OH and CH3OCH3to form the corresponding physisorbed CH2O and chemisorbed CH3OCH2OZ, and they can further couple with additional CH3OH and CH3OCH3to result in first C–C formation, verifying the proposed formaldehyde (CH2O) and methoxymethyl (CH3OCH2OZ) mechanisms.

Spin-polarized density functional theory calculations were performed to investigate N2dissociative adsorption on iron (100), (110), (111), (210), (211), (310), and (321) surfaces. An ordered c(2 × 2) structure was found on Fe(100) at 0.5 monolayer coverage, which is in excellent agreement with the experiment; and a c(4 × 2) ordered structure is also found at 0.75 monolayer saturation coverage. Strong surface reconstruction is found on Fe(110) upon nitrogen adsorption, where the densely packed (110) is reconstructed into (100)-alike. Under the consideration of temperature and N2partial pressure, the estimated N2desorption temperature on Fe(100) at 925 K agrees with the experimentally detected 920–950 K. In addition, N2pretreatment results in Fe(100) to be mostly exposed, while that of H2pretreatment favors Fe(110). Further direct comparison of N2and H2adsorptions has been made to show their differences and similarities.

The reactions of CO and H2O on the clean Fe(110) surface as well as surfaces with 0.25 monolayer O, OH, and H precoverage have been computed on the basis of density functional theory (GGA-PBE). Under the considerations of the reductive nature of CO as reactant and H2as product as well as the oxidative nature of CO2and H2O, we have studied the potential activity of metallic iron in the water-gas shift reaction. On the clean surface, CO oxidation following the redox mechanism has a similar barrier as CO dissociation; however, CO dissociation is much more favorable thermodynamically. Furthermore, surfaces with 0.25 monolayer O, OH, and H precoverage promote CO hydrogenation, while they suppress CO oxidation and dissociation. On the surfaces with different CO and H2O ratios, CO hydrogenation is promoted. On all of these surfaces, COOH formation is not favorable. Considering the reverse reaction, CO2dissociation is much favorable kinetically and thermodynamically on all of these surfaces, and CO2hydrogenation should be favorable. Finally, metallic iron is not an appropriate catalyst for the water-gas shift reaction.

Water dissociative adsorption on the clean and O-precovered Fe(111) surfaces at different coverage have been studied using the density functional theory method (GGA-PBE) and ab initio atomistic thermodynamics. On the clean p(3 × 3) Fe(111) surface, surface H, O, OH, and H2O species can migrate easily. Considering adsorption and H-bonding, the adsorbed H2O molecules can be dispersed or aggregated in close energies at low coverage, while in different aggregations at high coverage, indicating that the adsorbed H2O molecules might not have defined structures, as observed experimentally. On the O-precovered surface (nO= 1–8), the first dissociation step, nO + H2O = (n– 1)O + 2OH, has a very low barrier and is reversible; and the barriers of the sequential OH dissociation steps, (n– 1)O + 2OH = nO + H + OH and nO + H + OH = (n+ 1)O + 2H, are close (0.9–1.2 eV). All of these barriers are coverage independent. For OH and H adsorption at 1/3 ML coverage, surface OH forms a trimer (OH)3unit, and surface O forms a regular linear pattern. At one ML coverage, there are three dispersed (OH)3units for OH adsorption and three well-ordered parallel lines for O adsorption. The average adsorption energies for OH and O adsorption are coverage independent. Desorption temperatures of H2O and H2under ultrahigh vacuum conditions are computed. Systematic comparison among the Fe(110), Fe(100), and Fe(111) surfaces reveal their intrinsic differences in water dissociative adsorption and provide a basic understanding of water-involved reactions catalyzed by iron and interaction mechanisms of water interaction with metal surfaces.

Spin-polarized density functional theory computations have been carried out to investigate the surface morphology of Cunadsorption on the Fe5C2(100), Fe5C2(111), Fe5C2(510), Fe5C2(001), and Fe5C2(010) surface terminations in different surface Fe and C ratios. On the Fe5C2(100), and Fe5C2(510) surfaces, aggregation is thermodynamically more favored than dispersion, while dispersion is more favored than aggregation on the Fe5C2(111) surface for n= 2–4, on the Fe5C2(010) surface for n= 2 and on the Fe5C2(001) surface for n= 2–4. The difference in structures and stability at low coverage depends on the stronger Cu–Fe interaction over the Cu–Cu interaction as well as the location of the adsorption sites. The adsorption energies do not correlate with the surface Fe and C ratios. Comparison among the most stable Fe(110), Fe3C(001), and Fe5C2(100) surfaces reveals that the Fe(110) surface has higher Cu affinity than the Fe3C(001) and Fe5C2(100) surfaces; and the carbide surfaces have close Cu affinities; in agreement with the experimental observations. On all these iron and carbide surfaces, two-dimensional monolayer surface adsorption configurations are energetically more favored than the adsorption of three-dimensional Cunclusters, and it can be expected that the adsorbed Cu atoms should grow epitaxially as a layer-by-layer mode at the initial stage. On the metallic Fe(110), Fe(100), Fe(111), and Fe3C(010) surfaces, the adsorbed Cu atoms are negatively charged; while on the Fe3C(100), Fe5C2(100), Fe5C2(111), Fe5C2(010), and Fe5C2(001) surfaces, the adsorbed Cu atoms are positively charged. On the Fe3C(001) and Fe5C2(510) surfaces, the adsorbed Cu atoms mainly interacting with surface Fe atoms are very slightly negatively charged. This trend is in line with their difference in electronegativity. Our results build the foundation for further study of the Cu-promotion effect in Fe-based FTS in particular and for metal-doped heterogeneous catalysis in general.

Water adsorption and dissociation on the Fe(100) surface at different coverages have been calculated using density functional theory methods and ab initio thermodynamics. For the adsorption of (H2O)nclusters on the (3 × 4) Fe(100) surface, the adsorption energy is contributed by direct H2O–Fe interaction and hydrogen bonding. For n= 1–3, direct H2O–Fe interaction is dominant, and hydrogen bonding becomes more important for n= 4–5. For n= 6–8 and 12, structurally different adsorption configurations have very close energies. Monomeric H2O dissociation is more favored on the clean Fe(100) surface than that on H2O or OH precovered surfaces. O-assisted H2O dissociation is favorable kinetically (O + H2O = 2OH), and further OH dissociation is roughly thermo-neutral. With the increase of surface O coverage (nO, n= 2–7), further H2O dissociation has similar potential energy surfaces, and H2formation from surface adsorbed H atoms becomes easy, while the desorption energy is close to zero for n= 7. The calculated thermal desorption temperatures of H2O and H2on clean surface agree well with the available experiment data. The characteristic desorption temperatures of H2O and H2coincided at 310 K are controlled by the kinetics of disproportionation (2OH → O + H2O) and dissociation (2OH → 2O + H2) of surface OH groups. The dispersion corrections (PBE-D2) overestimate slightly the adsorption energies and temperatures of H2O and H2on iron surface. At 0.5 ML coverage (6 × OH), the adsorbed OH groups at the bridge sites do not share surface iron atoms and form two well-ordered parallel lines, and each OH group acting as donor and acceptor forms hydrogen bonding with the adjacent OH groups, in agreement with the experimentally observed surface structures. At 1 ML coverage of OH (12 × OH) and O (12 × O), the adsorbed OH groups at the bridge sites share surface iron atoms and form four well-ordered parallel lines; and the adsorbed O atoms are located at the hollow sites. Energetic analysis reveals that 1 ML OH coverage is accessible both kinetically and thermodynamically, while the formation of 1 ML O coverage is hindered kinetically since the OH dissociation barrier increases strongly with the increase of O pro-covered coverage. All these results provide insights into water-involved reactions catalyzed by iron and broaden our fundamental understanding into water interaction with metal surfaces.

Spin-polarized density functional theory computations have been carried out to investigate the adsorption configurations of Cun(n= 1–7, 13) on the most stable Fe(110) and Fe3C(001) surfaces. On both surfaces the adsorbed Cunclusters favor aggregation over dispersion, and monolayer adsorption configurations are more favored thermodynamically than the two-layer adsorbed structures because of the stronger Fe–Cu interaction over the Cu–Cu bonding. On the basis of the computed adsorption energies the Fe(110) surface has stronger Cu affinity than the Fe3C(001) surface, in agreement with the experimental results. The Fe(110) surface also has stronger Cunaggregation energies and more pronounced charge transfer from surface to adsorbed Cunclusters than the Fe3C(001) surface. Different Cungrowth modes have been discussed accordingly.

Spin-polarized density functional theory computations have been carried out to study the adsorption and dissociation of CO on clean as well as nCu-adsorbed and nCu-substituted Fe(100) surfaces (n= 1–3) at different coverage to explore the Cu promotion effect in CO activation. Increasing Cu content not only lowers CO dissociation energies but also increases CO dissociation barriers as well as making CO dissociation thermodynamically less favorable, and the clean Fe(100) surface is most active in CO adsorption and dissociation. The nCu-substituted Fe(100) surface can suppress CO adsorption and dissociation more strongly than the nCu-adsorbed Fe(100) surface. CO stretching frequencies at different coverages have been computed for assisting experimental investigations.

Hydrogen adsorption on the primarily exposed (001), (100), (101), and (201) surfaces of the hexagonal Mo2C phase at different coverage has been investigated at the level of density functional theory and using ab initio thermodynamics. On the Mo-terminated (001) and (100) as well as mixed Mo/C-terminated (101) and (201) surfaces, dissociative H2adsorption is favored both kinetically and thermodynamically. At high coverage, each surface can have several types of adsorption configurations coexisting, and these types are different from surface to surface. The stable coverage as a function of temperature and partial pressure provides useful information not only for surface science studies at ultrahigh vacuum condition but also for practical applications at high temperature and pressure in monitoring reactions. The differences in the adsorbed H atom numbers and energies of these surfaces indicate their different potential hydrotreating abilities. The relationship between surface stability and stable hydrogen coverage has been discussed.

Hydrogen adsorption structures and energetics on the (100), (110), (111), (210), (211), (310), and (321) iron surfaces up to saturation have been computed using spin-polarized density functional theory and ab initio thermodynamics. The computed hydrogen desorption temperatures and energies on the (100), (110), (111), and (211) surfaces as well as the Fe–H binding energies on the (110) and (111) surfaces agree well with the available experimental data. At typical hydrogen reduction temperature (675 K), the mainly exposed (110) and (310) facets represent the active surfaces, as supported by the transmission electron microscopy study. Our results offer an example of investigating and understanding surface structures and active facets of heterogeneous catalysts under experimental conditions.

CO adsorption and dissociation on the Mo and C terminations of the orthorhombic Mo2C(100) surface at different coverage were systematically investigated on the basis of density functional theory. On the Mo termination, only molecular adsorption is likely for nCO= 9–16. Mixed molecular and dissociative adsorption becomes possible for nCO= 8, while only dissociative adsorption is favorable for nCO= 1–7. This indicates the coverage-dependent CO dissociation and equilibrium between dissociation and desorption. On the C termination, there is no dissociative adsorption, and only molecular adsorption is favorable at all coverages (nCO= 1–16). The computed CO stretching frequencies as well as the predicted desorption states from ab initio thermodynamic analysis agree well with the available experimental findings. The stable coverage as a function of temperature and partial pressure provides useful information not only for surface science studies at ultrahigh vacuum conditions but also for practical applications at high temperature and pressure in exploring reactions.

CO activation on Fe(100) at different coverage was systematically computed on the basis of spin-polarized density functional theory. At the saturated coverage (11CO) on a p(3 × 4) surface size (24 exposed Fe atoms), top (1CO), bridge (3CO) and 4-fold hollow (7CO) adsorption configurations coexist. The stepwise adsorption energies and dissociation barriers at different coverage reveal equilibriums between desorption and dissociation of adsorbed CO molecules. It is found that only molecular adsorption is likely for nCO= 8–11, and mixed molecular and dissociative adsorption becomes possible for nCO= 5–7, while only dissociative adsorption is favorable for nCO= 1–4. The computed CO adsorption configurations and stretching frequencies as well as desorption temperatures from ab initio thermodynamic analysis agree well with the available experimental data.

The water–gas shift (WGS) reaction is widely used in mass hydrogen production for many chemical and energy industrial processes, where high-performance and ecofriendly catalysts play a crucial role in mitigating the urgent energy and environmental crisis. Excellent WGS catalysts require multifunctional active centers within multiple components, which remains unclear and controversial with respect to the understanding of the reaction mechanism, for instance, the favorable pathway and how the active sites work in synergy. One of the key constraints is the lack of knowledge on active centers at the atomic level, especially for the interface structure under the reaction conditions. In this paper, a FeOx–Cu model system was precisely constructed based on previous atomic-resolved scanning tunneling microscopy (STM) images, in which FeOx–Cu is known as the active phase in WGS commercial catalysts. The surface structure of the model under reaction condition was systematically investigated in order to fill the pressure gap during the STM experiment. Based on this model, the comprehensive reaction mechanism of WGS and the synergetic effect between components were revealed by the DFT calculation and two-dimensional energy surface. Compared with the Cu(100) surface, the FeOx–Cu(100) interface significantly promoted the adsorption of oxygen-containing species except CO and reduced the reaction barriers of H2O dissociation and CO2generation, thereby enhancing the activity. Based on this model system and the mechanism knowledge, Cr- and Mn-doped catalysts with superior activity were designed, which also provides evidence for the function of the Cr promoter in the Fe–Cr–Cu commercial catalyst.

Acrolein hydrogenation via allyl alcohol, propanal, and enol into propanol on the Ni(111) surface has been investigated using the spin-polarized periodic density functional theory method. On the basis of the computed adsorption energies and effective hydrogenation barriers, acrolein hydrogenation into propanal and allyl alcohol obeys the Langmuir–Hinshelwood mechanism and propanal formation is more favored kinetically and thermodynamically than allyl alcohol formation. Hydrogenation of propanal and allyl alcohol should follow the Eley–Rideal mechanism. The adsorption energies of acrolein, allyl alcohol, and propanal along with the partial hydrogenation selectivity on Ni, Au, Ag, and Pt catalysts have been compared and discussed.

Spin-polarized density functional theory method (GGA+U) and periodic supercell model have been used to study water adsorption properties on the Fe3O4(110) surface, which has A and B terminations in close surface energy. The adsorption of one and two water molecules is molecular on the A termination, while dissociative on the B termination. For the adsorption of three and four water molecules, mixed dissociative and molecular coadsorption is preferred on the A termination, and fully dissociative adsorption as well as mixed molecular and dissociative coadsorptions are preferred on the B terminations. The stepwise adsorption energies show that the full monolayer water adsorption on both terminations is thermodynamically possible. Further analysis shows that surface iron atoms and hydrogen bonding contribute to the adsorption energies. The adsorption mechanism has been analyzed on the basis of projected density of states (PDOS).

The reaction mechanisms and promoting effect in the water gas shift reaction on (Fe, Co, Ni)-MoS2have been computed at the level of density functional theory. The Mo-edge with 25% M substitution and 25% sulfur coverage is identified as the active site. On the redox reaction path (CO + H2O → CO + OH + H; OH → O + H; CO + O + 2H → CO2+ 2H), surface OH dissociation into O + H is the rate-determining step. On the carboxyl reaction path (CO + OH → COOH; COOH → CO2+ H), surface COOH dissociation into CO2+ H is the rate-determining step. On the basis of the computed effective barriers, the redox mechanism is kinetically more preferable than the carboxyl mechanism for Mo, Fe, and Co, whereas both mechanisms have very close effective barriers for Ni. Compared with pure MoS2, Ni has the largest promotion effect in reducing the effective barrier (0.72 eV), followed by Co (0.45 eV), whereas Fe has the smallest effect (0.04 eV). This promotion effect is tightly associated with the electronegativities and number of the valence electrons of the metals. As the best catalyst, Ni-MoS2has the lowest adsorption energies of the surface intermediates and also the lowest effective barrier.

For understanding the catalytic activity of Fe3O4-supported gold catalysts, the adsorption structures and energies of a single Au atom on the six terminations of the Fe3O4(111) surface have been computed at the level of density functional theory (GGA+U). For the most stable adsorption configurations, correlation has been found between the surface stability and the Au atom adsorption energy; that is, the more stable the surface, the lower the Au atom adsorption energy. It is also found that the adsorbed Au atom is reduced and has a negative charge on the iron-terminated surfaces, whereas it is oxidized and has a positive charge on the oxygen-terminated surfaces, and the latter is in agreement with the experimental observation. No correlation between the transferred charge and the adsorption energy has been found. Regarding the experimentally observed oxidation of gold nanoparticles on the iron oxide surface, it is possible to produce an oxygen-terminated surface for gold adsorption by synthetic tuning.

Adsorption and surface coverage of CO on the (001), (101), and (201) surfaces of β-Mo2C were computed at the level of density functional theory under the consideration of the temperature and CO partial pressure by using the ab initio atomistic thermodynamic method. On the basis of the computed Gibbs free energies, the relationship between CO coverage on the surfaces and temperature as well as CO partial pressure has been established, and excellent agreements have been found between the predicated CO desorption temperatures and the experimentally recorded temperature programmed desorption (TPD) spectra. These computed phase diagrams show that a stable CO coverage can be obtained within a range of temperature and partial pressure; different surfaces can have different coverage at the same conditions, and different partial pressure has a different desorption temperature. In addition, these phase diagrams provide useful information for adjusting the balance between temperature and CO partial pressure for a stable CO coverage and for identifying the active surface and the initial states under given conditions. These results should also be very interesting for surface science under ultra high vacuum conditions.

Spin-polarized density functional theory calculations have been performed to investigate formic acid dehydrogenation into carbon dioxide and hydrogen (HCO2H → CO2+ H2) on Ni(111). It is found that formic acid prefers the O (OC) atop adsorption on nickel surface and the H (H–O) atom bridging two neighboring nickel atoms, and formate prefers the bidentate adsorption with O atop on nickel surface. The computed stretching frequencies for deuterated formic acid (DCO2H) and deuterated formate (DCO2) on Ni(111) agree well with the experimentally observed IR spectra. Formic acid dehydrogenation into surface formate and hydrogen atom (HCO2H → HCO2+ H) has barrier of 0.41 eV and is exothermic by 0.35 eV. Formate dehydrogenation into carbon dioxide and hydrogen atom (HCO2→ CO2+ H) has an effective barrier of about 1.0 eV and is the rate-determining step. Our computed adsorption configurations and energetic data for formic acid dehydrogenation on Ni(111) are very close to the reported results for Pt(111), but in sharp contrast to the previously reported results for Pd(111). Our recalculated adsorption configurations and energetic data for formic acid dehydrogenation on Pd(111) are similar to those on Ni(111) and Pt(111), demonstrating the high similarities of these metals. These computed data show that Pd-catalyzed formic acid dehydrogenation has the lowest effective barrier (0.76 eV), followed by Ni (1.03 eV) and Pt (1.56 eV).

The stability of β-Mo2C surfaces has been computed at the level of density functional theory under the consideration of the temperature, pressure, and molar ratios of CH4/H2and CO/CO2gas mixtures by using ab initio atomistic thermodynamic calculations. It was found that the (001) surface is most stable at low temperature, whereas (101) becomes dominant at high temperature with CH4as carbon source, and the computed surface stability is supported by the experimental X-ray diffraction pattern and intensity. For CO as carbon source, the (101) surface has the smallest surface Gibbs free energy at temperatures up to 1000 K and is most stable. On the basis of the Wulff-type particle shapes from surface Gibbs free energies the (101) facet represents the largest surface area of β-Mo2C. Our findings are in perfect agreement with the results of high-resolution transmission electron microscopy.

On the basis of the adsorption of spillover H, H2O, and CO2, and their related surface dehydrations, the mechanism of γ-Al2O3in CO2reforming of CH4is studied at the level of density functional theory. On the dry γ-Al2O3(100) surface, the most favored surface dehydration is the surface OH from either surface bicarbonate (CO3+ H → HCO3; HCO3→ CO2+ OH) or surface carboxy (CO2+ H → COOH; COOH → CO + OH) dissociation. On the γ-Al2O3(110) surface, dehydration of the hydroxylated surface is most favorable. Compared with the highest CH4dissociation barrier on Ni(111) and Ni(211) in CO2reforming of CH4on Ni/γ-Al2O3catalyst, the first step CH4dissociation (CH4→ CH3+ H) on γ-Al2O3(100) should be the rate-determining step, while dehydration of the hydroxylated γ-Al2O3(110) surface should determine the overall reaction rate.

Spin-polarized density functional theory calculations have been performed to investigate the mechanisms for CxHyformation on Fe3C(100). It is found that H-assisted CO dissociation (CO + H → CHO; CHO → CH + O) has lower barrier than CO direct dissociation (CO → C + O), but surface Csatom hydrogenation to form surface CsH is the most favored pathway. As the first C2surface species, surface ketenylidene CsCO rising from CO adsorption is an important intermediate for C2Hxformation. Initial surface C2Hxforms from CsCO hydrogenation instead of direct dissociation. The formation of CsCH, CsHCH and CsH2CH has close effective barriers and depends on the CO/H2ratio. In addition, surface vacancy can activate CO strongly and lower the CO dissociation barrier considerably, and this regenerates the carburized active surface and closes a catalytic cycle.

The role of a Co promoter in a water-gas-shift reaction on Co-MoS2has been investigated on the basis of density functional theory computation. On the basis of the computed adsorption energy of the reaction intermediates and H2O dissociation barriers, the active catalyst is the Mo edge with 25% Co substitution and 25% sulfur coverage, while the S edge with 25% Co substitution and 50% sulfur coverage is not active. On the basis of the computed reaction barriers, the redox mechanism (CO + H2O → CO + O + 2H; CO + O + 2H → CO2+ H2) is the preferable reaction path, and the rate-determining step is the second step dissociation of OH into surface O and H, while the reaction path from carboxy (CO + OH → COOH; COOH → CO2+ H) is not favored due to its high dissociation barrier. In addition, formate (HCOO) is a side product from gas phase CO2and surface H and does not participate directly in the reaction mechanism. Detailed comparisons reveal that the Co promoter is not an active center in H2O dissociation and CO oxidation but changes the adsorption configuration of the reaction intermediates and reduces the reaction barriers. The Co promoter plays the role of a textual promoter in creating more active sites and accelerating the reaction rate.

The structure and stability of endohedral X@C60Fn(X = N, H; n= 1, 2) and (H@C60)2are computed at the B3LYP level of density functional theory. The most stable N@C60F has one endo N−C bond, while N@C60F2favors nitrogen atom at the cage center. Both H@C60F and H@C60F2favor isomers with one endo C−H bond. The most stable dimer structure, (H@C60)2, has one intercage C−C bond and two endo C−H bonds, and is stable below 650 K, but dissociates above 650 K.

The structures and energies of selected (CX)n(n= even, 6−60, X = F or H) polyhedra, computed at the B3LYP/6-311++G(d,p)//B3LYP/6-31G(d,p) level, show (CH)nisomers to be generally less strained than the corresponding (CF)nanalogues. For n= 6−24, both (CH)nand (CF)nfavor the same geometric patterns. The tube isomers (t-(CX)n) of large (n= 26−60) species with fused five-membered rings at the tube ends are thermodynamically more stable than the cage isomers (c-(CX)n), deduced from the most stable carbon cages. The stabilities of tube isomers are determined by the distortion of the bond angles (angle strain) and the nonbonded F···F repulsive interactions. The latter have been quantified by means of homodesmotic equations. Four structural types of t-(CX)nstructures are built on the basis of the number of fused five-membered rings, and the thermodynamically more stable t-(CX)nstructures have more fused five-membered rings, in opposition to the isolated pentagon rule governing the stability of carbon fullerenes.

Among all 1812 all-exo (SiH)60fullerene isomers, the most stable isomer has a tubelike configuration (D5d) with twelve five-membered rings fused and located at the ends of the tube. The most stable cage isomer can admit ten to twelve endo Si−H bonds (H10@Si60H50and H12@Si60H48), and they also represent the most stable (SiH)60isomers. It is found that (SiH)60isomers have much lower angle and torsion strains than (CH)60isomers.

The adsorption of CO at various coverage on the termination surfaces of the low-indexed (100), (110), and (111) surfaces of Fe4C has been performed at the level of density functional theory. It is found that CO prefers to adsorb around Fe atoms. It is also found that there is no significant lateral interaction between the adsorbed CO at 1/4 monolayer (ML) coverage, whereas there is stronger repulsive interaction at 1 ML for (100) and (110) surfaces, and 1/2 ML for (111) surface. It is also interesting to note that the surface ketenylidene (CCO) species on (100) TFe/C, (110) TFe/C, and (111) TCmay be the precursor for CO dissociation. On all surfaces, the adsorbed CO is partially negatively charged, indicating the enhanced electron transfer from the Fe4C surface to CO, and the more the electron transfer, the stronger the CO activation.

Spin-polarized density functional theory calculations have been performed to investigate the formation of CH xspecies from CO dissociation on the double-stepped Co(0001) surface. It is found that the steps not only favor CO dissociation but also are an advantage over CO and C/O hydrogenation. Furthermore, the step with the hcp and fcc sites in alternate arrangement is more active than the other consisting of the 4-fold site. Under realistic Fischer−Tropsch reaction conditions, surface CH xO are the most favored intermediates; and surface CH xare formed from CH xO dissociation, while the direct CO dissociation with carbide intermediate is not competitive.

The adsorption and reaction of carbon atoms on the Fe 5C 2(001) surface have been computed at four different levels of density functional theory. The formation of surface carbon clusters, carbon chains, and graphitic carbon is favored thermodynamically, and graphitic carbon is the final surface product. With hydrogen available on the surface, however, surface carbon hydrogenation (C + H) is more favored kinetically over its coupling reaction (C + C), and surface CH should be the most favorable species.

The structures and energies of the stepwise carburized Fetet1- and Feoct2-terminated Fe3O4(111) surfaces have been computed at the level of density functional theory. Carburization of both surfaces is favored thermodynamically, and these results agree with the experimental findings. It is to note that the Fetet1-terminated Fe3O4(111) surface has weak coverage effects, while the Feoct2-terminated Fe3O4(111) surface has strong coverage effects. On the basis of the calculated free energies of equilibrium, the carburization ability shows the order of C > CH > CH2> CH3> CO.

Spin-polarized density functional theory calculations were performed to investigate the oxidation of the Co(0001) surface and the effects of surface O on CO adsorption and dissociation as well as CHxcoupling and hydrogenation. Under the realistic Fischer−Tropsch synthesis conditions (493 K, PH2O/PH21−1.5), Co(0001) can be oxidized by H2O forming a 1/4monolayer O-covered surface, while subsurface oxidation or high oxidized surface (1/2monolayer) is thermodynamically not possible. Compared to clean Co(0001), the oxygen pre-covered surface lowers the ability for CO adsorption and activation, elevates the CO dissociation barrier, and favors CO2formation, as well as elevates the CH/CH coupling barrier and favors the CH hydrogenation thermodynamically. Consequently, the catalytic activity is reduced, and the monomeric building block for chain growth changes from CH to CH2. Interestingly, the formation of CO2will refresh the surface. Therefore, we suggest that either controlling H2O content to lower the possibility of oxidation or raising CO content to refresh the surface will maintain the surface activity.

On the basis of density functional theory calculations, the chemisorption of CO2on the transition metal surfaces was investigated to find out the key factors controlling its adsorption strength and activation degree. The interaction mechanism of CO2with the metal surfaces was discussed by analyzing the density of states. The adsorption strength of CO2is controlled by the d-band center of the metal surfaces and also affected by the charge transfer from the metal surfaces to the chemisorbed CO2. The degree of CO bond activation depends on the transferred charge. Therefore, both d-band center of the metal surfaces and the charge transfer should control the chemisorption of CO2.

The structures and stabilities of a set of carbon species on the Ni(111) surface, as well as the energy barriers of the initial stages of carbon growth, were computed by using density functional theory. It was found that the stability is on the order of graphitic monolayer > C in Ni bulk > carbon chain > small cluster > atomic carbon. The formation of adsorbed graphitic monolayer is most preferred on Ni(111) thermodynamically. The adsorbed carbon atoms exist at the coverage of 0.25 and 0.5 monolayer, and they reconstruct Ni(111) at the coverage of 0.75 and 1 monolayer. The reconstruction will be interesting for self-assembly of surface structures and needs experimental confirmation. The adsorbed C2and C3carbon clusters on Ni(111) have linear structures, in which the terminal carbons prefer occupying 3-fold hollow sites. In the stable adsorbed C3cluster, the middle carbon atom occupies a top site. The linear C4cluster and the branched C4cluster have very similar stability. Like the adsorbed graphitic monolayer, the branched C4cluster occupies similar surface sites, and this implies a simple carbon growth mechanism. On the basis of the computed energy barriers, the lateral transfer and clustering of carbon is favored kinetically, while the carbon atoms sinking into the Ni bulk have high energy barriers. During the adsorption processes, electrons transfer from the Ni surfaces to the adsorbed carbon species, while the spin densities of the adsorbed carbon species and the Ni surfaces decrease.

Spin-polarized density functional theory calculations have been carried out to characterize CO and H2co-adsorption on the Fe(111) surface. At 0.5 monolayer (ML) CO coverage, the most-stable adsorption configurations are close to those of individually adsorbed CO and H, and the interaction of adsorbed CO and H does not lead to any hydrogenated product. As the ratio of H2/CO increases, the most-table co-adsorption site of CO changes from the shallow-hollow site (sh) to the top site (t), while H maintains approximately at the top-shallow bridge site (tsb) or the quasi-fourfold site (qff). At 1 ML CO coverage, a similar trend is found where CO moves up to the first-layer Fe surface, while H prefers to diffuse into the subsurface when the H2/CO ratio increases. At very high H2/CO ratios, the most-stable configuration of CO has adsorption at the top site (t). In addition, formyl (CHO) is formed at H2/CO ratios of 1/2and 1/1, deriving from hydrogenation of adsorbed CO in the bridge-like site (brl). The interaction nature of CO−H is also analyzed by the DOS analysis.