Predicting CO₂ adsorption and reactivity on transition metal surfaces using popular density functional theory methods

ABSTRACT

In this work, with Ni (110) as a model catalyst surface and CO₂ as an adsorbate, a performance study of Density Functional Theory methods (functionals) is performed. CO being a possible intermediate in CO₂ conversion reactions, binding energies of both, CO₂ and CO, are calculated on the Ni surface and are compared with experimental data. OptPBE-vdW functional correctly predicts CO₂ binding energy on Ni (−62 kJ/mol), whereas CO binding energy is correctly predicted by the rPBE-vdW functional (−138 kJ/mol). The difference in computed adsorption energies by different functionals is attributed to the calculation of gas phase CO₂. Three alternate reaction systems based on a different number of C=O double bonds present in the gas phase molecule are considered to replace CO₂. The error in computed adsorption energy is directly proportional to the number of C=O double bonds present in the gas phase molecule. Additionally, both functionals predict similar carbon–oxygen activation barrier (40 kJ/mol) and equivalent C1s shifts for probe species (−2.6 eV for CCH₃ and +1.5 eV CO₃⁻), with respect to adsorbed CO₂. Thus, by including a correction factor of 28 kJ/mol for the computed CO₂ gas phase energy, we suggest using rPBE-vdW functional to investigate CO₂ conversion reactions on different metals.

KEYWORDS:

• Density functional theory
• carbon dioxide adsorption
• DFT–XPS
• DFT accuracy

Acknowledgments

The computational work for this article was partially performed on resources of the National Supercomputing Centre (NSCC), Singapore (https://www.nscc.sg), West Grid, Canada (www.westgrid.ca) and Compute Canada (www.computecanada.ca). We thank NSCC, Compute Canada and Westgrid for allowing us to use the high-performance computing resources. We would also like to thank Mr. Kartavya Bhola for many useful discussions.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Ojus Mohan http://orcid.org/0000-0003-0298-2725

Additional information

Funding

The authors would like to acknowledge the financial support provided by the Nanyang Technological University, Singapore. QTT acknowledges the support from the National Research Foundation (NRF), Prime Minister's Office, Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE) programme. SHM acknowledges funding from the Natural Sciences and Engineering Research Council of Canada (NSERC).