Carbon dioxide (CO2) is a significant greenhouse gas and a primary contributor to global warming. The transformation of CO2 into methane (CH4) presents a promising approach as it can serve as an alternative to traditional fossil fuels. Methane, with its high combustion efficiency, is a vital energy source for both heating and electricity generation. Thus, the photocatalytic reduction of CO2 emerges as a critical strategy for mitigating greenhouse gas emissions and alleviating the energy crisis.In our study, we employ monolayer graphitic carbon nitride (g-C3N4) as the subject of our simulation calculations. g-C3N4 has been extensively demonstrated as an exceptional photocatalytic material due to several advantageous properties. It possesses a suitable optical band gap and position (2.7 eV), making it highly efficient for light absorption and photocatalytic activity. Additionally, g-C3N4 can be readily fabricated at low cost, which enhances its feasibility for large-scale applications. The material is also biocompatible and nontoxic, ensuring its safety for environmental and biological systems. Furthermore, monolayer g-C3N4 exhibits a larger surface area and superior photogenerated carrier transport properties compared to bulk materials, which significantly improves its photocatalytic performance.

The physicochemical properties of materials are highly sensitive to the reaction mechanism and activity. Our study hypothesizes that the variation in Fermi levels is a key factor influencing reaction activity. Numerous methods have been developed in previous literature to modulate these properties effectively. For n-type materials, dopants such as phosphorus (P) and sulfur (S) can be integrated into carbon (C) or nitrogen (N) sites to fill intrinsic vacancies, resulting in an increase in free electrons and raising the Fermi level. This suggests that P,S,O-co-doped g-C3N4 would serve as a superior photocatalyst due to its reduced electrical resistance and greater number of free electrons compared to undoped g-C3N4. Conversely, for p-type materials, co-doping with elements such as potassium (K) and iron (Fe) can cause a downshift in band alignments. The lowered valence band can impart stronger oxidation abilities to photo-generated holes. In p-type semiconductors, holes, which possess strong oxidation capabilities due to the much lower valence band, act as the main carriers and can migrate to the surface to interact with reactants. Therefore, a synergistic effect exists between the low valence band and p-type behaviors. These approaches have collectively enhanced the performance and reduction capability of the catalyst, demonstrating significant improvements in the photocatalytic reduction of CO2 to methane. Furthermore, characterization of these surface-modified materials has revealed that their Fermi levels can vary, rather than being pinned at a specific value, further contributing to their improved catalytic properties.

Despite these advancements, the experimental elucidation of the complex reaction mechanisms remains unclear. To investigate the photocatalytic process of CO2 reduction, in situ diffuse reflectance Fourier transform infrared spectroscopy (DRIFTs) was employed to monitor the reaction intermediates during the photocatalytic reduction process of the modified materials. When CO2 gas and H2O vapor were introduced, numerous FTIR peaks were detected. The key intermediates identified in the photoreduction of CO2 to CH4 included COOH*, *CHO, and *OCH3. These intermediates play crucial roles in the reaction pathway, contributing to our understanding of the underlying mechanisms in the photocatalytic reduction of CO2.

We posit that previous computational models may have overlooked variations in the Fermi level during the photocatalytic process, leading to inaccuracies. Therefore, in this research, we designed intermediates with different valence states to assess changes in adsorption energy and the spontaneity of the reaction in thermodynamic and kinetic contexts. Our initial theoretical DFT calculations, incorporating Fermi level dependencies, revealed the relationship between adsorption energies of potential intermediates and changes in the Fermi level throughout the CO2 reduction process. Given the complexity of the reaction and the dual role of some intermediates as products, we proposed a kinetically favorable reaction pathway. By comparing conditions in the absence of light and under illumination, we provided a more detailed and persuasive explanation of the reaction mechanism.

Furthermore, we analyzed the electronic structures of intermediates with different valence states using three methods.Density of States provided insights into the overall distribution and changes in the system's electronic states, aiding in understanding the electronic structure and adsorption states.Differential Charge Density visualization technique demonstrated the redistribution of charges during the adsorption process, revealing the mechanism of charge transfer in adsorption.

Crystal Orbital Hamilton Population (COHP) analysis examined the bonding and antibonding characteristics of specific orbitals, quantitatively describing the formation and changes of chemical bonds, and explaining the impact of the electronic structure on the stability of chemical bonds.These methods collectively offered a comprehensive and detailed analysis of the photocatalytic process, enhancing our understanding of the reaction mechanism and the role of Fermi level variations.

Our surface modeling calculations show significant changes in adsorption energy for both positively and negatively charged intermediates. In the absence of light, the adsorption energy for positively charged intermediates becomes more positive with the rise in Fermi level, whereas for negatively charged intermediates, the adsorption energy becomes more negative. This results in substantial differences in reaction barriers under varying Fermi levels, highlighting the importance of Fermi level adjustments in understanding and designing high-performance photocatalysts. Under illumination, the electronic structure calculations reveal that at the N active sites, nitrogen contributes significantly to the valence band maximum (VBM). Photo-excited holes may destabilize negatively charged intermediates, reducing them to a neutral state, or photo-excited electrons may reduce positively charged intermediates to a neutral state. Such processes, which consume photo-excited electrons and holes, lower the reaction barriers and enhance the thermodynamic spontaneity of the reaction.

Combining our electronic structure calculations, we found distinct differences in the interaction and bonding contributions between the atomic orbitals of positively and negatively charged intermediates. The density of states (DOS) for negatively charged intermediates indicates charge transfer from the substrate surface to the adsorbate, resulting in charge accumulation. Projected density of states (PDOS) images show a significant increase in electron population on the bonding atoms of negatively charged intermediates compared to neutral states. Differential charge density visualizes the redistribution of charges during the adsorption process, revealing the mechanism of charge transfer. Crystal Orbital Hamilton Population (COHP) analyzes the bonding and antibonding characteristics of specific orbitals, quantitatively describing the formation and changes of chemical bonds and explaining the impact of electronic structure on the stability of chemical bonds.The absence of density of states peaks between the conduction band (CBM) and valence band (VBM) indicates that adsorption after adsorption does not introduce new impurity or defect states in the band gap and the Fermi energy levels are not pinned. This implies that the Fermi energy level can move in response to changes in external conditions (e.g., doping concentration, adsorption, etc.), and our assumptions are shown to be consistent by DOS.