Search

The electrochemical reduction of CO2 on single atom catalysts (SAC) has emerged as a highly promising yet intricate process, requiring an in-depth understanding of each elementary step of the reaction. Most of the theoretical studies pertaining the screening of new efficient catalysts neglect important effects such as the capability of the catalytic site to bind and activate CO2, the occurrence of competing reaction paths via formation of different isomers in the intermediate steps, and the role of water. In this work we will show that these are key aspects of the CO2 reduction reaction (CO2RR). Employing density functional theory, we investigated a series of Transition Metal atoms (Sc-Cu, Mo-Ag, W-Au) embedded in nitrogen doped graphene (TM@4N-Gr) and their activity in the CO2 reduction. Our results show that water is not only solvating the reaction intermediates, but it acts as a ligand itself, competing with CO2 when binding to the catalytic site. This coordination chemistry effect largely affects the stability of the chemical intermediates. Furthermore, only a small fraction of the investigated SACs can bind and activate CO2, and that in most cases the reaction proceeds via formation of the OCHO intermediate and not of the COOH one, as often assumed.

The Oxygen Evolution Reaction (OER) is a key part of water splitting. On metal and oxide surfaces it usually occurs via formation of three intermediates, M(OH), M(O), and M(OOH) (also referred to as OH*, O*, and OOH* species where * indicates a surface site). The last step consists of O2 release. So far, it has been generally assumed that the same path occurs on single atom catalysts (SACs). However, the chemistry of SACs may differ substantially from that of extended surfaces and is reminiscent of that of coordination compounds. This raises the question of whether on SACs the OER follows the classical mechanism or not. Using a DFT approach, we studied a set of 30 SACs made by ten metal atoms (Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Pd and Pt) anchored on three widely used 2D carbon-based materials, graphene, nitrogen-doped graphene and carbon nitride. In none of the cases examined the most favourable reaction path is the conventional one. In fact, in all cases other intermediates with higher stabilities form: M(OH)2, M(O)(OH), M(O)2, and M(O2) (OH* OH*, O* OH*, O* O*, O2* according to standard nomenclature). Therefore, the common assumption that on SACs the OER proceeds via formation of OH*, O*, and OOH* intermediates is not verified. Predictions of new catalysts based on the screening of large number of potential structures can lead to completely incorrect conclusions if these additional intermediates are not taken into consideration.

Direct Z-scheme designs perform well in light-to-fuel conversion. Here, an active and stable ternary Z-scheme core-shell heterojunction for photocatalytic hydrogen evolution (PHE) is fabricated consisting of hexagonal 2D α-Fe2O3 (as photocatalyst II) and 2D nitrogen-doped graphene (NGr as photocatalyst I) functionalized with the Pt single-atoms (SAs) cocatalyst. Under visible light, the 2D/2D α-Fe2O3 @NGr3[sbnd]PtSAs (NGr shell thickness of 3 nm and Pt loading of 0.5 wt%) achieves a remarkable PHE of 6.4 µmol mgcat−1 h–1, which is 16.4- and 3.28-times higher than those of free NGr (0.39 µmol mgcat−1 h–1) and binary α-Fe2O3 @NGr3 (1.95 µmol mgcat−1 h–1), thus outperforming the currently-advanced PHE catalysts. The outstanding performance is due to the superiority of a direct 2D/2D Z-scheme core-shell fabrication, including a large surface area for light harvesting, facile charge separation and transfer, and the workability of Pt SAs sites. Theoretical investigations provide additional insight into the mechanistic process of the ternary system for PHE reactions.

In this tutorial review we report the state-of-the-art of the modeling approaches of Single Atom Catalysts (SAC) for water splitting and fuel cells reactions. The discussion applies for Hydrogen Evolution Reaction (HER), Oxygen Reduction Reaction (OER), Hydrogen Oxidation Reaction (HOR), and Oxygen Reduction Reaction (ORR). The main scope of this work is to underline the relevant aspects of SACs modelling. On the one hand, the review could help computational chemists aiming to start the study of SACs. On the other hand, experimentalists could find the critical analysis of DFT results of interest to understand better the strengths and weaknesses of simulations, and how to interpret computational results. After an introductory section of SACs, we start by briefly presenting the state-of-the-art methodologies. Then, we analyze the critical aspects for a reliable prediction of the electronic properties, and we discuss the robustness of the structural models and ways to validate them. Furthermore, we discuss the main approaches to predict catalytic activity and selectivity, which is the final goal of the computational catalysis. We conclude this review with a critical analysis of the current challenges in the field, and the main limitations of the modeling approaches that are described.

In this brief account dedicated to the memory of Professor Renato Ugo I will trace the history of 40 years of research in cluster chemistry. I will describe how my research activity started in the early 1980s with studies inspired by the vision of Renato Ugo on the electronic structure of gas-phase Li clusters and their interaction with hydrogen molecules. Since then, research on cluster chemistry and physics has done enormous progress, thanks to the development of new experimental techniques, of advanced theoretical methods, and a continuous growth of computing power. The structure and properties of both naked clusters and ligand-stabilized organometallic clusters have been elucidated with the final aim to create strong links between cluster science, surface chemistry, and heterogenous catalysis. This resulted in a better understand of catalysis by supported nanoparticles and in the possibility to design new nanostructured catalysts with desired properties.

ZnO ultrathin films supported on Cu, Ag or Au are attracting increasing interest because they can form under special conditions on commercial Cu/ZnO/Al2O3 catalysts for methanol synthesis, or can be used as alternative materials for catalysis, or as prototypes of the so-called inverse catalysts. In the present study, we have investigated the thickness dependence of the structure and electronic properties of free standing and Cu(111) supported ZnO ultrathin films by means of DFT + U calculations. On the free-standing films, a flat graphitic-like morphology is observed up to 10 layers, where a transition to wurtzite-like structures occurs. On Cu(111), a slightly corrugated graphitic structure is obtained for ZnO films up to 4 layers. From 5 layers on, the rumpling becomes substantial and a phase transition to wurtzite takes place. The driving force behind the destabilization of flat graphitic films on Cu(111) is the charge transfer from the underlying Cu support to the film. This finding rationalizes the experimental observations that the transformation from graphitic to wurtzite polymorph of ZnO films on Cu takes place above 5-layer thickness.

The structural and electronic properties of bulk and ultrathin films of V 2 O 5 and MoO 3 layered oxides have been studied with first-principles density functional theory calculations including Van der Waals dispersion corrections. The U parameter in the DFT + U approach has been determined in order to properly reproduce geometry, band-gap, macroscopic dielectric constant, and formation enthalpies of the two materials. The mono-, and multi-layers are cleaved along the 〈0 0 1〉 and 〈0 1 0〉 stable crystallographic orientations for V 2 O 5 and MoO 3 , respectively. Three layers are needed in order to recover bulk-like properties of V 2 O 5 and MoO 3 . Spin-orbit effects have been incorporated in our simulations, and they result in marginal effects (∼20–30 meV) on the electronic band-gap of V 2 O 5 and a more pronounced (∼200 meV) effect on the band gap of bulk MoO 3 . We also discuss the importance of including local field effects for the reproduction of the anisotropy of the dielectric constant, which reflects the crystal structure of these materials

The properties of spherical 1-nm Pd43 nanoparticles confined between single layers of lepidocrocite-TiO2 are studied by means of Density Functional Theory calculations. Sandwich-like intercalation is then compared to the adsorption on a single oxide layer in terms of metal-oxide bond strength, structural distortion of the metal particle and charge transferred from the metal to the oxide. Then, a Pd(111) slab is put in contact to lepidocrocite-TiO2 on either one side or both sides, simulating the creation of extended interfaces. This permits to compare the behavior of small reactive metal particles to their more inert bulk counterpart. It results that intercalated Pd43 species are bound more strongly to the oxide support compared to those adsorbed on a single oxide layer, resulting in more pronounced structural changes and larger charge transfer, a fact with presumable direct consequences on the catalytic activity. The conclusions drawn for small Pd particles are confirmed also in the case of extended interfaces, even though up to a smaller extent.

In the present work, we have studied the effect of iso-valent dopants on the chemical reactivity of the stoichiometric SnO2 (110) surface. Si, Ge, Pb, Ti, Zr, Hf, and Ce ions have been placed at the Sn lattice sites (substitutional doping) in bulk, sub-surface, and surface positions. The results show that surface or sub-surface sites are clearly preferred over the bulk ones. Next, we studied the effect of the iso-valent dopants on the adsorption and decomposition reaction of HCOOH molecule. On this basis, four paths with desorption of: (a) H2, (b) CO, (c) H2O, or (d) CO2 can occur. Comparing the results of this decomposition reaction on the SnO2 surface with the same process on rutile TiO2 (110) having the same structure and morphology, reveals that SnO2 is considerably more reactive in all reaction steps of the formic acid decomposition.

The adsorption of methyl bromide on MgO ultrathin films supported on Mo(100) was studied by means of density functional theory calculations, in comparison to the MgO(100) and Mo(100) surfaces. The adsorption energy and geometry were shown to depend on the thickness of the supported oxide film. MgO films as thick as 2ML (or more) display adsorptive properties similar to MgO(100), i.e. the adsorption of CH3Br is mostly due to dispersion and the molecule lies in a tilted geometry almost parallel to the surface. The CH3Br HOMO-LUMO gap is almost unaltered with respect to the gas phase. On metallic Mo(100) surfaces the bonding is completely different with the CH3Br molecule strongly bound and the C–Br bond axis almost vertical with respect to the metal surface. The MgO monolayer supported on Mo exhibits somehow intermediate properties: the tilt angle is larger and the bonding is stronger than on MgO(100), due to the effect of the supporting metal. In this case, a small reduction of the HOMO-LUMO gap of the adsorbed molecule is reported. The results help to rationalize the observed behavior in photodissociation of CH3Br supported on different substrates.

Oxide-supported Cu nanoparticles and clusters catalyze a variety of important reactions, such as CO/CO 2 hydrogenation to methanol. Recent studies demonstrate that also sub-nanometer clusters consisting of only a few atoms can actively catalyze chemical reactions. In this study, we investigate the interaction between Cu 4 clusters and silica-surfaces, considering the de-hydroxylated and the fully hydroxylated α-quartz surfaces. We also considered various dopants such as Ti- and Nb-ions substitutional to Si, respectively, in order to see if an electronic change of the support has an effect on the reaction of the supported cluster. We find that hydroxyl groups can enhance the adsorption energy of the cluster, whereas the dopants have only little effects on the adsorption mode of the Cu cluster. On the fully hydroxylated surface, the cluster may react with the hydroxyl groups via reverse hydrogen spillover. Finally, we explore the reactivity of the silica-supported Cu 4 cluster in terms of acetylene trimerization, for which extended Cu surfaces have shown catalytic activity. We find that this reaction should occur with activation barriers below 0.8 eV; Nb-doping of the support does not seem to produce any direct effect on the reactivity of the Cu tetramer.

The adsorption and diffusion of ruthenium adatoms on the (101) surface of tetragonal zirconia was studied by means of periodic Density Functional Theory (PBE+U) calculations. The surface termination has a decisive role in determining the diffusion capability of the adsorbed Ru atoms. On the defect-free and fully dehydroxylated surface, Ru adatoms have several stable adsorption sites with adsorption energies as large as 2.5–2.9 eV However, the kinetic diffusion barriers between adjacent adsorption sites are around 0.5–0.6 eV, indicating a rather fast diffusion process. Surface oxygen vacancies, if present, strongly bind ruthenium adatoms and act as nucleation sites. On hydroxylated surfaces, the adsorption energy of Ru atoms is comparable to the dehydroxylated case, but the kinetic barriers for diffusion are remarkably higher, thus indicating that adsorbed species are less mobile in presence of surface OH groups. The effect is more pronounced for high concentrations of OH groups, since this results in hydrogen bonded hydroxyl units that further limit the diffusion process. These results indicate a possible way to increase the life-time of Ru ZrO2 heterogeneous catalysts by tuning the level of surface hydroxylation, in order to slow down sintering of metal particles via Ostwald ripening process.

The ketonization reaction of acetic acid on the (1 0 1) surface of tetragonal zirconia, a process relevant in the catalytic upgrade of cellulosic biomass has been studied by means of DFT+U calculations. The aim was to better understand the role of catalyst pre-reduction. Acetic acid adsorbs strongly on the zirconia surface, and deprotonation to acetate takes place easily. Then, a proton transfer from the methyl group of the acetate ion, CH 3 COO − , to the surface occurs to form an eno-species, CH 2 COO 2− (intermediate 1). In a parallel step, acetate ions convert into acyl fragments, CH 3 CO, by oxygen extraction (intermediate 2). Once formed, the acyl intermediate 2 is able to attack the eno-species intermediate 1, to form a β-keto-acid. On stoichiometric zirconia, the formation of the intermediate 1, CH 2 COO 2− , is unfavorable; oxygen vacancies on the reduced surface stabilize the reaction product and strongly reduce the activation energy. Reduced Zr 3+ centers are essential to stabilize the acyl intermediate 2. The present work shows at an atomistic level the beneficial role of O vacancies and reduced Zr 3+ centers for the ketonization process.

Materials for post Li-ion batteries are attracting an increasing interest. Here we have studied the fundamental properties of multi-valent Al and Ca ions intercalated into V2O5 and MoO3 layered oxides. To this end we performed Density Functional Theory calculations using a Hubbard U correction for the electronic structure of the oxides and including dispersion forces. We simulated very low loadings of the intercalated Al and Ca atoms, and we studied both stoichiometric and reduced oxides. Intercalation of the Al and Ca atoms within the oxide layers (intra-layer) and in the spaces between the layers (inter-layer) has been compared. In V2O5 inter-layer intercalation is preferred, and gives rise to theoretical voltages of 1.54 V for Al and of 3.08 V for Ca; for comparison, the theoretical voltage for a Li-V2O5 battery is 3.51 V. Al can also be incorporated within the V2O5 layers, but with a smaller binding energy, while Ca, bigger, is only stable in the inter-layer sites. The situation is partly different on MoO3. Here in fact Ca is stable only in inter-layer sites, and gives rise to theoretical voltages of 1.65 V. Al, on the contrary, prefers to be incorporated into the MoO3 layers, giving rise to a similar theoretical voltage of 1.60 V. For comparison, a Li- MoO3 battery has a computed voltage of 2.16 V. The reduced forms of V2O5 and MoO3 have also been considered. Here the O atoms that are more easily removed are those of the vanadyl, V=O, and molibdyl, Mo=O, groups. The only case where there an increase of the voltage is for Al and Ca intercalation into reduced V2O5, provided that the Al and Ca ions are near an oxygen vacancy, but not in the vacancy. In MoO3 the formation of oxygen vacancies is detrimental for the intercalation process.

We have performed a comparative study of the acid–base characteristics of the surfaces of anatase TiO2 and tetragonal ZrO2. To this end we performed DFT + U calculations on CO and CO2 probe molecules adsorbed both on terraces and steps of the two oxides. For titania, CO adsorption results in a moderate adsorption energy (about − 0.3 eV) and in a positive shift of the [Formula presented] stretching frequency (about + 40 cm− 1), typical of Lewis acid sites, with no clear difference in the acidity between terraces or steps. For zirconia we found a similar CO binding energy as for titania, and a CO vibrational shift that depends on the location of the Zr cation: negligible on terraces, similar to TiO2 on steps. We conclude that the acidic properties are similar in the two oxide surfaces. Things are different for CO2 adsorption. On titania the interaction is weak and surface carbonates compete with physisorbed CO2, indicating a weak basic character. On the contrary, on zirconia three types of stable carbonates have been identified. Their vibrational frequencies are consistent with IR measurements reported in the literature. The most stable species forms on steps of the t-ZrO2 surface and consists of a CO32 − unit which lies flat on the surface with the O atoms pointing towards three Zr ions. The species forms spontaneously by extraction of a lattice oxygen by an incoming CO2 molecule. The different reactivity points towards a much more pronounced basic character of zirconia compared to titania, at least if measured by CO2 adsorption.

The interaction of CO. CO2 and CH4, with the internal walls of p-phenylenesilica has been studied using periodic supercell DFT calculations and adding specific corrections to include dispersion forces Two models of p-phenylenesilica have been considered, with different number of silanols on the surface and characterized by a different sequence of silica rings, The three molecules interact preferentially with the silanol groups with adsorption energies typical of a physisorbed state (CO 14.3 kJ/mol, CO2 13.3 kJ/mol, CH4 9.2 kJ/mol). The dispersion part of the interaction is substantial, being about one half of the total value. We also considered possible ways to modify the adsorption properties by doping or modifying the inorganic or organic parts of the framework. In particular, we have introduced Ti-OH, Si-O- and Al--OH groups or altered the organic part of the framework, but in no case an increase of the adsorption energy has been observed. (C) 2009 Elsevier Inc. All rights reserved.

The structure of a rectangular TiO2 nanophase grown epitaxially on a Pt(111) substrate has been investigated by a combined experimental-theoretical approach. It is found that such nanophase is stoichiometric, incommensurate to the substrate and has the structure of a lepidocrocite layer. The film is weakly bound to the metal surface via the O atoms of the oxide layer and consequently it does not have a fully wetting behavior. Two almost iso-energetic structures have been found based on first principles DFT calculations, one characterized by a short and one by a long interface distance, this latter being energetically slightly preferred. However, when the strain due to lattice mismatch is accommodated on the Pt(111) substrate instead of the TiO2 film, only the long interface structure is found. The analysis of measured and computed valence band spectra and STM images supports the long interface, weakly interacting model. (C) 2007 Elsevier B.V. All rights reserved

Using all electron fully relativistic DFT calculations we, have computed the EPR properties of Au atoms bound to various sites of the MgO surface. Changes in g-tensor and hyperfine coupling constants provide a way to identify the gold adsorption site and to map the surface morphology by comparison of measured and calculated EPR spectra. We found a strong reduction of the isotropic hyperfine coupling constant, a(iso)(Au), for adsorbed gold compared to the free atom; this reduction, which is about 45\% for terrace sites, is more pronounced when Au interacts with low-coordinated sites like steps, edges and corners where it is about 60\%. The reduction of a(iso)(Au) is accompanied by a corresponding increase of the superhyperfine interaction with the surface oxygen sites, as measured by a(iso)(O-17). Large anisotropies in the g-tensor are computed for all sites.

The formation of an oxygen vacancy and the simultaneous re-adsorption of an oxygen atom on regular and low-coordinated (LC) sites (edges and corners) of the surface of alkaline-earth oxides with cubic rock salt structure, MgO, CaO, SrO, and BaO, has been investigated using DFT cluster model calculations. The process corresponds to the formation of a surface Frenkel defect when the vacancy formation energy is partially compensated by the energy gained in the formation of a peroxo group. The structural and electronic properties of vacancies and peroxo groups along the series of alkaline-earth oxides have been analyzed. We found that the role of low-coordinated sites on the surface chemistry of alkaline-earth oxides is of crucial importance for MgO, but decreases for the heavier members. For instance, on BaO the formation of a peroxo group is practically site-insensitive. This is not the case of the vacancy formation, which is always favored on the low-coordinated sites. Fil: Di Valentin, Cristiana. Università degli Studi di Milano; Italia Fil: Ferullo, Ricardo. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Bahía Blanca. Instituto de Química del Sur. Universidad Nacional del Sur. Departamento de Química. Instituto de Química del Sur; Argentina. Università degli Studi di Milano; Italia Fil: Binda, Riccardo. Università degli Studi di Milano; Italia Fil: Pacchioni, Gianfranco. Università degli Studi di Milano; Italia

The catalytic activity of pure and Ni-doped MgO surfaces in N2O decomposition has been investigated theoretically with DFT B3LYP cluster model calculations. The barrier to abstraction of O from N2O to form a surface peroxo group, O-2(2-), is found to be rate limiting on both pure and Ni-doped MgO. In the presence of Ni impurities, however, the barrier is reduced from 1.37 eV to 1.19 eV, thus accounting for the experimentally observed increase in catalytic activity as function of the Ni content. Dissociation reaction was found to take place also by direct interaction with surface Ni ion with a competing activation barrier of 1.22 eV. However, this latter process is less exothermic (1.45 eV vs 2.19 eV). The 0 abstraction by a Surface oxygen is followed by 0 diffusion and recombination with final desorption of an O-2 molecule from the surface. This process occurs preferentially on pure MgO while requires higher barriers in the presence of Ni. This can explain the observed decrease of catalyst activity when the Ni concentration becomes higher than 50%. (c) 2005 Elsevier B.V. All rights reserved

The structure and the vibrational properties of ultra-thin SiO2 films (one to three layers, 1-3 L) epitaxially grown on a Mo(112) substrate have been determined by first principle DFT calculations. For 2 L and 3 L films the structures considered are derived from those of beta-cristobalite, beta-tridymite, alpha- or beta-quartz; for 1 L films we considered regular arrays of isolated SiO4 tetrahedra, one-dimensional (SiO3)(n) chains, or an hexagonal two-dimensional (SiO2.5) structure. The 1 L films have the c(2 x 2) pattern observed experimentally. On the basis of the comparison of the computed frequencies with IR and HREEL spectra, of the existing data on film thickness, periodicity, and surface reactivity, we conclude that the film consists of a single silica layer with hexagonal c(2 x 2) structure. (c) 2005 Elsevier B.V. All rights reserved.

We have performed periodic supercell and embedded cluster DFT calculations on the formation of Pd 3 and Pd 4 clusters on regular and defect sites of the MgO(1 0 0) surface. The results show that when a Pd atom bound on a regular oxide anion of the surface adds to a supported Pd dimer, there is an energy gain which goes from 0.6 eV for Pd 2 formed at a divacancy site up to 1.3 eV for Pd 2 formed at a step site. For Pd trimers the ad-atom binding energy tends to further increase, sign of a developing metallic character. The results show on a first principle basis that there are two key steps in nucleation and grow of Pd clusters on the MgO surface: the trapping of a metal atom to a strong binding site (e.g. an oxygen vacancy, F or F + center) and the subsequent formation of the dimer. Once the dimer is formed, further growth to give the trimer and the tetramer is energetically favorable irrespective of the defect site where nucleation has started.

The properties of Ni-4(CO)(4) and Ni-8(CO)(4) model clusters supported on the MgO(001) surface have been studied by means of density functional all-electron calculations. Two- and three-dimensional Ni clusters have been adsorbed on regular terrace sites and on neutral and charged oxygen vacancies (F-s and F-s(+) centers) of the MgO surface; four CO molecules have been added in various sites of the supported clusters and their properties have been investigated. On the MgO(001) surface, the Ni clusters show an intrinsic instability when CO is added to the cluster in linear fashion atop Ni atoms; some clusters (Ni-4) fragmentate, while others (Ni-8) undergo a strong distortion which transforms the cluster from three-dimensional to a nearly planar configuration. This is the result of the weakening of the metal-metal bond induced by CO and of the preference for metal-oxide bonds under CO exposure. The effect is not observed when the Ni clusters are anchored at a point defect like an oxygen vacancy. In this case the electrons trapped in the vacancy are partially transferred to the metal cluster and reinforce the metal-metal bonds. Addition of CO to Ni-4 and Ni-8 on F centers does not alter the cluster topology, but results in measurable shifts of the CO frequency with respect to the unsupported complex. A comparison with the unsupported {[Ni-8(CO)(4)](-) cluster anion suggests that on F-s and F-s(+) centers the transfer of about one electron occurs from the surface to the metal particle. (C) 2004 Elsevier B.V. All rights reserved.}

We present a model of a new paramagnetic defect center which results from the interaction of atomic hydrogen with the MgO(1 0 0) surface. DFT calculations have been performed using periodic supercells and embedded cluster models where long-range polarization effects are included explicitly. The H atom promotes the creation of an oxygen vacancy (F center) by formation of the F S + (OH − ) defect where an hydroxyl group is adsorbed near an electron trapped in an oxygen vacancy. This new center has some characteristics similar to those of the classical F S + centers but a smaller formation energy; furthermore, being globally neutral, it can be treated also with supercell methods.

We report the results of first principles Hartree-Fock (HF) and density functional theory (DFT) calculations on the H-1, Si-29 and O-17 NMR chemical shifts of hydroxyl groups in silica. The structure of the isolated or =Si-OH of the geminal -Si(OH)(2) groups has been fully optimized from cluster models derived from crystalline alpha-quartz and the nuclear magnetic shielding properties have been determined according to the GIAO method. Quantitative agreement with the available experimental data has been obtained at the DFT level. (C) 1999 Elsevier Science S.A. All rights reserved

We report ab initio configuration interaction calculations of the optical properties of a neutral oxygen vacancy in Ge-doped silica. This defect center, V(SiGe), is assumed to involve a missing O atom between a Si and a Ge atom with formation of a direct Si-Ge bond, =Si-Ge=. Several calculations have been performed using cluster models to obtain a reliable estimate of the excitation energy. The lowest fully allowed singlet-singlet transition in V(SiGe) occurs in a spectral region around 7.3 eV according to the calculations, i.e. not too far from the absorption band at 7.6 eV attributed to a V(SiSi) center in pure silica. The emission properties of V(SiGe) and the reasons for the similar transition energy in V(SiSi) and V(SiGe) centers are discussed. (C) 1999 Elsevier Science B.V. All rights reserved

First-principle calculations based on cluster models have been performed to compute the observable electronic properties of a series of dia- and paramagnetic point defects in SiO2. Vibrational modes, hyperfine coupling constants, and core level binding energies have been determined at the Hartree-Fock and gradient corrected density functional theory levels and compared with the experimental infra-red, electron paramagnetic resonance and X-ray photoemission spectral features, when available. The results show the potential of the combined use of quantum mechanical calculations and spectroscopic measurements for the structural description of point defects in silica. (C) 1999 Elsevier Science B.V. All rights reserved.