Francesco Nattino

Ab Initio Molecular Dynamics Calculations versus Quantum-State- Resolved Experiments on CHD3 + Pt(111): New Insights into a Prototypical Gas−Surface Reaction

The dissociative chemisorption of methane on metal surfaces is of fundamental and practical interest, being a rate-limiting step in the steam reforming process. The reaction is best modeled with quantum dynamics calculations, but these are currently not guaranteed to produce accurate results because they rely on potential energy surfaces based on untested density functionals and on untested dynamical approximations. To help overcome these limitations, here we present for the first time statistically accurate reaction probabilities obtained with ab initio molecular dynamics (AIMD) for a polyatomic gas-phase molecule reacting with a metal surface. Using a general purpose density functional, the AIMD reaction probabilities are in semiquantitative agreement with new quantum-state-resolved experiments on CHD3 + Pt(111). The comparison suggests the use of the sudden approximation for treating the rotations even though CHD3 has large rotational constants and yields an estimated reaction barrier of 0.9 eV for CH4 + Pt(111).

Dissociative chemisorption of methane on metal surfaces: Tests of dynamical assumptions using quantum models and ab initio molecular dynamics

The dissociative chemisorption of methane on metal surfaces is of great practical and fundamental importance. Not only is it the rate-limiting step in the steam reforming of natural gas, the reaction exhibits interesting mode-selective behavior and a strong dependence on the temperature of the metal. We present a quantum model for this reaction on Ni(100) and Ni(111) surfaces based on the reaction path Hamiltonian. The dissociative sticking probabilities computed using this model agree well with available experimental data with regard to variation with incident energy, substrate temperature, and the vibrational state of the incident molecule. We significantly expand the vibrational basis set relative to earlier studies, which allows reaction probabilities to be calculated for doubly excited initial vibrational states, though it does not lead to appreciable changes in the reaction probabilities for singly excited initial states. Sudden models used to treat the center of mass motion parallel to the surface are compared with results from ab initio molecular dynamics and found to be reasonable. Similar comparisons for molecular rotation suggest that our rotationally adiabatic model is incorrect, and that sudden behavior is closer to reality. Such a model is proposed and tested. A model for predicting mode-selective behavior is tested, with mixed results, though we find it is consistent with experimental studies of normal vs. total (kinetic) energy scaling. Models for energy transfer into lattice vibrations are also examined.

Chemically Accurate Simulation of a Polyatomic Molecule-Metal Surface Reaction

Although important to heterogeneous catalysis, the ability to accurately model reactions of polyatomic molecules with metal surfaces has not kept pace with developments in gas phase dynamics. Partnering the specific reaction parameter (SRP) approach to density functional theory with ab initio molecular dynamics (AIMD) extends our ability to model reactions with metals with quantitative accuracy from only the lightest reactant, H2, to essentially all molecules. This is demonstrated with AIMD calculations on CHD3 + Ni(111) in which the SRP functional is fitted to supersonic beam experiments, and validated by showing that AIMD with the resulting functional reproduces initial-state selected sticking measurements with chemical accuracy (4.2 kJ/mol ≈ 1 kcal/mol). The need for only semilocal exchange makes our scheme computationally tractable for dissociation on transition metals.

Methane Dissociation on Pt(111) : Searching for a Specific Reaction Parameter Density Functional

The theoretical description of methane dissociating on metal surfaces is a current frontier in the field of gas-surface dynamics. Dynamical models that aim at achieving a highly accurate description of this reaction rely on potential energy surfaces based on density functional theory calculations at the generalized gradient approximation. We focus here on the effect that the exchange-correlation functional has on the reactivity of methane on a metal surface, using CHD3 + Pt(111) as a test case. We present new ab initio molecular dynamics calculations performed with various density functionals, looking also at functionals that account for the van der Waals (vdW) interaction. While searching for a semi-empirical specific reaction parameter density functional for this system, we find that the use of a weighted average of the PBE and the RPBE exchange functionals together with a vdW-corrected correlation functional leads to an improved agreement with quantum state-resolved experimental data for the sticking probability, compared to previous PBE calculations. With this semi-empirical density functional, we have also investigated the surface temperature dependence of the methane dissociation reaction and the influence of the rotational alignment on the reactivity, and compared our results with experiments.

Dissociation and recombination of D2 on Cu(111): Ab initio molecular dynamics calculations and improved analysis of desorption experiments

Obtaining quantitative agreement between theory and experiment for dissociative adsorption of hydrogen on and associative desorption of hydrogen from Cu(111) remains challenging. Particularly troubling is the fact that theory gives values for the high energy limit to the dissociative adsorption probability that is as much as two times larger than experiment. In the present work we approach this discrepancy in three ways. First, we carry out a new analysis of the raw experimental data for D2 associatively desorbing from Cu(111). We also perform new ab initio molecular dynamics (AIMD) calculations that include effects of surface atom motion. Finally, we simulate time-of-flight (TOF) spectra from the theoretical reaction probability curves and we directly compare them to the raw experimental data. The results show that the use of more flexible functional forms for fitting the raw TOF spectra gives fits that are in slightly better agreement with the raw data and in considerably better agreement with theory, even though the theoretical reaction probabilities still achieve higher values at high energies. The mean absolute error (MAE) for the energy E0 at which the reaction probability equals half the experimental saturation value is now lower than 1 kcal/mol, the limit that defines chemical accuracy, while a MAE of 1.5 kcal/mol was previously obtained. The new AIMD results are only slightly different from the previous static surface results and in slightly better agreement with experiment.

N2 dissociation on W(110): An ab initio molecular dynamics study on the effect of phonons

Accurately modeling the chemisorption dynamics of N2 on metal surfaces is of both practical and fundamental interest. The factors that may have hampered this achievement so far are the lack of an accurate density functional and the use of approximate methods to deal with surface phonons and non-adiabatic effects. In the current work, the dissociation of molecular nitrogen on W(110) has been studied using ab initio molecular dynamics (AIMD) calculations, simulating both surface temperature effects, such as lattice distortion, and surface motion effects, like recoil. The forces were calculated using density functional theory, and two density functionals were tested, namely, the Perdew-Burke-Ernzerhof (PBE) and the revised PBE (RPBE) functionals. The computed dissociation probability considerably differs from earlier static surface results, with AIMD predicting a much larger contribution of the indirect reaction channel, in which molecules dissociate after being temporally trapped in the proximity of the surface. Calculations suggest that the surface motion effects play a role here, since the energy transfer to the lattice does not allow molecules that have been trapped into potential wells close to the surface to find their way back to the gas phase. In comparison to experimental data, AIMD results overestimate the dissociation probability at the lowest energies investigated, where trapping dominates, suggesting a failure of both tested exchange-correlation functionals in describing the potential energy surface in the area sampled by trapped molecules.

Surface Reaction Barriometry: Methane Dissociation on Flat and Stepped Transition-Metal Surfaces

Accurately simulating heterogeneously catalyzed reactions requires reliable barriers for molecules reacting at defects on metal surfaces, such as steps. However, first-principles methods capable of computing these barriers to chemical accuracy have yet to be demonstrated. We show that state-resolved molecular beam experiments combined with ab initio molecular dynamics using specific reaction parameter density functional theory (SRP-DFT) can determine the molecule-metal surface interaction with the required reliability. Crucially, SRP-DFT exhibits transferability: the functional devised for methane reacting on a flat (111) face of Pt (and Ni) also describes its reaction on stepped Pt(211) with chemical accuracy. Our approach can help bridge the materials gap between fundamental surface science studies on regular surfaces and heterogeneous catalysis in which defected surfaces are important.

Modeling surface motion effects in N2 dissociation on W(110): Ab initio molecular dynamics calculations and generalized Langevin oscillator model

Accurately modeling surface temperature and surface motion effects is necessary to study molecule-surface reactions in which the energy dissipation to surface phonons can largely affect the observables of interest. We present here a critical comparison of two methods that allow to model such effects, namely, the ab initio molecular dynamics (AIMD) method and the generalized Langevin oscillator (GLO) model, using the dissociation of N2 on W(110) as a benchmark. AIMD is highly accurate as the surface atoms are explicitly part of the dynamics, but this advantage comes with a large computational cost. The GLO model is much more computationally convenient, but accounts for lattice motion effects in a very approximate way. Results show that, despite its simplicity, the GLO model is able to capture the physics of the system to a large extent, returning dissociation probabilities which are in better agreement with AIMD than static-surface results. Furthermore, the GLO model and the AIMD method predict very similar energy transfer to the lattice degrees of freedom in the non-reactive events, and similar dissociation dynamics.

Application of van der Waals functionals to the calculation of dissociative adsorption of N2 on W(110) for static and dynamic systems

The fundamental understanding of molecule-surface reactions is of great importance to heterogeneous catalysis, motivating many theoretical and experimental studies. Even though much attention has been dedicated to the dissociative chemisorption of N2 on tungsten surfaces, none of the existing theoretical models has been able to quantitatively reproduce experimental reaction probabilities for the sticking of N2 to W(110). In this work, the dissociative chemisorption of N2 on W(110) has been studied with both static electronic structure and ab initio molecular dynamics (AIMD) calculations including the surface temperature effects through surface atom motion. Calculations have been performed using density functional theory, testing functionals that account for the long range van der Waals (vdW) interactions, which were previously only considered in dynamical calculations within the static surface approximation. The vdW-DF2 functional improves the description of the potential energy surface for N2 on W(110), returning less deep molecular adsorption wells and a better ratio between the barriers for the indirect dissociation and the desorption, as suggested by previous theoretical work and experimental evidence. Using the vdW-DF2 functional less trapping-mediated dissociation is obtained compared to results obtained with standard semi-local functionals such as PBE and RPBE, improving agreement with experimental data at E(i) = 0.9 eV. However, at E(i) = 2.287 and off-normal incidence, the vdW-DF2 AIMD underestimates the experimental reaction probabilities, showing that also with the vdW-DF2 functional the N2 on W(110) interaction is not yet described with quantitative accuracy.