**4. Conclusions**

Au NPs do not seem to work directly, but we can assess that their presence enhance the ability of the sample to transfer charge with the solvent, and they slow down the decay process of the photocurrent, at least up to a certain load.

If the behavior of the samples is considered in detail, the most performing samples, in both the supporting electrolyte and in the presence of BPA, were the Au/TiO2/Au and TiO2/Au samples. On these samples the redox behavior of the M is well evident in the CV, recombination seems slowed down, the current decay is 3–5 times slower than that at the reference sample. On both samples, NPs of great dimensions are present at the bottom under the TiO2 layer, directly contacting the Ti support. The presence of these NPs at the bottom seems to be the crucial point for the working mechanism of the structure. Thanks to the scattering effects, it is actually possible on these samples to better exploit the radiation, and thus the active sites of the TiO2 layer.

Measurements for the Au/TiO2 and Au/TiO2/Au samples can be useful to interpret the role of the top surface layer of NPs. Considering the smaller dimensions of these NPs, they should be responsible both for possible plasmonic effects and for electron–hole separation, thus leading to reduced recombination. Due to plasmonic effects, localized energetic fields could generate e- /h+ couples: however, the low stability evidenced for the Au/TiO2 sample indicates that holes cannot be transferred neither to the electrolyte, nor to the SC, so that they remain in the NPs and oxidize them. However, corrosion seems to be correlated not only to plasmonic effects, but also to the distribution of the NPs. Of note is that the most important corrosion effects are measured, rather than for the TiO2–Au system, where the plasmonic effects were the most relevant, for the Au/TiO2/Au and Au/TiO2 samples, where an upper Au NPs layer is at direct contact with the electrolyte. Maybe the NPs in TiO2–Au are somehow protected, as they are embedded in the structure. Other works in the literature on similar SC/M coupling, suggested that a protecting layer could be realized in order to avoid corrosion [51].

Finally, the scarce performance of the TiO2–Au sample could be attributed mainly to a decreased surface area available for the charge transfer with the electrolyte (see [23]). The large content of Au NPs did not result in enhancement of the global performance of the structure. The photocurrent is lower, because the area is lowered. The low value of τ, and the rapid current decay, could indicate that most of the NPs are behaving as recombination centers for the photogenerated charges.

**Author Contributions:** Conceptualization, S.P. and A.V.; methodology, A.V.; investigation, L.M., E.M.U. and R.M.; writing—original draft preparation, S.P.; writing—review and editing, E.M.U., L.M., M.G., L.M., A.L.B., R.M. and I.N.; sample preparation, structural/optical characterization: M.G., L.M., B.R.B., V.R., C.S.C. and A.L.B.; supervision, A.V. and A.V.; Funding Acquisition, S.P. All the coauthors contributed to the data discussion.

**Funding:** This research was funded by Fondazione di Sardegna, CRP project F71I17000280002 – 2017.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
