Temperature‐Dependent Photoredox Catalysis for CO2 Reduction Coupled with Selective Benzyl Alcohol Oxidation over ZnIn2S4/In2O3 Heterostructure

CO2 reduction (CO2RR) with selective benzyl alcohol (BA) oxidation in a single photoredox reaction can simultaneously utilize photogenerated electrons and holes to realize efficient production of fuels and value‐added chemicals. Herein, a unique 2D/1D ZnIn2S4/In2O3 (ZIS/In2O3) heterostructure is developed displaying outstanding performance for photoredox catalysis. As is unequivocally illustrated by various advanced ex situ/in situ characterizations and theoretic calculations, the notable catalytic performances originate from the built‐in interfacial electric field within the ZIS/In2O3 heterostructure, which strongly ameliorates the separation and transport of charge carriers. Remarkably, the catalytic activity can further be boosted after coupling the additional thermal treatments, and the product selectivity is highly temperature dependent. Thereby, the precise formation of targeted products, which can serve as valuable industrial products and fuels can be controlled by changing the reaction temperatures, including obtaining the syngas with different H2/CO ratios from CO2 reduction, as well as benzaldehyde, hydrogenated benzoin, and dibenzyl ether (never reported for BA during photocatalysis) from the BA conversion. This study offers a constructive and inspiring contribution to reasonably developing photoredox catalysts, which can fully utilize photogenerated electrons and holes, as well as demonstrate how to controllably yield the targeted products by coupling thermo‐ and photoredox catalysis.


Introduction
The environmental crisis has gained global attention with the ever-increasing CO 2 emis-CO 2 reduction (CO 2 RR) with selective benzyl alcohol (BA) oxidation in a single photoredox reaction can simultaneously utilize photogenerated electrons and holes to realize efficient production of fuels and value-added chemicals.Herein, a unique 2D/1D ZnIn 2 S 4 /In 2 O 3 (ZIS/In 2 O 3 ) heterostructure is developed displaying outstanding performance for photoredox catalysis.As is unequivocally illustrated by various advanced ex situ/in situ characterizations and theoretic calculations, the notable catalytic performances originate from the built-in interfacial electric field within the ZIS/In 2 O 3 heterostructure, which strongly ameliorates the separation and transport of charge carriers.Remarkably, the catalytic activity can further be boosted after coupling the additional thermal treatments, and the product selectivity is highly temperature dependent.Thereby, the precise formation of targeted products, which can serve as valuable industrial products and fuels can be controlled by changing the reaction temperatures, including obtaining the syngas with different H 2 /CO ratios from CO 2 reduction, as well as benzaldehyde, hydrogenated benzoin, and dibenzyl ether (never reported for BA during photocatalysis) from the BA conversion.This study offers a constructive and inspiring contribution to reasonably developing photoredox catalysts, which can fully utilize photogenerated electrons and holes, as well as demonstrate how to controllably yield the targeted products by coupling thermoand photoredox catalysis.may concurrently generate new issues such as the increment of system cost and the formation of undesired oxidative products. [18]o this end, integrating CO 2 RR with selective organic oxidation in a single photocatalytic system has emerged as an effective alternative, in which the photogenerated holes are utilized for the selective oxidation of various biomass-derived organic molecules into value-added chemicals, accompanied by the cost drop and concurrent enhancement in the working efficiency of the overall photocatalytic process. [35]For example, oxidizing benzyl alcohol (BA) into value-added benzaldehyde (BAD) or higher value C-C coupled hydrogenated benzoin (HB) is highly appealing in the aspect of industrial application.The BAD is widely applied in industries related to dyeing, perfumery, and pharmaceuticals, [3] while the HB can serve as an important feedstock for the synthesis of additives and dyestuff. [4][38] Among them, ZnIn 2 S 4 (ZIS)-based materials have been extensively investigated, which particularly exhibit the exceptional ability to oxidize BA into BAD or C-C coupled products. [36,39]owever, it is necessary to point out that the corresponding photocatalytic activities over pristine ZIS are severely limited by its fast recombination rate of photogenerated electrons and holes.To overcome this obstacle, numerous efforts have been made, including constructing heterojunctions, [16] altering Zn/In ratio, [40] and doping foreign metal atoms. [18]In particular, the construction of heterojunctions can not only address the issues about the separation efficiency of photogenerated charge carriers but also broaden the light absorption capacity, which has been well illustrated by the previous reports on ZnS/ZnIn 2 S 4 , [15] and Ni 12 P 5 /ZnIn 2 S 4 . [41]Unfortunately, despite huge progress achieved last few years, for the heterostructured ZIS-based photocatalyst, 1) the BAD normally dominates among the oxidative products, while the performance of C-C coupled products (especially HB) by BA oxidation is still poor and far from industrial interest.Besides, 2) the selectivity of the products of BA oxidation has hardly been controlled.
[44][45][46][47] A thermocatalytically favorable catalyst is usually featured with a wide bandgap, limiting its ability of light adsorption, and thereby a declined photothermal conversion performance.In this context, indium sesquioxide (In 2 O 3 ) can be a reasonable choice and can serve as a characteristic thermocatalyst with good performance for the reduction of CO 2 , whereas the wide bandgap of its pristine phase distinctly hampers its application in photothermal catalysis. [44,46,48]To expand the limited light adsorption ability caused by the wide bandgap, various strategies have been developed, such as vacancy engineering, [46][47][48] foreign atom doping, [47] and heterostructure construction, [17,48] endowing the modified In 2 O 3 with obviously improved photothermal CO 2 RR performance.Especially, Ozin et al. reported an optimized In 2 O 3 heterostructure built by the crystalline stoichiometric In 2 O 3 core coated with the amorphous nonstoichiometric In 2 O 3Àx domains, thus performing a much-enhanced photothermal CO 2 RR ability, during which the high temperature is required. [48]Herein, it is worth mentioning that coupling In 2 O 3 with other functional ZIS has been substantiated to effectively elevate the photogenerated charge separation efficiency and therefore, an improved photocatalytic CO 2 RR activity is expected. [49]Bearing the above concerns into consideration, one can naturally envision that integrating ZIS (for selective oxidation of BA) and In 2 O 3 (for CO 2 RR) to construct ZIS/In 2 O 3 heterojunction could produce the valueadded products formation in a single photothermal redox cycle with high efficiency and selectivity.To the best of our knowledge, such a hybrid catalyst has never been investigated in the field of photothermal catalysis.
In this work, a 2D/1D ZIS/In 2 O 3 heterostructured catalyst was developed, and for the first time, used for the synchronous photothermal catalytic BA oxidation and CO 2 RR in a combined system.The combination of a series of ex situ/in situ and theoretical calculations unveiled that the coupling of In 2 O 3 with ZIS induced the formation of a built-in interfacial electric field (IEF), profoundly promoting the separation efficiency of the photogenerated electrons and holes, delaying the recombination of charge carriers, and thus enabling the efficient conversion of both BA and CO 2 in the absence of organic solvents and photosensitizers.More interestingly, ZIS/In 2 O 3 displayed distinct photothermal activity, which can also effectively control the selectivity of the products obtained from BA oxidation and CO 2 RR, respectively, by altering the reaction temperatures.Specifically, the highly valuable C-C coupling product HB can be obtained from BA oxidation within the mild temperature range (less than 200 °C), while BAD (a production rate of 90.88 mmol g À1 h À1 ) and dibenzyl ether (BE, obtained from the dehydration of BA [50] ; a production rate of 236.83 mmol g À1 h À1 in the present case) were the primary products at 300 °C.Note that BE is widely utilized in the industries of surface coating, rubber, and textile, [51] while producing BE from BA during photo/thermal catalysis has never been reported previously.In contrast, the ratio of H 2 and CO in the CO 2 RR product syngas was progressively regulated from 102 to 0.07 along with the increment of the reaction temperature from 100 to 300 °C, covering the entire ratio range of H 2 /CO where syngas can serve as industrial feedstocks. [28]We believe our current work could pave a novel route to fully utilize the thermo-photogenerated electrons and holes in one redox cycle for effective solar energy conversion, carbon emission reduction, as well as valueadded chemicals and fuel production.Also, the exciting discovery of the temperature-selectivity relation is expected to be inspiring for the follow-up works in the field of photothermal catalysis.The schematic illustration of the synthetic process for 2D/1D ZIS/In 2 O 3 hierarchical heterostructure is presented in Figure 1a.Initially, the In-based metal-organic framework (MOF), MIL-68(In), with a morphology of hexagonal prism was successfully synthesized at a moderate temperature (Figure 1b and S1, Supporting Information), [49] followed by the calcination of the as-prepared MIL-68(In), to produce 1D hollow In 2 O 3 tubes (Figure 1c   (Figure 2a).Hereafter, the optical properties of the ZIS/In 2 O 3 together with the reference samples were compared.In Figure 2b, UV-vis diffuse reflection spectroscopy (DRS) revealed that In 2 O 3 has limited light absorption in the visible region.Intriguingly, after growing ZIS nanosheets, the as-prepared composite exhibited strong visible light absorption at about 450 nm, probably due to the excellent visible light absorption of ZIS.Moreover, the corresponding bandgaps (E g ) of In 2 O 3 and ZIS were about 2.9 and 2.60 eV, respectively, which was uncovered by their respective Tauc plots (Figure 2c).Meanwhile, the conduction band (CB) positions of the ZIS and In 2 O 3 were determined by the associated Mott-Schottky (M-S) plots.As illustrated in Figure 2d,e, the flat band potentials (V fb ) of ZIS and In 2 O 3 were À0.77 and À0.64 V (vs NHE, pH = 7), respectively.For n-type semiconductors, the CB potential is generally more negative by about 0.1 V than the V fb .Therefore, the potentials of CB of In 2 O 3 and ZIS were estimated as À0.74 and À0.87 V (vs NHE, pH = 7), respectively.According to the formula: E g = E VB -E CB , the valence band (VB) position of the ZIS and In 2 O 3 were calculated as 1.73 and 2.16 V (vs NHE, pH = 7), respectively.Additionally, the VB positions of the ZIS and In 2 O 3 were also defined using VB -X-ray photoelectron spectroscopy (VB-XPS), with values of 1.71 and 2.14 eV, respectively (Figure 2f,g), consistent with the above calculations.Consequently, the positions of band edges for ZIS and In 2 O 3 semiconductors were estimated (Figure 2h).Obviously, the ZIS/In 2 O 3 heterojunction has a staggered electronic band structure, which would facilitate the separation and transport of the photoexcited electron-hole pairs based on the band-band transfer (P1) or the Z-scheme transfer (P2) mechanisms, therefore being able to drive CO 2 RR reaction by its suitable redox potentials. [49]fterward, the surface chemical states of ZIS, In 2 O 3, and ZIS/In 2 O 3 samples were investigated by XPS.As is shown in Figure 3a, the Zn 2p spectra of pristine ZIS exhibited peaks around 1,022.4 and 1,045.6 eV, corresponding to the Zn 2p 3/2 and 2p 1/2 of Zn 2þ , respectively.Figure 3b shows that the characteristic of In 3d 5/2 and 3d 1/2 peaks (assigned to the presence of In 3þ ) of bare ZIS can be observed at around 445.1 and 452.7 eV, respectively. [49]The S 2p XPS spectra of the ZIS can be deconvoluted into 2p 1/2 and 2p 3/2 contributions with a splitting energy of 1.2 eV (Figure 3c).Notably, all the characteristic peak positions in Zn 2p, In 3d, and S 2p XPS spectra of ZIS/In 2 O 3 exhibited negative shifts compared with those of the pristine ZIS.On the contrary, the binding energies of In 3d and O 1s XPS for ZIS/In 2 O 3 both positively shifted in comparison to those for the bare In 2 O 3 (Figure 3b,d).This phenomenon indicated a change in the chemical environments of Zn, In, and S for both ZIS and In 2 O 3 after their combination as the ZIS/In 2 O 3 heterojunction.Generally, the negative and positive shifts of the XPS peaks tend to represent the increase and decrease of electron density, respectively. [30,31,52]Therefore, the binding energy shifts manifested that the charge carriers (electrons for n-type semiconductors) migrated from In 2 O 3 into ZIS.When the equilibrium is achieved, the IEF can be built at the interfaces of ZIS/In 2 O 3 heterostructure with a direction of In 2 O 3 !ZIS.

Photothermal Catalytic Performance of ZIS/In 2 O 3
The catalytic performance of ZIS, In 2 O 3, and ZIS/In 2 O 3 for CO 2 RR coupled with BA oxidation was tested without solvent and additional photosensitizers in a fixed reactor (Figure 4a).Note that in the comparison experiments, no products were detected in the reaction system without catalyst, or with catalyst under the dark environments (at room temperature).When irradiated under visible light without additional heating treatment, the pristine In 2 O 3 and ZIS both exhibited photocatalytic redox activities for CO 2 RR and BA oxidation (Figure 4b-e).In detail, the photoexcited electrons of In 2 O 3 and ZIS can reduce CO 2 into syngas (H 2 mixed with CO) based on their band energy structures (Figure 2h).In contrast, the photoexcited holes of In 2 O 3 catalyzed the selective oxidation of BA into the product BAD only, while the bare ZIS could convert BA into both BAD and HB.The difference in the BA oxidation products between In 2 O 3 and ZIS might be owing to their different VB positions and thereby nonidentical oxidizing abilities of the photogenerated holes.Notably, the gaseous and liquid products of such a photocatalytic redox reaction driven by heterostructured ZIS/In 2 O 3 were boosted compared with those of pure ZIS and In 2 O 3 (Figure 4b-e), which probably originated from the more efficient separation of the photoexcited electron-hole pairs and more favorable transfer of charge carrier for ZIS/In 2 O 3 .Moreover, remarkably, the production rate of the BA oxidation products in the current case, HB and BAD, reached 7.23 and 1.75 mmol g À1 h À1 on ZIS/In 2 O 3 , respectively, which were around 2.92 and 1.65 times higher than those of HB (2.48 mmol g À1 h À1 ) and BAD (1.06 mmol g À1 h À1 ) catalyzed by bare ZIS.This implied that the separation-transfer mechanism of the photogenerated charge carriers over the ZIS/In 2 O 3 heterojunction was the one based on the band-band transfer (P1 pathway in Figure 2h).This is because, provided that the mechanism of ZIS/In 2 O 3 corresponded to the Z-scheme transfer, the photogenerated holes of ZIS would accordingly be recombined with the photogenerated electrons of In 2 O 3 .As a result, the photogenerated holes of In 2 O 3 should be reserved and a stronger oxidizing ability would be realized.Nevertheless, the experimental findings showed that the governing product of BA oxidation on ZIS/In 2 O 3 during photocatalysis was HB, and the BA molecules could be merely converted into BAD on bare In 2 O 3 (Figure 4d,e).Hence, it can be concluded that the band-band transfer mechanism dominated the ZIS/In 2 O 3 heterojunction.Then, we specially investigated the thermal effects on the performances of CO 2 RR and BA oxidation within the above photocatalysis system, where the reaction temperature was set as 100 °C.As expected, among the three catalysts, ZIS/In 2 O 3 heterojunction still generated the best production rate for all kinds of gaseous and liquid products (Figure 4b-e).Specifically, it can be found that bare ZIS tended to catalyze the production of H 2 by reducing the protons from the dehydrogenation of BA, while the product CO of CO 2 RR was greatly inhibited.Regarding the pristine In 2 O 3 , under photothermal conditions, the H 2 generation was inhibited to a certain extent while the CO product was significantly promoted, illustrating that heating treatment could especially facilitate photocatalytic CO 2 RR into CO on In 2 O 3 , in agreement with the previous report. [46]When it comes to the ZIS/In 2 O 3 heterostructure, although the H 2 production in our photocatalysis system was only slightly elevated by the additional heating treatment, the production of CO from CO 2 RR was substantially improved on this composite compared with the situations of mere photocatalysis (Figure 4b,c).In the aspect of BA oxidation products under photothermal conditions (Figure 4d, e), on one hand, the generation rate of both BAD and HB catalyzed by the ZIS/In 2 O 3 composite was enhanced compared with the cases without thermal treatments.On the other hand, the main BA oxidation product on heterostructured ZIS/In 2 O 3 was still HB.Combined with the fact that the photothermal catalytic oxidation of BA by bare In 2 O 3 still failed to produce any HB (Figure 4e), we could propose a conclusion that the ZIS/In 2 O 3 heterostructure behaved with excellent activity to catalyze both CO 2 RR and BA oxidation in a single photothermal catalytic system, in which the charge carriers primarily followed a band-band transfer mechanism.During the catalytic process, the ZIS and In 2 O 3 component in this heterostructured composite were mainly responsible for the BA oxidation and CO 2 RR, respectively.The above results were consistent with the selectivity findings on the production of H 2 , CO, HB, and BAD for these three catalysts, whose details were elaborated in Figure S7, Supporting Information.Besides, another point worth highlighting is that the ZIS/In 2 O 3 heterostructure owned an exceptional catalyst reusability, well retaining its catalytic activity after five successive cycles (Figure S8, Supporting Information).
Based on the above results and discussions, we also explored the influence of different temperatures on the activity of photothermal catalysis by the representative catalyst, heterostructured ZIS/In 2 O 3 .For the gaseous products, the production rate of CO increased gradually with the increment of the reaction temperature (when the temperature reached 300 °C, a maximum CO rate of 6.41 mmol g À1 h À1 was achieved), while an exactly opposite trend was observed for the H 2 production (Figure 5a).Correspondingly, the molar ratio of H 2 and CO was tuned from 102 to 0.07 by increasing the temperature from 100 to 300 °C, which apparently offered a new method to prepare syngas with different H 2 /CO ratios satisfying the requirements of various industrial applications.For the liquid products, the highly valuable C-C coupled product HB was the main product below 200 °C (Figure 5b), while a new product of BA conversion, BE, together with BAD started to surge from the temperature of 250 °C, and dominated when the temperature reached 300 °C.In addition, the above trends can be further proved by the selectivity results of the gaseous and liquid products from both CO 2 RR and BA oxidation catalyzed by ZIS/In 2 O 3 , as illustrated in Figure S9, Supporting Information.To shed more light on the thermal catalysis to enable the generation of BE, BA conversion was catalyzed at 300 °C in the absence of light irradiation (Figure S10, Supporting Information), from which a similar BE productivity can be obtained as that achieved from the photothermal conversion of BA.All the above results demonstrated that the product selectivity of the BA transformation during photothermal catalysis was temperature-dependent.At high temperatures, the thermocatalysis governed the reaction of BA conversion (where BE was predominant) on ZIS/In 2 O 3 composites.Moreover, in light of the established thermocatalytic properties of pristine In 2 O 3 , we conducted a rigorous evaluation of its redox catalytic activity, with a specific focus on its performance under photothermal conditions at an elevated temperature of 300 °C.Notably, the catalytic performance of bare In 2 O 3 was found to be significantly lower when compared to that of the heterostructured ZIS/In 2 O 3 (Figure S11, Supporting Information).Note that as is shown in Table S1, Supporting Information, the generation rates of HB and BE from BA conversion catalyzed by ZIS/In 2 O 3 in the current study were higher than those in most reported works.Additionally, the apparent quantum yield (AQY) of ZIS/In 2 O 3 at 400 nm was 6.91%, and the corresponding data and its calculation can be found in Figure S12, Supporting Information.Moreover, the ZIS/In 2 O 3 exhibited high stability in this photothermal redox catalysis system, which can be discovered by the fact that the activity of ZIS/In 2 O 3 did not decline visibly after five successive catalysis cycles at 300 °C (Figure 5c,d).The color (Figure S13, Supporting Information), XRD pattern (Figure S14, Supporting Information), and TEM image (Figure S15, Supporting Information) of the ZIS/In 2 O 3 before and after the photothermal catalysis uncovered that the phase, crystallinity, and morphology of this catalyst was retained well, further confirming the great reusability and huge practical application potentials of ZIS/In 2 O 3 .

Catalytic Mechanisms of ZIS/In 2 O 3
To gain deeper insights into the separation and transport of the photogenerated charge carriers of the targeted sample, ZIS/In 2 O 3 , transient photocurrent, electrochemical impedance spectroscopy (EIS), and time-resolved photoluminescence (TRPL) measurements were conducted. [38]Under light illumination, ZIS/In 2 O 3 hybrid exhibited a higher photocurrent density (Figure 6a) and a smaller charge transfer resistance (Figure 6b) than those of bare ZIS and In 2 O 3 , indicating that the ZIS/In 2 O 3 possessed much-enhanced charge separation efficiency and transfer ability. [49]For photocatalysts, the shorter emission lifetime represents that the photogenerated charge carriers can be better separated, [52,53] and as expected, Figure 6c showed that the average emission lifetime of ZIS/In 2 O 3 (3.71ns) was shorter than that of ZIS (6.72 ns) and In 2 O 3 (9.23 ns).These characterizations well certified that the construction of ZIS/In 2 O 3 heterostructure endowed this composite with a highly ameliorated charge separation and transport, therefore achieving the optimum catalytic activity.
To gain deeper insights into the differences in photocatalyst activity, the holes (h þ ) and electron (e À ) utilization rates during irradiation were further investigated using in situ electron paramagnetic resonance (EPR) measurements with or without benzyl alcohol (BA). [54,55]As shown in Figure 7a-f, in the characterization environments for detecting e À (or h þ , Figure 7g-l), a triple characteristic peak with the same intensity was observed in the  dark state, originating from the 2,2,6,6-Tetramethylpiperidine-1-oxyl (TEMPO) radical capture reagent.As TEMPO can effectively capture e À and h þ , the intensities of the signals decreased when extending the illumination time (Figure 7a-c and g-i). [41]n both cases of detecting the e À and h þ , an identical trend can be found: the lowest intensities of EPR signals were observed when ZIS/In 2 O 3 was subject to the light irradiations, followed by ZIS and In 2 O 3 , illustrating that the photogenerated e À and h þ on ZIS/In 2 O 3 can be separated more efficiently than on bare ZIS and In 2 O 3 .This observation was consistent with the results of photoelectrochemistry and TRPL tests.When the reactant BA was added into the test system (Figure 7d-f and j-l), for each probed catalyst, the signal intensity was distinctly lower than that in the absence of BA at each illumination time interval (Figure 7a-c and g-i), which was induced by the transfer of the photoexcited holes from photocatalysts to reactive BA molecules, in turn, the utilization of electrons was elevated (during the in situ EPR characterizations more of them were captured by the TEMPO).[41] Of note, the signal intensities of EPR against ZIS/In 2 O 3 were still the lowest (Figure 7f,l), clearly manifesting its excellent utilization efficiency of photogenerated e À and h þ during the catalytic process.
Furthermore, in situ Fourier transform infrared spectroscopy (FTIR) was employed to determine the reaction intermediates/ how fast the final products can be formed during the photothermal CO 2 RR and BA oxidation catalyzed by the ZIS/In 2 O 3 catalyst at 300 °C (Figure 8).On one hand, it can be found that even within the initial stage of the light irradiation, nearly all the presented FTIR peaks were obviously intensified, signifying the rapid occurrence of catalysis reactions.On the other hand, in the case of the reduction reaction, the observed peaks at 1,069, 1,273, 1,312, 1,654, and 1,720 cm À1 can be attributed to the presence of CHO* (intermediate of CO 2 reduction), carboxylate (CO 2 À ), monodentate carbonate (m-CO 3 2À ), and carbonate (CO 3 2À ), respectively. [12,18]The peak at 1,450 cm À1 was assigned to aromatic modes δ(C-C), while the sharp peak at 1,700 cm À1 was attributed to the vibration of carbonyl stretch mode ν(CO) in BAD.Moreover, the FTIR bands at 1,167 and 1,654 cm À1 were attributed to the COOH*, which is a crucial intermediate for the  formation of CO. [12,18] The reaction intermediates of BA conversion were also detected in the FTIR spectra.The peaks of 1,583 and 1,597 cm À1 were assigned to the C = C skeleton vibration of mononuclear aromatic hydrocarbons, whereas the peak at 3,071 cm À1 was assigned to C-H stretching vibration on the benzene ring. [56]These FTIR bands might originate from the presence of residual BA, as well as the intermediates of products (BAD, HB, and BE).The peak at 1,689 cm À1 belonged to vibrations of C = O groups of BAD. [56]The stretching vibrations of O-H and C-H in HB could be found at 3,511 and 2,870 cm À1 , respectively. [56]The characteristic peak of C-O-C in BE was presented at 1,113 cm À1 . [57]Moreover, 13 C isotope experiments were also carried out to confirm the carbon source.To this end, the photothermal redox catalysis of CO 2 RR and BA oxidation was tested at 300 °C.Especially, except the CO 2 was replaced by 13 CO 2 , the other conditions were kept the same.Figure S16, Supporting Information, clearly showed that the 13 CO was detected under the 13 CO 2 atmosphere, indicating the CO was generated from CO 2 RR.These results further corroborated the conversion of CO 2 and BA catalyzed by ZIS/In 2 O 3 heterostructure into the products that have been determined in photothermal chemistry.
Besides, to further confirm the photogenerated charge transfer within the ZIS/In 2 O 3 heterojunction, density functional theory (DFT) calculations were performed to estimate the Fermi level of ZIS and In 2 O 3 .The work functions (W F ) of ZIS and In 2 O 3 were calculated as around 4.7 and 4.5 eV, respectively (Figure 9a,b).To further demonstrate the W F magnitude between ZIS and In 2 O 3 , ultraviolet photoelectron spectroscopy (UPS) tests were also carried out, by which, the W F of ZIS and In 2 O 3 were determined as around 5.23 and 4.84 eV, respectively (Figure S17 and S18, Supporting Information).Although there were some differences between theoretical calculations and experimental measurements, the W F of ZIS was always larger than that of In 2 O 3 , indicating that the distance from the Fermi level to the vacuum level for ZIS was greater than that for In 2 O 3 (before contact).When the two components are in contact with each other, the majority of the carriers (electrons, as reflected by the XPS findings in Figure 3) would rapidly diffuse from In 2 O 3 into ZIS until the Fermi levels of these two components were equal.Also, this process led to an upward (In 2 O 3 ) and downward (ZIS) bending of the band edges, and a built-in IEF was established at the interface of the ZIS/ In 2 O 3 heterostructure, with a direction from In 2 O 3 to ZIS.Consequently, upon under the illumination, the photoexcited holes and electrons will transfer along and against the direction of the built-in IEF, respectively (Figure 9c).Based on these results, the charge transfer behavior in ZIS/In 2 O 3 interface was again verified to follow the band-band transfer mechanism.

Conclusion
In summary, we developed the 2D/1D ZIS/In 2 O 3 heterostructure as a photoredox catalyst, which displayed excellent performance for concurrent CO 2 RR (governed by In 2 O 3 ) and selective BA oxidation (governed by ZIS) in the absence of solvents and photosensitizers.Various advanced ex situ and in situ characterizations, as well as theoretic DFT calculations, confirm that its superior catalytic performance originated from the construction of heterostructure, which induced the emergence of a built-in IEF, conducive to improving the separation efficiency of the photogenerated electrons and holes, and accordingly delaying their recombination.Additionally, coupling the additional thermal treatment can further elevate the product production rate during such a photoredox catalysis.Remarkably, the selectivity of those products presented high-temperature dependence, thereby the targeted products can be controllably altered by changing the reaction temperatures.In detail, the CO/H 2 ratio in the CO 2 RR product syngas can be precisely tuned with the temperature increment, satisfying the requirements of various industrial applications.In contrast, the selectivity and production rate of value-added BAD and C-C coupled HB from BA oxidation can also be thermally controlled.Especially, BE, a significant industrial feedstock, was obtained from photothermal catalysis of BA conversion at a high reaction temperature for the first time.We believe our current work can contribute to inspiring the scientific community to more rationally develop photothermal redox catalysts, which can fully utilize the photogenerated electrons and holes for the effective conversion of solar energy, as well as the green synthesis of valuable fuels and industrial chemicals.
Synthesis of 1D Hexagonal MIL-68(In) Prisms: The 1D MIL-68(In) prisms were synthesized using a facile solvothermal method with some modifications as reported by Cho et al. [49,58] Typically, 60 mg In(NO 3 ) 3 •xH 2 O and 60 mg terephthalic acid were dissolved in 40 mL DMF and stirred for 2 min.The resulting solution was then placed in an oil bath at 120 °C for 30 min.After cooling down to room temperature, the white precipitate was filtrated and washed with ethanol three times.
Synthesis of 1D Hollow In 2 O 3 Tubes: The as-obtained MIL-68(In) prisms were annealed in the air at 120 °C for 2 h with a heating rate of 5 °C min À1 .Subsequently, they were further annealed at 500 °C for 2 h with the same heating rate to synthesize 1D hollow In 2 O 3 tubes.
[61][62] In a typical synthesis process, 10 mL H 2 O (pH = 2.5) and 75 mg as-prepared In 2 O 3 were added into a three-necked flask (20 mL volume) and stirred for 30 min.After that, 27.2 mg ZnCl 2 , 44.2 mg InCl 3 , and 30 mg thioacetamide were added and stirred for 5 min.The as-obtained mixture was subsequently put into an oil bath at 80 °C under stirring and left for 2 h.The product was washed with ethanol three times and dried at 60 °C in a vacuum to yield the ZIS/In 2 O 3 sample.The molar ratio of ZIS to In 2 O 3 was 0.7.For comparison, the bare ZIS nanosheets were synthesized under the same reaction conditions without the addition of In 2 O 3 as the substrate.
Characterizations: Powder XRD patterns of the probed samples were collected by an X-ray diffractometer (Bruker D8 Advance).UV-vis DRS was measured using a spectrophotometer (Shimadzu UV-3600) with Photocatalytic Activity Tests: The photothermal redox catalysis of CO 2 RR and BA oxidation was tested in a visible high-temperature and highpressure reactor.Typically, 50 mg photocatalyst and 10 mL benzyl alcohol were added to the reactor, which was then ultrasonicated for 10 min to ensure the even dispersion of the catalyst.The reactor was then vacuumed to remove air, and subsequently, 250 mL CO 2 gas was introduced.The reactor was stirred in the dark for 30 min to achieve a dynamic dissolution equilibrium of CO 2 .The Xe lamp (UV-vis) was used to simulate sunlight and provide the light source for the reactor, while the reaction temperature was controlled by a high-pressure reactor heater.Gaseous products were collected using a gas injection needle with the quantitative ring.CO and other products were detected by a flame ionization detector (FID), while H 2 was detected by a thermal conductivity detector (TCD) gas chromatography.The liquid products were collected and diluted using acetonitrile, followed by being detected with high-performance liquid chromatography (HPLC).
Photoelectrochemical Measurements: The photoelectrochemical tests were carried out on a three-electrode system (CHI-660E).A Pt wire and an Ag/AgCl were used as the counter electrode and reference electrode, respectively.The sample powder was deposited on the fluoride tin oxide (FTO, 14 Ω cm À2 ) substrate to serve as the working electrode.Typically, the 3 mg sample was dispersed in 500 μL deionized water by sonication to obtain a homogenously dispersed ink, and around 20 μL as-prepared ink was deposited on a 0.5 cm Â 0.5 cm FTO substrate as a catalyst film.After drying at room temperature, the working electrode was obtained.A quartz cell filled with 0.1 M Na 2 SO 4 or 0.1 M KCl electrolyte containing 0.05 M K 3 [Fe(CN) 6 ]/K 4 [Fe(CN) 6 ] was used as the reaction system for photoelectrochemical measurement, with a 300 W Xe lamp (CEL-HXF300, Beijing Au-light Co. Ltd., China) used as the light source.
In Situ EPR Tests: The EPR measurements were carried out at X-band frequency on a Bruker A300 spectrometer with a TEMPO radical capture reagent.Typically, the photocatalyst (10 mg) was added to an aqueous solution (for TEMPO-e À signal) or acetonitrile solution (for TEMPO-h þ signal) containing TEMPO (0.05 mM) within a container.The solution was dispersed by ultrasonication and loaded into a quartz tube for paramagnetic testing.The EPR spectra were measured for both dark and visible light conditions, with the instrument settings as follows: modulation frequency: 100.00 kHz; modulation amplitude: 1.00 G; time constant: 40.96 ms; conversion time: 40.00 ms; sweep time: 40.96 s.The microwave bridge power and frequency were 5.72 mW and 9.826 GHz, respectively.
In Situ FTIR Tests: To perform in situ FTIR measurements on ZIS/In 2 O 3 for photothermal redox catalysis of CO 2 RR coupled with BA oxidation, the catalyst was loaded onto the sample holder with a flat surface, and BA was added to the reactor.The reactor was then vacuumed and cleaned with argon.Afterward, an appropriate amount of CO 2 gas was introduced, then left for 30 min to allow the catalyst to better adsorb CO 2 gas.Meanwhile, the FTIR data were collected as the background.Subsequently, visible light was introduced to the reaction space through the observation window, and the sample was heated at 300 °C by the heating base.
Theoretic Calculations: DFT calculations for ZIS and In 2 O 3 were conducted via the Materials Studio (BIOVIA V2017, American) equipped with the CASTEP mode.Also, we utilized the Perdew-Burke-Ernzerhof (PBE) form exchange-correlation functional within the generalized gradient approximation (GGA).The structures of the (102) plane of ZIS and the (222) plane of In 2 O 3 were optimized.The average potential profiles were calculated to obtain the W F of ZIS and In 2 O 3 after the geometry optimization.Specifically, for ZIS, the cutoff energy of 348.3 eV and the Monkhorst-Pack grids of 1 Â 3 Â 1 were employed.The convergence thresholds for the geometry optimization were set as 2.0 Â 10 À6 eV atom À1 for energy, and 0.02 eV Å À1 for maximum force.Moreover, in the case of ZnO, the cutoff energy of 381 eV and the Monkhorst-Pack grids of 1 Â 2 Â 1 were adopted.The convergence thresholds for the geometry optimization were set as 1.0 Â 10 À5 eV atom À1 for energy, and 0.02 eV Å À1 for maximum force.

2. 1 .
Characterizations of 2D/1D ZIS/In 2 O 3 Heterostructure and S2, Supporting Information).Apparently, In 2 O 3 exhibited a similar surface morphology as that of the MIL-68(In) precursor.The high-resolution transmission electron microscopy (HRTEM) image displayed the lattice fringes with an interlayer spacing of 0.29 nm that could be attributed to (222) planes of cubic In 2 O 3 (Figure S3, Supporting Information), confirming the formation of In 2 O 3 .Next, the 2D/1D ZIS/In 2 O 3 was prepared by in situ growth of 2D ZIS nanosheets on the surface of 1D In 2 O 3 tubes, and its morphology was characterized by the field-emission scanning electron microscope (FESEM) and TEM (Figure 1d,e).As expected, 2D ZIS nanosheets uniformly covered the surface of 1D In 2 O 3 tubes.Additionally, as shown by the HRTEM image in Figure 1f, the lattice fringes displayed interlayer spacings of 0.32 and 0.29 nm associated with the (102) plane of hexagonal ZnIn 2 S 4 and the (222) plane of cubic In 2 O 3 , respectively, further verifying the successful preparation of ZIS/In 2 O 3 heterostructures.The energy-dispersive X-ray (EDX) spectra with the corresponding elemental mapping of Zn, In, S, and O for ZIS/In 2 O 3 heterostructures are demonstrated in Figure 1g,h and S4, Supporting Information, signifying that the homogenous composition distribution without impurity within the entire region of this heterostructure.Note that the distribution of O in Figure 1h was slightly different from other elements, which may be caused by the presence of the conductive glue at the bottom of the sample.For comparison, the single-phase 2D ZIS ultrathin nanosheet was also successfully fabricated in the absence of the 1D In 2 O 3 substrate, which was proved by its TEM and HRTEM characterizations (Figure S5 and S6, Supporting Information).The powder X-ray diffraction (XRD) was further carried out to investigate the crystalline phases of the probed samples.As is shown in Figure 2a, in the case of ZIS/In 2 O 3 heterostructure, the diffraction peaks at 21.5, 30.6, 35.5, 51.0, and 60.7°can be ascribed to the planes (211), (222), (400), (440), and (622) of cubic In 2 O 3 phase (JCPDS card No. 71-2195), respectively.At the same time, the left two characteristic peaks at 27.7 and 47.2°corresponded to the (102) and (110) facets of the hexagonal ZIS phase (JCPDS card No. 65-2023), respectively.In agreement with the findings of HRTEM, the successful synthesis of ZIS/In 2 O 3 heterostructure was clearly illustrated.Besides, the XRD patterns of single ZIS and In 2 O 3 also affirmed the formation of their respective pure phase

Figure 2 .
Figure 2. a) XRD and b) UV-vis DRS patterns of the prepared ZIS, In 2 O 3 , and ZIS/In 2 O 3 .c) The bandgap energies (E g ) of ZIS and In 2 O 3 .Mott-Schottky plots of d) ZIS and e) In 2 O 3 .VB-XPS spectra of f ) ZIS and g) In 2 O 3 .h) Schematic illustration of the band structure of ZIS and In 2 O 3 (P1 and P2 represent two possible charge transfer mechanisms of ZIS/In 2 O 3 under light irradiation: the band-band transfer and the Z-scheme transfer, respectively).

Figure 4 .
Figure 4. a) Catalysis mechanism of the reduction of CO 2 coupled with the selective BA oxidation.Activity comparison of ZIS, In 2 O 3, and ZIS/In 2 O 3 heterostructure for photocatalytic (noted as L) and photothermal conversion (at 100 °C, noted as L þ H) of CO 2 and BA into b) H 2 , c) CO, d) BAD, and e) HB.

Figure 5 .
Figure 5. a) Gaseous and b) liquid products of ZIS/In 2 O 3 from the photothermal conversion of CO 2 and BA at different temperatures.c,d) The reusability of ZIS/In 2 O 3 for photothermal conversion of CO 2 and BA at 300 °C.

Figure 7 .
Figure 7.In situ EPR spectra of photogenerated a-f ) electrons (e À ) and g-l) holes (h þ ) from ZIS, In 2 O 3, and ZIS/In 2 O 3 captured by TEMPO with or without the presence of BA under light irradiation for a different time.

Figure 8 .
Figure 8.In situ FTIR spectra of ZIS/In 2 O 3 for photothermal catalytic conversion of CO 2 and BA at 300 °C.

Figure 9 .
Figure 9. W F of a) ZIS and b) In 2 O 3 .c) The photoexcited charge separation and the transport mechanism of the ZIS/In 2 O 3 heterojunction.

BaSO 4
as the reference.Steady PL emission spectra were measured at room temperature on a PL spectrophotometer (Shimadzu RF530) excited by 370 nm light at room temperature.The morphology and element mapping of the ZIS, In 2 O 3, and ZIS/In 2 O 3 were detected using a TEM (FEI Tecnai G2 F20) and SEM (Hitachi SU8200) coupled with an EDX (ESCALAB 250) spectroscopy.The surface elemental composition and chemical valence states were conducted using an XPS (Thermo Scientific K-alphaþ) with Al Kα (1,484.6 eV) as the excitation source.The TRPL spectra of the samples were measured by a timeresolved fluorescence spectrometer (JASCO) with a nano-LED (374 nm) light as the excitation source.The work function was detected on UPS (Thermo ESCALAB 250XI).The 13 C isotope experiments were conducted on GC-MS (Agilent 7890; Column: DB-5 ms, 30 m* 250 μm* 0.25 μm; Temperature: 325 °C).