Brookite vs. rutile vs. anatase: What`s behind their various photocatalytic activities?

The results of investigating the morphological, optical and electronic properties of three synthesized TiO2 polymorphs (anatase (A and A2), brookite (B and B2) and rutile (R)) show that the origin of their different catalytic efficiency in the heterogeneous photocatalytic process lies in the depths of the charge carrier traps. The presence of shallow electron traps in brookite extends the number and “lifetime” of generated holes, so brookite is a good choice as a photocatalyst in such heterogeneous photocatalytic systems where these two variables are the reaction limiting factors. Low generation and high recombination rate of charge carriers in rutile are the consequence of too deep electron traps, therefore the generated electrons are not able to participate in reactive oxygen species generation reactions on the catalyst surface. The longer “lifetime” of charge carriers generated in anatase is due to the fact that anatase belongs to the indirect band gap semiconductors. It was further revealed that the increase of the anatase crystallite size (from 10 to 20 nm) can compensate the negative effect of decreasing specific surface area (from 129.5 to 65.0 m2/g) on the photocatalytic activity of anatase. The negative effect of decreasing specific surface area was very well expressed in the case of brookite, where the photocatalytic activity dropped with decreasing specific surface area (from 17.2 to 3.0 m2/g) while keeping constant brookite crystallite size (≈ 80 nm). Heterogeneous photocatalytic experiments with the model pollutant bisphenol A show that the deposition of bisphenol A degradation products on the catalyst surface can have a detrimental effect on photocatalytic activity of low specific surface area TiO2 polymorphs. The specifics (electronic properties, geometry of particles and photocatalytic activity of each TiO2 polymorph) of the investigated heterogeneous photocatalytic system for reactive oxygen species generation are the reason that the “synergistic effect” between the anatase and rutile phase when used as a physical mixture, regardless of the ratio between them, was not expressed. The anatase phase was shadowed by small and abundant rutile nanoparticles with low photocatalytic activity, which acted also as a harvester of UV-light due to a wider UV-light absorption range. We can conclude that the less studied TiO2 polymorph brookite is worthy of investigation, as the results of the present study show that the photocatalytic activity of brookite with the adequate specific surface area is comparable to the photocatalytic activity of the widely investigated anatase TiO2. In addition, it can be concluded that in the physical mixtures of different TiO2 polymorphs, the use of TiO2 particles of close diameters is of crucial importance to achieve a “synergistic” photocatalytic effect triggered by particle collisions.