Carrier dynamics govern photovoltaic-photocatalytic functionality in graphitic carbon nitride/transition metal dichalcogenide heterostructures.

The rapid growth of global energy demand and excessive fossil-fuel consumption have intensified the urgent need for efficient solar-energy-conversion materials. However, the vast configurational space of heterostructures makes reliable screening particularly challenging, especially when attempting to distinguish whether a given system functions as a photocatalyst or a photovoltaic material. Current design strategies predominantly rely on static electronic-structure descriptors, such as band alignment and interfacial charge transfer, which capture only equilibrium energetics and fail to reflect the nonequilibrium evolution of photogenerated carriers. Here, we combine first-principles calculations with nonadiabatic molecular dynamics (NAMD) simulations to establish a static-to-dynamic evaluation strategy for graphitic carbon nitride/transition metal dichalcogenide (g-C3N4@TMD) heterostructures constructed from experimentally realized monolayers. We demonstrate that static descriptors alone are insufficient to determine functionality, whereas carrier dynamics governs the functional behavior of these systems. Notably, g-C3N4@HfS2 and g-C3N4@TiS2 satisfy conventional Z-scheme criteria based on band alignment. However, NAMD simulations reveal that interlayer carrier transfer dominates over selective recombination, precluding the formation of an effective Z-scheme pathway. In contrast, g-C3N4@MoS2 and g-C3N4@WS2 exhibit ultrafast interlayer carrier transfer and strongly suppressed electron-hole recombination. Quantitatively, recombination times are approximately two orders of magnitude longer than the corresponding transfer times for g-C3N4@MoS2, indicating pronounced kinetic asymmetry that favors long-lived charge separation and further supports its photovoltaic potential, with an idealized Scharber-model PCE estimate approaching 16%. These findings demonstrate that the kinetic competition between carrier transfer and recombination provides a critical dynamic criterion for evaluating and refining the functional classification inferred from static descriptors. More broadly, this work provides a general strategy for integrating static and dynamic descriptors into the evaluation of low-dimensional energy materials, offering new insights into interfacial charge-transfer mechanisms and guiding the rational design of next-generation solar energy conversion systems.