The present manuscript presents an experimental investigation on reduced ("black") TiO2 nanomaterials for hydrogen production in photocatalytic experiments. The black TiO2 nanotubes are synthesized by anodization followed by different reducing thermal treatments to probe the specific annealing effect. The materials are characterized by SEM, TEM, XRD, XPS, ESR (or EPR), NMR, UV-vis spectroscopy, and two-point conductivity measurements. The open-circuit photocatalytic hydrogen generation is tested in water/methanol mixture (50/50 vol%) under AM 1.5G illumination.

The paper focuses on photocatalytic hydrogen production with semiconductor materials, which has attracted an enormous interest by researchers worldwide. The manuscript, presented in the form of a letter, sounds interesting starting from the title, is well written, and illustrates the results in a synthetic but accurate way. A clear comparison between this work and the one by Chen and Mao (Science 2011, ref. 1) is made to illustrate the possibility of producing hydrogen in photocatalytic experiments with TiO2 without the need of adding a co-catalyst (i.e. Pt nanoparticles), as done in ref. 1 and in many other works (relevant references are adequately reported). In this way, the authors give useful insights about this gap of knowledge in the state of the art, also disclosing the importance of the TiO2 crystalline phase. It is also interesting to note that the open-circuit photoactivity obtained with the high-pressure treatment is not strictly related to a better efficiency in photoelectrochemical tests. In fact, in this case a standard Ar/H2 reduction treatment leads to better results, ascribed to a higher conductivity of the film. Overall, only some minor flaws affect the manuscript, as reported below.

1. The illumination conditions for IMPS (Fig. 4b) are not completely elucidated.
2. The table in Fig. 4a presents a mistake (the resistivity of Ar/H2-treated nanotubes should be lower than that of high-P H2-annealed nanotubes, as reported in the text).
3. Hydrogen production tests have been performed in water/methanol (50/50 vol) mixture, i.e. in the presence of a hole scavenger. Would the present results still be valid without methanol?
4. The figures in the manuscript could have been better refined from a graphical or stylistic point of view (e.g. comma as numeric separator instead of the dot, small fonts). Especially Figure 4 does not appear well-finished.