Dataset from a Theoretical Computational Study on the Influence of Co-Cu Interface Structures on the Synthesis of Low-Carbon Alcohols from Synthesis Gas

1. Dataset Generation ProcessThis dataset was generated based on the theoretical calculation study of syngas-to-lower alcohols catalyzed by the Co-Cu bimetallic interface. It focused on the elementary reactions of syngas-to-ethanol and was completed in three steps: computational model construction, elementary reaction path design, and energy/structure data extraction, as detailed below:Model Construction Stage: Based on the FCC Co(111) and Cu(111) crystal planes, a p(5×4) supercell Co-Cu interface bimetallic model (containing 60 Co atoms and 20 Cu atoms) was constructed. The bottom two layers of Co atoms in the model were fixed, while all other atoms and surface adsorbates were allowed to relax. A 15 Å vacuum layer was set to avoid interactions between adjacent slabs, thus completing the structural modeling of the catalytic system.Reaction Path and Calculation Design Stage: For the core elementary reactions of syngas-to-ethanol, the full reaction paths including H₂ dissociation, CO adsorption and dissociation, CH*-CO* coupling, C species hydrogenation to ethanol formation, and methane side reaction were designed. The initial state (IS), transition state (TS), and final state (FS) structures of each reaction were determined, and the transition state configuration of each elementary reaction was identified by the transition state search method, forming a complete reaction calculation system.Data Calculation and Extraction Stage: Density functional theory (DFT) was used to optimize the energy and structure of each reaction site and reaction step, and microkinetic simulations were also carried out. Core data such as reaction energy, dissociation/coupling energy barrier, atomic charge distribution, d-band center, HOMO-LUMO energy gap, surface species coverage, turnover frequency (TOF), degree of rate control (DRC) of each site were extracted. Combined with the crystal structure parameters of reaction intermediates, the original dataset was formed.2. Data Processing Methods and StepsThe processing of this dataset was carried out around four core links: raw calculation data calibration, characteristic data extraction, data standardization, and validity verification. All processing steps were based on the original calculation results to ensure the authenticity and repeatability of the data, with the specific steps as follows:Raw calculation data calibration: The energy data obtained by DFT calculation were corrected for van der Waals (vdW) and spin polarization to eliminate the effects of non-bonding interactions and electron spin during the calculation. Taking the adsorbed state as the energy zero point, the relative energy of each reaction was recalibrated to unify the energy calculation benchmark.Core characteristic data extraction: From the optimized structure and energy files, the reaction energy (ΔE) and reaction energy barrier (Eᵦ) of each elementary reaction were extracted, with the calculation formulas of ΔE=E_FS-E_IS and Eᵦ=E_TS-E_IS. Meanwhile, structural and electronic property characteristic data such as Bader charge variation, d-band center value, C-O/C-C bond length, and HOMO-LUMO energy gap were extracted. From the microkinetic simulation results, TOF, DRC values and coverage data of each surface species under different temperature and pressure conditions were extracted.Data standardization processing: The barrier and energy data of different reaction sites and reaction paths were uniformly converted to the unit of electron volt (eV), the temperature and pressure data were unified to Kelvin (K) and bar respectively, and the rate data were unified to per second (s⁻¹). The structural parameters (bond length, atomic spacing) were uniformly converted to angstrom (Å) to realize the unit standardization of the dataset. The discrete structural and electronic property data were normalized to facilitate subsequent data analysis and model construction.Data validity verification and screening: The linear correlation between d-band center and CO dissociation barrier, as well as HOMO-LUMO energy gap and CH*-CO* coupling barrier was verified by linear fitting, and abnormal data with extremely low fitting degree were eliminated. Combined with the thermodynamic and kinetic laws of the reaction, the valid data of the dominant reaction path were screened out, and the redundant data of non-dominant paths were excluded. Convergence verification was performed on the microkinetic simulation results to ensure no convergence error in the rate and coverage data under temperature and pressure scanning, and finally a validated effective dataset was formed.3. Equipments and Tools Used(1) Computational EquipmentAll theoretical calculations in this study were completed using a High-Performance Computing (HPC) cluster with node configurations as follows: Intel Xeon Gold series processors, 64 cores/128 threads per node, 512 GB DDR4 memory, NVIDIA A100 GPU (40 GB HBM2) for graphics cards, and the compute nodes were interconnected via InfiniBand high-speed network, which met the computing power requirements for large-scale DFT calculations and microkinetic simulations.(2) Computational and Data Processing ToolsFirst-principles calculation tool: The Vienna Ab initio Simulation Package (VASP 5.4.4) was used to perform spin-polarized density functional theory calculations. The projector augmented-wave (PAW) method was adopted to describe the electron-ion interaction, the GGA-PBE functional was combined to treat the electron exchange-correlation energy, and the DFT-D3 correction was introduced to describe the van der Waals interaction. The Climbing Image Nudged Elastic Band (CI-NEB) method was used for transition state search and optimization.Microkinetic simulation tool: CATKINAS software was used to carry out microkinetic analysis of syngas-to-ethanol, realizing the simulation and data output of reaction rate, rate-determining steps, and surface species coverage under variable temperature and pressure conditions.Structure and charge analysis tool: VESTA software was used to complete the crystal structure visualization of the catalytic system and reaction intermediates and the extraction of structural parameters (bond length, atomic spacing). The Bader charge analysis program was used to calculate the charge distribution of adsorbed species and extract charge variation data. The d-band center and HOMO-LUMO energy gap were calculated and extracted through the band structure analysis module built in VASP.Data processing and analysis tool: OriginPro 2021 was used to complete linear fitting, normalization and visualization of data, and realize statistical analysis and chart drawing of barrier, energy, charge variation and other data. Microsoft Excel 2021 was used to sort out the original data, convert units and standardize, forming a structured dataset file.File format conversion and management tool: Tools such as VASP2XYZ and POSCAR Converter were used to complete the format conversion of calculation files (POSCAR, OUTCAR, CONTCAR) to facilitate the extraction of structure and energy data. Git was used for version management of calculation processes and data files to ensure the traceability of data.