Synthesis and Characterization of Rh/B–TNTs as a Recyclable Catalyst for Hydroformylation of Olefin Containing –CN Functional Group

The TiO2-based nanotubes (TNTs, B–TNTs) of different surface acidities and their supported Rh catalysts were designed and synthesized. The catalysts were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray photoelectron spectrometer (XPS), tempera–ture–programmed desorption of ammonia (NH3–TPD), atomic emission spectrometer (ICP), and Brunauer–Emmett–Tellerv (BET) surface-area analyzers. Images of SEM and TEM showed that the boron-decorated TiO2 nanotubes (B–TNTs) had a perfect multiwalled tubular structure; their length was up to hundreds of nanometers and inner diameter was about 7 nm. The results of NH3-TPD analyses showed that B–TNTs had a stronger acid site compared with TNTs. For Rh/TNTs and Rh/B–TNTs, Rh nanoparticles highly dispersed on B–TNTs were about 2.79 nm in average diameter and much smaller than those on TNTs, which were about 4.94 nm. The catalytic performances of catalysts for the hydroformylation of 2-methyl-3-butennitrile (2M3BN) were also evaluated, and results showed that the existence of B in Rh/B–TNTs had a great influence on the catalytic performance of the catalysts. The Rh/B–TNTs displayed higher catalytic activity, selectivity for aldehydes, and stability than the Rh/TNTs.


Introduction
Hydroformylation of olefins is one of the most important homogeneous catalytic reactions in the chemical industry with a worldwide oxoaldehyde production [1]. By hydroformylation, one more carbon aldehyde than the original olefin can be obtained. Aldehydes are an important chemical raw material and act as intermediates in the synthesis of drugs, pesticides, natural products, and so on [2][3][4]. According to statistics, the production of aldehydes globally is now over 6.0 × 10 6 t/a [5].
The typical hydroformylation of olefins is mainly catalyzed by homogeneous catalysis. Although the homogeneous catalyst system for hydroformylation has high catalytic activity, good selectivity, and other advantages, the transition metal complex used in the catalyst system may dissolve in the product, resulting in difficulties in the recovery of the catalyst [6,7]. Thus, there are two technical and scientific issues; one is the separation of the homogeneous catalyst from the product, and the other is preventing active components from loss. To resolve these issues, researchers have done much research treated with 60 mL 10 M NaOH aqueous solution in an autoclave for 12 h at 150 • C. The resulting material was washed with distilled water for it to become neutral and then dried at 60 • C, whereby a resulting white B-TNTs powder was obtained. The titanium sols containing 10 wt.% boron can be prepared by changing the amount of H 3 BO 3 to 2.02 g. The actual B contents of the two B-TNTs powders synthesized and labeled as B-TNTs (b) and B-TNTs (c) were 0.56 and 0.99 wt.%, respectively.

Synthesis of Rh/TNTs and Rh/B-TNTs
The Rh/TNTs and Rh/B-TNTs were prepared by using the impregnation-photoreducing procedure. A typical synthesis process is as follows: 1 g TNTs was added to 20 mL 2 wt.% RhAc2 aqueous solution, and then the suspension was vigorously agitated for 6 h. After low-energy sonication for 30 min, the suspension was centrifuged and the resulting solid was washed with a little deionized water to remove the ions adsorbed on the outer surface of TNTs, after which the solid was transferred into a 60 mL photo-reactor with 50 mL ethanol-water solution (Vethanol:Vwater = 9:1). The mixture was irradiated using a 300 W high-pressure mercury lamp, which ended with a color change of the mixture from blue to grey. Then, the mixture was centrifuged and the solid obtained was washed with deionized water and dried at 60 • C for 8 h in a vacuum. The catalyst obtained was labeled as Rh/TNTs (a 1 ). The same procedures were used to prepare Rh/B-TNTs (b 1 ), Rh/B-TNTs (c 1 ) to just change TNTs to B-TNTs (b) and B-TNTs (c). The preparation process of Rh/B-TNTs (c 2 ) and Rh/B-TNTs (c 3 ) is similar to that of Rh/B-TNTs (c 1 ), which had different Rh loadings (1 wt.% and 3 wt.% RhAc2).

Evaluation of Catalytic Performance of Catalysts for Hydroformylation
The catalytic activities of Rh/TNTs and Rh/B-TNTs for the hydroformylation of 2M3BN were evaluated. In a typical experiment, a 0.4 g catalyst and the required amount of substrate and solvent were placed in a 250-mL stainless steel autoclave reactor with a magnetic stirrer. The reactor was purged three times with H 2 and then pressurized to 6.0 MPa with CO and H 2 (CO/H 2 = 1:1). After this, the reactor was heated to the desired temperature while stirring. When the reaction was over, the stirring was stopped. The reactor was then cooled to room temperature and the pressure was released gradually. The product was analyzed by GC7890B-5977A MS (Agilent, Santa Clara, CA, USA) or GC (GC-2014 gas chromatograph equipped with a 30 m × 0.53 mm SE-30 capillary column and a FID, Shimadzu, Japan,). Recycling uses of catalysts were carried out under the same conditions after recovering the catalyst from the reaction solution via centrifugation.

Characterization
The crystal phase and structure of catalysts were detected by XRD, Rigaku D/Max-2500, (Rigaku, Japan), which was performed with Cu Ka radiation (λ = 1.54 Å) at 2θ from 10 • to 80 • . TEM images were recorded using a Tecnai G2 F20 instrument (FEI, Hillsboro, OR, USA) at an accelerating voltage of 200 kV. The morphologies were analyzed by SEM, Shimadzu SS-550, Shimadzu, Japan,). The X-ray photoelectron spectrometer (XPS, Kratos Axis Ultra DLD multi-technique X-ray photoelectron spectra, Kratos Analytical Ltd., Manchester, UK) was used to test the chemical states of Rh in catalysts, and all binding energies were calibrated using C1s (E b = 284.6 eV) for the reference. The contents of Rh and B were determined by ICP-AES (ICP-9000, USA Thermo Jarrell-Ash Corp, Franklin, MA, USA). N 2 adsorption/desorption isotherms were collected on an Autosorb-1-MP 1530VP (Quantachrome, Florida, FL, USA) automatic surface area and porosity analyzer. The sample was degassed at 473 K for 5 h and then analyzed at 77 K. The relative pressure (P/P 0 ) range used for the calculation of the BET surface area was from 0.05 to 0.30. NH 3 -TPD measurement was performed on an Autosorb-IQ-C-XR automatic surface area and a porosity analyzer equipped with a thermal conductivity detector (TCD, Quantachrome, Florida, FL, USA). The samples were pretreated at 150 • C for 1 h in a flow of Ar, and then subjected to NH 3 adsorption until saturation at room temperature. Desorbed NH 3 was monitored at a heating rate of 10 • C/min from room temperature to 800 • C.

BET and ICP Analysis
A Brunauer-Emmett-Teller (BET) analysis was carried out to investigate the specific surface area (SSA) of the catalysts ( Table 1). The actual contents of Rh and B in catalysts determined by the ICP were listed in Table 1. The SSA of TNTs (a), B-TNTs (b), and B-TNTs (c) were calculated to be 227.6, 238.6, and 275.8 m 2 /g, respectively. It can be observed that the BET surface area increased with the increasing content of B, an effect which can be attributed to the lower crystal size of the anatase phase in the B-TNTs supporter [28,29]. The SSA of all catalysts were significantly lower than that of pure supporters, which may be ascribed to the metal nanoparticles deposited on the outer and inner surface or which occupied the interspace between walls of TNTs [23,[30][31][32].  There is no peak related to the Rh (PDF: 05-0685), which means that Rh particles are small and highly dispersed on the TNTs and B-TNTs, which is a similar result to that observed by previous reports [15,23].

BET and ICP Analysis
A Brunauer-Emmett-Teller (BET) analysis was carried out to investigate the specific surface area (SSA) of the catalysts ( Table 1). The actual contents of Rh and B in catalysts determined by the ICP were listed in Table 1. The SSA of TNTs (a), B-TNTs (b), and B-TNTs (c) were calculated to be 227.6, 238.6, and 275.8 m 2 /g, respectively. It can be observed that the BET surface area increased with the increasing content of B, an effect which can be attributed to the lower crystal size of the anatase phase in the B-TNTs supporter [28,29]. The SSA of all catalysts were significantly lower than that of pure supporters, which may be ascribed to the metal nanoparticles deposited on the outer and inner surface or which occupied the interspace between walls of TNTs [23,[30][31][32].  There is no peak related to the Rh (PDF: 05-0685), which means that Rh particles are small and highly dispersed on the TNTs and B-TNTs, which is a similar result to that observed by previous reports [15,23].   Figure 2 shows the SEM images of the B-TNTs. As can be seen from Figure 2, the synthesized B-TNTs showed 1-D morphology with a length of up to hundreds of nanometers. We can see that  Figure 2 shows the SEM images of the B-TNTs. As can be seen from Figure 2, the synthesized B-TNTs showed 1-D morphology with a length of up to hundreds of nanometers. We can see that the single nanotube in the B-TNTs samples is clearly discernible, showing a uniform diameter and a smooth surface.   (Figure 3a,b). The nanotubes have a multiwalled structure with a spacing of about 7 nm. There is no obvious rupture or fracture caused by the light reduction process. These are attributable to the dispersed Rh-containing compound, which plays the same stable supporting role as Co and Fe compounds reported in the literature [16,27]. Rh nanoparticles are well-dispersed on the inner and outer surfaces of the nanotubes without agglomeration. Figure 3c,d show the size distributions of particles in different catalysts. The average diameter of the Rh particles in Rh/B-TNTs (c1) is about 2.79 nm (Figure 3d), which is obviously smaller than that of Rh/TNTs (a1) (4.94 nm, Figure  3c), and the size distributions of particles are in a relatively narrow range. These results mean that the presence of B in B-TNTs can effectively prevent the growth of Rh nanoparticles and improve the distributions of Rh particles, which is similar to the results found in literature [33].    (Figure 3a,b). The nanotubes have a multiwalled structure with a spacing of about 7 nm. There is no obvious rupture or fracture caused by the light reduction process. These are attributable to the dispersed Rh-containing compound, which plays the same stable supporting role as Co and Fe compounds reported in the literature [16,27]. Rh nanoparticles are well-dispersed on the inner and outer surfaces of the nanotubes without agglomeration. Figure 3c,d show the size distributions of particles in different catalysts. The average diameter of the Rh particles in Rh/B-TNTs (c 1 ) is about 2.79 nm (Figure 3d), which is obviously smaller than that of Rh/TNTs (a 1 ) (4.94 nm, Figure 3c), and the size distributions of particles are in a relatively narrow range. These results mean that the presence of B in B-TNTs can effectively prevent the growth of Rh nanoparticles and improve the distributions of Rh particles, which is similar to the results found in literature [33].  Figure 2 shows the SEM images of the B-TNTs. As can be seen from Figure 2, the synthesized B-TNTs showed 1-D morphology with a length of up to hundreds of nanometers. We can see that the single nanotube in the B-TNTs samples is clearly discernible, showing a uniform diameter and a smooth surface.   (Figure 3a,b). The nanotubes have a multiwalled structure with a spacing of about 7 nm. There is no obvious rupture or fracture caused by the light reduction process. These are attributable to the dispersed Rh-containing compound, which plays the same stable supporting role as Co and Fe compounds reported in the literature [16,27]. Rh nanoparticles are well-dispersed on the inner and outer surfaces of the nanotubes without agglomeration. Figure 3c,d show the size distributions of particles in different catalysts. The average diameter of the Rh particles in Rh/B-TNTs (c1) is about 2.79 nm (Figure 3d), which is obviously smaller than that of Rh/TNTs (a1) (4.94 nm, Figure  3c), and the size distributions of particles are in a relatively narrow range. These results mean that the presence of B in B-TNTs can effectively prevent the growth of Rh nanoparticles and improve the distributions of Rh particles, which is similar to the results found in literature [33].

XPS Analysis
XPS was performed to evaluate the chemical state of the catalysts. As seen in Figure 4a, there was only one peak at about 192.4 eV for B 1s. As compared to the standard binding energy for B 1s in TiB 2 (187.5 eV, Ti-B bonds) and in B 2 O 3 (193.1 eV, B-O bonds), the binding energy of B 1s was between that of B 2 O 3 and TiB 2 ; thus, the boron atom was probably incorporated into TiO 2 and the chemical environment surrounding the boron was likely to be Ti-B-O, which would be a similar result to those in previous reports [34,35].  [35,36]. Figure 4d,e show that the carbon peaks consisted of three components. The peak at 284.6 eV corresponds to the binding energies of C 1s, and the second and third smaller peaks at 286.2 and 288.6 eV observed for both samples are probably due to impurity and air absorbents [37].
As shown in Figure 4f,g, Rh 3d 5/2 and Rh 3d 3/2 show that the group peaks not only centered at 307.2 and 312.0 eV, but also at 309.1 and 313.8 eV. These indicate that Rh exists in two forms of Rh 0 and oxidation state. The higher binding energy state should have relevance to the incomplete reduction of Rh 2+ ions; however, the electron transfers from Rh adatoms to the titanate substrate and the formation of partially oxidized states cannot be excluded, which is consistent with the literature [38]. It is well-known that Rh 0 is the active site for the hydroformylation of olefins [15]. The percentage of Rh 0 in rhodium is calculated based on the area of the fitted Gauss peaks. When the Rh loading is almost identical, the proportions of Rh 0 in Rh/B-TNTs (c 1 ) is 89.3%, which is higher than that in Rh/TNTs (a 1 ). Thus, the incorporation of B can increase the proportion of Rh 0 in the catalyst, promoting the hydroformylation of 2M3BN.

XPS Analysis
XPS was performed to evaluate the chemical state of the catalysts. As seen in Figure 4a, there was only one peak at about 192.4 eV for B 1s. As compared to the standard binding energy for B 1s in TiB2 (187.5 eV, Ti-B bonds) and in B2O3 (193.1 eV, B-O bonds), the binding energy of B 1s was between that of B2O3 and TiB2; thus, the boron atom was probably incorporated into TiO2 and the chemical environment surrounding the boron was likely to be Ti-B-O, which would be a similar result to those in previous reports [34,35].  [35,36]. Figure 4d,e show that the carbon peaks consisted of three components. The peak at 284.6 eV corresponds to the binding energies of C 1s, and the second and third smaller peaks at 286.2 and 288.6 eV observed for both samples are probably due to impurity and air absorbents [37].
As shown in Figure 4f,g, Rh 3d5/2 and Rh 3d3/2 show that the group peaks not only centered at 307.2 and 312.0 eV, but also at 309.1 and 313.8 eV. These indicate that Rh exists in two forms of Rh 0 and oxidation state. The higher binding energy state should have relevance to the incomplete reduction of Rh 2+ ions; however, the electron transfers from Rh adatoms to the titanate substrate and the formation of partially oxidized states cannot be excluded, which is consistent with the literature [38]. It is well-known that Rh 0 is the active site for the hydroformylation of olefins [15]. The percentage of Rh 0 in rhodium is calculated based on the area of the fitted Gauss peaks. When the Rh loading is almost identical, the proportions of Rh 0 in Rh/B-TNTs (c1) is 89.3%, which is higher than that in Rh/TNTs (a1). Thus, the incorporation of B can increase the proportion of Rh 0 in the catalyst, promoting the hydroformylation of 2M3BN.

NH3-TPD Analysis
It has been reported that Lewis acid promoted CO insertion which favors hydroformylation over hydrogenation of alkenes [39]. Previous literature [40] reported that the incorporation of boron into Tisubstituted silicalite-2 zeolites (TS-2) could enhance both the number and strength of the acidic sites in TS-2.
NH3-TPD was used to measure the acidic properties of TNTs (a) and B-TNTs (c), and the results are shown in Figure 5. There are two NH3 desorption peaks at 135 and 335 °C in the TNTs ( Figure  5a), which are caused by the NH3 adsorbed on the acidic sites of the outer and inner surfaces of TNTs, respectively. It can be concluded that there are a lot of acidic sites on the inner and outer surfaces of TNTs, which is beneficial to the preferential adsorption of olefin to catalysts. Compared with TNTs (a), the B-TNTs (c) show three NH3 desorption peaks at around 110, 375, and 560 °C (Figure 5b), which is similar to the NH3-TPD results of Co-B/TNTs [41]. The peak at around 560 °C corresponding to NH3 adsorbed on stronger acidic sites implies that B may lead to the formation of a new kind of acidic site and enhance the acidity of catalysts.

NH 3 -TPD Analysis
It has been reported that Lewis acid promoted CO insertion which favors hydroformylation over hydrogenation of alkenes [39]. Previous literature [40] reported that the incorporation of boron into Ti-substituted silicalite-2 zeolites (TS-2) could enhance both the number and strength of the acidic sites in TS-2. NH 3 -TPD was used to measure the acidic properties of TNTs (a) and B-TNTs (c), and the results are shown in Figure 5. There are two NH 3 desorption peaks at 135 and 335 • C in the TNTs (Figure 5a), which are caused by the NH 3 adsorbed on the acidic sites of the outer and inner surfaces of TNTs, respectively. It can be concluded that there are a lot of acidic sites on the inner and outer surfaces of TNTs, which is beneficial to the preferential adsorption of olefin to catalysts. Compared with TNTs (a), the B-TNTs (c) show three NH 3 desorption peaks at around 110, 375, and 560 • C (Figure 5b), which is similar to the NH 3 -TPD results of Co-B/TNTs [41]. The peak at around 560 • C corresponding to NH 3 adsorbed on stronger acidic sites implies that B may lead to the formation of a new kind of acidic site and enhance the acidity of catalysts. respectively. It can be concluded that there are a lot of acidic sites on the inner and outer surfaces of TNTs, which is beneficial to the preferential adsorption of olefin to catalysts. Compared with TNTs (a), the B-TNTs (c) show three NH3 desorption peaks at around 110, 375, and 560 °C (Figure 5b), which is similar to the NH3-TPD results of Co-B/TNTs [41]. The peak at around 560 °C corresponding to NH3 adsorbed on stronger acidic sites implies that B may lead to the formation of a new kind of acidic site and enhance the acidity of catalysts.

Catalytic Activity Evaluation
Scheme 1 shows the possible aldehydes formed in the hydroformylation reaction of 2M3BN. In the real hydroformylation reaction of 2M3BN, an isomerization reaction of 2-methyl-2butennitrile (2M2BN) may take place. Theoretically speaking, the products should include at least three kinds of aldehyde in the hydroformylation of 2M3BN, one of which is linear aldehyde and the others are branched-chain aldehydes ( Figures S1-S3).
It is very interesting to study the selectivity for these aldehydes in the hydroformylation of 2M3BN and it can help us to understand how the -CN functional group affects the reaction process of the hydroformylation of olefin. The GC and MS fragments of products formed in the hydroformylation of 2M3BN are shown in the supporting information. Table 2 shows the influence of B content on the activity of the catalyst. As can be seen from Table  2, the conversion rate of 2M3BN over all catalysts can reach up to 100%.
However, the selectivity for aldehyde in the reaction over Rh/B-TNTs is higher than that over Rh/TNTs. The selectivity for product aldehydes is shown to increase from 72% to 81%. In addition, the isomerization of 2M3BN clearly reduces from 25.8% to 17.9%. These results imply that in the catalytic processes, the presence of B in the catalysts may increase the rate of migratory insertion of the CO group, which suppresses the isomerization of 2M3BN and improves reaction selectivity for aldehydes. When the amount of B in Rh/B-TNTs increases from 0.56 wt.% to 0.99 wt.%, both of the selectivity for aldehyde and the l:b ratio of the product aldehydes increase. Thus, the suitable amount of B doping should be 0.99 wt.% in the present work.  In the real hydroformylation reaction of 2M3BN, an isomerization reaction of 2-methyl-2-butennitrile (2M2BN) may take place. Theoretically speaking, the products should include at least three kinds of aldehyde in the hydroformylation of 2M3BN, one of which is linear aldehyde and the others are branched-chain aldehydes ( Figures S1-S3).
It is very interesting to study the selectivity for these aldehydes in the hydroformylation of 2M3BN and it can help us to understand how the -CN functional group affects the reaction process of the hydroformylation of olefin. The GC and MS fragments of products formed in the hydroformylation of 2M3BN are shown in the supporting information. Table 2 shows the influence of B content on the activity of the catalyst. As can be seen from Table 2, the conversion rate of 2M3BN over all catalysts can reach up to 100%.
However, the selectivity for aldehyde in the reaction over Rh/B-TNTs is higher than that over Rh/TNTs. The selectivity for product aldehydes is shown to increase from 72% to 81%. In addition, the isomerization of 2M3BN clearly reduces from 25.8% to 17.9%. These results imply that in the catalytic processes, the presence of B in the catalysts may increase the rate of migratory insertion of the CO group, which suppresses the isomerization of 2M3BN and improves reaction selectivity for aldehydes. When the amount of B in Rh/B-TNTs increases from 0.56 wt.% to 0.99 wt.%, both of the selectivity for aldehyde and the l:b ratio of the product aldehydes increase. Thus, the suitable amount of B doping should be 0.99 wt.% in the present work.  , n(2M2BN) is the amount of 2M2BN in the liquid after reaction, n(2M3BN) is the amount of reaction substrate. Table 3 shows the effects of Rh content on the catalytic performance of the Rh/B-TNTs. It can be seen from Table 3 that the conversion rate of 2M3BN, the turnover frequency (TOF), and the selectivity for aldehydes all increase with an increase in Rh loadings from 0.09 wt.% to 0.16 wt.%. However, when the amount of Rh increases from 0.16 wt.% to 0.19 wt.%, the selectivity for aldehydes does not increase, but the ratio of linear aldehyde to the branched one clearly decreases. Thus, it can be concluded that the better rhodium content is 0.16wt.% for the catalyst Rh/B-TNTs. , n(2M2BN) is the amount of 2M2BN in the liquid after reaction, n(2M3BN) is the amount of reaction substrate. Table 4 shows the effect of reaction temperature on the Rh/TNTs (a 1 ) and Rh/B-TNTs (c 1 ) catalyzed hydroformylation reaction of 2M3BN. Table 4 shows that the conversion of 2M3BN over Rh/TNTs, the TOF, and the ratio of l:b, all increased along with the increase in reaction temperature from 80 • C to 120 • C. For Rh/B-TNTs (c 1 ), when the reaction temperature increased from 80 • C to 120 • C, the conversion of 2M3BN, the total amount of aldehydes, and the ratio of l:b, all improved. Thus, it can be concluded that the suitable reaction temperature should be 120 • C. The difference is that under the almost identical Rh loading and reaction temperature, the activity of Rh/B-TNTs is significantly higher than that of Rh/TNTs. This clearly demonstrates that the presence of B in the catalysts is greatly beneficial for the hydroformylation reaction of 2M3BN. However, the isomerization of 2M3BN increases when the temperature increases from 80 • C to 120 • C, indicating that increasing the temperature is more favorable for the isomerization of 2M3BN.
The stability of catalysts is important for the hydroformylation of olefins for practical application. It was reported that the stability of Ni catalysts can be enhanced by boron doping [42]. Density functional theory calculations also suggested that the presence of boron could improve the stability of the catalyst by selectively blocking the deposition, nucleation, and growth of resilient carbon species of Co catalysts under Fischer-Tropsch synthesis conditions [43]. , n(2M2BN) is the amount of 2M2BN in the liquid after reaction, n(2M3BN) is the amount of reaction substrate.
To investigate the stability of the catalysts, Rh/TNTs (a 1 ) and Rh/B-TNTs (c 1 ) were selected for recycle catalytic experiments. The experimental results are listed in Table 5. As can be seen from this Table, Rh/B-TNTs (c 1 ) remained highly active and the yield of aldehyde maintained at around 67.8% in the fourth recycles. However, the Rh/TNTs (a 1 ) shows poor stability. We used ICP to test the rhodium content in the solution after the reaction (listed in Table 6). We can clearly see that the addition of B can significantly reduce the loss of rhodium. The Rh content is about 18.8 PPM in the solution after the first reaction used for the Rh/B-TNTs(c 1 ) catalyst and is much smaller than that in the solution used for the Rh/TNTs(a 1 ) catalyst, which is about 42.8 PPM. The experimental results show that boron doping can greatly improve the catalytic stability of the catalyst.   , n(2M2BN) is the amount of 2M2BN in the liquid after reaction, n(2M3BN) is the amount of reaction substrate.

Conclusions
This study showed that the high catalytic activity of B-modified Rh/TNTs catalysts for olefin hydroformylation could be improved by increasing the acid site of the catalyst. The Rh/B-TNTs catalysts not only display good catalytic performance, but also good circulation. The catalysts could be easily recovered and were used up to four times. Although we observed a modest decrease of activity with the recycling of the catalyst, the selectivity decreased quite a bit from 81% to 68%, and Rh leaching was also observed. However, the stability of the catalyst Rh/B-TNTs(c 1 ) was much better than that of Rh/TNTs(a 1 ). In conclusion, the introduction of B can not only improve the catalytic activity of the catalyst, but also improve its stability.