Precise Lattice‐Strain Modulation of Hematite Enabled by Gradient Doping of Mn for Enhanced Photoelectrocatalytic Oxidative C─C Bond Scission

The high‐value utilization of biomass feedstock is fascinating but limited by efficient C─H activation to break C─C bonds. Herein, F‐Fe2O3‐Mn photoanodes with modulable compressive strain are fabricated by gradient infusion of Mn into F‐doped hematite (F‐Fe2O3), which is illustrated to be highly efficient for oxidative C─C bond cleavage of various bio‐based 1,2‐diols to produce benzoic acids or aromatic ketones (94.5–97.2% yields) in photoelectrocatalytic (PEC) device, coupling with a high H2 production of 1180 μmol cm−2 (≈96% yield). The gradient doping of Mn species into the photoelectrode bulk results in improved photoexcited carriers separation and transfer efficiency of the photoelectrode (3.41 mA cm−2). On the other hand, the lattice distortion induced by Mn doping also leads to a strain effect on F─Fe2O3─Mn, which can precisely modulate the photoelectrode electronic structure. Control experiments, in situ characterization, and theoretical calculations elaborate that compressive strain is capable of adjusting the position of the d‐band center to facilitate C─H activation, remarkably enabling PEC oxidative C─C bond breaking of 1,2‐diol and the desorption of the oxidized product. This “one‐stone‐two‐bird” strategy presents a straightforward protocol for efficiently breaking C─C bonds in organic and biomass transformations via PEC oxidation.


Introduction
Photoelectrocatalytic (PEC) cells are being extensively researched as a potential device for direct utilization of solar energy for water splitting, H 2 O 2 production, and oxygen evolution reaction (OER) and hydrogen evolution reaction [1][2][3][4] In recent years, the selective C─C bond breaking of complex chemicals and biopolymers with tailored catalysts has contributed to the development of organic synthesis and green chemistry, typically enabled by conventional thermocatalysis [5,6] Among the bio-based oxygenated molecules, 1,2-diols (e.g., ethylene glycol, propylene glycol, and 1-phenyl-1,2-ethanediol) are mainly derived from (hemi-)cellulose and lignin [7] For organic synthesis and sustainable chemistry, the oxidative C─C bond cleavage of 1,2-diols is of great importance.One of the most classical methods for C─C bond breaking in 1,2-diols is the Criegee oxidation, in which 1,2-diols are oxidized to aldehydes in organic solvents (e.g., CH 3 CN, CH 3 OH, and CH 3 COOH) using oxidants such as lead tetraacetate [8][9][10] However, the used oxidant inevitably produces an equimolar amount of toxic waste, which is not compatible with the requirements of eco-friendliness and energy sustainability.Therefore, the construction of renewed protocols and fitted apparatus that can be used under mild conditions, with low toxicity and sustainable use of green energy (e.g., solar energy), is vital for the development of a low-carbon society [11,12] Photochemistry (PC) and electrochemistry (EC) have been utilized to investigate C─C bond breaking as clean and sustainable methods.Highly selective cleavage of C─C bonds by active species at metal-based electrodes in electrochemical catalysis results from the formation of metal oxide/hydroxide/hydroxylated species [13,14] The activation of free radicals to attack and break the C─C bond is the main mechanism for photochemically selective bond breaking [15,16] The high energy of C─H dissociation has led to the conclusion that C─H bond activation controls the C─C bond cleavage in the oxidation process [17][18][19][20] Also, the PEC cell, a technology that combines the strengths of PC and EC, has recently been explored in the field of C─H activation for the synthesis of organic small molecules (e.g., H 2 and phenol) [21][22][23][24] In 1972, Honda and Fujishima exhibited the initial use of PEC in the process of water splitting [25] Several semiconductors such as DOI: 10.1002/sstr.202300531 The high-value utilization of biomass feedstock is fascinating but limited by efficient C─H activation to break C─C bonds.Herein, F-Fe 2 O 3 -Mn photoanodes with modulable compressive strain are fabricated by gradient infusion of Mn into F-doped hematite (F-Fe 2 O 3 ), which is illustrated to be highly efficient for oxidative C─C bond cleavage of various bio-based 1,2-diols to produce benzoic acids or aromatic ketones (94.5-97.2%yields) in photoelectrocatalytic (PEC) device, coupling with a high H 2 production of 1180 μmol cm À2 (≈96% yield).The gradient doping of Mn species into the photoelectrode bulk results in improved photoexcited carriers separation and transfer efficiency of the photoelectrode (3.41 mA cm À2 ).On the other hand, the lattice distortion induced by Mn doping also leads to a strain effect on F─Fe 2 O 3 ─Mn, which can precisely modulate the photoelectrode electronic structure.Control experiments, in situ characterization, and theoretical calculations elaborate that compressive strain is capable of adjusting the position of the d-band center to facilitate C─H activation, remarkably enabling PEC oxidative C─C bond breaking of 1,2-diol and the desorption of the oxidized product.This "one-stone-two-bird" strategy presents a straightforward protocol for efficiently breaking C─C bonds in organic and biomass transformations via PEC oxidation.WO 3 , Fe 2 O 3 , BiVO 4 , and CdS have been used in PEC systems [26][27][28][29][30][31][32][33] Surface engineering, interface engineering, and bulk engineering techniques have been utilized to enhance the PEC performances of semiconductor photoelectrodes, attributed to the semiconductor bandgap, light absorption range, and self-limiting nature [34][35][36][37][38][39] The modulation of the electronic structure significantly influences the reactivity of the electrodes [40,41] The d-band center theory of electrode structure (i.e., the movement of the d-band center) can change the adsorption/desorption capacity of the electrode toward the reaction substrates and its intermediates [42,43] Shen et al. achieved modulation of the d-band center by constructing a heterojunction to optimize the electronic structure of Co through charge transfer at the interface, resulting in weak adsorption of *H at the electrode [44] Hu et al. quantitatively investigated the size dependence of the activity using d-band centers by controlling the size of Ru nanocrystals on the electrode surface [45] For bulk engineering, doping can lead to a lattice distortion-induced lattice strain effect in native phase crystals.The lattice strain effect has also been shown to be used to modulate the adsorption/desorption capacity between substrates and catalyst surfaces, which has become a common strategy for improving electrode performance.The activity of CO 2 reduction, OER, and water splitting can be controlled by modulating the stress relaxation properties of the electrode [46][47][48][49][50][51] There is limited research on the modulation of lattice strain by doping to improve the performance of photoelectrodes and regulate their reactivity toward organic small molecules (biomass derivatives) in PEC.Quantitative analysis is thus highly necessary to establish the correspondence among lattice strain effect, electronic structure, and substrate adsorption/desorption capacity, providing a thorough conformational elucidation for the efficient PEC oxidative breaking of C─C bond in 1,2-diols.
Inspired by the above discussions, herein, a series of surface Mn-doped F-Fe 2 O 3 photoanodes with the aim of precisely modulating the lattice strain of F-doped hematite (F-Fe 2 O 3 ) were fabricated by gradient doping of Mn using a hydrothermal and repeated annealing method.The changes in the crystal structure of the bulk phase induced by the surface doping operations were investigated for the improvement of the photoanode PEC performance and reactivity in the oxidative C─C bond breaking of 1,2-diols.The etched X-Ray photoelectron spectroscopy (XPS) and X-Ray fluorescence (XRF) uncovered that the surface doping of Mn elements on the hematite bulk phase resulted in a gradient distribution with respect to the hematite z-axis.Williamson-Hall analysis, based on X-Ray diffraction (XRD) and transmission electron microscope (TEM), revealed that compressive strain in the hematite generated by surface doping.Theoretical calculations and in situ attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) were adopted to elaborate the corresponding conformational relationship among the compressive strain, d-band center, and PEC oxidation activity.It was illustrated that the lattice strain effect could modify the surface strain of the photoelectrode, altering its d-band center for enhancing C─H bond activation to initiate C─C bond breaking.This one-stone-two-bird strategy based on precise lattice-strain modulation further broadens PEC systems for organic transformations and biomass valorization via enhanced and selective C─C bond breaking and relevant conversion processes.

Photoelectrode Characterization and Strain Analysis
F-Fe 2 O 3 (FF) and FFMx including F-Fe 2 O 3 -Mn-10 (FFM1), F-Fe 2 O 3 -Mn-20 (FFM2), and F-Fe 2 O 3 -Mn-30 (FFM3) photoelectrodes were fabricated by hydrothermal and annealing treatment (Figure S1, Supporting Information), where 10, 20, and 30 represent the molar percentage of Mn relative to Fe.The morphology of FF and FFMx photoelectrodes was characterized by scanning electron microscopy (SEM).All photoanodes presented uniform worm-like nanorods (Figure 1a-d).The cross-sectional view of FF displayed dense nanorod arrays with a longitudinal size of ≈1.4 μm (Figure 1e), which were uniformly arranged on the fuorine-doped tin oxide (FTO) substrate.Transmission electron microscopy (TEM) images confirmed a nanorod morphology of photoelectrodes (Figure 1f and S3, Supporting Information).The selected area electron diffraction (SAED) pattern of FFM2 exhibited a single-crystal performance (Figure S5, Supporting Information).The SEM and SAED results indicated that the Mn doping did not alter the morphology and crystallographic phase of FF.The TEM with energy-dispersive X-Ray spectroscopy analysis of FFM2 indicated the existence of Fe, Mn, and O species, implying that Mn was successfully doped into the FF.The Mn content showed a gradient distribution with respect to the substrate z-axis (Figure S6, Supporting Information).The high-resolution TEM (HRTEM) was used to investigate the crystal phase and microstructure of the photoanodes.Surface distortion on FFM2 was assessed by geometric phase analysis (GPA (Figure 1g,h), which proved to accurately represent strains and dislocations.Figure 1g visually illustrates a range of the crystallographic plane dislocations and defective states on the photoelectrodes (highlighted in the yellow box).These dislocations are responsible for the presence of strain effects, which can account for the fluctuations in the lattice distance.Color variations in GPA highlight multiple lattice dislocations in FFM2.In addition, the formation of dislocations and defects has the potential to enhance photoexcited electron-hole separation.The HRTEM images and the corresponding fast Fourier transformation (FFT) patterns clearly showed a high crystalline structure of the photoelectrodes.A lattice distance of ≈0.25 nm could be assigned to the (110) facet of FF, matching the standard values closely.Significant alterations in lattice distance were detected by HRTEM (Figure 1i-l), exhibiting 0.248 nm (FFM1), 0.245 nm (FFM2), and 0.242 nm (FFM3) for the (110) plane (Figure 1m-p).The observed decrease in lattice distance of FFMx can be attributed to the compressive strain that arises as a result of Mn doping.The values of compressive strain for FF, FFM1, FFM2, and FFM3 were counted as 0.8%, 2%, and 3.2%, respectively.The compressive strain might reduce lattice distances, thus modifying the d-band center.According to the d-band theory, such modification of the d-band center is closely associated with the electronic properties of the photoelectrode.Based on this, lattice strain is demonstrated to alter the surface geometry and electronic structure, thus possibly enhancing the catalytic activity of the photoelectrode.
XRD patterns (Figure 2a) were performed to analyze the microstrain and crystal structure of photoelectrodes.All photoanodes displayed nearly identical diffraction peaks, matching the hexagonal structure of the hematite phase (JCPDS No. 87-1165).The absence of any signals associated with Mn and MnO x phases implied that Mn doping did not influence the crystal structure of FF.The crystallinity along the (110) direction was significantly enhanced in FFM2, compared to FFM1 and FFM3.The enlargement of the (110) peaks of FFMx showed a shift toward a higher angle with increasing the amount of Mn (Figure 2b), which can be attributed to the lattice strain caused by the doping of Mn, in agreement with the TEM analysis (Figure 1i-l).Table S1, Supporting Information, shows the mean crystallite size of the photoelectrodes according to the Debye-Scherrer formula.The crystallite size of the photoelectrodes was reduced with decreasing lattice spacing in the order of FFM3 < FFM2 < FFM1 < FF.The lattice distortion and crystal size of the (110) Bragg peak of FFMx in the range of 20°-70°were further analyzed using Williamson-Hall (W-H) plots (Figure 2c and Table S1, Supporting Information).The negative slope in the fitted plots is a sign of compression strain [50,52] indicating that the strain existed in the lattice of FFMx.The slope value shows an inverse correlation with the Mn doping concentration, suggesting that higher Mn concentrations result in greater compressive strain [53] From the slope of the W-H plots, the compressive strain in the FF lattice followed the order of FFM1 (1.69%) < FFM2 (3.38%) < FFM3 (3.66%).It is consistent with the trend in grain size, implying that the decrease in grain size accounted for the compressive deformation, also in line with the results obtained from HRTEM.Generally, elemental doping can tailor the photoelectrode structure, causing lattice mismatch and distortion.Inducing interfacial interaction forces between components, albeit relatively small, can change the physicochemical properties of the photoelectrodes (e.g., atomic spacing and electronic structure).As shown in Figure S8, Supporting Information, the PEC performance was significantly improved after Mn doping.Compressive strain could be optimized by regulating the content of doped Mn species, thus adjusting the dband center to achieve balanced intermediate absorption and desorption energies for optimum catalytic performance.Overall, the gradient doping process is an effective way to tune photoelectrode properties by microengineering to increase PEC activity.
From the XPS survey spectra (Figure 2d), all photoelectrodes exhibited characteristic peaks matching the elements Fe and O.An additional peak of Mn was exhibited on the FFMx photoelectrodes.The fine XPS spectra of Fe 2p presented two characteristic peaks at 710.8 and 724.1 eV, belonging to Fe 2p 3/2 and Fe 2p 1/2 in photoelectrodes, respectively (Figure 2e).The characteristic peaks showed a slight shift compared to the FF.The Mn 2p peaks around 642.1 and 653.6 eV were attributed to the Mn 2p 3/2 and Mn 2p 1/2 of Mn 4þ (Figure 2f ).The Mn 2p peaks around 641.5 and 652.9 eV were assigned to the Mn 2p 3/2 and Mn 2p 1/2 of Mn 3þ , proving that the Mn was mainly present as Mn 4þ ions.respectively.From the XPS depth etching at a depth increasing from 0 to 100 nm, the peak intensity of Mn 2p was found to gradually decrease, while the Fe 2p peak diminished conversely (Figure S7, Supporting Information).The Mn/Fe atomic ratio of the photoelectrodes exhibited a gradient distribution (Figure 2g and Table S2, Supporting Information), and XRF disclosed a similar trend to XPS depth etching (Figure 2h).The phenomenon can be attributed to the doping of Mn into the FF lattice with different contents, causing a change in the FF lattice distances.As Mn 4þ (0.53 Å) has a smaller sized radius than Fe 3þ (0.55 Å), the lattice contraction caused by Mn substitution could be reasonable and became more evident with increasing Mn doping concentration.These results confirmed that Mn doping in the FF lattice was in a gradient distribution with respect to the substrate z-axis (Figure S6, Supporting Information).Additionally, two peaks at 530 and 531.1 eV were assigned to O 1s spectra (Figure 2i), corresponding to lattice oxygen (M─O) and oxygen vacancies, respectively.The lattice oxygen of the FFMx photoelectrode showed a shift toward higher values (Figure 2i), indicating that the presence of lattice strain enhanced the strength of M─O bonds.The alterations in the binding energies of Fe 2p and O 1s (Figure 2e-i), in comparison to FF, indicated that the incorporation of compressive strain altered the interatomic interactions in photoelectrodes.Therefore, the variation in lattice strain has the potential to regulate the d-band structure of the metals on the photoelectrodes, affect the electronic and geometric properties at the atomic scale, and influence the electronic structure of the photoelectrode.

The PEC Performance of FF and FFMx in Oxidative C─C Bond Breaking
The use of PEC devices to selectively break the C─C bond of bio-based 1,2-diols (taking hydrobenzoin (HBZ) as a model compound) for producing high-value chemicals has an important role in the development of sustainable and green chemistry.
Here, the PEC performance of the photoelectrodes was assessed, and the gradient doping was found to affect the photoelectrode performance.Prior to elucidating the impact of gradient doping on the PEC performance of FFMx nanorods, UV-vis absorption spectra were used to evaluate the optical properties of the photoelectrodes (Figure 3a).FF exhibited an absorption edge at ≈650 nm, and FFM1, FFM2, and FFM3 showed a slight redshift.The Tauc plots, which depict the computation of the optical bandgap for photoanodes, are presented in Figure 3b.All photoelectrodes exhibited bandgaps ranging from 1.93 to 1.99 eV.The values for FF, FFM1, FFM2, and FFM3 were 1.97, 1.99, 1.93, and 1.95 eV, respectively.These results manifested that varying the gradient doping did not significantly alter the optical properties of the photoelectrode.
The PEC performance of the photoelectrodes was subsequently conducted using a three-electrode system in 1 M NaOH.The linear sweep voltammetry (LSV) curves of all photoelectrodes are represented in Figure 3c and S8, Supporting Information.All photoelectrodes displayed a negligible dark current density from 0.80 to 1.50 V versus reversible hydrogen electrode (RHE).The photocurrent density of FF, FFM1, FFM2, and FFM3 was 1.48, 2.09, 2.87, and 2.30 mA cm À2 at 1.50 V versus RHE, respectively.The photocurrent density on FFMx exhibited a significant increase after doping the Mn species.The photocurrent of FFM2 was about 1.94 times higher than that of FF.The enhancement of photocurrent could be ascribed to reducing the recombination of electrons and holes originating from gradient doping.The transient photocurrent curves of FF and FFMx were evaluated at 1.50 V versus RHE under chopped light to assess the photoresponse and charge recombination behavior of the photoelectrode.The response of photocurrent to Mn doping showed a nonlinear increase.Instead, following a curvilinear pattern similar to a volcano, the photocurrent density peaked at FFM2 and maintained a steady current response (Figure S9, Supporting Information).Under simulated solar irradiation, FF showed sharp photocurrent peaks, which were significantly suppressed after gradient doping of Mn.It emphasized the importance of gradient doping in the enhancement of the photoresponse characteristics and charge separation.The active area of the FF and FFMx photoelectrodes can also be related to the increased photocurrent.The electrochemically active surface area (ECSA) of the photoelectrode was thus analyzed using cyclic voltammetry plots.The estimated active area was 2.1, 3.4, 5.2, and 4.2 μF cm À2 for FF, FFM1, FFM2, and FFM3 photoelectrode, respectively (Figure 3d).The catalytic activity area of FFMx showed a gradual increase with the rise of the Mn doping content.FFM2 exhibited the largest active area compared to other photoelectrodes, which was an indication of enhanced intrinsic performance.The results could be explained by the excellent photoelectronic properties of FFM2 resulting from the optimized gradient doping, which suppressed photoexcited electron-hole pair recombination.To further elucidate the enhanced PEC performance of hematite resulting from Mn doping, electrochemical measurements were utilized to assess the charge transfer behaviors of the photoelectrode.Figure S10, Supporting Information, displays the calculated charge transfer efficiency (η trans ) and separation efficiency (η sep ) (calculated details in the Experimental Section).It could be observed that the η trans and η sep of FF were 54.0% and 31.4% at 1.50 V versus RHE, with a marked increase after Mn doping.Both η trans (78.3%) and η sep (40.5%) in FFM2 photoelectrode were higher than that of FF, which could be caused by Mn doping to promote more efficient charge separation and transfer efficiency.
Electrochemical impedance spectroscopy (EIS) was used to further understand the effect of gradient doping on the charge transfer behavior in FF and FFMx.All photoelectrodes exhibited comparable system resistances, and the analog circuit diagram was constructed.The relevant parameters are displayed in Table S2, Supporting Information.A lower charge transfer resistance can be inferred from a smaller Nyquist radius.FFM2 recorded the minimum Nyquist radius (R ct = 162.5 Ω) in comparison to FF (395.6 Ω (Figure 3e and Table S3, Supporting Information).The significant reduction in R ct indicated that the FFM2 could reduce the charge transfer resistance, increase conductivity, and promote charge separation in the PEC system.Moreover, the flat potentials (E FB ) of the FFMx and the FF photoelectrodes ranged from À0.30 to À0.18 V (Figure S12, Supporting Information).The open-circuit photovoltage (OCP) of FFM2 exhibited the largest voltage drop (0.36 V).A higher OCP illustrated that a higher photovoltage at the photoelectrode surface promoted the separation of photoexcited electron-hole pairs (Figure 3f ).Steady-state photoluminescence (PL) spectra were utilized to uncover the significant role of lattice strain in reducing photoexcited carrier recombination and enhancing light emission characteristics in the PEC systems.Both FF and FFMx photoelectrodes in steady-state PL spectra held a wide emission peak centered at 525 nm (Figure 3g), known as the band-edge emission.Compared to FF, the PL emission intensity of FFMx significantly decreased.Hence, time-resolved PL (TRPL) spectra were employed to quantitatively analyze the carrier dynamics (Figure 3h).The TRPL was evaluated by a biexponential decay model and listed in Table S4, Supporting Information.The average PL lifetime for FF and FFMx was measured to be 708, 792, 826, and 795 ns, respectively (Table S4, Supporting Information).The extended lifetime in FFM2 indicated an increase in diffusion length, facilitating more efficient charge separation and transfer.FFM2 exhibited excellent intrinsic activity and electronic properties, which effectively facilitated the photogenerated charge separation and transport, resulting in outstanding PEC performance.
The PEC performance analysis showed that gradient doping was induced by Mn doping, which significantly improved the separation and transfer of photogenerated carriers.Moreover, the PEC oxidation of a bio-based HBZ to benzoic acid (BEN) was investigated on FF and FFMx.The photocurrent density of the photoelectrodes in a 1 M NaOH solution containing HBZ exhibited a significantly higher photocurrent than that in 1 M NaOH without HBZ over the whole range of applied potentials (Figure S8, Supporting Information).It meant that the oxidation potential required for the photoelectrodes in the NaOH solution containing HBZ was lower than that needed for water oxidation.The FFM2 exhibited a photocurrent density of 3.41 mA cm À2 in 1 M NaOH containing 2.5 mM HBZ, which was superior to that in 1 M NaOH.In addition, a higher photocurrent response could be observed in Figure S13, Supporting Information.This improvement is a result of the lattice strain induced by manganese doping, facilitating the separation and transport of charge carriers.The obtained HBZ conversion rate and BEN selectivity on FF, FFM1, FFM2, and FFM3 photoelectrodes are presented in Figure 4a and Table S5, Supporting Information.Clearly, the photoelectrode with lattice strain induced by Mn doping could improve HBZ conversion and BEN selectivity via PEC oxidation.According to highperformance liquid chromatography (HPLC) results, the optimized FFM2 achieved a remarkable conversion (97.8%) and yield (96.8%) in the PEC oxidation of HBZ to BEN, coupled with a high H 2 production outcome of 1180 μmol cm À2 (≈96% yield, Figure S13, Supporting Information).Furthermore, the general applicability of the FFM2 photoelectrode was evaluated by employing a range of aromatic 1,2-diols in this PEC oxidation system (Figure 4e).It was found that the FFM2 photoelectrode exhibited exceptional PEC oxidation performance in 20 h, achieving high performance for the tested 1,2-diols (≥96.3%yields) and superior selectivity toward corresponding aromatic carboxylic acids (≥95.1% yields) or aromatic ketones (≥94.5% yields).The stability of the FFM2 photoelectrode under prolonged light exposure in PEC oxidation was also examined.The results presented in Figure 3i indicated that the photocurrent density remained stable for a continuous 20 h of reaction time at 1.50 V versus RHE.After the durability tests, FFM2 was found to have the highest stability (96%) compared to the other photoelectrodes.In contrast, FF, FFM1, and FFM3 showed lower levels of stability, with activity remaining at ≈80%, 88%, and 80% of the initial activity, respectively.The SEM, XPS, XRD, and UV-vis spectra of FFM2 before and after PEC oxidation were recorded to provide evidence for excellent stability (Figure S14, Supporting Information).After the PEC oxidation, the morphology, crystal, and structure of FFM2 were not significantly altered.After five consecutive cycles, the HBZ conversion and BEN yield were not markedly reduced (99%/96.8% to 95.5%/94%, respectively), suggesting that FFM2 had significant structural stability.
The above experimental results demonstrated that the photoelectrode performance and PEC oxidation activity were significantly enhanced by gradient doping and the induced compressive strain.Density functional theory (DFT) calculations were conducted to reveal the relationship between compressive strain and 1,2-diol (HBZ) oxidation activity.Three distinct compressive strain structural models of FFMx were optimized and presented in Figure S15  down [42][43][44][45][46] As the compression strain increased, the center of the d-band decreased accordingly (Figure S16, Supporting Information).For different compressive strains, the d-band center of FFM1, FFM2, and FFM3 descended gradually from À1.02 toÀ1.46 eV.The downward shift of the d-band center led to higher electron occupancy in the antibonding states, consequently decreasing the interaction strength between the photoelectrode and the intermediate.A suitable d-band center was thus favorable for the PEC oxidation of 1,2-diol (HBZ) because the balance of adsorption/desorption energy of the intermediates played a decisive role in catalytic efficiency.The BEN yield of 77.8% on FFM1 and that of 81.1% on FFM3 corresponded to compressive strains of 0.8% and 3.2%, respectively.This revealed that there was an optimum activity interval for the reactivity of photoelectrodes in PEC oxidation of HBZ, and the compressive strain of FFM2 (2.0%) was exactly in this interval, hence making its performance optimum.In short, the compressive strain induced by the optimal doping can result in a d-band center with a suitable position, which has the optimal desorption energy for the crucial step of PEC oxidation of HBZ.

Structure-Activity Relationship of the Photoelectrode
In situ ATR-FTIR spectra were employed to elucidate the reaction mechanism of PEC oxidation of 1,2-diol (HBZ) over FFM2 (Figure 5a,b).The stretching vibrations associated with the distinctive skeletal structure of aromatic compounds were observed at wavenumbers of 1650, 1595, and 1230 cm À1 .The peaks at 2932 and 2850 cm À1 could be ascribed to the symmetric C─H vibrations of the phenyl group.The broad spectral region centered at around 3450 cm À1 was attributed to the vibrational mode associated with the stretching vibration of -OH and the hydrogen bonding of water molecules adsorbed on the surface.The presence of benzoin species could be detected by the signals at 1390 and 1230 cm À1 .The peak intensity tendency showed that benzaldehyde as the intermediate played a crucial role in the HBZ oxidation process.The strong absorption peak at 1080 cm À1 could be attributed to the HBZ.The peaks at 1650 cm À1 gradually increased as time progressed, indicating the formation and continuous accumulation of more BEN on FFM2 photoelectrode.These results demonstrated that the 1,2-diol (HBZ) could swiftly adsorb on FFM2 and its subsequent activation to produce BEN, indicating that the optimized FFM2 photoelectrode had a pronounced oxidation capacity.
The free energy profiles of PEC oxidation of the 1,2-diol (HBZ) and the related reaction intermediates on FFM1, FFM2, and FFM3 with different compression strains (0.8%, 2%, and 3.2%, respectively) are represented in Figure 5c.The adsorption of 1,2-diol (HBZ) on FFMx served as the outset of the PEC oxidation (Figure 4d), which was characterized as exothermic behavior.It was suggested that the adsorption of the 1,2-diol (HBZ) was a spontaneous thermodynamic process on FFMx with different lattice strains.The 1,2-diol was absorbed through a proton dehydrogenation, and the C─H activation energy (0.92 eV) of FFM2 was lower than that of FFM1 (1.04 eV) and FFM3 (1.19 eV).Furthermore, it was evidently observed that the BEN dissociation provided the maximum free energy variation (2.48-2.68eV), pointing out that the BEN dissociation determined the overall reaction rate during the HBZ oxidation.FFM2 with the compression lattice strain of 2% possessed the minimal desorption energy barrier of 2.48 eV, which was more beneficial for HBZ oxidation.Meanwhile, FFM1 (0.8%) and FFM3 (3.2%) needed to surpass the higher reaction barriers of 2.62 and 2.68 eV, respectively.The results indicated that optimized lattice strain was vital for decreasing the C─H activation and the rate-controlling free energy in HBZ oxidation, in line with the obtained experimental data.Therefore, a feasible reaction mechanism for PEC oxidative C─C breaking of 1,2-diol can be delineated.After adsorption of 1,2-diol (HBZ) to the photoelectrode interface, the C─H bond would be activated to produce benzoin.The photoexcited electron-hole pairs are generated at the photoelectrode interface under exposure to simulated solar irradiation, with the electrons being directed through the conductive substrate toward the electrode to produce H 2 .The photoexcited holes attack 1,2-diol (HBZ) to produce a portion of benzoin, followed by undergoing C─C bond breaking at the photoelectrode interface to give two molecules of BEN.In a word, the excellent PEC oxidation properties of FFM2 can be ascribed to the appropriate compressive strain that allows for optimal positioning of the d-band center, thereby enabling and facilitating both the C─H activation and the product desorption from the photoelectrode.

Conclusion
In summary, a surface doping protocol was developed to be markedly valid for inducing local lattice distortion in the bulk phase of the photoelectrode substrate, as demonstrated by control experiments, in situ characterizations, and theoretical calculations.The strain effect was quantitatively described, and gradient doping on the photoelectrode surface was closely correlated with the enhanced PEC performance, with optimal photocurrent observed at 3.41 mA cm À2 .Diverse lattice distortions, influenced by the Mn doping quantity, led to various compressive strains in the order of FFM1 (0.8%) < FFM2 (2%) < FFM3 (3.2%), hereby regulating the d-band center of the photoelectrode.The modulated d-band center of the photoelectrode enabled FFM2 to have an optimum adsorption/desorption capacity for 1,2-diol (HBZ), intermediates, and the product.The FFM2 was illustrated to be highly efficient for PEC oxidative C─C bond breaking various 1,2-diols, selectively furnishing aromatic carboxylic acids (≥95.1% yields) or aromatic ketones (≥94.5% yields), coupled with hydrogen production of 1180 μmol cm À2 .This work unveils the conformational association between lattice strain engineering and PEC activity of the photoelectrode, offering a new outlook for C─H activation and C─C bond cleavage involved in organic and biomass transformations.

Experimental Section
Preparation of FF Photoelectrode: The FF was synthesized using a previously reported hydrothermal method that is widely adopted [54] The precursor solution was made up of 0.15 M ferric chloride hexahydrate (FeCl 3 •6H 2 O), 1 M sodium nitrate (NaNO 3 ), and 0.037 M NH 4 F. Before adding the solution to a high-pressure vessel lined with polytetrafluoroethylene, the conductive FTO surface was positioned to face downward, and the pH was adjusted to 1.5 using HCl.The vessel containing 20 mL precursor solution was put into the oven and maintained at 120 °C for a duration of 10 h followed by gradual cooling to room temperature.A uniformly colored FeOOH coating formed on the FTO surface, which was then rinsed with deionized water to remove any remaining salt content, and then allowed to naturally dry in the air.FF was finally prepared by air annealing at 550 °C for 2 h, subsequently at 720 °C for 10 min.
Preparation of FFMx Photoelectrode: A specific amount of (CH 3 COO) 2 Mn (10 mol%, 20, and 30 mol% of Mn relative to Fe in FeCl 3 •6H 2 O) was added into 10 mL ethylene glycol monomethyl ether and stirred at room temperature for 30 min to act as the dopant source.Then, 50 μL of Mn and Fe precursors was spin-coated onto the as-fabricated FF photoelectrode at 1000 rpm for 15 s, and dried in the air.Finally, the three photoelectrodes with different Mn molar content of 10, 20, and 30 mol% (relative to Fe) were annealed in air at 350 °C for 2 h, denoted as FFM1, FFM2, and FFM3, respectively.
Photoelectrode Characterization: Detailed information is provided in the experiment section of the Supporting Information.
Product Analysis for PEC Oxidation of 1,2-Diols: The PEC oxidative conversion of 1,2-diols was performed in a two-chamber three-electrode cell with 10 mL 1 M NaOH solution containing 2.5 mM 1,2-diol in the anode and 10 mL NaOH in the Pt cathode.The double-chamber PEC reactor was fitted with a Nafion 117 proton exchange membrane.The simulated sunlight used a xenon lamp (PLS-FX300HU).The conversion of 1,2diol and the selectivity/yield of products were quantitatively performed by Agilent 1260 II HPLC equipped with a C18 column.Simultaneously, H 2 generated at the cathode was efficiently collected through vacuum dewatering.
Computational Method: All the DFT calculations were performed using the Vienna Ab Initio Package to provide information on the properties of the photoelectrodes.More details are available in the experiment section of the Supporting Information.

Figure 2 .
Figure 2. a) XRD patterns, and b) the magnified XRD patterns of (110) peak of FF and FFMx.c) W-H plots of FF and FFMx.d) Full XPS spectra of FF and FFMx.High-resolution XPS spectra of e) Fe 2p and f ) Mn 2p in FF and FFMx.g) XPS depth profiles of Mn/Fe atomic ratio from 0 to 100 nm.h) Relative atomic ratio of Mn/Fe analyzed by XPS and XRF.i) High-resolution XPS spectra of O 1s in FF and FFMx.

Figure 3 .
Figure 3. a) UV-vis spectra, b) Tauc plots, c) LSV curves, d) estimated ECSA, and e) EIS plots and equivalent circuit of FF and FFMx.f ) OCP decay profiles, g) PL absorption spectra, h) time-resolved PL decay plots, and i) stability test curves of FF and FFMx.
, Supporting Information.The variation of bond lengths for different lattice strains of FF is clearly represented in Figure 4c.According to the Sabatier principle and d-band center theory, the compressive strain has the ability to modify the d-band width, while the d-band center holds a steady band filling state and with the d-band shift up and

Figure 4 .
Figure 4. a) The activity of PEC oxidative C─C bond cleavage of HBZ on FF and FFMx.b) The oxidative C─C bond breaking of HBZ using photo-/electro-/ PEC systems.c) Different compressive strain and d-band center on FFMx.d) The adsorption energy of HBZ on FFMx.e) Catalytic results of PEC conversion of different 1,2-diols.Reaction conditions: HBZ (2.5 mM), water (1 M NaOH), and reaction time (20 h).

Figure 5 .
Figure 5.In situ FTIR spectra of PEC oxidation of HBZ over FFM2 as a function of a) external bias from 0.5 to 2.1 V versus RHE and b) reaction time.c) The free energy profiles for PEC oxidation of 1,2-diol (HBZ) to BEN on FFMx.