Photoelectrochemical reaction is emerging as a powerful approach for biomass conversion. However, it has been rarely explored for glucose conversion into value-added chemicals. Here we develop a photoelectrochemical approach for selective oxidation of glucose to high value-added glucaric acid by using single-atom Pt anchored on defective TiO2 nanorod arrays as photoanode. The defective structure induced by the oxygen vacancies can modulate the charge carrier dynamics and band structure, simultaneously. With optimized oxygen vacancies, the defective TiO2 photoanode shows greatly improved charge separation and significantly enhanced selectivity and yield of C6 products. By decorating single-atom Pt on the defective TiO2 photoanode, selective oxidation of glucose to glucaric acid can be achieved. In this work, defective TiO2 with single-atom Pt achieves a photocurrent density of 1.91 mA cm−2 for glucose oxidation at 0.6 V versus reversible hydrogen electrode, leading to an 84.3 % yield of glucaric acid under simulated sunlight irradiation.
Biomass conversion into building-block chemicals coupled with clean fuel generation using renewable energy is a promising strategy to alleviate the dependence on fossil fuels1. As one of the most important platform molecules of biomass, glucose has triggered growing attention due to its various value-added products such as glucaric acid (GLA), gluconic acid (GLU), sorbitol, and 5-hydroxymethylfurfural2. GLA as a building block chemical in a bio-based economy has been identified as one of the 12 top value-added chemicals by the US Department of Energy in 20043. GLA and its derivatives are widely used for the production of a variety of commodity products for healthcare including cholesterol reduction and cancer chemotherapy4. The high functionalization of GLA leads to high economic value and huge market demand, which is predicted to be USD 1.30 billion by 2025 million according to the report of Grand View Research, Inc2. However, the productions of GLA by current methods are insufficient to support the large market demand. Currently, the productions of GLA are obtained either by microbial fermentation or chemical oxidation5,6. The chemical oxidation is usually realized under harsh conditions, such as with HNO3 or oxidant, at high temperature, or under high pressure. Various by-products are formed during the chemical oxidation, and expensive separation for the extraction of GLA is required. Besides, selective catalytic conversion of glucose to GLA has also been achieved by electrochemical or thermal oxidation on noble metal catalysts (e.g., Au, Ag, Pd, and Pt) with additional oxidants, which are still cost-intensive and with limited efficiency for GLA production7.
Photoelectrochemical (PEC) oxidation provides another promising strategy for GLA production, which can be operated under mild conditions by eliminating the use of hazardous chemical oxidants or high-pressure O28. Besides, PEC cell is the way to drive reactions through the photo-induced electrons and holes directly from the sunlight with a very low applied potential, which shows good potential for energy saving and environmental protection9,10,11. Therefore, PEC GLA production is proposed to be a green strategy to produce value-added chemicals with clean fuel by combining the green energy source with renewable feedstock12. Although the noble metal-based catalysts have shown high selectivity towards GLA production during the anodic oxidation of glucose, the high cost restrains the actual application3,13,14,15. Single-atom catalysts with well-defined active centers and maximum atom utilization can greatly reduce the amount of noble metals, thus significantly reducing the cost16,17. Therefore, single-atom noble metals modified photoanodes are expected to achieve PEC glucose oxidation with high selectivity and yield.
In this work, selective oxidation of glucose to GLA is realized on the photoanode of defective TiO2 decorated with single-atom Pt via a PEC strategy (Fig. 1). The defective TiO2 is fabricated by electrochemical reduction of TiO2 nanorod arrays, and further deposition of single-atom Pt is achieved by atomic layer deposition. The disordered structure in the defective TiO2 has been verified to promote charge separation and transportation, along with the modulation of the oxidation products (Fig. 1). The single-atom Pt on defective TiO2 has been evidenced to regulate the selectivity of the GLA by accelerating the GLU oxidation. With the as-prepared photoanodes, a high yield of 84.3% for GLA is achieved at 0.6 VRHE under simulated sunlight illumination.
Results and discussion
Synthesis and characterizations of nanorod arrays with Pt atoms
TiO2 nanorod arrays (NRAs) were directly grown on fluorine-doped tin oxide (FTO) substrates via a typical hydrothermal method followed by annealing in air (Fig. 2a)18. Scanning electron microscopy (SEM) was first employed to characterize the morphology and structure of the sample. As illustrated in Fig. 2b, c, the obtained film on the FTO substrate is composed of NRAs, which are vertically and evenly distributed on the FTO substrate. The length of the nanorod is about 1.3 μm. The statistically determined diameter distribution (Inset of Fig. 2c) shows that the average diameter of the nanorod is 120 nm with peaks in the range of 90–150 nm. The x-ray diffraction (XRD) pattern indicates that all the diffraction peaks of (002), (211), and (101) from the NRAs are following the rutile phase (JCPDS No. 21-1276), confirming the formation of NRAs in the rutile TiO2 crystal structure (Fig. 2d).
The defective TiO2 (def-TiO2) NRAs were fabricated by electrochemical reduction of TiO2 NRAs in 0.5 M Na2SO4 electrolyte at −1.4 VRHE for 10 s. The single-atom Pt decorated def-TiO2 (Pt/def-TiO2) was obtained by atomic layer deposition (ALD) deposition of Pt single atoms (SAs) on def-TiO2 at 200 °C for 10 cycles (Fig. 2a). The reduction treatment did not induce any phase transition in def-TiO2 as suggested by the consistent XRD patterns of TiO2 and def-TiO2 (Fig. 2d). No additional diffraction peaks are observed in Pt/def-TiO2, indicating that no obvious clusters of Pt metal and compounds are formed on def-TiO2 NRAs. As shown in the diffuse reflectance UV–vis spectra (Fig. 2e), the TiO2 NRAs present a typical rutile TiO2 absorption with the edge at 420 nm, and its bandgap is determined to be 2.99 eV by the Tauc plots (Supplementary Fig. 1a). A slight red-shift is observed in the def-TiO2 absorption spectrum, while the light absorption of def-TiO2 is dramatically increased in the visible region, corresponding to a narrower bandgap of 2.70 eV (Supplementary Fig. 1a). This smaller bandgap can be attributed to the additional transitions caused by increased Ti3+ states, oxygen vacancies, hydroxyl group, and other defects at different energy levels19,20. Further deposition of Pt atoms does not induce any obvious changes in the absorption for def-TiO2, due to the small loading or the limited size of Pt.
The influences of reduction treatment on chemical states were characterized by X-ray photoelectron spectroscopy (XPS) (Fig. 2f). Only two Ti 2p peaks centered at 464.3 and 458.6 eV with a splitting of 5.7 eV are observed in TiO2, which are the typical Ti(IV) 2p1/2 and Ti(IV) 2p3/2 peaks in rutile TiO2. After electrochemical reduction, additional two peaks appear beside the Ti(IV) 2p peaks, which are located at 462.9 and 457.2 eV with a splitting of 5.7 eV. These two peaks are thus attributed to Ti(III) 2p1/2 and Ti(III) 2p3/2. The emerging Ti3+ signals verified that substantial oxygen vacancies are introduced into def-TiO2 during the reduction process19,21,22,23. Additionally, the electron spin resonance (ESR) spectrum of def-TiO2 shows a strong signal of trapped electrons in oxygen vacancies (g = 2.002) (Supplementary Fig. 1b)24, further confirming the introduction of oxygen vacancies by the reduction treatment. Numerous studies have revealed that the induced oxygen vacancies and disordered defects can induce an up-shift of the valance band, leading to a narrowed bandgap22,25,26. After deposition of Pt SAs, the Pt/def-TiO2 NRAs show slightly stronger Ti(III) 2p signals (Fig. 2f) and ESR intensity (Supplementary Fig. 1b) than def-TiO2, probably due to the electron transfer from the Pt atoms to the def-TiO2 support. This electron transfer between the def-TiO2 support and the Pt SAs can lead to a strong metal-support interaction27.
Characterizations of single atoms
Single-atom Pt was also deposited on TiO2 NRAs by ALD at 200 °C for 10 cycles (Pt/TiO2) for comparison. After deposition of Pt atoms, Pt/TiO2 shows no obvious change in the XRD pattern and UV-vis spectrum compared with TiO2 NRAs (Supplementary Figs. 2, 3). The transmission electron microscopy (TEM) revealed that the diameter of the individual Pt/TiO2 nanorod is about 120 nm (Supplementary Fig. 4a), consistent with the SEM results. High-resolution TEM (HRTEM) shows a single crystal structure of Pt/TiO2 with lattice spacings of 0.28 and 0.32 nm, corresponding to the d-spacing of rutile TiO2 in (101) and (110) planes, respectively (Supplementary Fig. 4b)28. Therefore, the axial direction of Pt/TiO2 or Pt/def-TiO2 nanorod is along the  direction (Fig. 3a). Following the XRD results, no nanoparticles are observed on the Pt/TiO2 nanorod in the HRTEM image, suggesting that the adsorbed Pt ions were not agglomerated during the ALD process. The inductively coupled plasma optical emission spectroscopy (ICP-OES) affirmed the existence of Pt ions in the Pt/TiO2 NRAs with a loading of 0.08 wt%. To have a direct view of the state of the Pt ions on the TiO2 nanorod, the high-angle annular dark-field (HADDF-STEM) was employed to characterize Pt/TiO2 (Supplementary Fig. 4c). The HADDF-STEM image shows that some isolated bright dots (indicated by the red dashed circles) can be distinguished from surrounding light atoms of the nanorod, indicating the existence of single-atom Pt. Furthermore, the STEM image and corresponding energy dispersive spectroscopy (EDS) mapping of the Pt/TiO2 nanorod demonstrate that the Pt atoms are evenly distributed over the whole nanorod (Supplementary Fig. 4d). After the reduction treatment, no obvious changes in the Pt/def-TiO2 NA can be observed by the low-resolution TEM (Fig. 3a). However, in the HRTEM image (Fig. 3b), the Pt/def-TiO2 NA shows a core-shell structure consisting of a crystalline core of rutile TiO2 with a disordered shell layer. This disordered shell layer can provide more active sites for the onset MeCpPtMe3 chemisorption, resulting in a huge increase in Pt loading. The loading is 0.7 wt% for Pt/def-TiO2 under the same ALD conditions as Pt/TiO2, and no agglomeration of adsorbed Pt ions is observed on the Pt/def-TiO2 nanorod (Fig. 3b). The HADDF-STEM image further reveals a large amount of isolated Pt atoms on the surface of def-TiO2 (Fig. 3c), suggesting that the disorder shell layer indeed provides sufficient active sites for single-atom Pt anchoring. The uniform distribution of Pt over the def-TiO2 nanorod is confirmed by the STEM image and corresponding EDS mapping (Fig. 3d).
As shown in Fig. 3e, the high-resolution Pt 4 f spectrum can be deconvoluted into four peaks, where the peaks centered at 73.87 and 70.54 eV with a splitting of 3.33 eV are attributed to the Pt(0) 4 f 7/2 and Pt(0) 4 f 5/2, and the peaks located at 75.33 and 72.0 eV with a splitting of 3.33 eV are corresponding to Pt(II) 4 f 7/2 and Pt(II) 4 f 5/2. This suggests that a rich chemical environment of Pt SAs with different oxidation and charge states coexists on the surface of defective TiO229. The existence of Pt(II) indicates that some Pt SAs are partially positively charged by electron transfer between the def-TiO2 support and the Pt SAs, which leads to an enhanced metal-support interaction27, consistent with the above Ti 2p XPS (Fig. 2f) and ESR (Supplementary Fig. 1b) results. The X-ray absorption near-edge structure (XANES) spectra were further performed to clarify the electronic structure. The Pt L3-edge XANES spectra show that the white-line intensity of Pt/def-TiO2 is lower than that of PtO2 and higher than that of Pt foil (Fig. 3f), further evidencing the partial oxidation state of the single-atom Pt27,30,31,32,33. Fig. 3g shows the Fourier transformed extended X-ray absorption fine structure (FT-EXAFS) spectra of Pt/def-TiO2 and references. Only one prominent peak at 1.6 Å is detected in Pt/def-TiO2, attributed to the first shell of Pt-O scattering. Furthermore, compared with the FT-EXAFS spectrum of Pt foil reference, no higher shell of Pt-Pt scattering is observed in Pt/def-TiO2, suggesting atomically dispersed Pt solely existing in Pt/def-TiO2.
In addition, EXAFS wavelet-transform (WT) analysis that can strongly distinguish the backscattering atoms was performed to further consolidate the above inference. As shown in Fig. 3h, the WT contour plots of Pt/def-TiO2 and Pt foil reference at the Pt edge were put together for comparison. In the WT contour plots of the PtO2 reference (Supplementary Fig. 5) and Pt foil reference (Fig. 3h), the first coordination shells show the intensity maximums at ∼4 and ∼8 Å−1, ascribed to the Pt-O and Pt-Pt contributions, respectively. As illustrated in the WT contour plots of Pt/def-TiO2 (Fig. 3h), only one Pt-O contribution at ∼4 Å−1 appears and no Pt-Pt contribution at ∼8 Å−1 is observed, further evidencing the atomically dispersed Pt atoms. Taken together with the abovementioned results, the Pt atoms are verified to be atomically dispersed in Pt/def-TiO2.
The PEC capabilities of the TiO2 photoanodes with different reduction times and ALD cycles for glucose oxidation were first evaluated through the linear sweep voltammetry (LSV) experiments in 1 M KOH with 10 mM glucose under simulated sunlight (Supplementary Figs. 6, 7). After optimization, the photocurrent densities for glucose (JG) and water oxidation (JW) of the TiO2, def-TiO2, and Pt/def-TiO2 photoanodes for glucose and water oxidation are shown in Fig. 4a. After reduction treatment, the def-TiO2 photoanode shows a low applied potential to reach the saturated photocurrent (Esat) for glucose oxidation (~0.4 VRHE) and a high JG of 1.86 mA cm−2 at 0.6 VRHE. While the TiO2 photoanode exhibits a much higher Esat for glucose oxidation (>0.8 VRHE) and a much lower JG (0.41 mA cm−2 at 0.6 VRHE). This implies that moderate oxygen vacancies can efficiently promote the charge dynamics for PEC glucose oxidation20,34. The further deposition of Pt SAs on def-TiO2 shows no effect on the Esat for glucose oxidation, but slightly enhances the JG of the Pt/def-TiO2 photoanode (1.91 mA cm−2 at 0.6 VRHE), demonstrating the contribution of single-atom Pt to glucose oxidation. Without the addition of glucose in the electrolyte, both the JGs from the def-TiO2 and Pt/def-TiO2 photoanodes were dramatically decreased, accompanied by obvious positive shifts of onset potentials and Esats (Fig. 4a). This is due to that the standard redox potential of glucose oxidation (0.05 V) is much lower than that of water oxidation (1.23 V)35,36. Importantly, the observed increments from Jw to JG in the def-TiO2 and Pt/def-TiO2 photoanodes are much higher than those in the TiO2 photoanode, suggesting that the reduction treatment can substantially suppress the water oxidation. The Ultraviolet Photoelectron Spectroscopy (UPS) results (Supplementary Fig. 8) exhibit a ~0.3 eV up-shit of the valance band maximum caused by the 10 s reduction treatment accounting for the suppressed water oxidation. Figure 4b shows the chopped photocurrent densities of the photoanodes at 0.6 VRHE with repeated on/off cycles correlated to the simulated solar light. All the photoanodes display stable photocurrents with fast and reproducible responses upon each cycle illumination.
To have a deep insight into the modulation of the defect structure on PEC glucose oxidation, the Mott−Schottky plots of the samples were collected in Fig. 4c and Supplementary Fig. 9. Compared to TiO2, the def-TiO2 and Pt/def-TiO2 photoanodes show greatly enhanced carrier densities and negative shifts of the flat potentials (Supplementary Fig. 9). As a result, the depletion regions (rd) of the def-TiO2 and Pt/def-TiO2 photoanodes are limited to 2.8–6.6 nm over the tested potentials (Supplementary Table 1). This limited rd will lead to a wider electron-conducting region for electron transportation20,26,37, which was further evaluated by intensity-modulated photocurrent spectroscopy (IMPS) (Supplementary Fig. 10)38,39,40,41. As a result, the Pt/def-TiO2 photoanode enables more efficient charge transport than the TiO2 photoanode, consistent with a shorter electron transit time (τd) (Fig. 4d). In the meantime, the Pt/def-TiO2 photoanode also shows an enhanced electron lifetime (τn), which was obtained from the small perturbation transient photocurrent measurements (Supplementary Fig. 11)42. The enhanced τn of the Pt/def-TiO2 photoanode could be mainly attributed to its gradually elevated valance band maximum in the reduction shell induced by the reduction treatment. The electron collection efficiency and (ηcol) diffusion length (Ln) can be further obtained based on the equations of Ln = L (τn/τt/2.35)1/2 and ηcol = 1 – τt/τn20,43. The Ln is thus determined to be around 1.0 μm for TiO2 photoanode in the potential range of 0.6–1.2 VRHE (Fig. 4e), which is slightly shorter than the NRA thickness (1.3 μm). For the Pt/def-TiO2 photoanode, Ln is increased to around 2.5 μm due to the significantly improved electron conductivity. The longer Ln of the Pt/def-TiO2 photoanode achieves an efficient ηcol around 90% in the potential range of 0.6–1.2 VRHE, while the smaller Ln of the TiO2 photoanode shows a ηcol less than 40% (Fig. 4f).
The products of the glucose oxidation from the Pt/def-TiO2 photoanode at 0.6 VRHE under AM 1.5 G, 100 mW cm−2 illumination were examined. The initial product was GLU, and it was then further oxidized into GLA (Fig. 4g). The glucose conversion (Xglucose) from the Pt/def-TiO2 photoanode was 98.8% after a 5.5 h reaction, with GLU and GLA yields (YGLU and YGLA) of 9.2 and 84.3%, respectively (Entry 7, Table 1), from which free-glucaric acid can be separated through ion exchange resin, separation by boronic acid affinity gel and azeotrope drying (Supplementary Fig. 12). Besides, the H2 evolution at both Pt/def-TiO2 photoanode and counter electrode was examined by online gas chromatography with Ar as carrier gas. The results show that H2 was undetectable at the Pt/def-TiO2 photoanode and only detected at the cathode electrode, meaning that the PEC oxidation of glucose on the photoanode could be a water-involved process. As shown in Supplementary Fig. 13, the faradaic efficiencies for H2 evolution during 5.5 h PEC glucose oxidation are >99%, indicating no side reaction occurred on the counter electrode. The faradaic efficiencies for GLA and GLU are 86.8 and 3.1%, respectively. Besides, the Xglucose and YGLA from the Pt/def-TiO2 photoanode at different applied potentials were also investigated (Fig. 4h). The Pt/def-TiO2 photoanode shows almost identical Xglucose and YGLA at different applied potentials, due to the limited depletion regions obtained from the Mott−Schottky results. The incident photon-to-GLA efficiency spectrum shows that the Pt/def-TiO2 photoanode can yield 70–80% photon-to-GLA conversion efficiencies in the UV region (Fig. 4i). Furthermore, the stability of the Pt/def-TiO2 photoanode was examined by 5 runs of recycled J-t test (27.5 h) of PEC glucose oxidation. As shown in Supplementary Fig. 14, the J-t plot of the Pt/def-TiO2 photoanode in the 1 first run shows that a stable photocurrent can be maintained for the first 3 h, while by prolonging the test to 5.5 h, a significant photocurrent attenuation was observed. This decreased photocurrent is mainly due to the depletion of glucose and GLU, which is verified by the recovered and reproduced photocurrents in the 2nd, 3rd, 4th, and 5th runs with the electrolyte refreshed every 5.5 h. Besides, after the 27.5 h test, no noticeable JG recession was observed, as presented in Supplementary Fig. 15a and b. These results suggest the good stability of the Pt/def-TiO2 photoanode. In addition, after the repeated PEC reaction, no obvious changes were observed in the Ti 2p and Pt 4 f XPS spectra of Pt/def-TiO2 photoanode (Supplementary Fig. 15c, d), further revealing the good stability of the Pt/def-TiO2 photoanode.
We also studied the influences of different photoanodes and different glucose concentrations (Cglucose) on the oxidation products (Entry 1–11, Table 1). The productions of GLU and CO2 with YGLU and YCO2 of ~29.7 and 24.2%, respectively, were observed by using the TiO2 photoanode, and no GLA was detected. (Entry 1, Table 1). The CO2 production is due to that the energy of photo-induced holes is too high for glucose oxidation, leading to cleavage of C-C bonds by most holes, which is supported by the diverse C1-C5 products detected by using TiO2 as a photoanode in previous reports44. The product distributions of glucose oxidation over the TiO2 photoanode from Entry 1, Table 1 demonstrate that various C1-C6 products were produced with only ~30% C6 products (Supplementary Fig. 16), further confirming the role of high-energy holes in the cleavage of C-C bonds. Since the GLU is the initial product for GLA generation, such a low concentration of GLU maybe not be sufficient for the production of GLA on the TiO2 photoanode. The deposition of Pt SAs on TiO2 didn’t significantly enhance the YGLU+GLA, however, the GLA emerged with a yield of 26% due to the selectivity of Pt SAs to GLA (Entry 2, Table 1). Besides, the YCO2 on the Pt/TiO2 photoanode is slightly decreased, probably due to that the promotion of Pt SAs to GLA leads to fewer holes involved in the cleavage of C-C bonds. With the reduction time increased from 2 s to 10 s, the defective TiO2 photoanodes show increased JGs and YGLU+GLA (Entry 3–5, Table 1). The increased JGs mainly result from the improved charge dynamics by the oxygen vacancies and the enhanced YGLU+GLA is due to the decreased energy of photo-induced holes by the elevated valance band maximum (Supplementary Figs. 17, 18). A further careful study of reduction degree by the electron energy-loss spectroscopy shows that the def-TiO2 nanorod possesses a ~21 nm-thick reduction shell and the content of oxygen vacancies is gradually increased from the inside to the surface, (Supplementary Fig. 19) corresponding to a gradually increased O distortion. Consequently, the defective amorphous layer and gradually increased O distortion would lead to obviously elevated valence band tail states and gradually elevated valance band maximums in the reduction shell (Supplementary Fig. 20)45, which are much higher than those determined by the absorption spectra. These gradually elevated valance band maximums and valence band tail states not only can promote charge carrier separation but also greatly reduce the energy of holes, suppressing the C-C bonds cleavage. Therefore, with the increase of reduction time, the YCO2s over the defective photoanodes are greatly decreased, and no detectable CO2 is observed after 2 s reduction (Entry 4–11 Table 1). Benefiting from this, the defective TiO2 photoanode with 10 s reduction (def-TiO2) shows a high photocurrent and a high YGLU+GLA of 91.5% (Entry 5, Table 1), however, its YGLA is still low (13.8%). While the further deposition of Pt SAs can greatly increase the YGLA to 84.3% (Entry 7, Table 1). TEM results show that the thicknesses of the disorder shell vary from ~0.5 to ~4 nm with a reduction time from 2 s to 10 s (Supplementary Fig. 21). Further prolonging the reduction time to 15 s, the thickness of the disorder shell is increased to ~8 nm (Supplementary Fig. 21), resulting in an over reduction of the photoanode for glucose oxidation.
The defective photoanode with 15 s reduction exhibits excessive oxygen vacancies acting as the recombination centers and insufficient oxidation capacity by the over elevated valance band maximum (Supplementary Figs. 17, 18). Xglucose and JG are consequently decreased (Entry 6, Table 1). When the ALD was prolonged to 50 cycles some Pt particles were formed on def-TiO2 (Supplementary Fig. 22a), while dense Pt particles with a diameter of ~2 nm were formed on def-TiO2 after 100 cycles (Supplementary Fig. 22b). The formed Pt particles are metallic Pt as suggested by the XPS spectrum (Supplementary Fig. 23). The increased metallic Pt particles can substantially decrease the JG (Entry 10–11, Table 1) due to that the oxygen vacancies between the disorder shell and the Pt particles act as the recombination centers. Importantly, with the increase of metallic Pt particles, the photoanodes show significantly decreased YGLA, implying the essence of Pt SAs for the selective production of GLA. These results demonstrate the key role of the interaction between the disordered TiO2 and Pt SAs in the selectivity of glucose to GLA. Besides, the Au and Pd SAs were also successfully deposited on the def-TiO2 (Au/def-TiO2 and Pd/def-TiO2) photoanode through an immersion method to investigate their PEC performances for glucose oxidation (Supplementary Fig. 24). As shown in Entry 12 and 13, Table 1, the Pd/def-TiO2 and Au/def-TiO2 photoanodes also show significantly enhanced YGLA of 65.3 and 69.9%, respectively, compared with the def-TiO2 photoanodes, implying the facilitations of Au and Pd SAs on the conversion of glucose to GLA. However, the YGLAs from the Pd/def-TiO2 and Au/def-TiO2 photoanodes are much lower than that from the Pt/def-TiO2 photoanode. Therefore, although the Pt SA is not the unique metal to promote the selective oxidation of glucose to GLA, it is the most active single atom for the selective GLA generation.
Reaction mechanism and path
Finally, the reaction pathway and mechanism over the Pt/def-TiO2 photoanodes were investigated by the 2D-HMBC NMR and rate constant of charge transfer (kt) from the IMPS. To obtain the 2D-HMBC NMR, the PEC oxidation of glucose was conducted in 20 ml deuteroxide as the solution with 1 M KOH and 10 mM glucose, and the final solution was adjusted to acidity with H2SO4 prior to NMR analysis. The possible glucose decomposition in 1 M KOH at 20 °C was investigated by 2D-HMBC NMR. No obvious changes were observed in the 2D-HMBC NMR spectra of the solutions before and after keeping for 10 h (Supplementary Fig. 25), suggesting no significant self-decomposition of glucose. This is consistent with previous studies of glucose oxidation in alkali solutions7,46,47. As shown in the 2D-HMBC NMR spectrum after 1 h oxidation (Supplementary Fig. 26), glucose, GLU, and some of their lactonization derivatives were observed. Only one kind of the lactonization derivative of GLU, i.e. D-glucono 1,4 lactone, was detected. After 3 h oxidation, except for the GLU and D-glucono 1,4 lactone, the GLA, the lactonization derivative of GLA (D-glucaro 1,4 lactone), and L-guluronic acid (GUR) emerged (Fig. 5a). It should be noted that the D-glucono 1,4 lactone is substantially declined and a small amount of D-glucaro 1,4 lactone is generated, indicating lactonization derivatives should also participate in a reaction. When oxidation duration was prolonged to 5.5 h, most of the glucose was converted into GLA equilibrium with a tiny amount of, D-glucaro 1,4 lactone, and a small quantity of GLU resided in the electrolyte (Supplementary Fig. 27). Therefore, the oxidation of glucose follows the path from glucose to GLU, then to GUR, and finally to GLA, meanwhile equilibrium with their lactonization derivatives (Supplementary Fig. 28), and their lactonization derivatives also should participate in the series of oxidation reactions. Furthermore, the Pt/def-TiO2 NRAs were separately immersed into the electrolytes of 1 M KOH with 10 mM glucose, 1 M KOH with 10 mM GLU, and 1 M KOH with 10 mM GUR without any applied voltage. After 5 h illumination of simulated sunlight at 20 °C, the electrolytes were analyzed by HPLC and the results show that trace amounts of GLA were detected in the electrolytes, suggesting that the oxidation of GLU to GLA proceeds over Pt SAs is indeed a PEC reaction.
To have a deep insight into the mechanism of serial oxidation of glucose to GLA, the kt for glucose, GLU, and GUR oxidation over the def-TiO2, Pt/def-TiO2, and Pt/def-TiO2 with 100 cycles (Pt particles/def-TiO2) photoanodes were examined (Fig. 5b–d). The corresponding rate constant of surface recombination (krec) and surface charge transfer efficiency (ηtran) for glucose, GLU, and GUR oxidation are computed, respectively (Supplementary Figs. 29–31). The kt for glucose, GLU, and GUR oxidation over the def-TiO2 photoanode are around 95, 45, and 190 s−1, respectively. The much lower kt for GLU but much higher kt for GUR indicates that the oxidation from GLU to GUR is the rate-limiting step for the selective oxidation of glucose to GLA. Additionally, the def-TiO2 photoanode also shows much lower ηtran for GLU oxidation than those for glucose and GUR oxidation, further evidencing the rate-limiting step of GLU oxidation (Supplementary Figs. 29–31). In addition, the much higher kt of all the samples for GUR indicates the GUR can be easily converted to GLA. This can explain why only a small amount of GUR was detected during the whole oxidation process.
The Pt SAs has an all-round effect on the promotion of the oxidation of aldehyde groups from glucose and alcohol hydroxyl groups from GLA, evidenced by their enhanced kts on the Pt/def-TiO2 photoanode (Fig. 5b, c). Importantly, the much lower kt for GLU is significantly increased to around 90 s−1, which can greatly accelerate the GLA conversion. This reveals that the selectivity of glucose to GLA over the Pt/def-TiO2 photoanode mainly stems from the accelerated kinetics of the GLU oxidation by the Pt SAs. Although the metallic Pt particles can accelerate the conversion of glucose to GLU (Fig. 5c), GLU oxidation is still limited, resulting in a low GLA selectivity. Besides, LSV curves for GLU, GUR, and GLA oxidation over the TiO2, def-TiO2, and Pt/def-TiO2 photoanodes were tested with and without illumination (Supplementary Fig. 32), which further confirmed the PEC reaction for glucose oxidation and fast kinetics for GUR oxidation. These LSV results also reveal the accelerated GLU oxidation over the Pt SAs and sluggish GLA oxidation, consequently resulting in a high selectivity of glucose to GLA over the Pt/def-TiO2 photoanode. Based on the above results, both the defective TiO2 and the Pt SAs are crucial for the selective oxidation of glucose to GLA (Fig. 5e). The def-TiO2 with a large electron conduction region and large surface band bending can greatly enhance the charge carrier extraction. The extracted electrons are transported to the counter electrode for H2 production, while the extracted holes are regulated to the appropriate energy for glucose oxidation by the valance band structure of the defective TiO2. Finally, the holes are collected by the glucose through the Pt SAs, realizing selective oxidation to GLA.
To elucidate the reaction pathway of glucose oxidation, electron paramagnetic resonance (EPR) spectroscopy and in-situ Fourier transform infrared spectra (FTIR) measurements were carried out to monitor the reaction intermediates. Figure 6a shows the in-situ FTIR of the Pt/def-TiO2 photoanode in the reaction of PEC glucose oxidation recorded at 0.6 VRHE from 0 to 180 min. The negative characteristic IR bands located at ~1350, 1415, 1585, 1640, and 1740 cm−1 are assigned to δ(CH2) in the C6-position of gluconate anion and glucose, symmetric νs(O–C–O) of COO− function in gluconate (C1-position), L-glucuronate (C1-position) and glucarate (C1-position and C6-position), asymmetric stretching vibration νas(O − C − O) of COO− function, δ(H2O) due to interfacial water absorbed on the photoanode, and ν(C = O) in gluconolactone or L-glucuronate (C6-position), respectively48,49,50,51,52,53. The νs(O–C–O) and νas(O − C − O) IR bands show obvious dependence on PEC duration, suggesting the continuous generation of the COO− function. Remarkably, no obvious C-C breaking occurred during the PEC glucose oxidation over the Pt/def-TiO2 photoanode as revealed by the undetected C − C bond cleavage compounds (CO: 1900 − 2100 cm−1, carbonate: 1396 cm−1, CO2: 2340–2350 cm−1)48. The stable δ(H2O) IR bands during the PEC glucose oxidation indicate the involvement of water in the reaction. Further EPR spectrum exhibits quartet signals with an intensity ratio of 1:2:2:1, assigned to the DMPO–•OH adduct (Fig. 6b)54. This means that the absorbed water molecules on the catalyst are oxidated by the holes to form •OH radicals. Meanwhile, •OH radicals and holes will separately attack the H atom of the C1-H, and the C1-O-H bonds, resulting in the formation of C1-OO- groups48,49. Besides, the peak intensities of δ(CH2) remain almost unchanged until 90 min and then decrease gradually with increased ν(C = O) IR bonds, suggesting that the oxidation process mainly occurred on the C6 position after 90 min. These results suggest that the transformation of glucose into GLA is achieved through the intermediate production of GLU and GUR.
A reaction mechanism for the PEC oxidation of glucose to GLA on the Pt/def-TiO2 photoanode can therefore be proposed (Fig. 6c). The holes from the photoanode will react with the absorbed water to form absorbed •OH radicals52,55, which can abstract the H atoms of the C1-H and C1-O-H bonds from absorbed C1 of glucose on the photoanode forming C1 = O bond in step 249,56. Subsequently, the C5-O bond is split by hydrolysis, and then the GLU is formed through desorption from the photoanode48,49. Previous study has shown that highly dispersed Pt catalysts are capable of selectively activating primary alcohols for further oxidation57. Therefore, our Pt SAs on the photoanode are proposed to selectively adsorb primary alcohols from the C6 position. After the co-adsorption of C6-OH on the Pt SAs and C6 on the photoanode (step 4), the absorbed •OH radicals will react with the H atoms of C6-OH and C6-H, separately, to form a C6 = O bond (step 5), where some products desorb from the catalyst to generate GUR and others will adjust their adsorption states for the subsequent reaction. During step 6, with the holes and hydroxyl groups, the C6 = O bond will be activated and a new C6-OH bond will be formed58. Subsequently, the groups from the C6 position will follow the same reaction pathway as the C1 position to yield GLA, eventually. Therefore, both the Pt SAs and absorbed •OH radicals can play key roles in the selective PEC oxidation of glucose to GLA.
In summary, a PEC method was developed to selectively convert glucose to high value-added GLA. The Pt/def-TiO2 photoanode derived from single-atom Pt anchored on defective rutile TiO2 nanorod arrays shows a high yield of GLA. The results suggest that the interaction between the disordered TiO2 and single-atom Pt is crucial in the selective production of GLA. The defective structure is revealed to facilitate charge separation and transport, leading to a huge increase in JG. Importantly, the energy of valance band holes can be modulated by the defect structure, realizing the optimized yield of C6 products. The selectivity of the GLA can be regulated by the Pt SAs through the acceleration of GLU oxidation. Therefore, a high of 84.3% for GLA at 0.6 VRHE under simulated sunlight illumination was achieved on the Pt/def-TiO2 photoanode. This PEC strategy for selective oxidation of organic compounds reveals a new strategy for the utilization of biomass feedstocks.
Growth of Rutile TiO2 NRAs
The TiO2 NRAs were grown on FTO substrates (7 Ω per square, Film: 700 nm, NSG Company) by a hydrothermal method18. In a typical synthesis, a tetrabutyl titanate precursor solution was first obtained by slowly mixing 0.7 ml tetrabutyl titanate (Alfa Aesar, 99%), 21 ml concentrated HCl (Alfa Aesar, 37%), and 30 ml deionized water. Before use, the FTO was ultrasonically cleaned with ethanol, the mixture of acetone, isopropanol, and water (1:1:1), acetone in sequence, and kept in the acetone solution. Subsequently, the precursor solution was transferred into a 50 mL sealed Teflon reactor with four pieces of cleaned FTO, which was kept at 180 °C for 4 h. Finally, the FTO substrates with as-synthesized NRAs were rinsed with deionized water, dried inair, and annealed in air at 500 °C for 1 h to obtain TiO2 NRAs.
Fabrication of defective TiO2 NRAs
The defective NRAs were obtained through an electrochemical reduction method59. In detail, a conventional three-electrode system was connected to a potentiostat electrochemical workstation (CHI 760E), where the TiO2 NRAs on the FTO, an Ag/AgCl (KCl saturated), and a Pt wire were used as the working, reference, and counter electrodes, respectively. The electrochemical reduction was carried out in 0.5 M Na2SO4 (pH = 6.8) electrolyte at an applied potential of −1.4 V vs reversible hydrogen electrode (VRHE) for 2, 5, 10, and 15 s to obtain defective TiO2 NRAs.
Fabrication of single-atom Pt, Pd, and Au on TiO2 and def-TiO2
The single-atom Pt was deposited on the TiO2 and def-TiO2 NRAs through an ALD method. During the ALD process, the trimethyl(methylcyclopentadienyl)-platinum (IV) (MeCpPtMe3), O3, and high purity N2 were used as precursors, reaction gas, and carrier gas, respectively. The TiO2 and def-TiO2 on the FTO substrates were placed into the ALD chamber. The MeCpPtMe3 precursor and gas lines were kept at 70 and 120 °C to supply continuous precursor vapor to the chamber. The Pt/TiO2 and Pt/def-TiO2 were obtained by the ALD at the temperature of 200 °C for 10 cycles. 50 and 100 cycles of the ALD were also conducted on the def-TiO2 NRAs for comparison.
The single-atom Pd and Au were deposited on the def-TiO2 NRAs through an immersion method. In detail, the def-TiO2 NRAs were immersed into 1 mmol/L HAuCl4 and H2PdCl4 solutions for 30 mins, respectively. Subsequently, the NRAs on the FTO substrates were annealed at 200 °C for 0.5 h in air to obtain the Pd/def-TiO2 and Au/def-TiO2 photoanodes.
The data that support the findings of this study are available within the article and its Supplementary Information files. All other relevant data supporting the findings of this study are available from the corresponding authors upon request.
Kuehnel, M. F. & Reisner, E. Solar hydrogen generation from lignocellulose. Angew. Chem. Int. Ed. Engl. 57, 3290–3296 (2018).
Moggia, G., Schalck, J., Daems, N. & Breugelmans, T. Two-steps synthesis of D-glucaric acid via D-gluconic acid by electrocatalytic oxidation of D-glucose on gold electrode: Influence of operational parameters. Electrochim. Acta 374, 137852 (2021).
Lee, J., Saha, B. & Vlachos, D. G. Pt catalysts for efficient aerobic oxidation of glucose to glucaric acid in water. Green. Chem. 18, 3815–3822 (2016).
Werpy, T. & Petersen, G. Top Value Added Chemicals from Biomass: Volume I-Results of Screening for Potential Candidates from Sugars and Synthesis Gas. (US Department of Energy, Washington, DC, 2004).
Mehltretter, C., Rist, C. J. J. O. A. & Chemistry, F. Sugar oxidation, saccharic and oxalic acids by the nitric acid oxidation of dextrose. J. Agric. Food Chem. 1, 779–783 (1953).
Mustakas, G., Slotter, R., Zipf, R. J. I. & Chemistry, E. Pilot plant potassium acid saccharate by nitric acid oxidation of dextrose. Ind. Eng. Chem. 46, 427–434 (1954).
Liu, W. J. et al. Efficient electrochemical production of glucaric acid and H2 via glucose electrolysis. Nat. Commun. 11, 265 (2020).
Cha, H. G. & Choi, K. S. Combined biomass valorization and hydrogen production in a photoelectrochemical cell. Nat. Chem. 7, 328–333 (2015).
Zhang, P., Wang, T. & Gong, J. Current mechanistic understanding of surface reactions over water-splitting photocatalysts. Chem 4, 223–245 (2018).
Zhao, E. et al. Advancing photoelectrochemical energy conversion through atomic design of catalysts. Adv. Sci. 9, e2104363 (2022).
Fang, Y., Hou, Y., Fu, X. & Wang, X. Semiconducting polymers for oxygen evolution reaction under light illumination. Chem. Rev. 122, 4204–4256 (2022).
Liu, D. et al. Selective photoelectrochemical oxidation of glycerol to high value-added dihydroxyacetone. Nat. Commun. 10, 1779 (2019).
Derrien, E. et al. Aerobic oxidation of glucose to glucaric acid under alkaline-free conditions: Au-based bimetallic catalysts and the effect of residues in a hemicellulose hydrolysate. Ind. Eng. Chem. Res. 56, 13175–13189 (2017).
Solmi, S., Morreale, C., Ospitali, F., Agnoli, S. & Cavani, F. J. C. Oxidation of D-glucose to glucaric acid using Au/C catalysts. ChemCatChem 9, 2797–2806 (2017).
Jin, X. et al. Synergistic effects of bimetallic PtPd/TiO2 nanocatalysts in oxidation of glucose to glucaric acid: structure dependent activity and selectivity. Ind. Eng. Chem. Res. 55, 2932–2945 (2016).
Zhang, T., Chen, Z., Walsh, A. G., Li, Y. & Zhang, P. Single-atom catalysts supported by crystalline porous materials: views from the inside. Adv. Mater. 32, 2002910 (2020).
Da, Y. et al. The applications of single-atom alloys in electrocatalysis: progress and challenges. SmartMat 4, e1136 (2023).
Liu, B. & Aydil, E. S. Growth of oriented single-crystalline rutile TiO2 nanorods on transparent conducting substrates for dye-sensitized solar cells. J. Am. Chem. Soc. 131, 3985–3990 (2009).
Tian, Z. et al. Hydrogen plasma reduced black TiO2-B nanowires for enhanced photoelectrochemical water-splitting. J. Power Sources 325, 697–705 (2016).
Tian, Z. et al. Highly conductive cable-like bicomponent titania photoanode approaching limitation of electron and hole collection. Adv. Funct. Mater. 36, 1803328 (2018).
Zhu, G. L. et al. Black brookite titania with high solar absorption and excellent photocatalytic performance. J. Mater. Chem. A 1, 9650–9653 (2013).
Yang, C. et al. Core-Shell nanostructured “black” rutile titania as excellent catalyst for hydrogen production enhanced by sulfur doping. J. Am. Chem. Soc. 135, 17831–17838 (2013).
Cui, H. et al. Black TiO2 nanotube arrays for high-efficiency photoelectrochemical water-splitting. J. Mater. Chem. A 2, 8612 (2014).
Naldoni, A. et al. Photocatalysis with reduced TiO2: from black TiO2 to cocatalyst-free hydrogen production. ACS Catal. 9, 345–364 (2019).
Chen, X., Liu, L., Yu, P. Y. & Mao, S. S. Increasing solar absorption for photocatalysis with black hydrogenated titanium dioxide nanocrystals. Science 331, 746–750 (2011).
Tian, Z. et al. Novel black BiVO4/TiO2−x photoanode with enhanced photon absorption and charge separation for efficient and stable solar water splitting. Adv. Energy Mater. 9, 1901287 (2019).
Chen, Y. et al. Discovering partially charged single-atom Pt for enhanced anti-markovnikov alkene hydrosilylation. J. Am. Chem. Soc. 140, 7407–7410 (2018).
Wang, G. et al. Hydrogen-treated TiO2 nanowire arrays for photoelectrochemical water splitting. Nano lett. 11, 3026–3033 (2011).
Daelman, N., Capdevila-Cortada, M. & Lopez, N. Dynamic charge and oxidation state of Pt/CeO2 single-atom catalysts. Nat. Mater. 18, 1215–1221 (2019).
Zhao, D. et al. MXene (Ti3C2) vacancy-confined single-atom catalyst for efficient functionalization of CO2. J. Am. Chem. Soc. 141, 4086–4093 (2019).
Chen, Y. et al. Engineering the atomic interface with single platinum atoms for enhanced photocatalytic hydrogen production. Angew. Chem. Int. Ed. Engl. 59, 1295–1301 (2020).
Macino, M. et al. Tuning of catalytic sites in Pt/TiO2 catalysts for the chemoselective hydrogenation of 3-nitrostyrene. Nat. Catal. 10, 873–881 (2019).
Chen, Z. et al. Addressing the quantitative conversion bottleneck in single-atom catalysis. Nat. Commun. 13, 1–8 (2022).
Zhang, K. et al. Overcoming charge collection limitation at solid/liquid interface by a controllable crystal deficient overlayer. Adv. Energy Mater. 7, 1600923 (2017).
Vogt, S., Schneider, M., Schäfer-Eberwein, H. & Nöll, G. J. A. C. Determination of the pH dependent redox potential of glucose oxidase by spectroelectrochemistry. Anal. Chem. 86, 7530–7535 (2014).
Komori, H. et al. In 2014 IEEE Biomedical Circuits and Systems Conference (BioCAS) Proceedings. 464–467 (IEEE).
Tian, Z. et al. Enhanced charge carrier lifetime of TiS3 photoanode by introduction of S22− vacancies for efficient photoelectrochemical hydrogen evolution. Adv. Funct. Mater. 30, 2001286 (2020).
Gao, Y. & Hamann, T. W. Quantitative hole collection for photoelectrochemical water oxidation with CuWO4. ChemComm 53, 1285–1288 (2017).
Cachet, H. & Sutter, E. M. M. Kinetics of water oxidation at TiO2 nanotube arrays at different ph domains investigated by electrochemical and light-modulated impedance spectroscopy. J. Phys. Chem. C. 119, 25548–25558 (2015).
Peter, L. M. Energetics and kinetics of light-driven oxygen evolution at semiconductor electrodes: the example of hematite. J. Solid State Electrochem 17, 315–326 (2013).
Ponomarev, E. A. & Peter, L. M. A Comparison of intensity-modulated photocurrent spectroscopy and photoelectrochemical impedance spectroscopy in a study of photoelectrochemical hydrogen evolution at P-Inp. J. Electroanal. Chem. 397, 45–52 (1995).
Snaith, H. J. et al. Charge collection and pore filling in solid-state dye-sensitized solar cells. Nanotechnology 19, 424003 (2008).
Chen, J. et al. Surface rutilization of anatase TiO2 nanorods for creation of synergistically bridging and fencing electron highways. Adv. Funct. Mater. 26, 456–465 (2016).
Fabrao, R. M., de Brito, J. F., da Silva, J. L., Stradiotto, N. R. & Zanoni, M. V. B. J. E. A. Appraisal of photoelectrocatalytic oxidation of glucose and production of high value chemicals on nanotube Ti/TiO2 electrode. Electrochim. Acta 222, 123–132 (2016).
Liu, L., Yu, P. Y., Chen, X., Mao, S. S. & Shen, D. Z. Hydrogenation and disorder in engineered black TiO2. Phys. Rev. Lett. 111, 065505 (2013).
Aoun, S. B. et al. Effect of metal ad-layers on Au (1 1 1) electrodes on electrocatalytic oxidation of glucose in an alkaline solution. J. Electroanal. Chem. 2567, 175–183 (2004).
Pasta, M., Ruffo, R., Falletta, E., Mari, C. & Pina, C. D. J. G. B. Alkaline glucose oxidation on nanostructured gold electrodes. Gold. Bull. 43, 57–64 (2010).
Holade, Y. et al. Electrocatalytic and electroanalytic investigation of carbohydrates oxidation on gold-based nanocatalysts in alkaline and neutral pHs. J. Electrochem. Soc. 165, H425–H436 (2018).
Men, Y.-L., Liu, P., Liu, Y., Meng, X.-Y. & Pan, Y.-X. Noble-metal-free WO3-decorated carbon nanotubes with strong W–C bonds for boosting an electrocatalytic glucose oxidation reaction. Ind. Eng. Chem. Res. 61, 4300–4309 (2022).
Neha, N. et al. Revisited Mechanisms for glucose electrooxidation at platinum and gold nanoparticles. Electrocatalysis 14, 121–130 (2022).
Bae, I., Yeager, E., Xing, X. & Liu, C. In situ infrared studies of glucose oxidation on platinum in an alkaline medium. J. Electroanal. Chem. 309, 131–145 (1991).
Luo, L. et al. Selective photoelectrocatalytic glycerol oxidation to dihydroxyacetone via enhanced middle hydroxyl adsorption over a Bi2O3-incorporated catalyst. J. Am. Chem. Soc. 144, 7720–7730 (2022).
Zhang, X., Chan, K.-Y., You, J.-K., Lin, Z.-G. & Tseung, A. C. J. J. O. E. C. Partial oxidation of glucose by a Pt|WO3 electrode. J. Electroanal. Chem. 430, 147–153 (1997).
Dikalov, S. I. & Mason, R. P. Reassignment of organic peroxyl radical adducts. Free Radic. Biol. Med. 27, 864–872 (1999).
Sato, S. & White, J. M. Photoassisted water-gas shift reaction over platinized titanium dioxide catalysts. J. Am. Chem. Soc. 102, 7206–7210 (1980).
Rafaïdeen, T., Baranton, S. & Coutanceau, C. Highly efficient and selective electrooxidation of glucose and xylose in alkaline medium at carbon supported alloyed PdAu nanocatalysts. Appl. Catal. B 243, 641–656 (2019).
Liang, D. et al. Selective oxidation of glycerol with oxygen in a base-free aqueous solution over MWNTs supported Pt catalysts. Appl. Catal. B 106, 423–432 (2011).
Li, X., Zhang, L., Wang, S. & Wu, Y. Recent advances in aqueous-phase catalytic conversions of biomass platform chemicals over heterogeneous catalysts. Front. Chem. 7, 948 (2019).
Yan, P. et al. Photoelectrochemical water splitting promoted with a disordered surface layer created by electrochemical reduction. ACS Appl. Mater. Interfaces 7, 3791–3796 (2015).
Authors acknowledge financial support by the National Research Foundation, Singapore, and A*STAR (Agency for Science, Technology and Research) under its LCER FI program Award No U2102d2002, and NUS R&G Postdoc Fellowship Program.
The authors declare no competing interests.
Peer review information
Nature Communications thanks Zeyan Wang and the other, anonymous, reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Tian, Z., Da, Y., Wang, M. et al. Selective photoelectrochemical oxidation of glucose to glucaric acid by single atom Pt decorated defective TiO2. Nat Commun 14, 142 (2023). https://doi.org/10.1038/s41467-023-35875-9
This article is cited by
pH controlled synthesis and photoelectrochemical kinetics of bismuth molybdate (Bi2MoO6) nanoplates as photoanodes for solar-driven water splitting
Journal of Solid State Electrochemistry (2023)
Development of non-enzymatic D-glucose and ethanol electroactive electrode based on ZnO–Cu/C multifunctional nanocomposite
Journal of Materials Science: Materials in Electronics (2023)
Synthesis of TiO2/ZnO/Au for application in a new type of paper-based photo-microfluidic fuel cell using human blood
Journal of Solid State Electrochemistry (2023)