Methane dissociative chemisorption and detailed balance on Pt(111): Dynamical constraints and the modest influence of tunneling

Abstract

A dynamically biased (d-) precursor mediated microcanonical trapping (PMMT) model of the activated dissociative chemisorption of methane on Pt(111) is applied to a wide range of dissociative sticking experiments, and, by detailed balance, to the methane product state distributions from the thermal associative desorption of adsorbed hydrogen with coadsorbed methyl radicals. Tunneling pathways were incorporated into the d-PMMT model to better replicate the translational energy distribution of the desorbing methane product from the laser induced thermal reaction of coadsorbed hydrogen and methyl radicals occurring near T(s) = 395 K. Although tunneling is predicted to be inconsequential to the thermal dissociative chemisorption of CH4 on Pt(111) at the high temperatures of catalytic interest, once the temperature drops to 395 K the tunneling fraction of the reactive thermal flux reaches 15%, and as temperatures drop below 275 K the tunneling fraction exceeds 50%. The d-PMMT model parameters of {E0 = 58.9 kJ/mol, s = 2, η(v) = 0.40} describe the apparent threshold energy for CH4/Pt(111) dissociative chemisorption, the number of surface oscillators involved in the precursor complex, and the efficacy of molecular vibrational energy to promote reaction, relative to translational energy directed along the surface normal. Molecular translations parallel to the surface and rotations are treated as spectator degrees of freedom. Transition state vibrational frequencies are derived from generalized gradient approximation-density functional theory electronic structure calculations. The d-PMMT model replicates the diverse range of experimental data available with good fidelity, including some new effusive molecular beam and ambient gas dissociative sticking measurements. Nevertheless, there are some indications that closer agreement between theory and experiments could be achieved if a surface efficacy less than one was introduced into the modeling as an additional dynamical constraint.

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

“…The results therefore strongly suggest that the PBE functional used underestimates the E b value for methane + Pt(111) by approximately 0.1 eV, this value being the energy with which the theoretical curve needs to be shifted in order to obtain agreement with experiment near the reaction threshold, while the E b value obtained with the RPBE functional (1.06 eV 46 ) is probably too high by 0.15 eV. These results are consistent with the finding that the PBE density functional typically over- 17 on the basis of calorimetry 28 and associative desorption experiments. 29 The agreement further improves if one considers that the barrier sampled in the desorption experiments (0.6 eV) is about 0.15 eV lower than the one sampled in adsorption (0.8 eV) due to lattice relaxation, 48 that is, the puckering of the Pt atom on which the methyl is adsorbed before desorption of methane.…”

“…3,8,12,17,25−29 There has been considerable debate 2,25 concerning the importance of tunneling in this and similar systems. Recent calculations 17,23,30 suggest only a minor role for tunneling under the conditions addressed by us. Research on Ni 31 and on Pt 17 surfaces suggests that the dissociation of methane should proceed through a direct mechanism under both thermal and hyperthermal conditions on these surfaces.…”

“…This suggests the possibility to accurately model experiments relevant to both kinetics and dynamics with a unified approach. Harrison and coworkers have indeed shown 17 that a dynamically biased semiempirical statistical model yields an accurate description of a large range of experiments. However, this model cannot describe that, as also observed for CH 4 + Pt(111), 8,26,27 the extent to which putting energy into vibration promotes reaction depends on which vibrational state is excited.…”

Quantum state resolved gas–surface reaction dynamics experiments: a tutorial review

“…More recently, Harrison et al have extended the PC-MURT model to allow energy to be transferred between the precursor complexes and the surface, 41 as well as to include the effects of tunneling through the activation barrier. 42 The resulting theory is referred to as the precursor mediated microcanonical trapping model (PMMT). While the PC-MURT and PMMT models are able to reproduce the measured sticking coefficients reasonably well, this cannot be interpreted as proof that the microscopic reaction mechanism is in fact statistical.…”

Quantum State–Resolved Studies of Chemisorption Reactions

“…Mode specificity and bond selectivity had been observed previously in reactions in the gas phase (125-127), but were not necessarily expected to occur for gas-surface reactions due to the availability of many degrees of freedom of the solid surface for energy redistribution and relaxation. Harrison and coworkers (77)(78)(79)(80)128) developed a statistical theory for chemisorption reactions, which assumed that complete IVR occurred in a physisorbed complex consisting of the reactant molecule and a certain number of surface atoms. The model was claimed to quantitatively describe the available experimental data on methane chemisorption using only the dissociation barrier height, the number of surface atoms in the complex, and a single phonon frequency characterizing the surface as input variables.…”