Origin of photoactivity in graphitic carbon nitride and strategies for enhancement of photocatalytic efficiency: insights from first-principles computations

The origin of the photoactivity in graphitic carbon nitride (g-C3N4) and the strategies for improving its photocatalytic efficiency were systematically investigated using first-principles computations. We found that g-C3N4 composed of tri-s-triazine units (g-CN1) is preferable in photocatalysis, owing to its visible-light absorption and appropriate band edge potentials. Despite the benefit of nanocrystallization of g-CN1, excessively minimized and passivated g-CN1 nanosheets (g-CN1NSs) should be inhibited, due to the intensely broadened band gaps in these structures. C- or N-vacancies in g-CN1NSs lead to gap states and smaller band widths, which should also be restrained. Compared with C substitution in B doped g-CN1NSs, N-substitution is favourable for enhancing the photoactivity of g-CN1NSs, due to the red-shift light absorption and the absence of gap states within this structure. Both WTe2 coupled and CdSe cluster loaded g-CN1NSs have decreased band gaps and directly separated carriers, which are beneficial to promote the photoactivity of g-CN1NSs. Among these modified g-CN1NS photocatalysts, WTe2 coupled g-CN1NSs are more preferable, as a result of their smaller band gap, free gap states and more rapid migration of excitons.