Oxidation of Cu(100) by CO2 dissociative adsorption

Alvaro Posada-Borbon,1 Benjamin Hagman,2 Andreas Schaefer,2,3 Chu Zhang,2 Mikhail Shipilin,2,4 Lindsay Merte,2,5 Anders Hellman,1 Johan Gustafson,2 and Henrik Grönbeck1
1Department of Physics and Competence Centre for Catalysis, Chalmers University of Technology, Sweden
2Synchrotron Radiation Research, Lund University, Sweden
3Department of Chemistry and Chemical Engineering and Competence Centre for Catalysis, Chalmers University of Technology, Sweden
4Department of Physics, AlbaNova University Center, Stockholm University, Sweden
5Department of Materials Science and Applied Mathematics, Malmö University, Sweden

Catalytic synthesis of methanol from CO2 is considered as one pathway for production of added-value chemicals by use of a greenhouse gas [1]. Already today, methanol is industrially synthesized by catalytic hydrogenation of CO2 over Cu/ZnO/Al2O3. However, the copper-based catalyst has shortcomings [2], where one is rapid deactivation [3]. This has been linked to the oxidation of the copper phase upon CO2 dissociation [4]. However, a detailed understanding of the reaction at the catalysts surface is still under debate [5]. In the present contribution, we use density functional theory (DFT) calculations and micro-kinetic modeling to investigate the dissociative adsorption of CO2 on the pristine and oxygen covered Cu(100). Surface stability with respect to oxygen coverage is calculated together with O 1s core level shifts. The results are compared to APXPS experiments. A reaction mechanism for CO2 dissociation at Cu(100) is proposed. As a structural model for monoatomic steps on Cu(100) we use Cu(611). We find that CO2 adsorption and dissociation on Cu (100) and Cu(611) occurs via CO2 adsorption in a bent CO2δ- mode over a hollow position. On the pristine surfaces, adsorption to the CO2δ- mode is associated with a barrier of 0.55 eV and 0.60 eV for Cu(100) and Cu(611), respectively. The corresponding barriers for dissociation from the adsorbed state are 0.83 eV [Cu(100)] and 0.57 eV [Cu(611)]. Only the first barrier is affected by the presence of co-adsorbed oxygen. However, the stability of adsorbed CO2 on the surface is critically dependent on the position of co-adsorbed oxygen atoms. Oxidation of Cu(100) is found to occur via the formation of a p(2x2) overlayer (0.25 ML) followed by a reconstructed (2√2x√2)R45 -O missing (MR) structure (0.50 ML). A c(4x6) structure with a 0.3 ML coverage is, however, close in stability for intermediate oxygen chemical potentials. We find that tensile strain (+2%) stabilizes the c(4x6) surface structure with respect to the p(2x2). The calculated shifts in the O 1s binding energy are in good agreement with the measured evolution of the binding energy. The shift to higher O 1s binding energies with increasing oxygen coverage is found to correlate with the charge on neighboring copper atoms. The work suggest a path for oxidation of Cu(100) from CO2 where CO2 adsorbs dissociatively exclusively on the step sites [6]. Oxygen diffusion and clustering initiates surface reconstruction, which eventually leads to growth of MR-patches. The computational results are in good agreement with experimental observations of the evolution of the O 1s core level shift as a function of oxygen coverage upon CO2 exposure [7].

  1. S. Saeidi et al. Journal of CO2 Utilization 5, 66–81 (2014).
  2. D. Cheng et al. ChemSusChem 6, 944–965 (2013).
  3. O. Martin et al. Angew. Chem. Int. Ed. 55, 6261-6265 (2016).
  4. B. Eren, et al. J. Am. Chem. Soc. 138, 8207-8211 (2016).
  5. I. C. Grabow and M. Mavrikakis. ACS Catal. 1, 4, 364-384 (2011).
  6. Hagman, B., Posada-Borbón, A. et al. J. Am. Chem. Soc. 140, 12974−12979 (2018).
  7. Posada-Borbón, A., Hagman, B. et al. Surf. Sci. 675, 64-69 (2018).