84 | December 2021 www.facs.website limited. Stability of molecular catalysts tends to become lower under the strong light illumination so that attention should also be paid to the stable ligand frameworks. In addition, molecular motifs having pyridyl anchor(s) showing relatively low solubility in water must be explored as a desirable target to be tightly anchored over the mesoporous TiO2 cathode surfaces. After our efforts, a platinum porphyrin having a single pyridyl anchor (PtP-py) was found to fulfill all these requirements (Figure 6).43 In the study, [Ru(dpbpy)2(qpy)] 2+ (Ru-dpqpy; dpbpy = 4,4′-diphenyl-2,2′-bipyridine), possessing a higher hydrophobicity with stabler adsorption characteristics, was also used to improve the photoanode performance. The TiO2-based photoanode (FTO/TiO2/Ru-dpqpy) and the TiO2-based cathode (FTO/TiO2/PtP-py) were prepared by submersing the pristine FTO/TiO2 electrodes into the solutions of Ru-dpqpy and PtP-py, respectively. Based on the absorbance change in each solution, the amounts of Ru-dpqpy and PtP-py adsorbed over the individual FTO/TiO2 electrode were estimated to be 0.12 and 0.10 μmol/cm2, respectively.44 The detailed studies using the FTO/TiO2/ PtP-py cathode unveiled its extremely small onset overpotential for HER, which is even smaller than 50 meV. Although it still adopts a precious element, the single-atom nature per catalyst clearly achieves drastically higher cost effectiveness. Thus, at the PtP-py-modified Figure 4. Schematic representations of possible binding modes for carboxylate (a), phosphonate (b), and pyridyl (c) anchors over the TiO2 surfaces. Figure 5. The FTO/TiO2/Ru-dmqpy photoanode and a platinum electrode connected with a simple conductive wire, exhibiting its high stability and effectiveness in H2 evolution at the dark cathode.42 TiO2 electrode, H2 production proceeds spontaneously without the need for any additional external bias, in the same manner as observed when platinum was adopted at the cathode.42 It must be noted here that the conduction band (CB) edge potential (i.e., flatband potential; EFB = -0.40 -0.059pH V vs. SCE45) possesses a driving force for H2 production larger than 50 meV (ca. 160 meV), which is somehow closely correlated with the driving force for the MV+•- driven HER at pH=5.0 (see above). We now postulate that the PtP-py anchored over the TiO2 surfaces cannot take the advantage of the above-mentioned dimerization pathway in order to lower the activation barrier for the often rate-limiting hydride formation process. Actually, the ideal p-p stacking distances of aromatic systems are ca. 3.4 Å, which clearly causes steric blockage to have a sufficiently strong Pt-Pt association (e.g., 2.8-3.2 Å) between the PtP-py units. An important insight gained in our recent study on the photocatalytic CO2 reduction by water-soluble cobalt porphyrins46 seems relevant to the reason why the filled Pt dz2 electron pair in PtP-py can raise its basicity without the aid of metal-metal association. Our DFT-based mechanistic study on the cobalt porphyrins unveiled that the filled dz2 orbital gradually increases its basicity upon successive porphyrin-based reduction processes.46 The injection of electrons into the vacant p* orbitals causes substantial congestion in the electron density surrounding the metal d orbitals, leading to cause the destabilization in some of the filled d orbitals. We thus speculate that the relatively low overpotential achieved by PtP-py is induced by the porphyrin-based reduction processes. The detailed study is now in progress. Why electrons flow between the two electrodes with bias-free conditions? To clarify the operation mechanism of our molecular-based PECs, linear potential sweep was made using a two-electrode conf iguration PEC made up of the FTO/ TiO2/Ru-dpqpy photoanode and FTO/TiO2/ PtP-py cathode.44 Using this setup, the photoanode potential was scanned versus the cathode potential with the reference terminal short connected to the cathode. The measurement was also combined with the light-on and -off switching cycles (Figure 7a). Thus, the potential axis has a description of V vs. cathode. The measurements were carried out using an acetate buffer solution (0.1 M, pH 5.0) containing Donor (EDTA). Control experiments were also carried out by suppressing either Ru-dpqpy or PtP-py in electrodes. One of the most remarkable results is that the photocurrent density reaches ca. 0.4 mA/cm2 even at 0 V vs. cathode, indicating that a sufficient current flows even with this bias-free condition (Figure 7a).44 In addition, upon holding the anode potential at 0 V for 1 h with the light-on condition, ca. 4.2 μmol of H2 evolved at the dark cathode with a near quantitative Faradaic efficiency. The results clearly indicated that electrons injected into the CB of TiO2 at the photoanode flow over to the cathode even without applying any external bias. Indeed, even by the lack of PtP-py in the cathode, the FTO/TiO2 cathode shows a color change into blue due to the substantial charge accumulation at the CB (Figure 7b,c). It was also confirmed that the electrons reaching the cathode can effectively drive the HER catalyzed by PtP-py with vigorous evolution of bubbles (Figure 7d,e). By disconnecting the electrochemical analyzer from the PEC, we further ascertained that a relatively small photoinduced potential shift is given between the two electrodes. The observed shift is rather small (ca. 20 μV) but certainly caused the electromotive force (EMF) required to transfer electrons from the photoanode to the cathode. The origin of EMF was further investigated by observing the charge accumulation into the CB at the photoanode by simply illuminating the FTO/ TiO2/Ru-dpqpy electrode soaked in the solution containing Donor (EDTA) (Figure 7f). The in-situ absorption spectroscopy clearly evidenced the growth of a broad visible to NIR band ascribable to the charge accumulation at the CB of TiO2 (Figure 7f). The electron
RkJQdWJsaXNoZXIy NDU2MA==