Insight into Dimethyl Ether Carbonylation Reaction over Mordenite

Mar 1, 2013 - Methanol carbonylation over copper-modified mordenite zeolite: A solid-state NMR study. Lei Zhou , Shenhui Li , Guodong Qi , Yongchao Su...
0 downloads 0 Views 879KB Size
Article pubs.acs.org/JPCC

Insight into Dimethyl Ether Carbonylation Reaction over Mordenite Zeolite from in-Situ Solid-State NMR Spectroscopy Bojie Li, Jun Xu,* Bing Han, Xiumei Wang, Guodong Qi, Zhengfeng Zhang, Chao Wang, and Feng Deng* State Key Laboratory Magnetic Resonance and Atomic and Molecular Physics, Wuhan Center for Magnetic Resonance, Key Laboratory of Magnetic Resonance in Biological System, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan 430071, China S Supporting Information *

ABSTRACT: Carbonylation of dimethyl ether (DME) with CO over Hmordenite (H-MOR) zeolites from 423 to 573 K was studied by in-situ 13 C solid-state NMR spectroscopy. The reaction was monitored separately in the 8-membered ring (MR) and in the 12-MR channels of the zeolite under identical conditions. The experimental results indicated that both 8MR and 12-MR channels were capable of producing methyl acetate product but with remarkably different selectivities. At low-reaction temperature, surface acetyl species (CH3CO−) as stabilized acylium cation was solely identified in the 8-MR channels, which was confirmed by measurement of its characteristic chemical shift anisotropy (CSA) parameters. Additionally, the intermediate role of the acetyl species was evidenced by the fact that it could react with DME to form methyl acetate. This demonstrated the theoretically proposed specificity of the 8-MR channels for generation and stabilization of the acetyl intermediate which was considered as the ratelimiting step of carbonylation reaction. While in the 12-MR channels, the acetyl intermediate was not found during the reaction, and formation of hydrocarbons was favored. The absence of acetyl intermediate might account for the lower reactivity of the 12MR channels to generate methyl acetate product.

1. INTRODUCTION Among various heterogeneous catalysts, zeolites are of great importance because of their distinct physicochemical properties and potential in a broad scope of catalytic reactions, including alkylation, carbonylation, cracking, and so forth. Acidity combining with the microporous framework structure constructed by parallel or intersected channels provides zeolites with unique catalytic reactivity and selectivity.1−7 Molecular level understanding of adsorbates, transition state, and intermediates confined by the specific channels (cavities) of zeolites during catalytic reactions is quite challenging for elucidation of the reaction mechanism and the rational design of zeolite catalysts. Carbonylation of alcohol or ether with CO is an industrial interest reaction to produce carboxylic acids and esters. Rh or Ir organometallic complex catalysts containing halide promoters are commercially utilized for the production of acetic acid from methanol.8,9 However, the use of precious metals and halide components in the carbonylation reaction may conflict with the modern industrial expectations on the cost control and environmental impact. Since the pioneer work of Fujimoto et al., catalytic methanol carbonylation using environmentally benign solid acid catalysts has attracted intensive attention.10−14 Metal-exchanged heteropoly acids (such as H3P12WO4) and various types of zeolites are the most studied catalysts. However, high-reaction temperatures are required, and the selectivity for methyl acetate product from methanol is © 2013 American Chemical Society

significantly hindered by the formation of hydrocarbons as byproducts.12,15 For example, the highest selectivity of methyl acetate over iridium-exchanged 12-tungstophosphoric catalysts is 40% at 498 K.15 Instead of methanol, carbonylation of dimethyl ether (DME) with CO was realized over rhodiumpromoted cesium salts of 12-tungstophosphoric acid with a selectivity to methyl acetate higher than 90%.16 Remarkably, unprecedented selectivity toward methyl acetate (>99%) under relatively low temperatures (423−463 K) was demonstrated in H-MOR zeolite.17−19 With the aim of rational design of the catalysts for practical application, there has been an increasing interest in the carbonylation mechanism. By using kinetic study, Cheung and co-workers proposed that the formation of methoxy species and the subsequent attack by CO to form an acetyl intermediate is the rate-limiting step for the whole reaction on H-MOR zeolites.17,18 The following reaction between the acetyl intermediate and DME gives rise to methyl acetate, and simultaneous regeneration of the methoxy species initially formed from DME on the Brønsted acid site in the induction period facilitates the next cycle of reaction.13,20 On the other hand, the methanol derived from DME along with the formation of methoxy species would Received: January 11, 2013 Revised: February 27, 2013 Published: March 1, 2013 5840

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

Scheme 1. Mechanism of DME Carbonylation with CO in the 8-MR Channels of H-MOR

carbonylation on Cu-MOR zeolites and DME carbonylation on Cs salts of 12-tungstophosphoric acid which paves the way for the mechanism elucidation.11,14 To the best of our knowledge, there is still a lack of unambiguous identification of acetyl intermediates in the carbonylation of DME over unmodified HMOR zeolite. It is interesting to experimentally correlate the acetyl intermediates with the product selectivity in the different channels of H-MOR zeolite, which may show insight into the mechanism of DME carbonylation reaction over H-MOR zeolite. In this contribution, the carbonylation of DME with CO over H-MOR zeolite was studied by using in-situ solid-state NMR spectroscopy. The reactivity of 8-MR and 12-MR channels was separately explored. Surface acetyl species as stabilized acylium cation and acetaldehyde were identified in the 8-MR and 12MR channels, respectively. Their intermediate roles as well as their relationship with the reactivity of carbonylation reaction were discussed.

react with the acetyl intermediate to form methyl acetate and to restore a Brønsted acid site (Scheme 1). Compared with H-MFI (consisting of 10-membered ring (MR)), H-FAU zeolites (consisting of 12-MR channels), and H-FER zeolite (consisting of 8-MR and 10-MR channels), it was found that the 8-MR channels rather than the 12-MR channels of H-MOR zeolite exhibited distinct methyl acetate selectivity.17,18 Because the carbonlyation rates was demonstrated to be correlated with the H+ density in the 8-MR channels but not with that in the 12-MR channels, the more pronounced confinement effect within the 8-MR channels was considered to be responsible for the remarkably different reactivity in H-MOR. According to this mechanism, the smaller 8-MR channels favor the formation and stabilization of a transition state that involves insertion of CO into methoxy species because of the stronger interaction between the transition state and the framework lattice oxygen of the 8MR channels (the interaction is quite weaker in the larger 12MR channels). By theoretical calculations, Boronat and coworkers21,22 further showed that in addition to the confinement effect of the smaller channels, a special acid site plays a key role in determining the reactivity. It was found that the T3-O33 position in the 8-MR channels is the only site favorable for stabilizing the transition state formed by the attack of CO to methoxy species. This is explained by the perfect fit of the transition state in the 8-MR channels because of the special orientation of the methoxy species in the channels. All the literature proposed that the transition state leads to formation of a stable acetyl intermediate and a consequent methyl acetate product. Therefore, the high selectivity of carbonylation of DME to methyl acetate is strictly related to the formation of acetyl intermediate particularly in the 8-MR channels of H-MOR. However, the direct experimental observation of the acetyl intermediate during the carbonylation reaction is quite limited. Indeed, the acetyl species (CH3CO−) rather than the acylium cation (CH3(CO)+) representing a stable surface species has been identified during methanol

2. EXPERIMENTAL SECTION 2.1. Sample Preparation. H-MOR. H-MOR was prepared from Na-MOR (Si/Al = 10.3). Typically, Na-MOR was first converted into NH4-MOR by 4-fold ion exchange with 2.0 M aqueous solution of NH4NO3 at 353 K for 12 h. After each exchange, the sample was washed with excessive deionized water and was isolated by filtration. The sample was finally dried at 353 K overnight, and NH4-MOR was obtained. NH4MOR was then calcinated at 773 K in the tube furnace flowing dry air for 3 h to get the H-MOR catalyst. The sample was dehydrated by the following steps: heating the sample from room temperature to 373 K at a rate of 1 K/min and maintaining that temperature for 3 h and then heating from 373 to 723 K at a rate of 1 K/min and maintaining the final temperature for 12 h under a pressure below 10−3 Pa. H-MOR with Brønsted Acid Sites Solely in 12-MR Channels (Denoted as H-MOR-12MR). Na+ ions could replace the protons in the 8-MR channels without any influence on protons 5841

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

in the 12-MR channels of the H-MOR zeolite. The H-MOR12MR catalyst was prepared according to the previous report.23 Typically, NH4-MOR was partially exchanged with NaNO3 solution (0.014−2.44 M) to achieve a 45% Na+ exchange. Then, the sample was washed with excessive deionized water and was isolated by filtration. After drying at 353 K overnight in air, it was calcined at 773 K for 3 h in the flowing dry air to obtain the desired catalyst. The sample was dehydrated with the same procedure as for H-MOR. H-MOR with Brønsted Acid Sites Solely in 8-MR Channels (Denoted as H-MOR-8MR). Pyridine was used to selectively block the 12-MR channels without influence on the 8-MR channels.24 Typically, on a vacuum line, a known amount of pyridine (pyridine/protons in H-MOR = 1.2:1) was introduced and was frozen by liquid N2 into a predehydrated H-MOR zeolite. The sample was then heated to 473 K at a rate of 5 K/ min and was kept at that temperature for 10 min to remove the excessive pyridine. Then, the sample was cooled to 353 K within 1 h under a pressure below 10−3 Pa. The dehydrated HMOR-8MR was obtained and was sealed under vacuum. 2.2. Solid-State NMR Spectroscopy. All solid-state NMR experiments were carried out at 7.03 T on a Varian Infinityplus300 spectrometer with resonance frequencies of 299.78 and 75.39 MHz for 1H and 13C, respectively. 1H magic-angle spinning (MAS) NMR spectra were recorded using a 4 mm MAS probe and a spinning rate of 14 kHz. 13C MAS NMR spectra were recorded using a 7.5 mm MAS probe and a spinning rate of 4 kHz. The chemical shifts were referenced to tetramethylsilane (TMS) for 1H and to hexamethylbenzene (HMB) for 13C. Repetition times of 5 s for 1H and of 10 s for 13 C were used for single-pulse NMR experiments. For the 1 H→13C CP/MAS NMR experiments, the Hartmann−Hahn condition was achieved using HMB with a connect time of 5 ms and a repetition time of 2 s. To determine the chemical shift anisotropy (CSA) parameters, a slower spinning rate of 2 kHz was used to obtain 13C CP/MAS NMR spectra. The anisotropy of chemical shift (ΔCSA) and the asymmetry factor (η) are defined by ΔCSA = δ33 − δiso and η = (δ22 − δ11)/(δ33 − δiso), respectively. Chemical shift principle components δ11, δ22, and δ33 are defined as |δ33 − δiso| ≥ |δ11 − δiso| ≥ |δ22 − δiso|, and δiso = 1/3(δ11 + δ22 + δ33). The simulations of the NMR spectra were performed using the DMFIT software.25 2.3. In-Situ Solid-State NMR Experiments. The predehydrated H-MOR, H-MOR-12MR, and H-MOR-8MR samples were directly used for the carbonylation reaction. A known amount of DME and 13CO (Al:DME:CO = 1:1:1.2 for H-MOR and Al:DME:CO = 1:0.5:0.6 for H-MOR-12MR and H-MOR-8MR) was introduced into the sample in a glass ampule at room temperature and was frozen onto the sample by liquid N2, and then the ampule was flame-sealed. The carbonylation reaction was performed in the sealed ampule under successively elevated temperatures for a specific period reaction, was quenched by liquid N2, and then the sealed ampule was transferred into a 7.5 mm rotor for NMR measurements. All the NMR experiments were performed at room temperature (298 K).

Figure 1. 1H MAS NMR spectra of (a) H-MOR, (b) H-MOR-12MR, and (c) H-MOR-8MR zeolites. The spinning rate is 14 kHz.

NMR spectra of H-MOR, H-MOR-12MR, and H-MOR-8MR catalysts. Two major signals at 4.2 and 2.4 ppm are observed in the 1H MAS NMR spectra of H-MOR and H-MOR-12MR (Figure 1a and b), which can be assigned to bridging Si(OH)Al groups (Brønsted acid sites) and nonacidic SiOH groups, respectively.27 In comparison with H-MOR, the concentration of Brønsted protons reflected by the signal at 4.2 ppm is significantly reduced in H-MOR-12MR corresponding to ca. 45% of the total protons. This coincides with the concentration of the protons in the 12-MR channels19,23 indicating that the protons in the 8-MR channels have been completely replaced by exchanged Na+ ions. For the H-MOR-8MR (Figure 1c), besides the two intrinsic signals at 4.2 and 2.4 ppm, two new signals at 8.1 and 15.0 ppm appear because of the interaction between the pyridine molecule and the protons of the zeolite. On the basis of the chemical shift, they can be assigned to the pyridine adsorbed on nonacidic SiOH groups (8.1 ppm) via hydrogen bond and protonated pyridine ions adsorbed on Brønsted acid sites (15.0 ppm), respectively.28 By integration of the signals at 15 and 4.3 ppm, it is found that about 45% of the original Brønsted acid sites have interacted with adsorbed pyridine molecules implying that 55% of the Brønsted acid protons in the 8-MR channels remain inaccessible and stay uninfluenced by adsorbed pyridine molecules. Because the acid sites of the 8-MR channels has been completely removed by exchanged Na+ ions, the reaction can only take place in the 12-MR channels of H-MOR-12MR. On the other hand, in H-MOR-8MR, the block of the 12-MR channels by pyridine molecules makes the acidic protons only available in the 8-MR channels for the carbonylation reaction. 3.2. Surface Acetyl Species Formed in H-MOR. A previous study showed that surface acetyl species can be selectively produced by reaction of CH3I and CO on heteropoly acid and its salt.13,14 Following this strategy, we prepared the surface acetyl species on H-MOR zeolites. Figure 2a shows the 13C CP/MAS spectra after coadsorption of 13 CH3I and 13CO on H-MOR catalyst and after heating the sample at 473 K for 1 h. The spectra were recorded at a spinning rate of 4 kHz. In addition to the adsorbed 13CH3I at −20 ppm, the surface acetyl species is unambiguously identified characterized by the signals at 185 ppm and 21 ppm because of the carbonyl and methyl groups, respectively. The formation of

3. RESULTS AND DISCUSSION 3.1. Catalyst Characterization. Brønsted acidity is prerequisite for the carbonylation reaction in the two channels of H-MOR.26 1H MAS NMR can provide direct information about the acidity of zeolites.27 Figure 1 shows the 1H MAS 5842

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

Figure 2. 13C CP/MAS NMR spectra of product formed from (a) reaction of 13CO and 13CH3I on H-MOR at 473 K for 1 h; (b) reaction of 13CO and 13CH3I on H-MOR-12MR at 473 K for 0.5 h; and (c) reaction of 13CO, 13CH3I, and DME on H-MOR at 473 K for 1 h. The spinning rate is 4 kHz. The asterisks denote spinning sidebands.

Figure 3. 13C MAS NMR spectra obtained at a spinning rate of 2 kHz from (a) reaction of 13CH3I and 13CO on H-MOR at 473 K and (b) CH3COOCH3 adsorbed on H-MOR.

The reactivity of surface acetyl species was further explored by coreaction of 13CH3I, 13CO, and DME on H-MOR. As shown in Figure 2c, after heating the sample at 473 K for 1 h, new signals at 177 and 53 ppm appear, and the signal at 185 ppm remarkably increases indicating the occurrence of a carbonylation reaction. The former two signals can be ascribed to the carbonyl and methoxy group of methyl acetate formed in the 12-MR channel. Because no acetyl intermediate species was detectable in the 12-MR channel, it can be deduced that the methyl acetate products were first formed inside the 8-MR channel and then some of them diffused into the 12-MR channel. This experiment definitely demonstrates that the acetyl species is involved in the carbonylation reaction and can be converted into methyl acetate. 3.3. Carbonylation of DME over H-MOR, H-MOR12MR, and H-MOR-8MR Zeolites. In-situ solid-state 13C CP/MAS NMR spectra of product formed from DME and 13 CO on H-MOR at 423−573 K are shown in Figure 4. In addition to unreacted DME (60 ppm), a weak signal at 50 ppm is observed at 423 K (Figure 4a). The appearance of the signal at 50 ppm indicates the formation of methanol because of decomposition of DME on Brønsted acid sites.18,29 Methoxy species could be simultaneously formed having a typical chemical shift of ca. 60 ppm,20 which may be overlapped with the signal of DME at 60 ppm and thus may be undistinguishable. Indeed, this situation usually occurs for the in-situ NMR observation of methoxy species derived from methanol, which could be identified by stopped-flow MAS NMR technique.29−31 The low-field weak signals (at 175−200 ppm) fall in the chemical shift range of carbonyl groups.32 By heating the sample at 453 K for 0.5 h, three signals at 177, 185, and 197 ppm become pronounced along with the decrease of DME (60 ppm) and its derivatives, such as methanol and methoxy species (Figure 4b). The relatively weak signal at 197 ppm can be assigned to the carbonyl group of aldehyde such as acetaldehyde, which has a chemical shift of ca. 200 ppm in solution and zeolite.32,33 The signals at 177 and 185 ppm are due to methyl acetate formed in the 12-MR and 8-MR channels, respectively. The existence of acetyl species in the 8MR channel and its contribution to the signal at 185 ppm can

methoxy species from 13CH3I is also evident at 58 ppm, which can act as an intermediate for the attack of CO to generate surface acetyl species. Since no trapping agent such as water or DME exists in the system, the formed surface acetyl species is present as the final product without further conversion to other species such as methyl acetate. Reaction of 13CH3I and 13CO was also performed at 473 K for 1 h over H-MOR-12MR in which Brønsted acid sites are only available in the 12-MR channels. As can be seen from Figure 2b, no acetyl species (185 ppm and 21 ppm) is formed except for methoxy species at 58 ppm and 13CH3I at −20 ppm. This unambiguously indicates that the acetyl species can be solely formed in the 8-MR channels of H-MOR zeolite. The fact that the acetyl species was not found in the 12-MR channels but only in the 8-MR channels agrees well with the previous assumption of the specificity of 8-MR channels for the acetyl intermediates.21,26 As a surface-bounded species, the mobility of acetyl species is strictly limited, which gives rise to a large chemical shift anisotropy (CSA) characterized by the strong spinning sidebands in the spectrum (Figure 2a). For a more detailed analysis, the 13C CP/MAS spectrum was also recorded at a low MAS spinning rate of 2 kHz (Figure 3a) to determine the CSA parameters of acetyl species, which are listed in Table 1 (sample 1). For comparison, we also measured the CSA parameters of methyl acetate (Figure 3b and Figure S1 of the Supporting Information) absorbed on H-MOR. Two signals at 185 and 177 ppm because of the carbonyl carbons were observed when methyl acetate was adsorbed on H-MOR. As will be seen in the following, methyl acetate formed in the H-MOR-8MR solely gives rise to the 13C NMR signal at 185 ppm. Thus, we assign the two signals at 185 and 177 ppm to methyl acetate adsorbed in the 8-MR and 12-MR channels, respectively. The CSA parameters of methyl acetate adsorbed in the 8-MR channels are also listed in Table 1 (sample 2). It can be seen from Table 1 that although the methyl acetate in the 8-MR channel has a similar isotropic chemical shift (δiso, 184.2 ppm) to surface acetyl species (185.3 ppm) formed in the same channel, their CSA parameters, such as ΔCSA and η, are quite different. 5843

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

Table 1. CSA Parameters for the 185 ppm Signal from Different Species sample

surface species

δiso (±0.1)

δ11 (±0.1)

δ22 (±0.1)

δ33 (±0.1)

ΔCSA (±0.1)

η (±0.01)

1 2 3 4

acetyl speciesa adsorbed CH3OOCH3b acetyl species and CH3COOCH3c acetyl speciesd

185.3 184.2 184.6 185.3

251.8 224.3 233.4 251.9

185.5 198.9 195.6 185.5

118.5 129.4 124.9 118.5

−66.8 −54.8 −59.7 −66.8

0.99 0.46 0.63 0.99

a

Reaction product of CO and 13CH3I on H-MOR at 473 K for 1 h. bCH3COOCH3 adsorbed on H-MOR. cReaction product of 13CO and DME on H-MOR at 453 K for 0.5 h. dAcetyl species remaining after vacuum treatment of sample 3 to remove CH3COOCH3.

Figure 5. 13C MAS NMR spectra obtained at a spinning rate of 2 kHz from reaction of 13CO and DME on H-MOR at 453 K. The signal at 101 ppm on spectrum b is due to the spinning sideband of DME at 60 ppm (not shown).

product, vacuum treatment was used to isolate the surfacebounded acetyl species. After reaction of 13CO and DME at 453 K, 20 h vacuum treatment (10−2 Pa) was performed on the catalyst at room temperature (Figure 6a). In comparison with

Figure 4. 13C CP/MAS NMR spectra of product formed from reaction of DME and 13CO on H-MOR at (a) 423 K, (b) 453 K, (c) 493 K, and (d) 523 K for 0.5 h. The spinning rate is 4 kHz. The asterisks denote spinning sidebands.

be confirmed by CSA analysis (see the following). According to a previous report, the methyl group of methyl acetate or acetyl species is originated from DME.17 Here, 13C unlabeled DME makes its signal (ca. 20 ppm) invisible. Further increasing the temperature to 493 K results in a complete consumption of DME reactant and in an increase of product signals at 185 and 177 ppm (Figure 4c). The acetaldehyde signal at 197 ppm disappears at 493 K opposite to the signals of methyl acetate product. This suggests that the acetaldehyde might act as an intermediate in the carbonlyation reaction. In agreement with previous works,19 side reactions occur at high temperature (523 K) reflected by the formation of hydrocarbons which is evidenced by the broad signals in the chemical shift range of 120−150 ppm because of aromatics and the weak signals at 20−40 ppm because of alkanes (Figure 4d). The signal at 185 ppm corresponding to carbonyl groups formed at 453 K (Figure 4b) is different from that of surface acetyl species reflected by its much weaker spinning sidebands. To gain insight into the species, its CSA was measured at a MAS spinning rate of 2 kHz (Figure 5). As shown in Table 1 (sample 3), the CSA parameters (principle components, ΔCSA and η) of the 185 ppm signal are between those of surface acetyl species and adsorbed methyl acetate. Therefore, we conclude that both acetyl species and methyl acetate were formed at 453 K and contributed to the signal at 185 ppm. As a result, averaged CSA parameters were obtained. Moreover, we tried to isolate the acetyl species formed in the reaction of 13CO with DME on H-MOR. As mentioned above, because the acetyl species could be masked by the coexisting methyl acetate

Figure 6. 13C CP/MAS NMR spectra of product formed from reaction of DME and 13CO on H-MOR: (a) at 453 K for 0.5 h and then evacuating for 20 h at room temperature; (b) heating sample a at 453 K for 0.5 h. The spinning rate is 4 kHz. The asterisks denote spinning sidebands.

the original spectrum (Figure 4b), the signal at 177 ppm because of methyl acetate in the 12-MR channel significantly decreased, and the signals at 197 and 185 ppm also decreased to some extent. This indicates that the methyl acetate product in the 12-MR channel was almost removed. We analyzed the CSA of the 185 ppm signal at a spinning rate of 2 kHz (Figure 7). Its CSA parameter was identical to that of acetyl species (sample 4, Table 1). This indicates that methyl acetate was completely removed and that only acetyl species remained in the 8-MR channel. The reactivity of surface acetyl species was further examined. Heating the vacuum-treated sample at 453 K for 0.5 h led to a remarkable increase of the 177 ppm signal, a slight decrease of the 185 ppm signal (Figure 6b), and simultaneous decrease of signals of DME and methanol (60 and 50 ppm). The further consumption of the reactant and the 5844

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

initially formed inside the 12-MR channel could diffuse into the 8-MR channel giving rise to the 185 ppm signal. Additionally, the low reactivity toward carbonylation of DME is also reflected by the rapid deactivation of the catalyst at high temperatures (493 and 523 K, Figure 8c and d). A large quantity of hydrocarbons at 120−150 ppm because of aromatics and at 10−20 ppm because of alkanes dominate the spectrum at 523 K (Figure 8d). Similar to H-MOR, acetaldehyde at 197 ppm could be observed at 453 and 473 K in H-MOR-12MR, which might play an intermediate role for formation of methyl acetate. Our experimental results indicate that the carbonylation reaction can proceed in the 12-MR channels without the contribution from the 8-MR channels. In addition, the results also support the theoretical prediction that formation of methyl acetate is kinetically impeded in the 12-MR channels, but formation of hydrocarbons is favored.21,22 Figure 9 shows 13C CP/MAS NMR spectra of reactions of DME with 13CO on H-MOR-8MR zeolite at 423−523 K for

Figure 7. 13C MAS NMR spectra obtained at a spinning rate of 2 kHz after 20 h vacuum treatment of sample from reaction of 13CO and DME on H-MOR at 453 K.

formation of methyl acetate product demonstrated the restart of the carbonylation reaction. Because initially introduced 13CO reactant had been completely removed by the vacuum treatment (as evidenced by 13C MAS NMR), the newly formed methyl acetate was definitely due to the reaction of acetyl species with the remaining DME and its derivatives. Moreover, the significant reduction of the MAS spinning sidebands of the 185 ppm signal (Figure 6a and b) points to the decrease and conversion of acetyl species, which evidence its intermediate role in the carbonylation reaction. Figure 8 shows the in-situ 13C CP/MAS NMR spectra of product formed from DME and 13CO on H-MOR-12MR at

Figure 9. 13C CP/MAS NMR spectra of product formed from reaction of DME and 13CO on H-MOR-8MR at: (a) 423K; (b) 453K; (c) 493K; (d) 523K for 0.5 h. The spinning rate is 4 kHz. The asterisks denote spinning sidebands.

0.5 h. At 423 K, only signals from DME at 60 ppm and pyridine at 125−150 ppm were observable (Figure 9a). The signals of pyridine molecules adsorbed in the 12-MR channels show no obvious change throughout the reaction indicating that the pyridine molecules were not involved in the carbonylation reaction (Figure S2 of the Supporting Information). At 453 K, carbonylation reaction occurs in the 8-MR channel evidenced by appearance of the 185 ppm signal and consumption of the 60 ppm signal (Figure 9b). The much weaker spinning sideband of the 185 ppm signal compared to that of acetyl species (Figure 2a) suggests that apart from the acetyl species, methyl acetate was formed in the 8-MR channels. Further increasing the reaction temperature to 493 and 523 K (Figure 9c and d) results in a gradual increase of the 185 ppm signal (its MAS spinning sidebands remains almost unchanged) indicating that more and more methyl acetate products were generated. Because the 12-MR channels were blocked by adsorbed pyridine molecules, the methyl acetate initially formed in the 8-MR channel was unable to diffuse into the 12-MR channel, and no signal at 177 ppm because of methyl acetate in this channel was visible. Interestingly, acetaldehyde (at 197 ppm)

Figure 8. 13C CP/MAS NMR spectra of product formed from reaction of DME and 13CO on H-MOR-12MR at (a) 423 K, (b) 453 K, (c) 493 K, and (d) 523 K for 0.5 h. The spinning rate is 4 kHz. The asterisks denote spinning sidebands.

423−523K for 0.5 h. In comparison with H-MOR, the HMOR-12MR catalyst exhibits much lower activity for carbonylation reaction evidenced by the lower concentration of carbonyl groups (185 and 177 ppm) at 423 and 453 K (Figure 8a and b). Because acetyl species cannot be formed in the 12MR channels, the signals at 185 and 177 ppm should be resulted from the carbonyl groups of methyl acetate in the 8MR and 12-MR channels, respectively. For H-MOR-12MR, the absence of Brønsted acid sites in the 8-MR channel excludes the carbonylation of DME in this channel. The methyl acetate 5845

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

acetate is highly selective. At low temperature (453 K), surface acetyl species due to attack of CO on methoxy species derived from DME was identified in the 8-MR channels by its characteristic CSA parameters. The high reactivity of this species with DME to form methyl acetate demonstrated its intermediate role in the carbonylation of DME. Because the formation of acetyl species has been considered as the ratelimiting step, the result provides direct experimental evidence of the specificity of the 8-MR channels for the high carbonylation rate and selectivity to methyl acetate. Although methyl acetate could be produced in the 12-MR channels as well, the 12-MR channels notably favored the formation of hydrocarbons. The absence of acetyl species in the 12-MR channels may account for the low reactivity of the channels for formation of methyl acetate. When the two channels of MOR zeolite simultaneously function in the carbonylation reaction, the process for the formation of hydrocarbons in the 12-MR channels is significantly suppressed by the acetyl-mediated carbonylation process that occurs preferentially in the 8-MR channels.

which could be formed inside the 12-MR channel was absent in the 8-MR channels. At 523 K, a small amount of formate species at ca. 170 ppm was observed as well. In comparison with the H-MOR-12MR catalyst, hydrocarbons (characterized by signals at 10−20 ppm because of alkanes and signals at 120− 150 ppm because of aromatics) formed in the 8-MR channels of H-MOR-8MR catalyst are negligible throughout the reaction. Although the signals of aromatics are overlapped with those of adsorbed pyridine molecules, their concentration seems to be extremely low as the signals of pyridine molecules remain almost unchanged throughout the reaction. Therefore, our 13C NMR experimental data demonstrate that methyl acetate could be formed in both 8-MR and 12-MR channels, which have chemical shifts of 185 and 177 ppm, respectively, whereas surface acetyl species could be solely generated in the 8-MR channels, which mediates the carbonylation reaction. 3.4. Mechanism of Carbonylation Reaction. The existence of surface acetyl species and its intermediate role in the carbonylaton of DME unambiguously confirm the proposed mechanism for the formation of methyl acetate in the 8-MR channels of H-MOR zeolite (Scheme 1). Because the formation of acetyl intermediates has been considered as the rate-limiting step of the carbonylation reaction,17 the observation of acetyl intermediates in the 8-MR channels of H-MOR zeolite under low temperatures can be rationally related to its distinct reactivity and selectivity to methyl acetate product. On the other hand, previous works claimed that the 12-MR channel has low reactivity for the carbonylation of DME, which is confirmed by our work. Although methyl acetate with low concentration could be generated in the 12-MR channel at low temperature, the formation of hydrocarbons was favored at high reaction temperatures. Because the acetyl intermediates could not be formed in the 12-MR channel during the reaction, the formation of methyl acetate in this channel suggests that the carbonylation reaction might proceed through a pathway different from that in the 8-MR channel. Because acetaldehyde was solely observed in the 12-MR channel, we propose that it might mediate the formation of methyl acetate. However, further work is still needed to clarify its detailed role in the catalytic reaction. During the carbonylation process, the initial step is similar for the two channels of H-MOR zeolite, which is characterized by the formation of methoxy species via DME on Brønsted acid sites. As demonstrated by theoretical calculations,21,22 the formation of methoxy species has no remarkable difference in the two types of channels. However, the facile interaction of methoxy species with DME and methanol to form hydrocarbons is strongly favored in the 12-MR channels. Thus, starting from the methoxy intermediates, the formation of acetaldehyde intermediate and consequent methyl acetate in the 12-MR channels is a less competitive reaction pathway. In contrast, in the 8-MR channels, acetyl species intermediate is preferentially generated by attack of CO on methoxy species, which mediates the carbonylation of DME to form methyl acetate.



ASSOCIATED CONTENT

S Supporting Information *

Additional information about CH3COOCH3 and pyridine adsorbed on H-MOR (Figure S1 and Figure S2). This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Telephone: +86-27-87198820. Fax: +86-27-87199291. E-mail: [email protected] (F. D); [email protected] (J. X). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank Natural Science Foundation of China (20933009, 21210005, 21221064, and 21173254) and Wuhan Science and Technology Bureau (Chen Guang project for young scientists) for financial support (201271031383).



REFERENCES

(1) Petrovic, I.; Navrotsky, A.; Davis, M. E.; Zones, S. I. Chem. Mater. 1993, 5, 1805. (2) Smit, B.; Maesen, T. L. M. Nature 1995, 374, 42. (3) Vanwell, W. J. M.; Wolthuizen, J. P.; Smit, B.; Vanhooff, J. H. C.; Vansanten, R. A. Angew. Chem., Int. Ed. 1995, 34, 2543. (4) Schenk, M.; Smit, B.; Vlugt, T. J. H.; Maesen, T. L. M. Angew. Chem., Int. Ed. 2001, 40, 736. (5) Schenk, M.; Calero, S.; Maesen, T. L. M.; van Benthem, L. L.; Verbeek, M. G.; Smit, B. Angew. Chem., Int. Ed. 2002, 41, 2500. (6) Anderson, M. W.; Klinowski, J. J. Am. Chem. Soc. 1990, 112, 10. (7) Maesen, T. L. M.; Schenk, M.; Vlugt, T. J. H.; de Jonge, J. P.; Smit, B. J. Catal. 1999, 188, 403. (8) Paulik, F. E.; Roth, J. F. Chem. Commun 1968, 1578. (9) Sunley, G. J.; Watson, D. J. Catal. Today 2000, 58, 293. (10) Fujimoto, K.; Shikada, T.; Omata, K.; Tominaga, H. Chem. Lett. 1984, 2047. (11) Blasco, T.; Boronat, M.; Concepcion, P.; Corma, A.; Law, D.; Vidal-Moya, J. A. Angew. Chem., Int. Ed. 2007, 46, 3938. (12) Ellis, B.; Howard, M. J.; Joyner, R. W.; Reddy, K. N.; Padley, M. B.; Smith, W. J. Stud. Surf. Sci. Catal. 1996, 101, 771. (13) Luzgin, M. V.; Kazantsev, M. S.; Wang, W.; Stepanov, A. G. J. Phys. Chem. C 2009, 113, 19639.

4. CONCLUSION The carbonylation of DME with CO in the 8-MR and 12-MR channels of H-MOR zeolite was separately investigated by insitu solid-state NMR spectroscopy. Different reaction pathways were identified in the different channels for production of methyl acetate. In the 8-MR channels, the formation of methyl 5846

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847

The Journal of Physical Chemistry C

Article

(14) Luzgin, M. V.; Kazantsev, M. S.; Volkova, G. G.; Wang, W.; Stepanov, A. G. J. Catal. 2011, 277, 72. (15) Wegman, R. W. J. Chem. Soc., Chem. Commun. 1994, 947. (16) Volkova, G. G.; Plyasova, L. M.; Salanov, A. N.; Kustova, G. N.; Yurieva, T. M.; Likholobov, V. A. Catal. Lett. 2002, 80, 175. (17) Cheung, P.; Bhan, A.; Sunley, G. J.; Law, D. J.; Iglesia, E. J. Catal. 2007, 245, 110. (18) Cheung, P.; Bhan, A.; Sunley, G. J.; Iglesia, E. Angew. Chem., Int. Ed. 2006, 45, 1617. (19) Bhan, A.; Iglesia, E. Acc. Chem. Res. 2008, 41, 559. (20) Jiang, Y. J.; Hunger, M.; Wang, W. J. Am. Chem. Soc. 2006, 128, 11679. (21) Boronat, M.; Martinez-Sanchez, C.; Law, D.; Corma, A. J. Am. Chem. Soc. 2008, 130, 16316. (22) Boronat, M.; Martinez, C.; Corma, A. Phys. Chem. Chem. Phys. 2011, 13, 2603. (23) Veefkind, V. A.; Smidt, M. L.; Lercher, J. A. Appl. Catal., A: General 2000, 194, 319. (24) Maache, M.; Janin, A.; Lavalley, J. C.; Benazzi, E. Zeolites 1995, 15, 507. (25) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calve, S.; Alonso, B.; Durand, J. O.; Bujoli, B.; Gan, Z. H.; Hoatson, G. Magn. Reson. Chem. 2002, 40, 70. (26) Bhan, A.; Allian, A. D.; Sunley, G. J.; Law, D. J.; Iglesia, E. J. Am. Chem. Soc. 2007, 129, 4919. (27) Freude, D.; Hunger, M.; Pfeifer, H.; Schwieger, W. Chem. Phys. Lett. 1986, 128, 62. (28) Hunger, M. Solid State Nucl. Magn. Reson. 1996, 6, 1. (29) Wang, W.; Seiler, M.; Hunger, M. J. Phys. Chem. B 2001, 105, 12553. (30) Michael, H. Prog. Nucl. Magn. Reson. Spectrosc. 2008, 53, 105. (31) Wang, W.; Hunger, M. Acc. Chem. Res. 2008, 41, 895. (32) Breitmaier, E.; Voelter, W. 13C NMR spectroscopy: methods and applications in organic chemistry; Verlag Chemie: Weinheim, Germany, 1978. (33) Lee, S. O.; Kitchin, S. J.; Harris, K. D. M.; Sankar, G.; Dugal, M.; Thomas, J. M. J. Phys. Chem. B 2002, 106, 1322.

5847

dx.doi.org/10.1021/jp400331m | J. Phys. Chem. C 2013, 117, 5840−5847