Highly Efficient Selective Hydrogenation of Cinnamaldehyde to Cinnamyl Alcohol over CoRe/TiO2 Catalyst

Allylic alcohols typically produced through selective hydrogenation of α,β-unsaturated aldehydes are important intermediates in fine chemical industry, but it is still a challenge to achieve its high selectivity transformation. Herein, we report a series of TiO2-supported CoRe bimetallic catalysts for the selective hydrogenation of cinnamaldehyde (CAL) to cinnamyl alcohol (COL) using formic acid (FA) as a hydrogen donor. The resultant catalyst with the optimized Co/Re ratio of 1:1 can achieve an exceptional COL selectivity of 89% with a CAL conversion of 99% under mild conditions of 140 °C for 4 h, and the catalyst can be reused four times without loss of activity. Meanwhile, the Co1Re1/TiO2/FA system was efficient for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols. The presence of ReOx on the Co1Re1/TiO2 catalyst surface was advantageous to the adsorption of C=O, and the ultrafine Co nanoparticles provided abundant hydrogenation active sites for the selective hydrogenation. Moreover, FA as a hydrogen donor improved the selectivity to α,β-unsaturated alcohols.


Introduction
Selective hydrogenation of α,β-unsaturated aldehydes to unsaturated alcohols is an important process used to obtain a great deal of valuable chemicals [1][2][3][4]. Cinnamaldehyde (CAL), a typical α,β-unsaturated aldehyde, could be selectively hydrogenated to generate cinnamyl alcohol (COL), which is regarded as one of the most promising building blocks in pharmaceutical, agrochemical and fragrance industries [5,6].
Generally, the selective hydrogenation of CAL leads to the reduction of different functional groups, including C=O and C=C, and produces COL, 3-phenylpropionaldehyde (HCAL) and 3-phenylpropanol (HCOL) [7][8][9][10]. Owing to a higher binding energy of C=O bonds than C=C bonds (715 vs. 615 KJ·mol −1 ), the hydrogenation of C=O bonds is more unfavorable in thermodynamics [11][12][13]. Therefore, it is essential to develop highperformance catalysts to improve the selectivity hydrogenation of C=O bonds and avoid the hydrogenation of C=C bonds.
Supported metal nanoparticle catalysts have been widely used in the industry due to their merits of easy separation and recovery [7,14]. The radial expansion of a metal

Catalyst Characterization
The structural parameters of the samples are summarized in Table 1. Deposition of relatively low Co and Re contents on TiO2 only slightly influenced its specific surface area, and the SBET values of all catalysts are close to TiO2 and remain near 50 m 2 ·g −1 . On the other hand, the average pore diameter Dp of metal-supported catalysts Co/TiO2, Re/TiO2 and Co1Re1/TiO2 were larger than TiO2, and the pore volumes Vp of them were reduced, which could be explained by partial blocking of the narrowest pores by the metallic phase, indicating that the metals were embedded in the carrier pores [44].  b The pore size and pore volumes were derived from the adsorption branches of isotherms by using the BJH model.

Catalyst Characterization
The structural parameters of the samples are summarized in Table 1. Deposition of relatively low Co and Re contents on TiO 2 only slightly influenced its specific surface area, and the S BET values of all catalysts are close to TiO 2 and remain near 50 m 2 ·g −1 . On the other hand, the average pore diameter D p of metal-supported catalysts Co/TiO 2 , Re/TiO 2 and Co 1 Re 1 /TiO 2 were larger than TiO 2 , and the pore volumes V p of them were reduced, which could be explained by partial blocking of the narrowest pores by the metallic phase, indicating that the metals were embedded in the carrier pores [44].  b The pore size and pore volumes were derived from the adsorption branches of isotherms by using the BJH model.
The morphologies of Co/TiO 2 and Co 1 Re 1 /TiO 2 catalysts were characterized by TEM. As illustrated in Figure 2a, no clear Co nanoparticles were observed on the Co/TiO 2 catalyst, which is in agreement with Cheng's work [45]. As seen from Figure 2c, no diffraction spot was observed in the fast Fourier transformation (FFT) image of the CoRe particles, which is Molecules 2023, 28, 3336 4 of 13 similar to our earlier work [46], implying its amorphous structure. The size-distribution histogram substantiates that the average size of the CoRe nanoparticles is about 1.7 nm (Figure 2d), which is similar to the size (1.8 nm) measured by CO chemisorption, and is much smaller than monometallic Co (8.1 nm, Table 1). This observation was further demonstrated by the results of Co dispersion measured by CO chemisorption ( Table 1). The dispersion of Co on Co 1 Re 1 /TiO 2 is much higher than that of Co/TiO 2 . These phenomena manifest that the introduction of the second metal Re significantly reduced the particle size of the Co nanoparticles and improved Co dispersion, thus providing more hydrogenation active sites for the selective hydrogenation of CAL.
Molecules 2023, 28, x FOR PEER REVIEW 4 of 14 The morphologies of Co/TiO2 and Co1Re1/TiO2 catalysts were characterized by TEM. As illustrated in Figure 2a, no clear Co nanoparticles were observed on the Co/TiO2 catalyst, which is in agreement with Cheng's work [45]. As seen from Figure 2c, no diffraction spot was observed in the fast Fourier transformation (FFT) image of the CoRe particles, which is similar to our earlier work [46], implying its amorphous structure. The size-distribution histogram substantiates that the average size of the CoRe nanoparticles is about 1.7 nm (Figure 2d), which is similar to the size (1.8 nm) measured by CO chemisorption, and is much smaller than monometallic Co (8.1 nm, Table 1). This observation was further demonstrated by the results of Co dispersion measured by CO chemisorption ( Table 1). The dispersion of Co on Co1Re1/TiO2 is much higher than that of Co/TiO2. These phenomena manifest that the introduction of the second metal Re significantly reduced the particle size of the Co nanoparticles and improved Co dispersion, thus providing more hydrogenation active sites for the selective hydrogenation of CAL. The chemical state and surface composition of Co/TiO2, Re1/TiO2 and Co1Re1/TiO2 catalysts were assessed by XPS (Figure 3), and the calculated abundances of different surface Co and Re species are summarized in Table S3. As shown in Figure 3a, the peaks in the Co 2p spectrum of Co/TiO2 at 778.2 and 793.4 eV are assigned to Co 0 , the peaks at 781.0 and 796.8 eV are attributed to CoO, and the peaks at 796.1 and 802.5 eV are satellite peaks [47]. The observation of CoO was due to the oxidation of surface metallic Co nanoparticles in the air. As shown in Figure 3b, the binding energy in the range of 39-50 eV belonged to The chemical state and surface composition of Co/TiO 2 , Re 1 /TiO 2 and Co 1 Re 1 /TiO 2 catalysts were assessed by XPS (Figure 3), and the calculated abundances of different surface Co and Re species are summarized in Table S3. As shown in Figure 3a, the peaks in the Co 2p spectrum of Co/TiO 2 at 778.2 and 793.4 eV are assigned to Co 0 , the peaks at 781.0 and 796.8 eV are attributed to CoO, and the peaks at 796.1 and 802.5 eV are satellite peaks [47]. The observation of CoO was due to the oxidation of surface metallic Co nanoparticles in the air. As shown in Figure 3b, the binding energy in the range of 39-50 eV belonged to the Re 4f region, which was deconvoluted into doublet peaks for 4f 5/2 and 4f 7/2 orbits, implying the presence of ReO (4f 5/2 = 44.2 eV, 4f 7/2 = 41.9 eV), ReO 2 (4f 5/2 = 46.3 eV, 4f 7/2 = 44.0 eV), and Re 2 O 5 (4f 5/2 = 48.3 eV, 4f 7/2 = 46.0 eV) [48] in Re/TiO 2 with an atomic ratio of 65:16:19 (Table S2), and no metallic Re was detected, which was contributed to the high affinity of Re for oxygen [49]. Compared to their monometallic counterparts, the content of ReO in Co 1 Re 1 /TiO 2 significantly increased to 71%, which may promote the spillover of hydrogen on its surface and therefore be beneficial to the hydrogenation process [46].  [48] in Re/TiO2 with an atomic ratio of 65:16:19 (Table S2), and no metallic Re was detected, which was contributed to the high affinity of Re for oxygen [49]. Compared to their monometallic counterparts, the content of ReO in Co1Re1/TiO2 significantly increased to 71%, which may promote the spillover of hydrogen on its surface and therefore be beneficial to the hydrogenation process [46].

Effects of Reaction Conditions
As mentioned in Figure 1, using isopropanol as a hydrogen donor over the Co1Re1/TiO2 catalyst can cause a high yield of COL to be obtained. However, a higher temperature (160 °C) and a longer reaction time (12 h) were required, and a great amount of HCOL (17%) was generated, thus decreasing COL selectivity. As shown in Table S3 (entry 4), compared to other common hydrogen donors (isopropanol, triethyl silicane and ammonium formate), when using FA as the hydrogen donor, a high COL yield could be achieved under mild reaction conditions in the inert solvent tetrahydrofuran (THF). Meanwhile, FA is green, sustainable, atom economical and easy to operate. Therefore, it is a good choice of hydrogen donor for the selective hydrogenation of CAL to COL. Thus, we used FA as a hydrogen donor to further study the hydrogenation of CAL to COL. Based on the results in Table 2, the combination of FA and triethylamine (NEt3) was found to be essential for achieving high activity and selectivity in the hydrogenation of CAL to COL over Co1Re1/TiO2, and a CAL conversion of 99% and a COL selectivity of 89% were obtained at a FA:NEt3 molar ratio of 1:1 ( Table 2, entry 2), which displayed obvious advantages over molecular hydrogen (Table 2, entry 5). The addition of NEt3 to provide basic sites can substantially facilitate the crucial FA deprotonation process, which appears to be a key factor for achieving high activity of CAL hydrogenation to COL [50]. Too much or too little amounts of NEt3 in the reaction system, however, led to a decrease in the selectivity of COL (Table 2, entries 3 and 4). In addition, the CAL:FA molar ratio also affected the reactivity of the hydrogenation of CAL to COL, as shown in Table 3. Low amounts of FA led to the deficiency of active hydrogen and to the decrease in the conversion of CAL, while excess FA led to an increase in by-products, and the optimal molar ratio of CAL:FA was 1:2 (Table 3, entry 2).

Effects of Reaction Conditions
As mentioned in Figure 1, using isopropanol as a hydrogen donor over the Co 1 Re 1 /TiO 2 catalyst can cause a high yield of COL to be obtained. However, a higher temperature (160 • C) and a longer reaction time (12 h) were required, and a great amount of HCOL (17%) was generated, thus decreasing COL selectivity. As shown in Table S3 (entry 4), compared to other common hydrogen donors (isopropanol, triethyl silicane and ammonium formate), when using FA as the hydrogen donor, a high COL yield could be achieved under mild reaction conditions in the inert solvent tetrahydrofuran (THF). Meanwhile, FA is green, sustainable, atom economical and easy to operate. Therefore, it is a good choice of hydrogen donor for the selective hydrogenation of CAL to COL. Thus, we used FA as a hydrogen donor to further study the hydrogenation of CAL to COL. Based on the results in Table 2, the combination of FA and triethylamine (NEt 3 ) was found to be essential for achieving high activity and selectivity in the hydrogenation of CAL to COL over Co 1 Re 1 /TiO 2 , and a CAL conversion of 99% and a COL selectivity of 89% were obtained at a FA:NEt 3 molar ratio of 1:1 ( Table 2, entry 2), which displayed obvious advantages over molecular hydrogen (Table 2, entry 5). The addition of NEt 3 to provide basic sites can substantially facilitate the crucial FA deprotonation process, which appears to be a key factor for achieving high activity of CAL hydrogenation to COL [50]. Too much or too little amounts of NEt 3 in the reaction system, however, led to a decrease in the selectivity of COL ( Table 2, entries 3 and 4). In addition, the CAL:FA molar ratio also affected the reactivity of the hydrogenation of CAL to COL, as shown in Table 3. Low amounts of FA led to the deficiency of active hydrogen and to the decrease in the conversion of CAL, while excess FA led to an increase in by-products, and the optimal molar ratio of CAL:FA was 1:2 (Table 3, entry 2).  The effects of the reaction temperature and catalyst dosage for the hydrogenation of CAL to COL were investigated over Co 1 Re 1 /TiO 2 , and the results are shown in Figure 4. Figure 4a shows the results of the hydrogenation of CAL at varied temperatures (from 100 to 160 • C). During the reaction, no HCAL was detected. The catalyst showed low activity for the reaction at a low temperature of 100 • C. Increasing the reaction temperature to 140 • C achieved a remarkable increase in the catalytic performance. However, the excessive reaction temperature (160 • C) led to a significant increase in the excessive hydrogenation of HCOL and other by-products. Thus, 140 • C was the appropriate reaction temperature for CAL-selective hydrogenation to COL. Fixing the reaction temperature at 140 • C, the hydrogenation of CAL was carried out over the Co 1 Re 1 /TiO 2 catalyst at a dosage in the range from 40 to 100 mg, and the results are shown in Figure 4b. It can be seen that the yield of COL was significantly improved with the increase in catalyst dosage from 40 to 80 mg. However, an excessive catalyst dosage reduced the selectivity of COL increased the excessive hydrogenation by-product HCOL and other by-products. As expected, our rationally designed Co 1 Re 1 /TiO 2 catalyst exhibited higher activity and selectivity for the hydrogenation of CAL to COL compared with TiO 2 , Co/TiO 2 , and Re/TiO 2 catalysts (Table S4), and it showed significant advantages compared with the relevant literature (Table S5) [24,27,30,[32][33][34][35]45,[51][52][53][54][55][56][57][58][59][60][61]. The results of the selective hydrogenation of CAL to COL suggest that the synergistic effect among the TiO 2 support and Co and Re metals is the main reason for the enhanced performance of the Co 1 Re 1 /TiO 2 catalyst. Subsequently, we investigated the stability of the Co1Re1/TiO2 catalyst for the hydrogenation of CAL to COL. During each cycle, after a complete reaction at 140 °C for 4 h, the catalyst was centrifuged, washed with THF for five times, and reused for the next runs. As shown in Figure 5, the Co1Re1/TiO2 catalyst kept its good performance during recycling, the conversion of CAL was 97%, and the selectivity of COL was 87% after four runs. The XRD pattern of the spent catalyst showed no noticeable morphology changes compared with the fresh one ( Figure S1). Moreover, the comparable metal contents in the fresh and spent catalysts determined by ICP-OES indicated no obvious metal leaching (Table  S6). These results suggest that the Co1Re1/TiO2 catalyst has excellent stability for the hydrogenation of CAL to COL.  Subsequently, we investigated the stability of the Co 1 Re 1 /TiO 2 catalyst for the hydrogenation of CAL to COL. During each cycle, after a complete reaction at 140 • C for 4 h, the catalyst was centrifuged, washed with THF for five times, and reused for the next runs. As shown in Figure 5, the Co 1 Re 1 /TiO 2 catalyst kept its good performance during recycling, the conversion of CAL was 97%, and the selectivity of COL was 87% after four runs. The XRD pattern of the spent catalyst showed no noticeable morphology changes compared with the fresh one ( Figure S1). Moreover, the comparable metal contents in the fresh and spent catalysts determined by ICP-OES indicated no obvious metal leaching (Table S6). These results suggest that the Co 1 Re 1 /TiO 2 catalyst has excellent stability for the hydrogenation of CAL to COL.  Subsequently, we investigated the stability of the Co1Re1/TiO2 catalyst for the hydrogenation of CAL to COL. During each cycle, after a complete reaction at 140 °C for 4 h, the catalyst was centrifuged, washed with THF for five times, and reused for the next runs. As shown in Figure 5, the Co1Re1/TiO2 catalyst kept its good performance during recycling, the conversion of CAL was 97%, and the selectivity of COL was 87% after four runs. The XRD pattern of the spent catalyst showed no noticeable morphology changes compared with the fresh one ( Figure S1). Moreover, the comparable metal contents in the fresh and spent catalysts determined by ICP-OES indicated no obvious metal leaching (Table  S6). These results suggest that the Co1Re1/TiO2 catalyst has excellent stability for the hydrogenation of CAL to COL.   In addition, the applicability of the Co 1 Re 1 /TiO 2 /FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,βunsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co 1 Re 1 /TiO 2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co 1 Re 1 /TiO 2 /FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- 2 In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- 3 Molecules 2023, 28, x FOR PEER REVIEW 8 of 14 In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path-

99
In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- In addition, the applicability of the Co1Re1/TiO2/FA system was tested for the selective hydrogenation of various α,β-unsaturated aldehydes to the corresponding α,β-unsaturated alcohols under the optimized reaction conditions, except for the reaction time, which was adjusted to obtain high yields. The results are listed in Table 4. Similar to COL, α-methyl cinnamaldehyde also achieved high conversion (99%, entry 1) and selectivity (86%), and the remaining 14% selectivity was attributed to simultaneous hydrogenation of C=C and C=O by-product. Moreover, aliphatic crotonaldehyde and citral could be converted well, and the target products were offered with selectivities of 74% and 96%, respectively (entries 2, 3). It is clear that substituent groups greatly affected the performance of the Co1Re1/TiO2 catalyst. In general, the selectivity to α,β-unsaturated alcohols is closely related to the steric prohibition of the substituent groups. The more steric prohibition of C=C by the substitutes at γ-carbon, the higher the C=O selectivity, which is in good agreement with the literature [62]. Furthermore, a high conversion of 99% was obtained after the hydrogenation of cycloaliphatic isophorone (entry 4), with a product selectivity of 84% accompanied by excessive hydrogenation by-product of 16%. These results well verify that the Co1Re1/TiO2/FA system is effective for the selective hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols. To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co1Re1/TiO2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt3, cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt3 is as follows [34,64,65] NEt3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination path- To speculate the possible reaction pathway, the time course experiments of the hydrogenation of CAL over the Co 1 Re 1 /TiO 2 catalyst was carried out under optimal reaction conditions ( Figure 6a). As shown in Figure 6a, CAL was rapidly converted (73%), and a 42% yield of COL was achieved at the initial 1 h, accompanied by 26% of cinnamyl formate and 5% of HCOL. Under the weak alkaline condition of NEt 3 , cinnamyl formate could be easily formed through esterification of COL with formic acid [63], which then smoothly decreased and disappeared within 4 h with the decomposition of formic acid. CAL was almost completely converted at 4 h, and COL and HCOL were offered with yields of 88% and 10%, respectively. It is recognized that the decomposition mechanism of FA in the presence of NEt 3 is as follows [34,64,65] NEt 3 acts as a proton scavenger to facilitate the O-H bond cleavage, thus forming metal-formate species during the initial step of the reaction. Then metal-formate produces molecular hydrogen through a β-elimination pathway. This hydrogen desorption step is irreversible, indicating that it is feasible to use renewable FA as a convenient hydrogen donor instead of molecular H 2 for sustainable and green organic synthesis. The proposed mechanism is shown in Figure 6b. The Co 1 Re 1 /TiO 2 catalyst and FA produce metal-formate species in the presence of NEt 3 , which are then decomposed to generate metal-hydride species and CO 2 . Meanwhile, the C=O bond on CAL is adsorbed on the ReO x species and is coordinated with metal-hydride. Finally, it is neutralized with + HNEt 3 to form COL, thereby realizing the regeneration of the Co 1 Re 1 /TiO 2 catalyst and NEt 3 .

Preparation of Various CoRe/TiO2 Catalysts
A series of TiO2-supported CoRe catalysts with variable Co:Re molar ratios (Co:Re = 2:1, 1:1 or 1:2) were prepared by an incipient wetness impregnation method. The loading of Co in the catalysts was kept at 2 wt%. Taking the CoRe/TiO2 catalyst with a Co:Re molar ratio of 1:1 as an example, the preparation method was as follows: firstly, TiO2 was calcined at 500 °C for 4 h to remove the impurities prior to impregnation of the metal precursor. Then, 0.0998 g of Co(NO3)2·6H2O and 0.0920 g of NH4ReO4 were dissolved in ca. 1 mL deionized H2O. An appropriate amount of TiO2 (ca. 0.98 g) was slowly added to the aqueous solution under ultrasound. After being impregnated at room temperature for 24 h, the mixture was dried at 110 °C overnight and finally reduced at 500 °C in a tubular furnace under hydrogen flow for 3 h to obtain the target catalyst, which was denoted as Co1Re1/TiO2. Two monometallic catalysts 2 wt% Co/TiO2 and 2 wt% Re/TiO2 were prepared by using the same method for comparison. The information on the content of Co and Re in the corresponding catalyst is described in Table S7.

Catalyst Characterization
Brunauer-Emmett-Teller (BET) surface areas and pore structures of the catalyst samples were measured by pulsed N2 adsorption-desorption method at −196 °C using Micromeritics ASAP 2460 analyzer. Before N2 physisorption, the samples were degassed under vacuum at 250 °C for 3 h. Transmission electron microscopy (TEM) images were ob-

Preparation of Various CoRe/TiO 2 Catalysts
A series of TiO 2 -supported CoRe catalysts with variable Co:Re molar ratios (Co:Re = 2:1, 1:1 or 1:2) were prepared by an incipient wetness impregnation method. The loading of Co in the catalysts was kept at 2 wt%. Taking the CoRe/TiO 2 catalyst with a Co:Re molar ratio of 1:1 as an example, the preparation method was as follows: firstly, TiO 2 was calcined at 500 • C for 4 h to remove the impurities prior to impregnation of the metal precursor. Then, 0.0998 g of Co(NO 3 ) 2 ·6H 2 O and 0.0920 g of NH 4 ReO 4 were dissolved in ca. 1 mL deionized H 2 O. An appropriate amount of TiO 2 (ca. 0.98 g) was slowly added to the aqueous solution under ultrasound. After being impregnated at room temperature for 24 h, the mixture was dried at 110 • C overnight and finally reduced at 500 • C in a tubular furnace under hydrogen flow for 3 h to obtain the target catalyst, which was denoted as Co 1 Re 1 /TiO 2 . Two monometallic catalysts 2 wt% Co/TiO 2 and 2 wt% Re/TiO 2 were prepared by using the same method for comparison. The information on the content of Co and Re in the corresponding catalyst is described in Table S7.

Catalyst Characterization
Brunauer-Emmett-Teller (BET) surface areas and pore structures of the catalyst samples were measured by pulsed N 2 adsorption-desorption method at −196 • C using Micromeritics ASAP 2460 analyzer. Before N 2 physisorption, the samples were degassed under vacuum at 250 • C for 3 h. Transmission electron microscopy (TEM) images were obtained using a Tecnai G2 F30 S-Twin instrument (FEI Co., Columbia, SC, USA). The particle size distribution of metal nanoparticles in the sample was determined by measuring approximately 100 particles randomly selected from the TEM micrographs. The metal dispersion and particle size were measured by CO chemisorption at 35 • C using a Micromeritics ASAP 2020 system. X-ray diffraction (XRD) was conducted using an X' Pert PRO X-ray diffractometer equipped with Cu Kα radiation at 40 kV and 40 mA (λ = 0.15405 nm). Samples were scanned from 20 • to 80 • with a scanning rate of 4 • ·min −1 and a step size of 0.02 • . The content of the metals was measured by inductively coupled plasma-optical emission spectroscopy (ICP-OES) using a Thermo Fisher iCAP PRO. X-ray photoelectron spectroscopy (XPS) spectra were obtained using an Escalab Mark II X-ray spectrometer (VG Co., Manchester, UK) equipped with a magnesium anode (Mg Kα = 1253.6 eV). Energy corrections were performed using a C 1s peak of the pollutant carbon at 284.8 eV.

Catalytic Performance
The catalytic hydrogenation of α,β-unsaturated aldehydes to α,β-unsaturated alcohols was performed in a 25 mL stainless-steel autoclave equipped with magnetic stirring. In a typical experiment, 3 mmol of α,β-unsaturated aldehyde, 6 mmol of FA, 6 mmol of NEt 3 , 80 mg of catalyst and 10 mL of THF were added into the reactor. The reactor was sealed and purged with N 2 five times. The autoclave was then heated to the required temperature and kept at this temperature for the required time under the continuous stirring speed of 1000 rpm. After the reaction, the autoclave was quickly cooled to room temperature, and the reaction products were separated from the catalyst by centrifugation and quantitatively analyzed with an Agilent 7890 A gas chromatographer equipped with an HP-5 capillary column (30.0 m × 0.32 mm × 0.25 µm) and a flame ionization detector (FID) using ndodecane as an internal standard. The conformation of the products was performed on an Agilent 6890 GC system coupled to a mass spectrometer equipped with an Agilent 5973 quadrupole mass analyzer. CAL conversion and selectivity and the yield of the products were calculated by the following equations.

Conclusions
In summary, CoRe bimetallic catalysts supported on TiO 2 were achieved and were first reported in the selective hydrogenation of CAL to COL using FA as a hydrogen donor to replace the traditional molecular hydrogen. Especially, the Co 1 Re 1 /TiO 2 catalyst performed excellent activity, selectivity and stability, with a 99% conversion of CAL and 89% COL selectivity, and no obvious deactivation was observed after using it four times. Under similar reaction conditions, using α-methyl cinnamaldehyde, crotonaldehyde, citral and isophorone as feedstocks, high conversions and excellent selectivities to allylic alcohols were also achieved. The technique established in this work provides a green, mild and efficient process for the selective hydrogenation of α,β-unsaturated aldehydes to allylic alcohols.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.