Photocatalytic hydrogen production depends on more than a semiconductor’s ability to absorb light. The material must also separate photo-generated electrons and holes, direct electrons to hydrogen-forming sites and remain active during the reaction. Studies of carbon nitride (C3N4) combined with metal sulfides AgIn5S8 and MoS2 illustrate how catalyst structure and component placement can influence these processes.
For engineers and technical decision-makers assessing solar-driven hydrogen technologies, the central lesson is that a composite’s composition alone does not explain its performance. The quality of contact between its phases, the location of the cocatalyst and the reaction medium all affect how the system behaves.
Photocatalytic hydrogen production and charge management
When a semiconductor absorbs light with sufficient energy, electrons move from the valence band to the conduction band, leaving positively charged holes behind. These charge carriers can participate in surface reactions, but they can also recombine before reaching an active site. Recombination releases absorbed energy as heat or light rather than contributing to hydrogen formation.
A photocatalyst therefore needs suitable light absorption and band positions, along with effective charge separation and transfer. In the systems discussed here, carbon nitride is paired with a metal sulfide to form a semiconductor heterojunction. The intended role of the interface is to provide a pathway for charge transfer that limits recombination and supports hydrogen evolution.
The experiments used platinum as a cocatalyst and sacrificial reagents to consume photogenerated holes. This is an important distinction: hydrogen evolution measured with a sacrificial reagent is not the same as overall water splitting, which requires both hydrogen and oxygen production.
AgIn5S8/C3N4: why platinum location matters
In tests with carbon nitride and platinum, the initial hydrogen formation rate was high when triethanolamine (TEOA) served as the sacrificial reagent, but the rate declined during the reaction. Characterization indicated that platinum nanoparticles became incorporated into the carbon nitride bulk, reducing their accessibility for proton reduction. When oxalic acid was used instead, this behavior was not observed; platinum nanoparticles were enriched at the outer surface of carbon nitride.
The location of platinum also differed in AgIn5S8/C3N4 composites. Platinum preferentially deposited on the AgIn5S8 phase, which helped keep the particles from becoming incorporated into the carbon nitride and left them available for hydrogen evolution. The comparison shows why catalyst preparation and reaction conditions should be considered together: they can affect where a cocatalyst ends up, not just how much is present.
In situ electron paramagnetic resonance (EPR) and photoluminescence (PL) measurements were used to examine charge behavior. As the AgIn5S8 content increased, the EPR signal associated with electrons decreased. PL emission intensity also declined as the composite’s AgIn5S8 content increased. These observations were consistent with electron transfer from carbon nitride to AgIn5S8. With platinum present, the reduced PL emission was also associated with transfer of excited electrons toward platinum, either directly or through the sulfide phase.
MoS2/C3N4: the interface depends on synthesis
MoS2/C3N4 composites were prepared using three approaches: in situ photodeposition, a sonochemical method and thermal decomposition. The 2H-phase MoS2 composite made by photodeposition produced the highest hydrogen evolution rate among the materials tested under UV-visible irradiation.
Microscopy and optical characterization helped explain the difference. In less active samples, larger MoS2 particles partly covered the carbon nitride. In the more active 2H MoS2 photodeposited composite, a sufficiently thin crystalline MoS2 layer formed intimate contact with carbon nitride. That interface was associated with more favorable charge separation and higher photocatalytic activity.
Unlike the AgIn5S8 composites, platinum was observed on both carbon nitride and MoS2 in the MoS2-based materials. In situ EPR measurements showed a lower conduction-band electron signal for the platinum-loaded 2H MoS2 photodeposited composite than for the other catalysts. This was interpreted as evidence of more efficient electron transfer from carbon nitride toward platinum. Together, the findings indicate that synthesis route, sulfide phase and interface morphology can matter as much as the identity of the sulfide.
Sacrificial reagents are part of the system
The choice of sacrificial reagent affects how photogenerated holes are consumed and can therefore influence hydrogen production. Comparing reagents with the same semiconductor and cocatalyst offers a degree of consistency, but the reagent’s oxidation potential is not the only variable. Solubility, degradation pathways, solution pH and the escape of solvent from the reaction system may also influence results.
Consequently, a hydrogen production rate should be interpreted in the context of the complete test conditions. Comparisons are more meaningful when catalyst composition, cocatalyst loading, illumination and sacrificial reagent are all accounted for. The reported work also cautions that the effects of different reagents require further investigation before broad conclusions can be drawn.
Practical implications for catalyst development
These results point to several considerations for designing and evaluating semiconductor photocatalysts:
- Engineer the interface: intimate contact between carbon nitride and a sulfide phase can support charge transfer.
- Control cocatalyst placement: accessible platinum sites are important for proton reduction, and their location can vary with the catalyst and reaction conditions.
- Connect characterization to activity: microscopy, EPR and PL can help relate morphology and charge behavior to hydrogen evolution.
- Report the reaction environment: sacrificial reagent choice affects interpretation and comparability of results.
Photocatalytic hydrogen production remains a materials-design challenge in which optical absorption, charge transfer, surface accessibility and reaction chemistry interact. The AgIn5S8 and MoS2 examples show that the best-performing composite is not defined simply by combining two semiconductors. Its interface must be formed effectively, its cocatalyst must remain accessible, and its performance must be assessed within the conditions of the reaction. These principles provide a practical framework for evaluating candidate photocatalysts without treating laboratory hydrogen evolution as equivalent to complete water splitting.