The investigation into polymer semiconductor photoelectrodes for light-driven water splitting

Photoelectrochemical (PEC) photosynthesis has been regarded as a promising method to convert inexhaustible solar energy to fuels, in which polymer semiconductors have been identified to be promising photocatalysts because they are abundant, environmentally benign, and has easily tuneable band structure. Graphitic carbon nitride (g-C3N4) as one of the most promising polymer photocatalysts has realized efficient water splitting in a suspension system. However, its PEC application is limited by its low efficiency due to both unclear understanding of reaction kinetics in the film and low quality g-C3N4 film fabricated. Herein, a novel film fabrication method has been developed to prepare dense, uniform and highly crystallized g-C3N4 films as a photoanode with a controllable thickness. Comparing with g-C3N4 films prepared by other methods, the relationship between crystallinity, deep trap states and PEC performance of g-C3N4 photoelectrodes was investigated. I found that longer-lived charge carriers were present in more poorly crystalline samples, due to more deeply trapped states, which inversely correlated with photoelectrochemical performance. The electron diffusion length in such a crystalline g-C3N4 film was determined to be ca. 1000 nm, which is the first report for this promising photoelectrode. Although g-C3N4 has a long electron diffusion length, its PEC performance was still minimal due to the low charge carrier density and severe charge recombination. A novel one-step construction approach has then been developed to solve this problem, by synthesising a nanojunction metal-free photoanode, composed of B-doped g-C3N4 nanolayer on the surface and bulk carbon nitride. This type of nanojunction overcame a few intrinsic drawbacks of g-C3N4 film, e.g. severe charge recombination and slow charge transfer. For the optimum sample, the top layer of the nanojunction has a depth of ca. 100 nm and the bottom layer is ca. 900 nm. This nanojunction photoanode resulted into a 10 fold higher photocurrent than bulk g-C3N4 photoanode with a record photocurrent density of 103.2 μA/cm2 at 1.23 V vs RHE under one sun condition and a high incident photon-to-current efficiency (IPCE) of ca. 10% at 400 nm. The EIS, MS and IMPS spectroscopies proved such enhancement was mainly due to more than 10 times faster charge transfer rate at electrode/electrolyte interface and nearly 3 times higher conductivity due to the nanojunction structure. Based on the progress in g-C3N4 photoanode, a more challenging topic was explored which is to change g-C3N4 semiconductor from a photoanode to a photocathode. Developing g-C3N4 photocathode is of great importance because its negative conduction band position is favorable for water reduction, which has been widely proved in a suspension system. However, the intrinsic property of the g-C3N4 film as an n-type semiconductor limits hydrogen generation at the electrode/electrolyte interface as n-type semiconductors exhibit upward band bending and holes accumulate on the surface. In this thesis, it is demonstrated that surface trap states could effectively contribute to the photocathodic performance of a g-C3N4 film. Introducing nitrogen defects and C-OH terminal groups in the structure of g-C3N4 created a large portion of shallow trap states with 1000 times extended lifetime that could trap electrons for the water reduction reaction. This was further validated by hot H2O2 treatment that could transfer the g-C3N4 photoanode to a photocathode, confirming that shallow trap states are the critical reason for synthesising a photocathode from an n-type polymer semiconductor. Overall this thesis provides an effective strategy for g-C3N4 polymer to be efficient photoanode and photocathode, forming strong basis for its application in solar to H2 fuel production.