Periodic Density Matrix Embedding for CO Adsorption on the MgO(001)Surface

The adsorption of simple gas molecules to metal oxide surfaces is a primary step in many heterogeneous catalysis applications. Quantum chemical modeling of these reactions is a challenge both in terms of cost and accuracy, and quantum-embedding methods are promising, especially for localized chemical phenomena. In this work, we employ density matrix embedding theory (DMET) for periodic systems to calculate the adsorption energy of CO to the MgO(001) surface. Using coupled-cluster theory with single and double excitations and second-order Møller-Plesset perturbation theory as quantum chemical solvers, we perform calculations with embedding clusters up to 266 electrons in 306 orbitals; the largest embedding models agreeing to within 1.2 kcal/mol of the non-embedding references. Due to the need of large impurity clusters for surface chemistry, we present a memory-efficient procedure of storing and manipulating electron repulsion integrals in the embedding space within the framework of periodic DMET.