Haiyan
An
*,
Yujiao
Hou
,
Shenzhen
Chang
,
Jie
Zhang
and
Qingshan
Zhu
College of Chemistry, Dalian University of Technology, Dalian 116023, P. R. China. E-mail: anhy@dlut.edu.cn; Fax: +86-411-84657675; Tel: +86-411-84657675
First published on 17th October 2019
The selective oxidation of thioethers is among the most straightforward and common methods to obtain the synthetic intermediates sulfoxides for application in chemical industry, medicinal chemistry and biology. In this study, we prepared four hybrid compounds, i.e. Cs4[M(H2O)4][PMo6O21(PABA)3]2·nH2O 1–4 (M = Co, Mn, Ni, Zn; PABA = p-aminobenzoic acid), based on carboxylic acid-modified polyoxomolybdates and metal cations with wonderful catalytic performance for the selective oxidation of thioethers. The compounds 1–4 comprise PABA ligand-modified polyoxomolybdates [PMo6O21(PABA)3]3− as building units and the Co2+/Mn2+/Ni2+/Zn2+ cations as linkers to generate novel dimeric architectures. Using compound 1 as a pre-catalyst, methyl phenyl sulfides were nearly entirely converted to sulfoxides with the selectivity of 98% at room temperature within 20 minutes. Satisfactory catalytic effects were also observed in the selective oxidation of various phenyl sulfides with substituent groups; in particular, another typical thioether, i.e. the chemical warfare agent simulant 2-chloroethyl ethyl sulfide (CEES), could be completely degraded to nontoxic 2-chloroethyl ethyl sulfoxide (CEESO) with the selectivity of 99% within 12 minutes at room temperature. These pre-catalysts could be recycled at least 10 times by simple filtration with a negligible decrease in conversion and selectivity.
Polyoxometalates (POMs), as well-defined inorganic metal oxide clusters, possess significant applications in catalysis, optics, medicine, etc.3–18 Among these applications, oxidation catalysis has been one of the most attractive aspects in the last few years owing to the resistance of POMs to oxidation. As an important branch, the selective oxidation of various sulfides catalyzed by POMs has been explored by Hill, Hu, Song, Niu and other groups.19–35 In the early years, some homogeneous materials, such as [(n-C4H9)4N]4(α-Mo8O26), [{CH2CH(CH2)6Si}xOySiWwOz]4−, and TBA4H2[BW11Mn(H2O)O39]·H2O, based on POMs have been developed for the selective oxidation of a wide range of thioethers.21–23 Subsequently, to solve the problem of recyclability, supported heterogeneous materials based on POMs and some solid materials, such as zeolites, polymers, or silica, have been extensively studied and have gained ever-increasing attention. In 2016, Zhao et al. have reported that the POM anion {SiV2W10} can be combined with a linear positively charged polymer, poly(diallyldimethylammonium chloride) (PDDA), to obtain the composite material PDDA-SiV2W10 through electrostatic interaction, which can efficiently catalyze the oxidation of various thioethers.24 The same year, Song and coworkers prepared a supported POM pre-catalyst by employing the facile condensation reaction between polytungstate {P2W15} and γ-Al2O3 spheres.25 This supported material P2W15-γ-Al2O3 is active for the oxygenation of thioethers, and displays high stability, excellent reactivity and selectivity. Recently, Hu's group successively obtained several composite materials based on POMs, [CnH2n+1N(CH3)3]7HNb6O19, Zn2Cr-LDH-PW11M and Mg3Al-LDH-Nb6, which exhibited good catalytic activities for the oxidative degradation of CEES.26–28 However, most of these supported materials always suffer from a low loading efficiency, ill-defined structure and inhomogeneous pre-catalyst binding sites, which result in differences in both activity and selectivity. In addition, previous investigations mostly centered on the design and synthesis of heterogeneous POM materials, which can selectively catalyze various thioethers into their corresponding sulfones. Although sulfone compounds are of great importance, high selectivity to obtain sulfoxides in the oxidation reaction is also expected. Current catalytic systems based on POMs displayed low selectivity for the oxidation of sulfides to sulfoxides, such as PW12@Al-MCFen, {As4W40O140(Ru2(Ac)2)} and VV17VIV12(C6H8O4)8.19,29,30 Therefore, it is still highly desirable to exploit alternative catalytic materials based on POMs with high reactivity and selectivity to obtain sulfoxide compounds.
As a special subclass of hybrid POMs that can be rationally designed and synthesized, and increase the stability and activity of the precursors, organic ligand covalently modified POMs have recently attracted great attention in catalysis. Owing to their intrinsic nature and flexible tunability, a few hybrid POMs, such as a hybrid polymer based on polyoxovanadates and 1,3,5-benzenetricarboxamide and mixed carboxylic acid-modified polyoxomolydates, have been evidenced to be privileged pre-catalysts that enable the selective oxidative degradation of CEES.31,32 Inspired by this idea, we predicted that the obtained carboxylic acid-modified POMs with the exposed metal cation linkers Cs4[M(H2O)4][PMo6O21(PABA)3]2·nH2O 1–4 (M = Co, Mn, Ni, Zn) would also be good candidates for catalyzing the selective oxidation of thioethers. As expected, these compounds displayed excellent ability to selectively oxidize various sulfide substrates. Within 20 minutes, various phenyl sulfides could be rapidly converted to the corresponding sulfoxide with high selectivity using these compounds as pre-catalysts. More importantly, the sulfur mustard simulant CEES can also be highly selectively and rapidly degraded to nontoxic CEESO within 12 minutes.
Single crystal X-ray diffraction analyses demonstrated that 1 and 2 were isostructural and crystallized in the same space group P constructed from the same polyoxoanion unit {PMo6O21}. The asymmetric unit of 1 contained one crystallographically independent [PMo6O21(PABA)3]3− anion, half of one Co2+ cation, one Na+ cation and two Cs+ cations (Fig. S1†). The unique Co(1) ion displayed a distorted octahedral geometry, defined by two terminal O atoms derived from two polyoxoanions [Co–O 2.136(4)–2.151(5) Å] and four water molecules [Co–OW 1.962(3)–2.144(4) Å]. First, the {PMo6O21} units were covalently coordinated to the carboxylic acid ligand PABA to produce the single-side modified POMs [PMo6O21(PABA)3]3− (Fig. 1a). Then, the two modified POMs were jointed together to form a dimer via the Co–O–Mo bond. Such dimers were next linked together to generate a 1D supramolecular chain through the hydrogen bonds (N1–H⋯O25 = 2.954 Å in 1) between the N atoms of PABA and the O atoms of the {PMo6O21} units (Fig. 1c). Finally, these 1D chains were further joined together to obtain a 2D supramolecular layer via the hydrogen bonds (N2–H⋯O20 = 2.744 Å, N2–H⋯O13 = 2.938 Å in 1) existing between the N atoms of PABA and the O atoms of the {PMo6O21} units (Fig. 1d).
Bond valence sum (BVS) calculations indicated that the oxidation states of Mo, and P were +6, and +3, respectively, and those of Co/Mn were +2.
Fig. S2† displays the IR spectra of compounds 1–4, indicating the isostructural characteristic of these compounds. The terminal Mo–O bonds and the bridging M–O–Mo units (M = P and Mo) were distributed at 850–950 cm−1 and 580–690 cm−1, respectively. The peaks from 1000 and 1610 cm−1 are ascribed to the carboxylic acid ligands. The absorption maxima at 1605, 1541, and 1415 cm−1 for 1 are assigned to the vas(COO) and vs(COO) stretching frequencies, which indicates the bidentate coordination manner of the carboxylate group of PABA. The diffuse reflectance spectra of 1–4 are shown in Fig. S5† to spot the absorption bands distributed in the ultraviolet and visible regions. Taking compound 1 as an example, the UV region (200–400 nm) exhibits two intense absorption peaks, 234 and 315 nm, which are attributed to the O → Mo charge transfer for the {PMo6O21} polyoxoanions. In addition, one absorption band located at 525 nm in the visible region is assigned to 4T1g → 4T2g and 4A2g → 4T1g for the low-spin Co2+ cations.
Fig. S6† displays the TGA of 1 and 2, which demonstrated multistep weight loss processes. The total weight loss of 34.9% for 1 and 35.2% for 2 are in agreement with the calculated weight losses of 35.4% and 35.6% from 50 to 850 °C. The PXRD patterns of these samples match the calculated peaks, which manifested the phase purities of 1 and 2. In addition, the similar diffraction patterns of 1–4 further indicate the isostructural characteristic (Fig. S7 and S8†).
The selective oxidation of methyl phenyl sulfide was chosen as a model catalytic reaction to evaluate the performance of compounds 1–4. To implement the catalytic oxidation reaction, methyl phenyl sulfide (0.25 mmol), the pre-catalyst (2.5 μmol), H2O2 (oxidant, 0.3 mmol), and naphthalene (internal standard, 0.25 mmol) were performed in ethanol (0.5 ml) at room temperature, and the results of this catalytic system were monitored by gas chromatography (GC). As expected, methyl phenyl sulfide was efficiently catalyzed to the corresponding sulfoxide within 20 min, and compound 1 displayed a relatively high catalytic effect (Fig. 2b). Indeed, compound 1 displayed more notable activity than most reported POM-based materials, and could oxidize 98.7% of methyl phenyl sulfide into sulfoxide with 98% selectivity within 20 min. The performances of these compounds were compared with the heterogeneous catalytic materials based on POMs in recent years, such as the organic–inorganic hybrid POM materials {Cu3(ptz)4(Co2Mo10)}, {[Cu(mIM)4]V2O6}, VV17VIV12(C6H8O4)8, {As4W40O140(Ru2(Ac)2)} and the supported materials PDDA-SiV2W10, P2W15-Al2O3, PW12@Al-MCFen, and SBA-15/K6P2W18O62. As summarized in Table S1,† the compound 1 displayed significantly higher activity than the first three hybrid POM compounds and the first composite material (obtained using TBHP or UHP as the oxidant or long reaction time); this catalytic system also showed a little higher activity than other pre-catalysts that were obtained using a little longer reaction time and displayed relatively low selectivity. Fig. 2c shows the recorded progress of the oxidation of methyl phenyl sulfide in the presence of 1 through GC signals, which illustrated that methyl phenyl sulfide was almost entirely oxidized to the corresponding sulfoxide. To further verify the accuracy of the catalytic process, 1H NMR spectroscopy was employed to monitor the conversion and selectivity (Fig. 3a). The catalytic reaction at 4, 8, 16 and 20 minutes was investigated via1H NMR spectroscopy in CDCl3. In addition, the pure methyl phenyl sulfide, methyl phenyl sulfoxide, and methyl phenyl sulfone were also tested to further illuminate the reaction. The comparison of the 1H NMR spectra of these actual processes and the pure substrate or products demonstrates that methyl phenyl sulfide can be almost converted to methyl phenyl sulfoxide within 20 minutes.
Then, to understand the active center of compound 1 for the selective oxidation of methyl phenyl sulfide, we further investigated the catalytic performance of PABA, CoCl2 salt, and compound 5 (Cs3[PMo6O21(PABA)3]·nH2O). Compound 5 was confirmed by IR spectroscopy and EDS data because of the poor crystal quality (Fig. S3†). Fig. 3b reveals that both PABA and CoCl2 have converted very little methyl phenyl sulfide, and the compound 5 originating from the PABA-modified POMs can catalyze 72% of methyl phenyl sulfide with a selectivity of 89%, which is far behind that in the case of 1. Thus, it can be seen that integrating the Co site and PABA-modified POMs together can improve the catalytic oxidation reaction with rapid and excellent selectivity for the oxidation of methyl phenyl sulfide. Furthermore, a comparison of the catalytic oxidation of compound 1 with those of {CoAsMo6(PABA)3} and {CoTeMo6(PABA)3}, which possessed different central atoms but similar architectures to the compound 1, was conducted to explore how the {ZMo6} unit impacted the catalytic activity. {CoAsMo6(PABA)3} and {CoTeMo6(PABA)3} converted 96.8% and 95.5% of methyl phenyl sulfide with a selectivity of 96.1% and 96.3% under similar conditions, respectively, which indicates that the central heteroatoms played a certain role for the selective oxidation of methyl phenyl sulfide and the corresponding order is PMo > AsMo > TeMo (Fig. 3c). A similar phenomenon has been investigated by the Mizuno and Su's group.36,37
In view of previous research studies and the abovementioned results, we speculate that the peroxo-metal species originating from 1 and H2O2 play a crucial part in the catalytic oxidation process, which have been widely investigated based on POMs.26–28,38 In addition, to exclude the free radical mechanism, radical scavengers (such as diphenylamine and 1,4-benzoquinone) were added into the catalytic reactions, and no change of the conversion was observed. Hence, we proposed a possible mechanism of methyl phenyl sulfide selective oxidation in Fig. 4a. To verify the interaction between 1 and H2O2, UV-Vis spectra of the mixed solution were measured before and after the addition of H2O2 into solution 1. Indeed, two red shifts in the ultraviolet region (276 → 284 nm) and in the visual region (540 → 544 nm), and a new band at 760 nm confirm the interaction between compound 1 and H2O2 (Fig. 4b and c). As shown in Fig. 4a, the active peroxomolybdenum and peroxocobalt species derived from compound 1 and H2O2 worked on the S atom of the sulfides to yield the corresponding products.
Subsequently, we also investigated the recyclability and stability of this pre-catalyst. Just as in Fig. 5a, the conversion of methyl phenyl sulfide remained almost steady in the following 10 cycles, and only a negligible decrease occurred using the recovered pre-catalyst by simple filtration. No catalytic activity was observed when the pre-catalyst was filtered. Furthermore, no distinct change could be detected from the IR spectra and PXRD patterns before and after the catalytic reaction (Fig. 5b and c). From the abovementioned results, we can conclude that these compounds are excellent heterogeneous oxidation pre-catalysts.
Given that compounds 1–4 exhibited excellent catalytic activities in the oxidation of methyl phenyl sulfide, and compound 1 outperformed other compounds with respect to the catalytic activity, various organic sulfides with different electronic and steric effects were continuously investigated using compound 1 (Table 1). The conversion of these substrates did not obviously decrease despite the introduction of the electron donor or acceptor units with less steric hindrance into monophenyl sulfides. We were also delighted to find that even the sulfides with large steric hindrance, such as dibenzyl sulfide and diphenyl sulfide, did not greatly influence the catalytic activities. The above results demonstrate that these compounds possessed superior catalytic capability in the oxidation of sulfides. In addition, to further clarify the function of the central heteroatoms of the {ZMo6} units, the selective oxidation of these organic sulfides were also performed in the presence of {CoAsMo6(PABA)3} and {CoTeMo6(PABA)3} (Tables S2–S7†), which indicated that the heteroatoms indeed made a difference in the catalytic activity. We also investigated the catalytic oxidation reaction kinetics of methyl phenyl sulfide. The time profiles for the selective oxidation reaction process of methyl phenyl sulfide are shown in Fig. 2. A plot of the reaction time against ln(Ct/C0) displays a linear correlation, which shows that the oxidation of methyl phenyl sulfide is in accordance with the first order reaction. The first-order kinetics constants k1–k4 for 1–4 are 0.23842 min−1, 0.1624 min−1, 0.20015 min−1, and 0.17483 min−1, respectively (Fig. S9†).
Entry | Substrates | Products | Con. (%) | Sel.b (%) |
---|---|---|---|---|
a Reaction conditions: Sulfide (0.25 mmol, 1 equiv.), pre-catalyst (1 mol%), H2O2 (1.2 equiv.), ethanol (0.5 mL), for 20 min, at room temperature. b Selectivity to sulfoxides, the byproduct was sulfone. | ||||
1 |
![]() |
![]() |
99.5 | 98 |
2 |
![]() |
![]() |
98.4 | 97.5 |
3 |
![]() |
![]() |
98 | 96.5 |
4 |
![]() |
![]() |
97.8 | 95.7 |
5 |
![]() |
![]() |
98.6 | 96.5 |
6 |
![]() |
![]() |
98.1 | 94.3 |
7 |
![]() |
![]() |
95.1 | 92.9 |
Footnote |
† Electronic supplementary information (ESI) available: ORTEP drawing of 1; IR spectra, UV–Vis diffuse reflective spectra, TG plots, PXRD figures for 1–4; kinetic analysis of methyl phenyl sulfide; comparison of methyl phenyl sulfide oxidation in recent years; oxidation of various sulfides to the corresponding sulfoxides and sulfones; crystal data and structure refinement for 1 and 2; selected distances and angles for 1 and 2. CCDC 1949665 and 1949666. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9qi01098j |
This journal is © the Partner Organisations 2020 |