Consistent LDA’ + DMFT—an unambiguous way to avoid double counting problem: NiO test

We present a consistent way of treating a double counting problem unavoidably arising within the LDA + DMFT combined approach to realistic calculations of electronic structure of strongly correlated systems. The main obstacle here is the absence of systematic (e.g., diagrammatic) way to express LDA (local density approximation) contribution to exchange correlation energy appearing in the density functional theory. It is not clear then, which part of interaction entering DMFT (dynamical mean-field theory) is already taken into account through LDA calculations. Because of that, up to now there is no accepted unique expression for the double counting correction in LDA + DMFT. To avoid this problem we propose here the consistent LDA’ + DMFT approach, where LDA exchange correlation contribution is explicitly excluded for correlated states (bands) during self-consistent band structure calculations. What is left out of Coulomb interaction for those strongly correlated states (bands) is its non-local part, which is not included in DMFT, and the local Hartreelike contribution. Then the double counting correction is uniquely reduced to the local Hartree contribution. Correlations for strongly correlated states are then directly accounted for via the standard DMFT. We further test the consistent LDA’ + DMFT scheme and compare it with conventional LDA + DMFT calculating the electronic structure of NiO. Opposite to the conventional LDA + DMFT our consistent LDA’ + DMFT approach unambiguously produces the insulating band structure in agreement with experiments.     

Selective surface chemistry of allyl alcohol and allyl aldehyde on Si(1 0 0)2×1: Competition of [2 + 2] CC cycloaddition with O-H dissociation and with [2 + 2] CO cycloaddition in bifunctional molecules

Competition between the CC functional group with the OH group in allyl alcohol and with the CO group in allyl aldehyde in the adsorption and thermal chemistry on Si(1 0 0)2×1 has been studied by X-ray photoelectron spectroscopy (XPS) and temperature-programmed desorption (TPD), as well as density-functional theory (DFT) calculations. The similarities found in the C 1s and O 1s spectra for both molecules indicate that the O-H dissociation product for allyl alcohol and [2 + 2] CO cycloaddition product for allyl aldehyde are preferred over the corresponding [2 + 2] CC cycloaddition products. Temperature-dependent XPS and TPD studies further show that thermal evolution of these molecules gives rise to the formation of ethylene, acetylene, and propene on Si(1 0 0)2×1, with additional CO evolution only from allyl alcohol. The formation of these desorption products also supports that the [2 + 2] CC cycloaddition reaction does not occur. In addition, the formation of SiC at 1090 K is observed for both allyl alcohol and allyl aldehyde. We propose plausible surface-mediated reaction pathways for the formation of these thermal evolution products. The present work illustrates the crucial role of the Si(1 0 0)2×1 surface in selective reactions of the Si dimers with the O−H group in allyl alcohol and with the CO group in allyl aldehyde over the CC functional group common to both molecules.     

Ab initio study of atomic geometry and electronic states of GaSb(0 0 1) reconstructions

We have employed the pseudopotential method and the density functional scheme to study the atomic geometry and electronic states of the GaSb(0 0 1) surface such as (1 × 3), c(2 × 6) and (4 × 3) reconstructions. It is found that both of (1 × 3) and c(2 × 6) reconstructions are characterised by metallic band structures, and thus violate the so-called electron counting rule, one of the main building principles of the stability of compound semiconductor surfaces. We establish that the stability of these reconstructions results from significant elastic deformation in the top atomic layers of the surface, a process which overcomes the penality incurred by the violation of the electron counting rule. The atomic geometry and electronic states for the two reconstructions are compared and contrasted with each other. The α and β phases of the (4 × 3) reconstruction also show large atomic relaxations but are semiconducting and obey the electron counting rule.     

Interaction of 2,2,6,6-tetramethyl-3,5-heptanedione with the Si(1 0 0)-2 × 1 surface: Scanning tunneling microscopy and density functional theory study

Room temperature adsorption and reaction of 2,2,6,6-tetramethyl-3,5-heptanedione (dpmH) on the Si(1 0 0)-2 × 1 surface has been studied with ultra-high vacuum scanning tunneling microscopy (UHV-STM) and temperature programmed desorption (TPD). The molecule is found to chemisorb as a mixture of at least five distinct species. Density functional theory (DFT) was used to calculate the structures and adsorption energies of 12 possible addition products. Unique bonding assignments for each experimental feature are proposed by consideration of a common intermediate reaction network, and a comparison of possible reaction pathways leading to the final products. These assignments are: OH inter-dimer dissociation, OH intra-dimer dissociation, 1,5 intra-dimer addition, 1,5 inter-dimer addition, and intra-dimer [2 + 2]CO addition with OH dissociation on an adjacent dimer. TPD and STM results show that the molecule dissociates completely upon annealing to 700 °C with formation of the c(4 × 4) phase at low exposures, and SiC islands for exposures exceeding 0.15 L.     

A theoretical study of c-C5H8 adsorption on Ge (0 0 1)-2 × 1 and on dimer vacancies on the surface: Electronic structure and bonding

In this work we analyzed the geometry and the chemical interactions for c-C₅H₈ adsorption on Ge (0 0 1), using density functional theory calculations (DFT). We examined the changes in the atomic interactions using a slab model. We considered two cases, the cyclopentene adsorption on Ge(0 0 1) and on dimer vacancies on the surface. We found an average distance H-Ge, -C-Ge and C-Ge of 1.50, 1.70 and 1.65 Å, respectively, on dimer vacancies; and an average C-Ge distance of 2.05 Å on Ge-Ge dimer. We also computed the density of states (DOS) and the DOS weighted overlap populations (OPDOS) corresponding to C-C, C-Ge, C-H, and Ge-Ge bonds. During adsorption the main contribution are the CC double bond in both cases, and the next C and the H's belonging to this bonds in the case of adsorption on dimer vacancies. The orbital contribution includes participation of the 2p_y and 2p_z orbitals corresponding to unsaturated C atoms, 2p_z corresponding to side saturated C, and the 4p orbitals of Ge for the adsorption on dimer vacancies; 2s and 2p_z orbitals corresponding to double bond C atoms, 4s and 4p_z orbitals of Ge for the adsorption on Ge(0 0 1).     

Coadsorption of CN and O on Cu (1 0 0) surface: A density functional study

The adsorption of cyanide (CN) or oxygen atom, as well as the coadsorption of CN + O on Cu (1 0 0) surface is studied by using density functional theory (DFT) and the cluster model method. Cu₁₄ cluster is used to simulate the surface. Perpendicular and parallel bonding geometries of CN adsorbed on Cu (1 0 0) surface are considered, respectively. The present calculations show that the CN may be absorbed on top and bridge sites by carbon atom of cyanide (C-down), and C-down on top site is the most favorable. The adsorbed C-N stretch frequencies compared with that of the gaseous CN species are all red-shifted, except the C-down on top site. The charge transfer from the surface to the CN species leads to an increase in work function for the Cu surface. The oxygen atom adsorbed on the four-fold hollow site of Cu (1 0 0) is the most favorable, and is consistent with the experimental study. The coadsorption of O at a four-fold hollow site tends to block adsorption of CN at the nearby sites. If O coverage increases, the CN may be adsorbed on the top and bridges sites with the C-down model. The reaction CN + O → OCN on the Cu (1 0 0) is predicted to be exothermic, and formed OCN species may be stably absorbed on the Cu (1 0 0).     

Configuration of pentacene (C22H14) films on Si(1 0 0)-2 × 1 studied by NEXAFS

Pentacene films on Si(1 0 0)-(2 × 1) surface at 300 K were investigated using near edge X-ray absorption fine structure (NEXAFS) at the carbon K-edge. NEXAFS spectra show that pentacene molecules are chemisorbed on the Si(1 0 0)-(2 × 1) surface for monolayer with flat-laying and predominantly physisorbed on the Si(1 0 0)-(2 × 1) surface for multilayer films with an upright molecular orientation. Absorption angle of pentacene molecules were measured through π^∗ transition. The angles between the double bond and the silicon surface were 35-55°, 65° and 76° at monolayer, 24 and 48 nm pentacene deposited on the Si(1 0 0) surface, respectively. We observed that the intermediate flat-laying phase is favored for monolayer coverage, while the films of molecules standing perpendicular to the Si(1 0 0) surface are favored for multilayer coverage.     

A first-principles study of C + O reaction on NiCo(1 1 1) surface

A density-functional theory method has been conducted to investigate the association of C + O on (1 1 1) facets of ordered NiCo alloy and the results have been compared with those obtained on pure Ni(1 1 1) surface. In reaction of C + O, the favorable reaction path is that C adsorbed on HCP-1 site moves to the nearest Ni-Co bridge site, and associates with O migrating from FCC-1 site to result in CO adsorbed on the bridge site of Ni-Co. However, the reaction barrier is higher by 0.35 eV than that on pure Ni(1 1 1), which indicates that the incorporation of Co into the Ni crystal is not in favor of the reaction of carbon delimination.     

Structural analysis of the c(4 × 2) reconstruction in Si(0 0 1) and Ge(0 0 1) surfaces by low-energy electron diffraction

The c(4 × 2) structures in (0 0 1) surfaces of Si and Ge have been studied by low-energy electron diffraction (LEED). Using a proper cleaning method for the Si surface, we were able to observe clear c(4 × 2) LEED patterns up to incident energy of ∼400 eV as well as the Ge surface. Extensive experimental intensity-voltage curves allowed us to optimize the asymmetric dimer model up to the eighth layer (including the dimer layer) in depth in the dynamical LEED calculation. Optimized structural parameters are almost the same for the Si and Ge except for the height of the buckled-up atom of the asymmetric dimer. For the Ge surface, the structural parameters are in excellent agreement with those obtained by a previous theoretical calculation. The tilt angle and bond length of the dimer are 18 ± 1 (19 ± 1)° and 2.4 ± 0.1 (2.5 ± 0.1) Å for the Si(0 0 1) (Ge(0 0 1)), respectively.     

Study of ammonia molecule adsorbing on diamond (1 0 0) surface

The diamond (1 0 0) surface with amino terminations is investigated based on density function theory within the generalized gradient approximation. Our calculated negative electron affinity of diamond (1 0 0) surface with hydrogen termination provides a necessary condition for initiating radical reaction. The results display that the ammonia molecule can form stable C-N covalent bonds on the diamond surface. In addition, due to the lower adsorption energy of one amino group binding on diamond surface, single amino group (SAG) model is easy to be realized in experiment with the comparison of double amino group (DAG) model. The adsorbed ammonia molecule will induce acceptor-like gap states with little change of the valence and conduction band of diamond in SAG model. The adsorption mechanism in the formation of ammonia monolayer on H-terminated diamond (1 0 0) surface, and two possible adsorption structures (SAG and DAG) were especially studied.     

Vibrationally-assisted dissociative adsorption of oxygen on Ru(0 0 0 1)-p(2 × 1)-O

Supersonic molecular beam technique combined with high resolution X-ray photoelectron spectroscopy using synchrotron radiation was applied to the study of the dynamics of dissociative adsorption of oxygen on Ru(0 0 0 1) surface in high coverage region. The Ru(0 0 0 1) surface pre-covered with oxygen atoms of 0.5 monolayer, which corresponds to the p(2 × 1)-O structure, was dosed to oxygen molecules with translational energy of 0.5 eV. Oxygen uptake was compared between the cases with and without the beam source heated in order to verify the effects of internal energy of oxygen. We found drastic enhancement in initial sticking probability of oxygen when the beam source was heated to 1400 K. We concluded that the enhancement of sticking probability is mainly caused by molecular vibrational excitation, indicating that dissociation barrier is located in the exit channel on potential energy surface.     

A theoretical study of the water gas shift reaction mechanism on Cu(1 1 1) model system

The water gas shift (WGS) reaction is an important reaction system and has wide applications in several processes. However, the mechanism of the reaction is still in dispute. In this paper we have investigated the reaction mechanism on the model Cu(1 1 1) system using the density functional method and slab models. We have characterized the kinetics and the thermodynamics of the four reaction pathways containing 24 elementary steps and computed the reaction potential energy surfaces. Calculations show that the formate (HCOO) intermediate mechanism (CO + OH → HCOO → CO₂ + H) and the associative mechanism (CO + OH → CO₂ + H) are kinetically unlikely because of the high formation barrier. On the other hand, the carboxyl (HOCO) intermediate mechanism (CO + OH → HOCO → CO₂ + H) and the redox mechanism (CO + O → CO₂) are demonstrated to be feasible. Our calculations also indicate that surface oxygen atoms can reduce the barriers of both water dissociation and HOCO decomposition significantly. The calculated potential energy surfaces show that the water dissociation which produces OH groups is the rate-determining step at the initial stage of the reaction or in the absence of surface oxygen atoms. With the development of the reaction or in the presence of oxygen atoms on the surface, CO + OH → HOCO and CO + O → CO₂ become the rate-limiting step for the carboxyl and redox mechanisms, respectively.     

Competition between associative and dissociative adsorption of 1,2-dihalogenated benzenes on Si(1 0 0)2 × 1: Formation of dihalocyclohexadiene, halophenyl and phenylene adstructures

The room temperature (RT) adsorption of 1,2-difluorobenzene (1,2-DFB), 1,2-dichlorobenzene (1,2-DCB) and 1,2-dibromobenzene (1,2-DBB) on Si(1 0 0)2 × 1 have been investigated by X-ray photoelectron spectroscopy (XPS) and temperature programmed desorption (TPD). Both XPS and TPD data show that the relative degree of dissociative to associative adsorption of the dihalogenated benzene (DXB) appears to increase with decreasing electronegativity of the halogen atom (X). In particular, the C 1s intensity ratios for the C-H and C-Si components to the C-X component are found to be 2, 3 and 9.6 for 1,2-DFB, 1,2-DCB and 1,2-DBB, respectively. These results indicate that 1,2-DFB, like benzene, exclusively adsorbs molecularly as a difluorocyclohexadiene adspecies on Si(1 0 0)2 × 1 while 1,2-DBB adsorbs predominantly with double debromination to form 1,2-phenylene. The majority of 1,2-DCB (75%) is found to adsorb molecularly, with the rest (25%) undergone single or double dechlorination to form chlorophenyl and phenylene, respectively. All three DXB molecules appear to have similar coverage as benzene. The two molecular desorption features for 1,2-DFB and 1,2-DCE are observed with desorption maxima at 460 K and 540 K similar to those found for benzene, which suggests that the dihalocyclohexadiene adstructures involve similar bonding through the benzene ring. In accord with the XPS data, no molecular desorption feature is observed for 1,2-DBB on the 2 × 1 surface. Further decomposition of the resulting phenylene adstructures is evident from the desorption fragment, C₂H₂, found at 610 K and 740 K. Recombinative desorption of HCl and HBr above 880 K are also found for 1,2-DCB and 1,2-DBB, respectively. The observed differences between associative and dissociative adsorption for the three DXB adsorbates could be attributed not only to the large difference in the C-X bond strength but also to the relative contributions from inductively withdrawing and resonantly donating electrons exerted by the halogen (X) atoms to the benzene ring.     

Significance of the HO2 + CO reaction during the combustion of CO + H2 mixtures at high pressures

This paper constitutes an experimental and numerical study, using uncertainty analysis of the most important parameters, to evaluate the mechanism for the combustion of CO + H₂ mixtures at high pressures in the range 15-50 bar and temperatures from 950 to 1100 K. Experiments were performed in a rapid compression machine. Autoignition delays were measured for stoichiometric compositions of CO + H₂ containing between 0 and 80% CO in the total fuel mixture. The experimental results showed an unequivocal monotonic increase as the proportion of CO in the mixture was raised. Comparisons were made also with the measured ignition delays in mixtures of H₂ with increasing dilution by N₂, corresponding to the proportions of CO present. These times also increased monotonically, albeit with a greater sensitivity to the extent of dilution than those measured in the CO + H₂ mixtures. By contrast, numerical simulations for the same mixtures, based on a kinetic model derived by Davis et al. displayed a qualitative discrepancy as there was virtually no sensitivity of the ignition delay to the changing ratio of CO + H₂, certainly up to 80% replacement. No exceptions to this trend were found, despite tests being made using seven other kinetic models for CO + H₂ combustion. Global uncertainty analyses were then applied to the Davis et al. model in order to trace the origins of this discrepancy. The analyses took into account the uncertainties in all rate parameters in the model, which is a pre-requisite for evaluation against ignition delay data. It is shown that the reaction rate constant recommended by Baulch et al. for the HO₂ + CO reaction, at T ∼ 1000 K, could be up to a factor of 10 too high and that lowering this rate corrected the qualitative anomaly between experiment and numerical prediction.     

Li ion neutralization on Ag layers grown on Cu(1 1 1)

Electron transfer processes in the neutralization of Li⁺ ions on Ag layers grown on Cu(1 1 1) are investigated in quest of quantum confinement effects. Neutralization probabilities in the scattering of Li⁺ for incident ion energies in the 300 eV to 2 keV range are reported for Ag coverages ranging from 0.15 ML to 5 ML. Results are compared to those for Ag(1 1 1) and Cu(1 1 1) surfaces of bulk crystals. Although existing studies of the characteristics of Ag layers on Cu(1 1 1) indicate significant differences in electronic structure as a function of film thickness, the electron transfer probabilities we measure are found to be very close to those for bulk Ag(1 1 1). These results are commented on the basis of existing models and earlier studies of Li ion neutralization on various metals.     

Layer-resolved photoelectron diffraction from Si(0 0 1) and GaAs(0 0 1)

Photoelectron diffraction in the layer-resolved mode brings more detailed information about local atomic arrangement than is obtained in the standard mode. This is demonstrated in crystals with diamond and zinc-blende structures, both for unpolarized photon excitation as well as for circularly polarized excitation. The full angular distributions of photoemission intensities are evaluated for large atomic clusters representing ideally truncated surfaces of Si(0 0 1) and GaAs(0 0 1). Highly structured layer-resolved patterns enable a more detailed understanding of the standard mode outcomes. Photoelectron intensities from atomic layers placed at different depths under the crystal surface provide direct evidence about electron attenuation and its anisotropy in crystals.     

Electronic structure study of reconstructed Au-SiC(0 0 0 1) surfaces

A study of surface and interface properties of reconstructed Au-SiC(0 0 0 1) surfaces is reported. Two reconstructions were prepared on SiC(0 0 0 1), a √3 × √3R30° and a Si-rich 3 × 3, before Au deposition and subsequent annealing at different temperatures. For the Si-rich 3 × 3 surface the existence of three stable reconstructions 2√3 × 2√3R30°, 3 × 3 and 5 × 5 are revealed after deposition of Au layers, 4-8 Å thick, and annealing at progressively higher temperatures between 500 and 950 °C. For the 2√3 surface two surface shifted Si 2p components are revealed and the Au 4f spectra clearly indicate silicide formation. The variation in relative intensity for the different core level components with photon energy suggests formation of an ordered silicide layer with some excess Si on top. Similar core level spectra and variations in relative intensity with photon energy are obtained for the 3 × 3 and 5 × 5 phases but the amount of excess Si on top is observed to be smaller and an additional weak Si 2p component becomes discernable.For the √3 surface the evolution of the core level spectra after Au deposition and annealing is shown to be distinctly different than for the Si-rich 3 × 3 surface and only one stable reconstruction, a 3 × 3 phase, is observed at similar annealing temperatures.     

Kinetics of the H + NCO reaction

Ab initio transition state theory (TST) based master equation simulations are used to predict the temperature and pressure dependence of the H + NCO reaction rate and product branching. The barrierless entrance channels to form singlet HNCO and NCOH are studied with variable reaction coordinate TST employing a potential energy surface obtained from multi-reference configuration interaction ab initio calculations. The remaining channels, including reactions on the triplet surface, are studied with standard TST methods employing high level electronic structure results. The energy transfer parameters for the master equation simulations arise from a fit to the experimentally observed HNCO dissociation rate. The lowest energy threshold to formation of bimolecular products, ³NH + CO, lies well below the reactants. The bottleneck for intersystem crossing, which precedes the formation of ³NH + CO from the singlet adducts, becomes the dominant bottleneck for that channel at quite low energies relative to reactants. The effect of this bottleneck is studied with model calculations designed to reproduce detailed experimental observations of photolysis branching ratios. This bottleneck greatly reduces the flux from H + NCO to ³NH + CO via the singlet adducts. As a result, stabilization and reaction on solely the triplet surface are significant components of the overall rate. The present predictions for the high pressure and collisionless limit rate coefficients are accurately reproduced over the 200-2500 K range by the expressions, 1.53 × 10⁻⁵T^−1.86exp(−399/T) + 1.07 × 10³T^−3.15exp(−15219/T) and 5.62 × 10⁻¹²T^0.493exp(148/T) cm³ molecule⁻¹ s⁻¹, respectively, where T is in K. These predictions are in reasonably satisfactory agreement with the somewhat discordant experimental rate measurements.     

Growth and morphology of the epitaxial Fe(1 1 0)/MgO(1 1 1)/Fe(1 1 0) Trilayers

Growth and surface morphology of epitaxial Fe(1 1 0)/MgO(1 1 1)/Fe(1 1 0) trilayers constituting a magnetic tunnel junction were investigated by low-energy electron diffraction (LEED) and scanning tunneling microscopy (STM). STM reveals a grain-like growth mode of MgO on Fe(1 1 0) resulting in dense MgO(1 1 1) films at room temperature as well as at 250 °C. As observed by STM, initial deposition of MgO leads to a partial oxidation of the Fe(1 1 0) surface which is confirmed by Auger electron spectroscopy. The top Fe layer deposited on MgO(1 1 1) at room temperature is relatively rough consisting of clusters which can be transformed by annealing to an atomically flat epitaxial Fe(1 1 0) film.     

Interdiffusion of adsorbed Pb and Bi on Cu(1 0 0)

The impingement and interdiffusion of adsorbed Pb and Bi layers spreading from separated 3D pure bulk sources on Cu(1 0 0) has been studied, at T = 513 K, by in situ scanning Auger microscopy. When the leading edges of the pure Pb and Bi diffusion profiles impinge, they both consist of low-coverage lattice gas surface alloyed phases. In these low-coverage phases, Pb displaces surface alloyed Bi and the point of intersection of the profiles drifts towards the Bi source. These features lead to the conclusion that Pb atoms are more strongly bound at surface alloyed sites in Cu(1 0 0) than Bi atoms. Once the total coverage (Pb + Bi) on the substrate reaches about one monolayer, Pb and Bi are dealloyed from the substrate, and the interdiffusion profiles become essentially symmetric. Pb and Bi mix in all proportions, with an interdiffusion coefficient of ∼10⁻¹³ m²/s. This is considerably smaller than the self-diffusion coefficients previously observed for pure Pb and Bi in their respective high-coverage phases, indicating that the mechanism of interdiffusion is different from that of self-diffusion. As interdiffusion proceeds, the point of intersection of the Pb and Bi profiles reverses its drift direction, leading to the conclusion that binding of Bi atoms to the Cu(1 0 0) substrate is stronger than that of Pb atoms in the highest-coverage surface dealloyed layers.