Characterization of CsPMo/meso-Y
The powder X-ray diffraction (XRD) was carried out to investigate the structures of CsPMo/meso-Y with different CsPMo loadings as shown in Fig. 1a. The diffraction peaks of CsPMo/meso-Y coincided exactly with that of meso-Y, indicating that the incorporation of CsPMo doesn’t affect the structure of meso-Y32. Moreover, the diffraction peaks of CsPMo were not observed in the XRD patterns of CsPMo/meso-Y, indicating that CsPMo was uniformly high-dispersed in the meso-Y33. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images in Supplementary Figs. 1 and 2 show that both meso-Y and CsPMo/meso-Y exhibit blocky morphology without obvious difference, indicating that the structure and morphology of meso-Y don’t change after the loading of CsPMo, which is in agreement with the result of XRD analysis. The Fourier-transform infrared (FT-IR) spectra of CsPMo/meso-Y with different CsPMo loadings are shown in Fig. 1b, which show typical infrared bands corresponding to Y zeolite. Besides, two extra characteristic peaks are detected at 970 and 911 cm− 1, which are assigned to the stretching vibrations of Mo = O and Mo-Ob-Mo of CsPMo, respectively34. This result indicates that CsPMo is successfully loaded on meso-Y35.
N2 adsorption-desorption measurements were conducted on meso-Y and CsPMo/meso-Y, which are shown in Supplementary Fig. 3 and Supplementary Table 1. In comparison to the original meso-Y, the N2 adsorption capacity, specific surface area, pore volume and average pore diameter of CsPMo/meso-Y decrease. Among them, the specific surface area of 7.5% CsPMo/meso-Y decreases from 662 m2/g to 558 m2/g, the pore volume decreases from 0.20 cm3/g to 0.13 cm3/g and the average pore diameter decreases from 4.94 nm to 2.92 nm, which also demonstrate that CsPMo is successfully loaded into the pores of the meso-Y36. It is worth noting that CsPMo/meso-Y maintains the abundant void space even after loading CsPMo, which can be conducive to the adsorption of thioether into meso-Y channel to improve the catalytic activity for oxidation reaction26, 37. Supplementary Fig. 4 shows the energy-dispersive X-ray spectroscopy (EDS) images of 7.5% CsPMo/meso-Y, which manifest the uniform dispersion of Cs, P, Mo, Si and Al elements in CsPMo/meso-Y. In other words, CsPMo is well distributed within meso-Y. Inductively coupled plasma-optical emission spectrometry (ICP-OES) was used to determine the Mo content in the CsPMo/meso-Y catalysts (Supplementary Table 1). By calculation, the actual loadings of CsPMo are 3.3%, 4.6%, 6.0%, 7.5% and 9.3%, respectively, which increase with the incorporation of CsPMo. The results further confirm that CsPMo is successfully loaded on the meso-Y.
The electronic structure of the CsPMo/meso-Y was further investigated by X-ray photoelectron spectroscopy (XPS) analysis. The Mo 3d spectrum of CsPMo (Fig. 2a) shows two peaks at 233.0 and 236.1 eV, which correspond to MoVI 3d5/2 and MoVI 3d3/2, respectively. Compared with CsPMo, two additional peaks at ca.232.5 and 235.4 eV are observed in the spectra of CsPMo/meso-Y, which can be attributed to MoV 3d5/2 and MoV 3d3/2, respectively
$$\:(PM{o}_{\text{12}}^{\text{VI}}{O}_{40}{)}^{3-}+\text{pe}+\text{q}{\text{H}}^{+}\to\:{H}_{q}(PM{o}_{12-p}^{VI}M{o}_{p}^{V}{O}_{40}{)}^{(q-p-3)}$$
1
(Fig. 2b–f)38. The results clearly demonstrate that MoV formed in the CsPMo/meso-Y, which can be further confirmed by the phenomenon that the products changed from yellow (CsPMo) to green (CsPMo/meso-Y) (Supplementary Fig. 5). It is worth noting that the peaks of MoVI 3d5/2, MoVI 3d3/2 and O 1s (Fig. 3a) in the 7.5% CsPMo/meso-Y shift towards higher binding energies in contrast to CsPMo. Furthermore, compared with meso-Y, the binding energies of O 1s and Si 2p of 7.5% CsPMo/meso-Y decrease (Fig. 3a, b). These results indicate that electrons transfer from CsPMo to meso-Y, which demonstrates that the interactions formed between CsPMo and meso-Y7. The formation of MoV is likely to be caused by the incorporation of CsPMo into meso-Y. It is well-known that the presence of electrons and H+ at the same time is conducive to the formation of MoV (Eq. (1))5, 39. On the one hand, electron transfer occured between CsPMo and meso-Y during the preparation of the CsPMo/meso-Y. On the other hand, meso-Y could provide a lot of H+ due to its abundant Bronsted acid sites. In other words, the incorporation of CsPMo into meso-Y can offer both electrons and H+, which facilitate the formation of MoV. Above data demonstrate that meso-Y with abundant Bronsted acid sites is an excellent support which can regulate the valence of Mo of CsPMo.
To further investigate the interaction between CsPMo and meso-Y, we performed 31P solid state nuclear magnetic resonance (31P NMR) analyses of CsPMo and CsPMo/meso-Y. As seen in Fig. 4, a typical resonance at -6.35 ppm is observed for pure CsPMo, which can be attributed to the P atom
of the CsPMo Keggin structure. For the CsPMo/meso-Y, the resonance peak is nearly the same with that of CsPMo, suggesting that the Keggin structure of CsPMo was well-preserved. It is worth noting that a slight shift of P signal to downfield is observed after the incorporation of CsPMo into meso-Y, proving that there could be strong interactions between CsPMo and meso-Y, resulting in a lower electron density about the P atoms7. This feature further confirms the XPS results that there are electrons transfer between CsPMo and meso-Y.
Catalytic oxidation of thioether
The activities of selective oxidation of sulfides over CsPMo/meso-Y were tested with methyl p-tolyl sulfide as a model substrate and tert-butyl hydroperoxide (TBHP) as an oxidant (Table 1). In order to explore the optimal conditions for the reaction, the effects of CsPMo loadings, solvent types, catalyst dosages, reaction temperature and molar ratio of TBHP/substrate (O/S) on the catalytic performance were investigated.
The conversion of methyl p-tolyl sulfide by meso-Y alone was just 32.34% (Table 1, entry 7), while the conversion was 29.93% without catalyst (Table 1, entry 8), suggesting that meso-Y is not responsible for the reaction. Pure CsPMo also showed relatively low conversion (80.64%), which could be attributed to the insufficient exposure of active sites. The conversion of the prepared composites (CsPMo/meso-Y) was increased to more than 96% within 30 min (Table 1, entries 1–5), manifesting the higher activity for oxidation of thioether after embedding CsPMo into meso-Y. This might be attributed to the formation of mixed-valence species ((PMoVI, VO40)x−), which has proven to be conducive to improve catalytic activity40, 41. Further, the effect of CsPMo loading amount on the catalytic performance for oxidation of sulfides was
Table 1| Activities of different catalysts towards the oxidation of methyl p-tolyl sulfidea.
Entry
|
Catal.
|
Catal.
(mg)
|
O/Sb
|
Con.
(%)
|
Sel.c
(%)
|
1
|
3.3% CsPMo/meso-Y
|
10
|
1
|
96.81
|
98.74
|
2
|
4.6% CsPMo/meso-Y
|
10
|
1
|
97.29
|
97.95
|
3
|
6.0% CsPMo/meso-Y
|
10
|
1
|
98.2
|
98.05
|
4
|
7.5% CsPMo/meso-Y
|
10
|
1
|
99.14
|
98.31
|
5
|
9.3% CsPMo/meso-Y
|
10
|
1
|
98.62
|
98.56
|
6
|
CsPMo
|
0.8
|
1
|
80.64
|
98.79
|
7
|
meso-Y
|
9.2
|
1
|
32.34
|
98.32
|
8
|
Blank
|
0
|
1
|
29.93
|
98.01
|
aReaction conditions: 0.5 mmol substrate, 1 mL ethanol, 0.5 mmol TBHP, 60 oC, 30 min.
bO/S = the molar ratio of oxidant to substrate.
cThe selectivity of sulfoxide, the byproduct is sulphone.
investigated by increasing the CsPMo loadings from 3.3–9.3% (Table 1, entries 1–5). The conversion of methyl p-tolyl sulfide rose as the CsPMo loadings increased. While the conversion decreased when the loadings of CsPMo exceeded 7.5%. As seen in Table 1, 7.5% CsPMo/meso-Y showed the highest activity (99.14%) due to its highest MoV content, which can be confirmed by XPS (Fig. 2b–f). Figure 5a shows the solid-state UV-visible absorption spectra, a peak at ca.700 nm is observed, which corresponds to heteropoly blue42. And 7.5% CsPMo/meso-Y shows more significant absorption, which indicates that 7.5% CsPMo/meso-Y has the highest MoV content42, 43. This result is consistent with the results of XPS and the fact that 7.5% CsPMo/meso-Y shows a darker green (Supplementary Fig. 5). Meanwhile, according to the molar ratio of MoVI/MoV in Supplementary Table 2, 7.5% CsPMo/meso-Y has the highest MoV content. These results implied that MoV played a crucial role in oxidation of methyl p-tolyl sulfide. Additionally, the cyclic voltammetry (CV) further demonstrated the oxidative capability of 7.5% CsPMo/meso-Y was increased compared with either meso-Y or CsPMo (Fig. 5b), which may due to the higher active area after the incorporation of CsPMo into meso-Y, causing the higher activity for selective oxidation of thioether41.
As seen in Fig. 6, conversion and selectivity could be affected by various solvents. When ethanol was used as solvent, 7.5% CsPMo/meso-Y showed the highest conversion (99.14%). Interestingly, the conversion with various solvents followed the order of ethanol > n-propanol > n-butanol > n-pentanol > cyclohexane, which could be related to the polarity of the solvents. Polarity can be expressed with dielectric constant ε, polarity/polarizability index π* and Dimroth-Reichardt’s polarity parameter ET(30) (Supplementary Table 3)44–48. It can be seen that the more polar the solvent is, the easier it is to provide protons. These protons originating from solvent contribute to facilitating the oxidation reaction according to Eq. (2), where [HPA]ox and [HPA]red represent the oxidized and reduced forms of the heteropolyanion, respectively5. Similarly, meso-Y could also contribute to this oxidation process by supplying H+ for the process. While the relatively low conversion (96.33%) was obtained with methanol as solvent, whose polarity is highest. This may be due to the low solubility of methyl p-tolyl sulfide in methanol.
$$\:{\left[HPA\right]}_{ox}+substrate+{iH}^{+}\rightleftarrows\:{H}_{i}{\left[HPA\right]}_{red}+oxidized\:substrate\:\:\:\:\:\:\:\:$$
2
The dosage of catalysts, reaction temperature and the molar ratio of TBHP/substrate (O/S) also have influences on the catalytic performance of the reaction (Fig. 7a–c). The results revea-
led that the optimum amount of catalyst is 10 mg, which led to a conversion (99.14%) and selectivity (98.31%) of thioether within 30 min (Fig. 7a). The conversion of methyl p-tolyl sulfide increased as the reaction temperature rose to 60 oC; when the temperature reached 70 oC, the conversion of thioether decreased slightly. Therefore, the optimum temperature for the reaction is 60 oC. In addition, the exorbitant O/S could cause the selectivity decrease, which is due to the fact that excess oxidant leads to further oxidation of methyl p-tolyl sulfoxide to methyl p-tolyl sulfone49. Based on the above results, under the optimal reaction conditions, methyl p-tolyl sulfide can be converted to methyl p-tolyl sulfoxide within 30 minutes (Fig. 7d).
Due to the excellent results obtained above, a series of sulfides were selected as substrates to investigate the catalytic activity of the 7.5% CsPMo/meso-Y (Supplementary Table 4). The results show that the conversions of these substrates are all above 98% within 30 min, which clearly demonstrates that 7.5% CsPMo/meso-Y shows outstanding catalytic capability in the oxidation of sulfides.
In addition, the cycling stability of the 7.5% CsPMo/meso-Y catalyst was investigated. The used catalyst was centrifuged and washed three times with ethanol, and dried at 60 oC. As shown in Fig. 8a, the conversion and selectivity of the reaction don’t decrease significantly after three cycles (the conversion of methyl p-tolyl sulfide is 93.42% and the selectivity of methyl p-tolyl sulfoxide is 99.59% in the third cycle), indicating that the 7.5% CsPMo/meso-Y catalyst has excellent cycling stability. The XRD, FT-IR and EDS spectra of recovered 7.5% CsPMo/meso-Y after 3 cycles are consistent with those of fresh 7.5% CsPMo/meso-Y (Supplementary Figs. 6 and 7), which further proves the excellent stability and reusability of 7.5% CsPMo/meso-Y under reaction conditions.
Furthermore, we also investigated the oxidation reaction mechanism of methyl p-tolyl sulfide by free radical trapping experiments and Raman (Fig. 8b and Supplementary Fig. 8). It has been found that the introduction of scavengers, including 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), isopropanol (IPA), p-benzoquinone (PBQ) and diphenylamine (Ph2NH), has little effect on the conversion of thioether, suggesting that there is no involvement of free radicals in this catalytic process50, 51. Therefore, based on previous literatures, we speculated that the peroxo-metal group, which ariose due to the interaction between POM and TBHP, played a key role in the catalytic oxidation reaction52, 53. To confirm the inference, Raman spectra of 7.5% CsPMo/meso-Y were measured before and after use (Supplementary Fig. 8). Indeed, the difference of the peaks at 886 cm− 1 was observed after catalysis, assigned to the -O-O- vibrations, indicating the formation of peroxo-metal species54, 55. The mechanism for the catalytic oxidation of sulfides was proposed as shown in Fig. 8c, the active peroxo-molybdenum derived from 7.5% CsPMo/meso-Y and TBHP catalyzed the S atoms of the sulfides to obtain the corresponding sulfoxides and sulfones. Compared with CsPMo, the binding energy of mixed-valence species ((PMoVI, VO40)x−) with TBHP is higher, which is conducive to the formation of peroxo-metal species, causing a higher sulfide oxidation performance7.