Photocatalytic Conversion of Fructose to Lactic Acid by BiOBr/Zn@SnO 2 Material

: Photocatalysis provides a prospective approach for achieving high-value products under mild conditions. To realize this, constructing a selective, low-cost and environmentally friendly photocatalyst is the most critical factor. In this study, BiOBr/Zn@SnO 2 is fabricated by a one-pot hydrothermal synthesis method and BiOBr: SnO 2 ratio is 3:1; this material is applied as photocatalyst in fructose selective conversion to lactic acid. The bandgap structure can be regulated via two-step modiﬁcation, which includes Zn doping SnO 2 and Zn@SnO 2 coupling BiOBr. The photocatalyst shows excellent conversion efﬁciency in fructose and high selectivity in lactic acid generation under alkaline conditions. The conversion rate is almost 100%, and the lactic acid yield is 79.6% under optimal reaction conditions. The catalyst is highly sustainable in reusability; the lactic acid yield can reach 67.4% after ﬁve runs. The possible reaction mechanism is also proposed to disclose the photocatalysis processes. are 3.8 nm, 17.0 nm and 17.1 nm, respectively (Table The results indicate that the combination of BOBr and SnO 2 , as well as the doping of Zn element, signiﬁcantly enhance the speciﬁc surface areas and pore sizes of the products. All the above-mentioned improve the opportunity for catalyst participation in the reaction, thus enhancing the catalysis performance of the product.


Introduction
The irrational exploitation and use of non-renewable resources have caused a significant negative impact on the production and development of human society. Heavy ecological and environmental damage have seriously affected the survival quality and future of humanity. Biomass energy has been deemed as a promising alternative to solve this predicament. Among all the converted products, lactic acid has attracted considerable attention due to its wide application ranging from food to the cosmetic industries [1,2]. It is also one of the top-12 high-value bio-based platform molecules identified by the US Department of Energy [3,4]. In recent years, lactic acid has been used to synthesize high-value-added anilines via ruthenium-catalyzed amination, which provides a fast and efficient chemical method for the synthesis of amino acids from biomass. In addition, lactic acid can be used for the synthesis of polylactic acid, which is one of the most promising biodegradable plastics [5,6].
So far, there are two main approaches to lactic acid production: biological and chemical processes. The biological method mainly uses enzymes to hydrolyze carbohydrates to produce lactic acid. However, a series of defects such as slow reaction speed, high energy consumption, and rigorous raw material requirements make this method difficult to apply in large-scale utilization [7]. In recent years, there has been significant interest in the photoreforming (PR) of biomass, which relies on inexhaustible light energy to harness biomass to meet the low-carbon, high-sustainability goals. Biomass PR typically involves a hydrogen evolution reaction (HER) and biomass oxidation reactions (BOR), where HER occurs over a metal-loaded photocatalyst, reducing protons in water and producing H 2 through the conduction band (CB) electrons [8]. However, the controllability and selectivity of BOR processes are challenging, due primarily to there being the existence of various possible pathways. Different reactive oxygen species (ROS), which including hydroxyl radicals, superoxide radicals, and electron holes are all considered to play a vital role in the process. It is imperative to improve the selective transformation in the process of BOR.
During the processes of photorefinery, efficient charge separation of the photocatalyst and suitable redox potential is critical to its catalysis performance. Duan et al. [9] proposed coupling heterojunctions formed by two different semiconductor materials to modify charge transfer, and Chaves et al. [10] suggested that bandgap engineering should be combined with homojunction as a means to modulate the redox potential and thus achieve efficient selection for sugar oxidation. Zhao et al. [11] chose Zn 1-x Cd x S to perform a series of experiments on the modulation of bandgap and redox potential. They found that the reduction in bandgap and the increase in redox potential could, to some extent, reduce the hydrogen precipitation reaction of sugars while increasing the oxidation reaction and improving the selectivity of lactic acid, but all of their studies suffer from low lactate yields. Meanwhile, the experiments of Zhao et al. [11] and Sun et al. [12] revealed that ·OH plays a key role in the photocatalytic conversion of sugars under weak alkaline conditions. Therefore, in the investigation of the photocatalytic conversion of fructose to lactic acid, the bandgap, redox potential, and hydroxyl radical production of the photocatalyst are essential conditions for the experiment's success [13].
In previous research, many biomasses were utilized to generate liquid lactic acid through chemical processes-three-carbon sugars such as propanone [14,15], and sixcarbon sugars such as hexose [16,17] and polysaccharide cellulose [18,19]. Fructose is a simple monosaccharide; it is a tautomer of glucose, mostly white crystal or crystalline powder, odorless, sweet, hygroscopic, very soluble in water, soluble in ethanol, methanol, insoluble in methane and ether, etc. [20]. In this research, it was selected as the material to generate lactic acid.
The coupling of SnO 2 with Bi-containing semiconductor materials has been studied as a new trend for composites that can improve photocatalytic performance [21]. Due to the traditional preparation methods, which are time-consuming, energy-intensive and have low purity [22], some other methods were referenced in this study, such as modified SnO 2 with BiOBr in one pot [23,24]. In this research, BiOBr and SnO 2 were chosen as original semiconductor materials. The bandgap of the composite can further be adjusted and the redox potential was also regulated by loading Zn on SnO 2 . Both processing operations are beneficial to the selective transformation of fructose to lactic acid. The heterojunction was synthesized via coupling of BiOBr with the Zn-doped SnO 2 ; the results indicated the catalyst had excellent selectivity toward lactic acid from fructose; the lactic acid selectivity was about 79.6%.

Results and Discussion
2.1. Characterization of BiOBr, BiOBr/SnO 2 , BiOBr/Zn@SnO 2 SEM and TEM analyses were carried out to observe the microstructure and morphology of the fabricated BiOBr/Zn@SnO 2 . Figure 1A,B depict the layer-stacked structure of the product; the comparison with the pre-synthetic material can be found in Figure S1, and obvious structural changes can be found in the figure. Figure 1C-G show the results of the distribution of each element in the material taken by energy dispersive X-ray spectroscopy (EDS). It can be seen that related elements, such as Bi, Br, Zn, Sn, and O, are well distributed on the composite, indicating BiOBr and SnO 2 have co-existed in the composite structure successfully. SnO 2 nanoparticles were dispersed on the surface of BiOBr and Zn was also well loaded on the resulting material. TEM characterization was carried out to further investigate the relationship between Zn@SnO 2 and BiOBr. Figure 1H exhibits the relatively uniform dispersion of the two materials; Figure 1I displays the existence of two lattice stripes, which corresponded to SnO 2 (110, 3.4 Å) lattice stripes [25] and BiOBr (102, 2.8 Å) lattice stripes [26], respectively. The chemical structural aspects of BiOBr, BiOBr/SnO2, and BiOBr/Zn@SnO2 are investigated by Fourier transform infrared (FT-IR) and powder X-ray diffractometry (XRD). Figure 2 shows the FT-IR spectra of the materials, from which it can be seen that the characteristic peak appears at 501 cm −1 and belongs to the vibrational peak of Bi-O [27]. While 1636 cm −1 and 3433 cm −1 can be attributed to the vibrational peaks of O-H [28], the peak emerging at 2361 cm −1 is associated with the vibrational peak of the C = O bond in the absorbed CO2, and the presence of a vibrational peak near 650 cm −1 is caused by the contraction vibration of O-Sn-O.  The chemical structural aspects of BiOBr, BiOBr/SnO 2 , and BiOBr/Zn@SnO 2 are investigated by Fourier transform infrared (FT-IR) and powder X-ray diffractometry (XRD). Figure 2 shows the FT-IR spectra of the materials, from which it can be seen that the characteristic peak appears at 501 cm −1 and belongs to the vibrational peak of Bi-O [27]. While 1636 cm −1 and 3433 cm −1 can be attributed to the vibrational peaks of O-H [28], the peak emerging at 2361 cm −1 is associated with the vibrational peak of the C=O bond in the absorbed CO 2 , and the presence of a vibrational peak near 650 cm −1 is caused by the contraction vibration of O-Sn-O. The chemical structural aspects of BiOBr, BiOBr/SnO2, and BiOBr/Zn@SnO2 are investigated by Fourier transform infrared (FT-IR) and powder X-ray diffractometry (XRD). Figure 2 shows the FT-IR spectra of the materials, from which it can be seen that the characteristic peak appears at 501 cm −1 and belongs to the vibrational peak of Bi-O [27]. While 1636 cm −1 and 3433 cm −1 can be attributed to the vibrational peaks of O-H [28], the peak emerging at 2361 cm −1 is associated with the vibrational peak of the C = O bond in the absorbed CO2, and the presence of a vibrational peak near 650 cm −1 is caused by the contraction vibration of O-Sn-O.    crystal planes, respectively. To further investigate the influence of Zn-doping on the crystal structural of the product, XRD patterns of ZnO, SnO 2 and Zn@SnO 2 were conducted (as could be seen in Figure S2).  Figure S2).  There were almost no typical peaks of zinc or other zinc compounds that could be observed in the pattern of Zn@SnO2, and almost no diffraction peaks shift could be found after the Zn-doping, which should be ascribed to the similar ionic radius of Zn(II) (0.73 Å ) and Sn(IV) (0.71 Å ) [29]. While the diffraction peaks of BiOBr/SnO2 and BiOBr/Zn@SnO2 showed similar characteristic peaks as BiOBr, there were almost no typical diffraction There were almost no typical peaks of zinc or other zinc compounds that could be observed in the pattern of Zn@SnO 2 , and almost no diffraction peaks shift could be found after the Zn-doping, which should be ascribed to the similar ionic radius of Zn(II) (0.73 Å) and Sn(IV) (0.71 Å) [29]. While the diffraction peaks of BiOBr/SnO 2 and BiOBr/Zn@SnO 2 showed similar characteristic peaks as BiOBr, there were almost no typical diffraction peaks of SnO 2 observed in the patterns. This might be due to the intensities of SnO 2 peaks being much weaker compared with that of BiOBr as well as the peak positions being just overlapped by the peaks of BiOBr [21]. However, the peak that emerged at 33.54 should belong to the (101) facet of SnO 2 , which further proved the coexistence of BiOBr and SnO 2 . Figure 3B exhibits the full survey of BiOBr/Zn@SnO 2 , Bi, Br, Sn, and Zn of BiOBr/Zn@SnO 2 all emerged in the spectrum, which indicates SnO 2 and Zn are well loaded onto BiOBr. Figure 3C shows the spectrum of Bi 4f, the two typical peaks located at 159.50 eV and 164.70 eV attributed to Bi4f 7/2 and Bi4f 5/2 , respectively, which indicate Bi is in the valence state of +3 in this catalyst. In Figure 3D, the two peaks located at 68.39 eV and 69.42 eV may belong to Br3d 3/2 and Br3d 5/2 , respectively. The corresponding valency of Br is −1 [30]. The peak at 442.86 eV in Figure 3E is Zn 3d. The peaks at 487.35 eV and 495.86 eV in Figure 3F are Sn3d 5/2 and Sn3d 3/2 , respectively. Both of them attribute to Sn (IV) [31].
The specific surface area and pore size distribution of BiOBr, BiOBr/SnO 2 and BiOBr/Zn@SnO 2 are analyzed using a BET analyzer. The adsorption and desorption isotherms of all the samples are shown in Figure 4A-C, respectively. It can be found that the desorption return line has a clear hysteresis return line, which indicates the hysteresis loop is type h3 and possesses mesopores structure [32].
The specific surface areas of BiOBr, BiOBr/SnO 2 , and BiOBr/Zn@SnO 2 are calculated as 5.8 m 2 ·g −1 , 42.7 m 2 ·g −1 and 61.5 m 2 ·g −1 , respectively. Furthermore, the average pore sizes of BiOBr/SnO 2 and BiOBr/Zn@SnO 2 are 3.8 nm, 17.0 nm and 17.1 nm, respectively ( Table 1). The results indicate that the combination of BOBr and SnO 2 , as well as the doping of Zn element, significantly enhance the specific surface areas and pore sizes of the products. All the above-mentioned improve the opportunity for catalyst participation in the reaction, thus enhancing the catalysis performance of the product.

R PEER REVIEW 5 of 14
peaks of SnO2 observed in the patterns. This might be due to the intensities of SnO2 peaks being much weaker compared with that of BiOBr as well as the peak positions being just overlapped by the peaks of BiOBr [21]. However, the peak that emerged at 33.54 should belong to the (101) facet of SnO2, which further proved the coexistence of BiOBr and SnO2. Figure 3B exhibits the full survey of BiOBr/Zn@SnO2, Bi, Br, Sn, and Zn of Bi-OBr/Zn@SnO2 all emerged in the spectrum, which indicates SnO2 and Zn are well loaded onto BiOBr. Figure 3C shows the spectrum of Bi 4f, the two typical peaks located at 159.50 eV and 164.70 eV attributed to Bi4f7/2 and Bi4f5/2, respectively, which indicate Bi is in the valence state of +3 in this catalyst. In Figure 3D, the two peaks located at 68.39 eV and 69.42 eV may belong to Br3d3/2 and Br3d5/2, respectively. The corresponding valency of Br is −1 [30]. The peak at 442.86 eV in Figure 3E is Zn 3d. The peaks at 487.35 eV and 495.86 eV in Figure 3F are Sn3d5/2 and Sn3d3/2, respectively. Both of them attribute to Sn (IV) [31].
The specific surface area and pore size distribution of BiOBr, BiOBr/SnO2 and Bi-OBr/Zn@SnO2 are analyzed using a BET analyzer. The adsorption and desorption isotherms of all the samples are shown in Figure 4A-C, respectively. It can be found that the desorption return line has a clear hysteresis return line, which indicates the hysteresis loop is type h3 and possesses mesopores structure [32].  Table  1). The results indicate that the combination of BOBr and SnO2, as well as the doping of Zn element, significantly enhance the specific surface areas and pore sizes of the products. All the above-mentioned improve the opportunity for catalyst participation in the reaction, thus enhancing the catalysis performance of the product. The visible absorption as well as the bandwidth of BiOBr, BiOBr/SnO2 and Bi-OBr/Zn@SnO2 were investigated using UV-vis spectroscopy. In Figure 5A, the absorption edges of BiOBr, BiOBr/SnO2, and BiOBr/Zn@SnO2 are at 400 nm, 380 nm and 410 nm, respectively. The optical band edges in Figure 5B are determined by the Tauc equation as ahv = A(hv-Eg) n/2 , where A is a constant, a is the absorption coefficient, h is Planck's constant, v is the photon energy, Eg is the bandgap and n is defined by the optical leap in the semiconductor (n = 1 for a direct leap and n = 4 for an indirect leap).  The visible absorption as well as the bandwidth of BiOBr, BiOBr/SnO 2 and BiOBr/ Zn@SnO 2 were investigated using UV-vis spectroscopy. In Figure 5A, the absorption edges of BiOBr, BiOBr/SnO 2 , and BiOBr/Zn@SnO 2 are at 400 nm, 380 nm and 410 nm, respectively. The optical band edges in Figure 5B are determined by the Tauc equation as ahv = A(hv-Eg) n/2 , where A is a constant, a is the absorption coefficient, h is Planck's constant, v is the photon energy, Eg is the bandgap and n is defined by the optical leap in the semiconductor (n = 1 for a direct leap and n = 4 for an indirect leap).  Table  1). The results indicate that the combination of BOBr and SnO2, as well as the doping of Zn element, significantly enhance the specific surface areas and pore sizes of the products. All the above-mentioned improve the opportunity for catalyst participation in the reaction, thus enhancing the catalysis performance of the product.

Optical Characterization of BiOBr, BiOBr/SnO2, BiOBr/Zn@SnO2
The visible absorption as well as the bandwidth of BiOBr, BiOBr/SnO2 and Bi-OBr/Zn@SnO2 were investigated using UV-vis spectroscopy. In Figure 5A, the absorption edges of BiOBr, BiOBr/SnO2, and BiOBr/Zn@SnO2 are at 400 nm, 380 nm and 410 nm, respectively. The optical band edges in Figure 5B are determined by the Tauc equation as ahv = A(hv-Eg) n/2 , where A is a constant, a is the absorption coefficient, h is Planck's constant, v is the photon energy, Eg is the bandgap and n is defined by the optical leap in the semiconductor (n = 1 for a direct leap and n = 4 for an indirect leap).  It is understood from a review of the data that BiOBr and SnO 2 have an n value of 1.24 and are direct leaps [33,34]. The bandgaps of BiOBr, BiOBr/SnO 2 , and BiOBr/Zn@SnO 2 can be measured as 2.72 eV, 2.57 eV, and 2.21 eV can be measured in Figure 5B by plotting (ahv) 1/2 versus hv. It can be seen that, as the catalyst is modified, the bandgap decreases and the range of visible light that can be absorbed increases, resulting in a reduction in the hydrogen precipitation reaction of sugars, the catalysts having been modified to reduce the hydrogen precipitation of sugars, increasing the oxidation reaction and improving the selectivity of lactic acid.
The photoelectric properties of BiOBr, BiOBr/SnO 2 and BiOBr/Zn@SnO 2 were investigated. The light and dark periodic photocurrents, alternating current (AC) impedance and Lsv of the three materials were measured using a Shanghai Chenhua model 660 e electrochemical workstation. Figure 6A shows the light and dark periodic photocurrents of the three catalysts. It can be seen that the photocurrent response values of BiOBr/Zn@SnO 2 are higher than the other two catalysts, indicating that BiOBr/Zn@SnO 2 has better photosensitivity and can participate in the photocatalytic reaction more efficiently. Figure 6B shows the AC impedance profiles of the three catalysts. It can be seen that the impedance of the catalysts decreases with the doping of SnO 2 and Zn, indicating that the new materials once synthesized can maintain a lower electron transfer resistance to the electrolyte and achieve a higher photocatalytic efficiency. In addition, L SV scans of the three photocatalysts under bright and dark-light conditions were also carried out, the results of which are shown in Figure 6C. It can be seen that the photocurrent response values of BiOBr/Zn@SnO 2 are higher than those of the other two catalysts under both visible and dark conditions, indicating that the catalysts have a higher photogenerated charge capability. Testing of these three photovoltaic properties suggests that the newly synthesized catalyst has higher photocatalytic activity than the original single catalyst.
It is understood from a review of the data that BiOBr and SnO2 have an n value of 1.24 and are direct leaps [33,34]. The bandgaps of BiOBr, BiOBr/SnO2, and BiOBr/Zn@SnO2 can be measured as 2.72 eV, 2.57 eV, and 2.21 eV can be measured in Figure 5B by plotting (ahv) 1/2 versus hv. It can be seen that, as the catalyst is modified, the bandgap decreases and the range of visible light that can be absorbed increases, resulting in a reduction in the hydrogen precipitation reaction of sugars, the catalysts having been modified to reduce the hydrogen precipitation of sugars, increasing the oxidation reaction and improving the selectivity of lactic acid.
The photoelectric properties of BiOBr, BiOBr/SnO2 and BiOBr/Zn@SnO2 were investigated. The light and dark periodic photocurrents, alternating current (AC) impedance and Lsv of the three materials were measured using a Shanghai Chenhua model 660 e electrochemical workstation. Figure 6A shows the light and dark periodic photocurrents of the three catalysts. It can be seen that the photocurrent response values of Bi-OBr/Zn@SnO2 are higher than the other two catalysts, indicating that BiOBr/Zn@SnO2 has better photosensitivity and can participate in the photocatalytic reaction more efficiently. Figure 6B shows the AC impedance profiles of the three catalysts. It can be seen that the impedance of the catalysts decreases with the doping of SnO2 and Zn, indicating that the new materials once synthesized can maintain a lower electron transfer resistance to the electrolyte and achieve a higher photocatalytic efficiency. In addition, LSV scans of the three photocatalysts under bright and dark-light conditions were also carried out, the results of which are shown in Figure 6C. It can be seen that the photocurrent response values of BiOBr/Zn@SnO2 are higher than those of the other two catalysts under both visible and dark conditions, indicating that the catalysts have a higher photogenerated charge capability. Testing of these three photovoltaic properties suggests that the newly synthesized catalyst has higher photocatalytic activity than the original single catalyst.

Investigating the Effect of Different Conditions on Fructose Photoproduction of Lactic Acid
The effect of various factors on the lactic acid production from fructose was investigated by designing different reaction sets. Firstly, the photoreformation of fructose was carried out with catalysts under neutral and acidic conditions. It was found that no lactic acid was produced under acidic conditions, while the yield of the three catalysts only reached about 10% under neutral conditions, suggesting that photoreformation of lactic acid is dependent on alkaline conditions, and therefore experiments were designed to produce lactic acid from photoreformation of fructose using NaOH as the solvent. This was followed by dark-light experiments with all three catalysts under alkaline conditions, which showed no lactic acid production, indicating that light is an exciting condition for the reaction. As shown in Figure 7A, 100 mg of fructose was subjected to light catalysis at room temperature under the same light conditions and pH with different amounts of catalyst, and it can be seen that as the amount of catalyst increased, the lactic acid yield reached the highest at a catalyst dosage of 20 mg, and the reaction was basically flat when the amount of catalyst continued to increase, so the optimal amount of catalyst was concluded to be 20 mg.
(B). LSV scans of the three catalysts under bright and dark-light conditions (C).

Investigating the Effect of Different Conditions on Fructose Photoproduction of Lactic Acid
The effect of various factors on the lactic acid production from fructose was investigated by designing different reaction sets. Firstly, the photoreformation of fructose was carried out with catalysts under neutral and acidic conditions. It was found that no lactic acid was produced under acidic conditions, while the yield of the three catalysts only reached about 10% under neutral conditions, suggesting that photoreformation of lactic acid is dependent on alkaline conditions, and therefore experiments were designed to produce lactic acid from photoreformation of fructose using NaOH as the solvent. This was followed by dark-light experiments with all three catalysts under alkaline conditions, which showed no lactic acid production, indicating that light is an exciting condition for the reaction. As shown in Figure 7A, 100 mg of fructose was subjected to light catalysis at room temperature under the same light conditions and pH with different amounts of catalyst, and it can be seen that as the amount of catalyst increased, the lactic acid yield reached the highest at a catalyst dosage of 20 mg, and the reaction was basically flat when the amount of catalyst continued to increase, so the optimal amount of catalyst was concluded to be 20 mg. The NaOH concentration variable experiment was designed immediately afterwards, as shown in Figure 7B, with 100 mg of fructose and 20 mg of catalyst, and the variable experiment was made on the solution concentration under the same light and room temperature conditions, and it could be seen that the lactic acid yield increased continuously as the concentration became larger, with the best yield being obtained when it The NaOH concentration variable experiment was designed immediately afterwards, as shown in Figure 7B, with 100 mg of fructose and 20 mg of catalyst, and the variable experiment was made on the solution concentration under the same light and room temperature conditions, and it could be seen that the lactic acid yield increased continuously as the concentration became larger, with the best yield being obtained when it reached 3.0 mol·L −1 , where the increase in lactic acid yield was more related to the presence of -OH. Finally, an experiment was designed with the duration of light as the variable, as shown in Figure 7C, in which 100 mg of fructose with 20 mg of catalyst was selected and the experiment was carried out at room temperature as well as in 3.0 mol·L −1 NaOH solution, and the lactic acid yield became higher as the duration of light increased and remained constant after 3 h. The optimum conditions for catalytic lactic acid and the maximum yield were obtained by the above experiments, i.e., 100 mg of fructose with 20 mg of catalyst added to 3.0 mol·L −1 NaOH solution, and 79.6% lactic acid was obtained by 3 h of light exposure at room temperature.
The stability and reusability of the newly synthesized catalyst was tested and the results are shown in Figure 7D. In total, 100 mg of fructose and 20 mg of BiOBr/Zn@SnO 2 were added to 10 mL of NaOH solution at a concentration of 3 mol/L. The catalyst was filter dried and recovered after 3 h of light prodding to test the recovery each time as well as the conversion of lactic acid. After five cycles of the experiment, it can be seen that the yield of lactic acid decreased slightly with the recycling of the catalyst, but still remained high; at the same time, and comparing this experiment with previous studies, it can be seen that the conversion rate has been greatly improved (Table 2). Thus, the prepared BiOBr/Zn@SnO 2 has good stability and reusability, and can save materials while being more cost-effective. The influence of various radicals and hole on the synthesis of lactic acid was also explored. It is well known that various reactive oxidants are produced during the photocatalytic process, and that these oxidative active free radicals play different roles in the overall reaction system. In order to investigate the effect of these oxidation reactive substances on the reaction, reactive substance inhibitors were used to determine the effect on the reaction if the production of a particular reactive substance was inhibited. Three reactive inhibitors, 1,4-benzoquinone (BQ), disodium ethylenediaminetetraacetate (EDTA-Na 2 ) and isopropyl alcohol (IPA), were used as inhibitors of the superoxide radical (·O 2 − ), hole (H + ) and hydroxyl radical (OH − ) radicals [35]. These inhibitors were added separately under the same conditions to investigate the effect on the photopromoted reforming of lactic acid from fructose, and the results are shown in Figure 8A. It can be seen that the effect of cavity H + on the overall reaction is not significant, but the yield of lactic acid decreases significantly when superoxide radicals (·O 2 − ) and hydroxyl radicals (·OH) are inhibited by the inhibitors, so it can be judged that superoxide radicals and hydroxyl radicals play an important role in the phototactic reforming of lactic acid from fructose. Therefore, ESR spin labeling of the superoxide radical (·O 2 − ) and hydroxyl radical (·OH) in the reaction system was carried out with DMPO as the trapping agent, as shown in Figure 8B. As far as we know, DMPO, ·O 2 − or ·OH alone do not produce a signal in the solution system, while the combination of DMPO and ·O 2 − has a quadruple signal peak with a relative intensity of 1:1:1:1 [36], which can be clearly seen in Figure 9B, and is trapped by the trapping agent to produce a signal. The same combination of DMPO and ·OH showed a quadruple signal peak with a relative intensity of 1:2:2:1 [37], and no hydroxyl radicals were produced under dark conditions. Thus, both superoxide radicals and hydroxyl radicals were generated under light conditions in the catalyst system and played an oxidative role in the subsequent photoreactivation of fructose. tem, while the combination of DMPO and ·O2 has a quadruple signal peak with a relative intensity of 1:1:1:1 [36], which can be clearly seen in Figure 9B, and is trapped by the trapping agent to produce a signal. The same combination of DMPO and ·OH showed a quadruple signal peak with a relative intensity of 1:2:2:1 [37], and no hydroxyl radicals were produced under dark conditions. Thus, both superoxide radicals and hydroxyl radicals were generated under light conditions in the catalyst system and played an oxidative role in the subsequent photoreactivation of fructose.

Possible Pathways for Lactic Acid-Lactate Production by Fructose Photorefinery
The reaction pathways of fructose convert to lactic acid using BiOBr/Zn@SnO2, as a photocatalyst was proposed in this research. A possible mechanism was proposed based on the experiment results and previous literature [21]. As can be seen in Figure 9, the heterojunction of BiOBr/Zn@SnO2 with a suitable bandgap is activated under visible light irradiation, and the photogenerated electrons generated in the conduction band reduce the absorbed oxygen to form superoxide radicals. Furthermore, the produced superoxide radicals can be continuously transformed into other reactive substances under light irradiation, such as hydroxyl radicals, which promote the conversion of fructose to lactic acid [11]. The formation of lactic acid from fructose under alkaline conditions in the photorefinery start from the ring-opening of fructose, followed by hydrolysis (trans-aldol) of fructose in the presence of reactive oxygen species to produce glyceraldehyde and 1,3-dihydroxyacetone. Through further dehydration, pyruvic aldehyde is formed and, finally, lactic acid is produced through a series of steps. However, for the small number of by-products such as formic acid produced after the reaction, this could be produced from the α/βoxidation of fructose, as he detected yields of lactic acid were much higher than the substances such as formic acid. This suggests that lactic acid has a greater selectivity in the reaction.
BiOBr/Zn@SnO2 were prepared by the one-pot method that was modified based on the previous literature [21]. Weigh 0.97 g of Bi (NO3)3·5H2O (0.0025 mol) and 0.238 g of KBr (0.002 mol) into 20 mL of ethylene glycol as solution A, and 0.235 g of SnCl4·5H2O (0.0007 mol) and 0.029 g Zn(CH3COO)2·2H2O (0.001 mol) into mix 10 mL of ethanol and 0.107 g of NaOH (0.0027 mol), as solution B. After sonication for 1 h, solution A was slowly poured into solution B and also mixed with solution C. Solution C was put into a PTFE reactor and reacted at 120 °C for 12 h. The mixture was washed three times with anhydrous ethanol and water in sequence, then filtered. The target product was dried in a vacuum oven at 60 °C overnight.
Scheme 1 demonstrates the preparation process of one-pot chemistry to synthesis Bi-OBr/Zn@SnO2. Briefly, a certain amount of SnCl4·5H2O is dissolved in sodium hydroxide and ethanol solution to form a white flocculent Sn(OH)4, while adding Zn(CH3COO)2·2H2O in a hydrothermal reaction can load Zn onto the SnO2 molecule, then

Possible Pathways for Lactic Acid-Lactate Production by Fructose Photorefinery
The reaction pathways of fructose convert to lactic acid using BiOBr/Zn@SnO 2 , as a photocatalyst was proposed in this research. A possible mechanism was proposed based on the experiment results and previous literature [21]. As can be seen in Figure 9, the heterojunction of BiOBr/Zn@SnO 2 with a suitable bandgap is activated under visible light irradiation, and the photogenerated electrons generated in the conduction band reduce the absorbed oxygen to form superoxide radicals. Furthermore, the produced superoxide radicals can be continuously transformed into other reactive substances under light irradiation, such as hydroxyl radicals, which promote the conversion of fructose to lactic acid [11]. The formation of lactic acid from fructose under alkaline conditions in the photorefinery start from the ring-opening of fructose, followed by hydrolysis (transaldol) of fructose in the presence of reactive oxygen species to produce glyceraldehyde and 1,3-dihydroxyacetone. Through further dehydration, pyruvic aldehyde is formed and, finally, lactic acid is produced through a series of steps. However, for the small number of by-products such as formic acid produced after the reaction, this could be produced from the α/β-oxidation of fructose, as he detected yields of lactic acid were much higher than the substances such as formic acid. This suggests that lactic acid has a greater selectivity in the reaction.
BiOBr/Zn@SnO 2 were prepared by the one-pot method that was modified based on the previous literature [21]. Weigh 0.97 g of Bi (NO 3 ) 3 ·5H 2 O (0.0025 mol) and 0.238 g of KBr (0.002 mol) into 20 mL of ethylene glycol as solution A, and 0.235 g of SnCl 4 ·5H 2 O (0.0007 mol) and 0.029 g Zn(CH 3 COO) 2 ·2H 2 O (0.001 mol) into mix 10 mL of ethanol and 0.107 g of NaOH (0.0027 mol), as solution B. After sonication for 1 h, solution A was slowly poured into solution B and also mixed with solution C. Solution C was put into a PTFE reactor and reacted at 120 • C for 12 h. The mixture was washed three times with anhydrous ethanol and water in sequence, then filtered. The target product was dried in a vacuum oven at 60 • C overnight.
Scheme 1 demonstrates the preparation process of one-pot chemistry to synthesis BiOBr/Zn@SnO 2 . Briefly, a certain amount of SnCl 4 ·5H 2 O is dissolved in sodium hydroxide and ethanol solution to form a white flocculent Sn(OH) 4 , while adding Zn(CH 3 COO) 2 ·2H 2 O in a hydrothermal reaction can load Zn onto the SnO 2 molecule, then Bi(NO 3 ) 3 ·5H 2 O is poured into the ethylene glycol solution, and KBr is added to slowly form a yellow solution, after which the white flocculent solution is poured into the yellow solution and stirred well before being put into the reactor for heating and reaction, and the target product of BiOBr/Zn@SnO 2 is obtained by separation and precipitation and drying. Two comparative catalysts, BiOBr and BiOBr/SnO 2 , are also prepared, with both having been synthesized by hydrothermal reaction in an autoclave, for use in some subsequent experimental comparisons. Batch experiments were carried out under the irradiation of a xenon lamp (300 w). A total of 100.0 mg of fructose was immersed in 10.0 mL of NaOH solution (0.5-3.0 M). Effect of catalyst sorts (BiOBr, BiOBr/SnO2, BiOBr/Zn@SnO2), NaOH concentration, catalyst dosage, and light duration were all investigated. Results of the final lactic acid yield were determined via high-performance liquid chromatography (HPLC). The reusability of the catalyst was explored under optimal reaction conditions. after each run, the solution was withdrawn to detect the final lactic acid yield by high performance liquid chromatography (HPLC). Simultaneously, 100 mg of fructose and 10 mL NaOH solution with appropriate concentration was added into the reactor and the next run was started. Scheme 1. One-pot process for the preparation of BiOBr/Zn@SnO2 Fourier transform infrared (FT-IR), powder X-ray diffractometer (XRD), UV-Vis diffuse reflectance spectrometer (UV-Vis), Scanning electron microscope (SEM), high magnification transmission electron microscope (HRTEM), X-ray photoelectron spectrometer (XPS), specific surface area analyzer (BET) was conducted to characterize the products. Photocurrents, linear scanning voltammetry (LSV), and AC impedance were measured using a Shanghai Chenhua model 660e electrochemical workstation. Electron spin resonance (ESR) was used to test the free radicals in the solution system to study their effect on the reaction. Details of the detect conditions could be found in the supplementary data.
To investigate the effect of various active substances on lactic acid yield, a series of comparative experiments with the addition of inhibitors were designed. Inhibitors such as IPA, EDTA, and BQ were added to the reaction system to assess the effect of the active species on the reaction. Batch experiments were carried out under the irradiation of a xenon lamp (300 w). A total of 100.0 mg of fructose was immersed in 10.0 mL of NaOH solution (0.5-3.0 M). Effect of catalyst sorts (BiOBr, BiOBr/SnO 2 , BiOBr/Zn@SnO 2 ), NaOH concentration, catalyst dosage, and light duration were all investigated. Results of the final lactic acid yield were determined via high-performance liquid chromatography (HPLC). The reusability of the catalyst was explored under optimal reaction conditions. after each run, the solution was withdrawn to detect the final lactic acid yield by high performance liquid chromatography (HPLC). Simultaneously, 100 mg of fructose and 10 mL NaOH solution with appropriate concentration was added into the reactor and the next run was started.

Conclusions
Fourier transform infrared (FT-IR), powder X-ray diffractometer (XRD), UV-Vis diffuse reflectance spectrometer (UV-Vis), Scanning electron microscope (SEM), high magnification transmission electron microscope (HRTEM), X-ray photoelectron spectrometer (XPS), specific surface area analyzer (BET) was conducted to characterize the products. Photocurrents, linear scanning voltammetry (LSV), and AC impedance were measured using a Shanghai Chenhua model 660e electrochemical workstation. Electron spin resonance (ESR) was used to test the free radicals in the solution system to study their effect on the reaction. Details of the detect conditions could be found in the supplementary data.
To investigate the effect of various active substances on lactic acid yield, a series of comparative experiments with the addition of inhibitors were designed. Inhibitors such as IPA, EDTA, and BQ were added to the reaction system to assess the effect of the active species on the reaction.

Conclusions
BiOBr/Zn@SnO 2 was synthesized and applied as photocatalyst in the selective conversion of fructose to lactic acid. The catalyst showed excellent fructose and high lactic acid selectivity. The conversion rate was almost 100%, and the yield of lactic acid was about 79.6%. Possible reaction pathways were proposed based on the results of active oxygen species poisoning experiments, which included superoxide radicals (O 2 − ), vacancies (H + ), and hydroxyl radicals (OH) [38]. The results of reusability show the catalyst possesses excellent performance. This result demonstrates the scientific value of the experiment and provides a new direction for the high value utilization of biomass, with the catalyst being simple to synthesize, simple to process, and also effective in reducing costs, which has a more significant advantage in the high value utilization of biomass.