3.2. XRD Analysis
Figure 1 shows the XRD patterns of Rh/TNTs, Rh/B–TNTs. It can be seen from
Figure 1 that all of the peaks can be perfectly indexed to anatase titania (PDF: 21-1272). The peaks at 2θ = 25.4°, 37.8°, 48.1°, 54.6°, 55.0°, 62.7°, 68.8°, and 75.0° can be assigned to diffractions of anatase TiO
2 (101), (004), (200), (105), (211), (204), (116), and (215). 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].
3.4. TEM Analysis
Figure 3 shows the TEM images of the Rh/TNTs and Rh/B–TNTs. It is clear from
Figure 3 that TNTs and B–TNTs show perfect nanotubular morphology with a length of about 200–300 nm and diameter of about 8–10 nm (
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].
3.5. 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
2O
3 (193.1 eV, B–O bonds), the binding energy of B 1s was between that of B
2O
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].
Figure 4b,c display the O 1s spectra comparison between Rh/TNTs (a
1) and Rh/B–TNTs (c
1). From
Figure 4b,c, the O 1s XPS spectra of the Rh/B–TNTs (c
1) sample could be fitted by three peaks, corresponding to Ti–O (530.0 eV), the hydrogen group O–H (531.6 eV), and the B–O (533.0 eV) bond, respectively, while there were only two O 1s XPS peaks assigned to the Ti–O and O–H bonds, respectively of the Rh/TNTs (a
1) sample, which also proved the presence of the B–O bond [
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.
3.6. 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 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.
3.7. 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-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.
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.
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].
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.