AVS 66 Session SS+HC-MoA: CO2, CO, Water, and Small Molecule Chemistry at Surfaces

Calculations of electrochemical CO2 reduction to formate, alcohols and hydrocarbons will be presented. The mechanism for the formation of the various products is established, the rate evaluated and comparison made with experimental measurements. The rate of the main side reaction, the hydrogen evolution reaction, is also estimated. The calculations are based on a detailed atomistic model of the electrical double layer (metal slab and water layer) and density functional theory calculations to evaluate not only the free energy of intermediates as a function of applied voltage but also the activation energy for each elementary step, both Heyrovsky and Tafel reactions [1]. Comparison is also made with calculations using an implicit solvation model [2]. A range of close packed metal surfaces are compared, including Cu, Ag, Au, Ni, Fe, Rh, Ir and Pt. The resuls are in remarkably good agreement with the reported experimental measurements. A two parameter descriptor is established that can help identify improved catalysts for CO2 electrochemical reduction.

[1] J. Hussain, H. Jónsson and E. Skúlason, ACS Catalysis 8, 5240 (2018).

[2] M. Van den Bossche, E. Skúlason, C. Rose-Petruck and H. Jónsson, J. Phys. Chem. C 123, 4116 (2019).

Copper oxide catalysts are promising candidates for reducing CO₂ into useful fuels, such as ethanol, but the mechanism remains obscure. Studying the O-Cu(100) surface, which represents the initial transition of the oxidation of copper to copper oxide, and the adsorption process of CO₂ has the potential to elucidate the CO₂ reduction mechanism.

We performed a DFT/STM theoretical and experimental probe of the properties of CO­­_{2 ­}adsorption on the O-Cu(100) surface._­ The Cu(100) surface was repeatedly sputtered with Ar⁺ and annealed at 550°C in an ultra-high vacuum chamber, and subsequent Auger spectroscopy revealed the lack of surface contamination. The O-Cu(100) surface was obtained by exposing the Cu(100) face to 10^-6 mbar of oxygen for 5 minutes at 300°C. The sample was then cooled and transferred into an attached low temperature (5K) ultra-high vacuum (10^-11 mbar) STM chamber.

The atomic resolution STM revealed the (2√2× √2) R45°O-Cu(100) reconstruction, in good agreement with the DFT calculations. The reconstruction consists of an O-Cu-O row structure separated by missing Cu rows. Additionally, there are two equivalent domains which result from nucleation along the [001] and [010] directions of the Cu(100). Differential conductance spectroscopy reveals an increase in the work function of the O-Cu(100) surface, and two additional unoccupied states generated by the oxygen atoms, in agreement with the DFT calculations.

CO­₂was adsorbed onto the O-Cu(100) surface via in situ dosing in the STM. The CO₂ adsorbed exclusively between two oxygen atoms in the missing row reconstruction, which has the largest predicted adsorption energy. The lack of point defects on the surface indicates that the CO­₂ does not dissociate, and the CO­₂ molecules are easily disturbed by the tip under all tunneling conditions, which is consistent with the theoretically predicted low diffusion barrier.

This work acknowledges funding from the NSF (1809837).

Metal oxide semiconductors are commonly researched materials for solid-state gas sensors; however, several limitations (i.e., operating temperatures of ≥300 °C and poor selectivity) impede wide-spread commercialization of these devices. Plasma processing offers a desirable alternative route to traditional methods, such as doping, because of the tunability of treatment parameters and the ability to modify the surface of the material while maintaining bulk properties. Previous work using plasma modification to enhance tin dioxide (SnO₂) gas sensor performance has mainly focused on oxygen or oxygen/argon plasma systems because these systems are thought to etch oxygen from the SnO₂ lattice creating oxygen vacancies that can lead to lower operating temperatures and improved sensor selectivity. Thus, further work needs to be done with other precursor gases to determine an effective strategy for fabricating improved gas sensors.

CO₂ chemistry has received significant attention recently, due to the greenhouse effect of CO₂ emissions and the resulting climate change. CO₂ reduction reactions, such as methanol synthesis and reverse water-gas shift, provide routes for recycling of CO₂ and thus limiting the CO₂ emissions. These reactions are commonly performed over Cu-based catalysts, making the interaction of CO₂ and Cu, on the atomic scale, of importance for a fundamental understanding and the development of new and more efficient catalysts.

We have previously studied the dissociative adsorption of CO₂ on Cu(100) using APXPS and DFT. In summary, exposure of the Cu surface to CO₂ in the mbar range at temperatures above room temperature results in dissociation of CO₂ into CO, that desorbs, and O that stays on the surface. The rate of the increase in O coverage, however, was not consistent with what one would expect from adsorption on the flat Cu(100) surface. Instead, we propose a model where the dissociation happen at atomic steps. The steps were found to both lower the activation barrier for the dissociation and separate the products, such that the probability for recombination is lowered.

As an obvious follow-up of this study, we have studied the dissociative adsorption of CO₂ on Cu(911), which exposes five atoms wide (100) terraces separated by monatomic steps. In contrast to what we expected, the O coverage did not increase significantly faster on this stepped surface. Our preliminary analysis suggests that diffusion of O from one step to another reduces the effect of the steps separating O and CO, but also that the steps facilitate O diffusion to the subsurface region and possibly stabilisation of CO₂ or CO₃ species on the surface.

In this presentation we will report how we conclude that the steps control the dissociation and, especially, the present status of the studies of Cu(911).

Commercial steam reforming reactors operate at temperatures of 1000K or higher, and methane dissociation on the Ni catalyst is generally believed to be the rate-limiting step in this industrially important process. Despite the commercial importance of this reaction, nearly all studies probing the dynamics of methane dissociation have focused on surface temperatures of 600K or lower. Here, we use energy and vibrationally state selected methane molecules in a supersonic molecular beam to quantify the impact of surface temperature on methane activation over a wide range of surface temperatures. Our use of methane molecules with a precisely defined energy highlights provides a clear view of how surface temperature impacts reactivity.

Vibrationally state-resolved reactivity measurements reveal details of fundamental processes that impact reactivity in the field of heterogeneous catalysis. Non-statistical, mode-specific, and bond-selective enhancements observed for methane and its isotopologues on transition metal surfaces provide insights into energy flow during reactions. Reactive gas molecules with strictly-defined energy in well-defined energetic coordinates used in state-selective experiments have also proven to be valuable probes of how surface atom motion affects overall reactivity. For this work, vibrationally state-resolved data was collected via infrared (IR) laser excitation of the anti-symmetric stretch of supersonically-expanded methane (CH₄) gas molecules impinging on a lightly-stepped Ni(997) surface. Measurements on the single crystal were investigated over a broad range of surface temperatures (82 K ≤ T_S ≤ 1000 K) while utilizing varying incident energies (E_i = 20 kJ/mol to >140 kJ/mol). A comparison with prior data on Ni(111) surface reveals the role that steps may play in methane activation.