Producing hydrogen through water splitting using particulate photocatalysts is a promising green technology for large-scale solar energy conversion. Graphitic carbon nitride (g-C3N4) shows potential in hydrogen generation, particularly when small molecule organics act as hole scavengers. However, single-phased g-C3N4 has been considered unsuitable for overall water splitting due to its insufficient oxygen evolution reaction (OER) ability. This inability has often been attributed to the weak oxidation capacity of photo-induced valence band holes. While g-C3N4 meets the thermodynamic requirements for water splitting with a band gap exceeding 2.0 eV and appropriate band positions, the lack of O2 production indicates an additional, unidentified factor hindering the OER. This research aims to pinpoint this bottleneck and explore strategies to overcome it for efficient overall water splitting using single-phased g-C3N4 under visible light.

Previous studies have demonstrated the effectiveness of g-C3N4 in hydrogen evolution reactions, especially when using sacrificial reagents. However, achieving overall water splitting with single-phased g-C3N4 has proven challenging. Researchers have explored composite photocatalysts incorporating g-C3N4 with OER co-catalysts to overcome the limitations of the single-phased material. While some success has been achieved through heterojunction construction, the inherent limitations of single-phased g-C3N4 in OER remain a significant obstacle. The literature generally points to the weak oxidation capacity of the valence band holes as the primary reason for the poor OER performance, but this research seeks to uncover more nuanced factors.

The study employed in situ diffuse reflection infrared Fourier transform spectroscopy (DRIFTS) with isotopic labeling (¹⁶O/¹⁸O) and near-ambient pressure X-ray photoelectron spectroscopy (NAP-XPS) to monitor the OER intermediate at the H2O/g-C3N4 interface during the reaction. Fluorinated g-C3N4 (F-CN) samples were prepared via hydrothermal treatment with varying concentrations of NaF. The morphological and structural characteristics of the samples were analyzed using techniques like XRD, TEM, and N2 adsorption-desorption. Photocatalytic water-splitting experiments were performed under white light and AM1.5G simulated solar irradiation to assess H2 and O2 evolution rates. Apparent quantum yield (AQY) measurements determined the efficiency of light utilization. Transient fluorescence decay spectroscopy and valence band XPS were used to investigate the exciton lifetime and valence band position. Density functional theory (DFT) calculations were performed to simulate the OER pathway on the CN and F-CN surfaces and analyze the impact of intermediate OER configurations. Electrochemical OER and HER measurements were conducted to assess the effect of fluorination on the half-reactions. Hydrogen peroxide (H2O2) production was quantified using a Ce⁴⁺ back titration method.

In situ DRIFTS and NAP-XPS revealed the accumulation of C=O bonding at the H2O/g-C3N4 interface during OER. This accumulation was identified as the primary bottleneck preventing overall water splitting. Fluorination effectively suppressed C=O formation, leading to a dramatic increase in both H2 and O2 evolution rates. The H2 evolution rate of the optimal F-CN catalyst (F0.1-CN) was 14.21 times higher than pristine CN under white light illumination and 9.63 times higher under AM1.5G simulated solar irradiation. Continuous O2 evolution was also observed for F-CN catalysts, unlike pristine CN. The study demonstrated that fluorination's positive effect wasn't due to enhanced light absorption, exciton lifetime, or valence band position. DFT calculations indicated that fluorination optimizes the OER pathway by activating neighboring N atoms, lowering energy barriers, and avoiding strong C-O interactions. Electrochemical measurements confirmed the improved OER activity of F-CN compared to pristine CN. The H2/O2 ratio was higher than the stoichiometric 2:1 ratio, which is likely due to additional reduction of O2 into H2O2. The rate of H2O2 production was almost equivalent to the difference between H2 and O2 production. The optimal F0.1-CN catalyst exhibited significantly enhanced stability during the water splitting reaction compared to the pristine CN.

The findings directly address the research question by identifying C=O bond accumulation as the previously unknown bottleneck in overall water splitting for single-phased g-C3N4. The significant improvement in water-splitting efficiency achieved through surface fluorination demonstrates a viable strategy for activating this material for this application. The results highlight the importance of considering surface chemistry and reaction intermediates when designing photocatalysts for water splitting. The insights from DFT calculations offer a deeper understanding of the mechanism underlying the enhanced OER activity. This research contributes significantly to the field by providing a solution to a long-standing challenge in g-C3N4-based photocatalysis.

This study successfully identified the accumulation of C=O bonds as the primary bottleneck hindering overall water splitting in single-phased g-C3N4 photocatalysts. A simple fluorination strategy effectively addresses this bottleneck by minimizing C=O formation and enhancing OER activity. This work demonstrates an order-of-magnitude improvement in overall water-splitting efficiency, paving the way for the development of efficient and cost-effective hydrogen production technologies. Future research could explore other surface modification strategies, investigate different fluorination methods, and optimize the catalyst structure for even higher performance and stability.

While the study demonstrates significant improvements in water-splitting activity through fluorination, the overall solar-to-hydrogen efficiency remains relatively low (0.00195%). Further optimization of the catalyst synthesis and reaction conditions is necessary to enhance efficiency. The study primarily focused on a specific fluorination method; exploring other surface modification techniques could reveal additional avenues for improving performance. The long-term stability of the F-CN catalyst under continuous operation requires further evaluation.