III- Computational methods The Density-Functional based Tight-Binding (DFTB) method was employed to optimize conventional cells for various metals, with lattice vectors being optimized prior to geometry optimization. The optimized conventional cells were used as input data to generate four layers of (111) 2D slabs for different metals. Grimme's GFN1-xTB served as the model Hamiltonian for the DFTB calculations108. Subsequently, single-point calculations were performed in the periodic DFT program (BAND)109-110. to obtain the formation energies of the slabs. Highquality k-space grids were employed for both DFTB and BAND calculations. The generalized gradient approximation (GGA) with PBE as the default exchange-correlation functional was used for the BAND calculations. A Slater-type basis set of triple- valence quality with a single polarization function (TZP) was used for all atoms in the calculations90. The same settings were used to calculate the formation energy of negatively charged metal slabs. To investigate the interaction between the metal slabs and CO2, a single CO2 molecule that had already been optimized in ADF was introduced to the calculations. The CO2 molecule was positioned at a distance of 220 picometers from an on-top metallic atom of the slab. A subsequent calculation was then performed using the same settings to obtain the formation energy and charge distribution for a negatively charged slab+CO2 system. It should be noted that when a negative charge is introduced to the slab+CO2 system, the engine considers the slab and CO2 as a whole to obtain the lowest energy charge distribution. The same protocol was also utilized to determine the formation energies of negatively charged slabs+CO systems. The CO desorption energy was subsequently calculated as follows:
RkJQdWJsaXNoZXIy MTk4NDMw