Xia Changjiu; Zhao Yi; Zhu Bin; Lin Min; Peng Xinxin; Dai Zhenyu; Shu Xingtian
(State Key Laboratory of Catalytic Materials and Reaction Engineering, Research Institute of Petroleum
Processing, SINOPEC, Beijing 100083)
Abstract: Two clean liquid–phase cyclohexanone oxidation routes catalyzed by DHBEA and HTS zeolites, in the absence of organic solvents, have been developed for producing high value-added chemical intermediates. Under optimized conditions, the cyclohexanone conversion reaches up to 60%, and the selectivity of total target products (ε-caprolactone, 6-hydroxyhexanoic acid and adipic acid) is over 90% achieved by the HTS zeolite; while both cyclohexanone conversion and the 6-hydroxyhexanoic acid selectivity are over 95% obtained on the DHBEA zeolite. Both the Lewis and Br?nsted acid sites of DHBEA zeolite can preferentially activate the carbonyl group of cyclohexanone without any impact on H2O2 molecules. Meanwhile, the HTS zeolite can predominantly make H2O2 more reactive, which agrees well with the molecular calculation results. Hence, two different Baeyer-Villiger oxidation mechanisms based on the activation of H2O2 and cyclohexanone are proposed. Then, 6-hydroxyhexanoic acid is formed via the ring-opening of ε-caprolactone. However, C-OH groups cannot be reactivated by DHBEA zeolite, leading to insignificant adipic acid formation, while the selectivity of adipic acid is 28.5% obtained on the HTS zeolite. Consequently, the higher catalytic performance of the DHBEA zeolite is ascribed to its larger amount of active sites and greater diffusion features than those of HTS zeolite.
Key words: hollow titanium silicate; cyclohexanone oxidation; zeolite; Baeyer-Villiger; molecular calculation
Catalytic transformation of cyclohexanone to high value–added chemical intermediates, such as adipic acid, cyclohexanone oxime, and ε-caprolactone, is ultraimportant in the commercial production of plastics and fabrics[1]. In conventional oxidation routes, highly toxic and strong acidic solutions, e.g. concentrated HNO3or peracids, are extensively used as oxidants and catalysts for synthesizing adipic acid and ε-caprolactone, respectively[2]. Since these reagents are corrosive for reactors and harmful to the human being and the environment, many efforts have been devoted to finding some green and sustainable routes. To achieve this goal, the low concentration (30%—60%) H2O2solution has been selected as an environmentally-friendly oxidant to substitute for the stoichiometric peroxoacids. Unfortunately, the aqueous H2O2solution of low concentration has weaker oxidative reactivity than acidic peracids or HNO3, therefore a series of catalysts have been developed to enhance the nucleophilic attack capability of H2O2molecules. Particularly, heterogeneous catalysts can have much better advantages than homogeneous ones, due to the convenience for separating these catalysts from products and high stability for regeneration[3]. For example, the Sn-β zeolite shows great selectivity in the Baeyer-Villiger (BV) oxidation of cyclohexanone, as reported by A. Corma and coworkers[4]. It benef its from the tetrahedral framework of Sn atoms, which can accept electron pairs of carbonyl group of cyclohexanone, making it easily be attracted by nucleophilic reagents. Meanwhile, the Sn-β zeolite has the three-dimensional 12-membered channel networks with a cross section of 0.76 nm×0.66 nm along the (100) direction, which possesses a great mass diffusion performance. For the production of adipic acid, the solid acidic H2WO4and Na2WO4materials show great catalytic performance in the cyclohexanone oxidation reaction under mild conditions. It is attributed to the formation of highly active H2[WO(O2)2(OH)2] species, which are produced from the oxidation of H2WO4by H2O2solution[5-8]. Above all, these heterogeneous catalysts usually have poor stability or are difficult to be synthesized in industry. Thus, it is ultra-necessary to develop new efficient and stable heterogeneous catalysts for commercial cyclohexanone oxidation processes.
Herein, the commercially prepared nano-sized DHBEA and hollow TS-1 (HTS) zeolites are introduced as catalysts to promote the cyclohexanone oxidation reactions with H2O2solution serving as the oxidant. Nano-sized HBEA zeolite, which is synthesized via the surface-wettability method, exhibits good catalytic activity and stability in the catalytic cracking (FCC), alkylation, and isomerization reactions[9-11]. In order to increase the amount of Lewis acid sites, this material is subjected to dealumination with an aqueous 68.6% H2SO4solution. Hollow titanium silicate (HTS) zeolite with abundant intracrystalline mesopores is prepared by the crystalline rearrangement method. Furthermore, the HTS zeolite has been applied commercially in many catalytic oxidation processes, such as phenol hydroxylation, cyclohexanone ammoximation, and propylene epoxidation reactions[12-15]. It is assigned to the activation of H2O2molecules catalyzed by the tetrahedral framework of Ti species, which can accept the lone electron pairs of H2O2molecules. According to their catalytic results, the corresponding mechanisms are proposed and identified by the DFT methods, which can provide important guidance for cyclohexanone transformation. As a result, this study can provide rational understanding on the green and efficient cyclohexanone oxidation routes for producing the value-added chemical intermediates, as evidenced by both the experimental and theoretical methods.
Highly crystalline nano-sized HBEA zeolite (with a Si/Al molar ratio of 22.24) was supplied by the Research Institute of Petroleum Processing (RIPP), SINOPEC[16]. This zeolite was suspended in an aqueous 68.6% H2SO4solution at 80 °C for 6 hours in air under continuous stirring. The product was washed and f iltered for several times, and then it was dried at 100 °C for 12 hours and calcined at 550 °C for 6 hours. The final product was labeled as DHBEA zeolite, with a Si/Al molar ratio of 69.12 and a well crystalline BEA topological structure. The DHBEA zeolite was analyzed by multiple methods, including XRD, BET, SEM, TEM, NH3-TPD, pyridine adsorbed IR, and27Al MAS NMR. The characterization results are illustrated in the supporting information.
The HTS zeolite is made by the post-synthesis of conventional TS-1 zeolite under hydrothermal conditions, as reported in the previous work[17]. Firstly, a mixture of tetraethyl orthosilicate (TEOS) and tetrabutyl orthotitanate (TBOT) was dissolved in an aqueous TPAOH solution, with a SiO2: TiO2: TPAOH: H2O molar ratio of 1: 0.03: 0.25: 40. This mixture was stirred and heated at 80 °C for 6 hours to remove the alcohol. Then, this sol was put into a PTFE-lined autoclave, and then heated at 170 °C for 3 days under static condition to obtain the conventional TS-1 zeolite. In the second step, the calcined templatefree TS-1 zeolite was put in the aqueous TPAOH solution, with a TS-1 (counted as SiO2): TPAOH: H2O molar ratio of 1:0.1:10. The post-synthesis was carried out at 170 °C for 3 days under static condition, and then the hollow TS-1 zeolite (labeled as HTS) was obtained. Finally, the HTS zeolite was characterized by XRD, BET, Ti2/3p XPS, FTIR and UV-Vis spectrometric analyses, as illustrated in the supporting information. Table 1 shows the XRD, BET and SEM analysis results of the DHBEA and HTS zeolites.
Catalytic testing was operated in the three-necked glass reactors under continuous stirring at atmosphere pressure. Firstly, 0.63 g of catalyst and 9.8 g of cyclohexanone were put into glass reactors, and heated up to 80 °C for 5 minutes. Then, 11.3 g of 30% H2O2solution was injected to the mixture, and the product was collected from the mixture at regular intervals. The products were identified by GC-MS spectroscopy using an Agilent 5977A series GC/MSD system. And the samples were quantified by an Agilent 6890 gas chromatograph, equipped with a 3-m-long HP-5 column and a hydrogen f lame ionization detector.
In order to understand the adsorption and activation behavior of cyclohexanone and H2O2molecules on the active sites of DHBEA and HTS zeolites, the trigonal Al(OSiH3)3species and tetrahedral Ti(OSiH3)4clusters have been proposed[18-20]. These two clusters are cut from the unit cell of DHBEA and HTS zeolites, which are optimized by using the DMol3and PW91/DND techniques. The adsorption and activation of cyclohexanone and H2O2, and furthermore the transition states on the active Ti and Al sites are calculated by the Adsorption Locator module in MS software. In addition, the Ti(OSiH3)3OH cluster is considered as the active site for the BV catalytic oxidation on the TS-1 zeolite, and its geometry optimization of reaction pathway is carried out by the B3PW91 method.
To reflect the catalytic performance difference between the DHBEA and HTS zeolites, they are characterized by multiple methods, as illustrated in Table 1. Nitrogen adsorption-desorption (BET) results show that DHBEA zeolite has larger surface area, pore volume, and its mesopore volume, which can be attributed to the intercrystalline voids in different small zeolite crystals. And it is observed that the particle size of DHBEA zeolite is smaller than that of HTS zeolite, as shown in Figure 1. Moreover, the pore mouth of HBEA zeolite (0.76 nm×0.64nm) is much wider than that of HTS zeolite (0.55 nm×0.51nm)[21-22]. Thus, it is inferred that the DHBEA zeolite shows high mass diffusion property of reagent and product molecules inside its 12-membered ring channel.
Table 1 XRD and BET analysis of DHBEA and HTS zeolites
Figure 1 SEM and TEM (insert) images of DHBEA and HTS zeolites
To quantify the amount of the Br?nsted and Lewis acid sites of two zeolites, the pyridine adsorbed IR characterization was carried out at 250 °C and 350 °C, respectively, as given in Table 2. It is evident that the HTS zeolite only has very few Lewis acid sites, while the DHBEA zeolite has large amount of both Lewis and Br?nsted acid sites simultaneously. The former ones are attributed to the tetrahedral Ti species, which can accept the lone electron pairs by using the3dempty orbital of framework Ti atoms[23-26]. And the Br?nsted acid sites are attributed to ≡Si (OH)+Al-≡ species, due to the substitution of Si atom by one Al atom, which needs one proton ion (H+) to keep the charge balance[27-29]. And the Lewis acid sites are associated with the three-coordinated Al species in both the framework and extraframework positions, which originate from the cleavage of Si-O-Al bonds[30-33].
Figure 2 shows the catalytic performance of DHBEA and HTS zeolites in cyclohexanone oxidation reaction under the same conditions, with 30% H2O2solution used as the oxidant. As shown in Figure 2 (a), the maximum cyclohexanone conversion in the DHBEA zeolite catalyzed system (over 90%) is much higher than that in the HTS one (about 60%), which means that the DHBEA zeolite has higher capability for activation of cyclohexanone or H2O2molecules during the catalytic oxidation process. Generally, the activation of substrates is dependent on the amount of acid sites, and hence the catalytic performance is in good agreement with the results of pyridine adsorbed IR analysis. Moreover, the maximum cyclohexanone conversion is obtained within 0.5 h over the DHBEA zeolite, while that obtained over HTS zeolite is achieved in 2 h, respectively. It is indicated that the DHBEA zeolite has greater mass diffusion performance than HTS zeolite, which is closely related to their crystal size and microporous channels that can match well with the results shown in Figure 1 and Table 1.As illustrated in Figure 2 (b), the total target product selectivity catalyzed by the DHBEA zeolite is much higher than that catalyzed by the HTS zeolite. It is inferred that many side-products are generated in the HTS zeolite catalyzed system, which is caused by the highly reactive TiOOH species (as demonstrated by many experimental and simulation researches in literature reports)[34-36]. To verify this conclusion, Figure 3 shows the distribution of target product selectivity as a function of reaction time in both DHBEA and HTS catalyzed systems. We can observe that a small amount of ε-caprolactone is produced at the initial stage in both systems, and it decreases gradually along with the increase in reaction time. At the same time, the selectivity of 6-hydroxyhexanoic acid, formed via the hydrolysis of ε-caprolactone, is increasing until it reaches a maximum conversion in 2 h. However, there is almost no adipic acid generated in the DHBEA catalyzed system, while the selectivity of adipic acid is over 20% in the HTS catalyzed reaction. Scheme 1 gives the reaction pathways of cyclohexanone oxidation catalyzed by the DHBEA and HTS zeolites, which reveals that the DHBEA zeolite does not cause activation on 6-hydroxyhexanoic acid and H2O2molecules without forming thereby the deep-oxidation product (adipic acid). As a result, different product distributions are observed for two zeolites, which are associated with the different mechanisms caused by various active sites. And the detailed mechanisms will be demonstrated and discussed in the next section.
Table 2 Amount of Lewis and Br?nsted acid sites of DHBEA and HTS zeolites detected by pyridine adsorbed IR method
Figure 2 Catalytic activity (a) and target product selectivity (b) of cyclohexanone oxidation catalyzed by HTS and DHBEA zeolites under the same reaction conditions
From the viewpoint of application, high selectivity of single product is preferential for its separation and purification to meet the demand for low energy consumption and capital cost. The selectivity of 6-hydroxyhexanoic acid is more than 95% on the DHBEA zeolite, while the cyclohexanone conversion is over 94%. Thus, the DHBEA zeolite serves as a better catalyst candidate for cyclohexanone oxidation than the HTS zeolite under the same conditions, showing great industrial potential in a large scale of production.
Figure 3 Target product selectivity distribution of cyclohexanone oxidation catalyzed by DHBEA zeolite and HTS zeolite under the same reaction conditions
Scheme 1 Reaction pathways of cyclohexanone oxidation catalyzed by DHBEA and HTS zeolites with 30% aqueous H2O2 serving as oxidant
To optimize the cyclohexanone oxidation catalyzed by the DHBEA and HTS zeolites, different reaction parameters are investigated, as shown in Table 3. It is worthy of mentioning that large amount of organic solvents (MTBE and dioxane) is employed in the Baeyer-Villiger oxidation of cyclohexanone, as reported by A. Corma and his coworkers, to keep the product ε-caprolactone from hydrolyzed ring-opening reaction[37-38]. Therefore, organic solvents, such as methanol and acetone, were added to the both catalytic systems, as shown in the entries 3—4 and 15—16. It is observed that when organic solvent is added, both cyclohexanone conversion and total target product selectivity are becoming dramatically less in the two zeolite-catalyzed systems. Organic solvents are helpful for the selectivity of ε-caprolactone, but they reduce the concentration of cyclohexanone and H2O2molecules in the mixtures. Moreover, some side reactions may occur between solvents and products, since 6-hydroxyhexanoic acid and adipic acid can react with methanol to form corresponding esters. With respect to addition of extra H2O used as solvent in both systems, as shown in the entries 2 and 17 in Table 3, it is observed that the catalytic performance becomes weaker for the HTS catalyst, while it is favorable to the selectivity of 6-hydroxyhexanoic acid. Thus, we infer that no organic solvents addition can be the best choice for obtaining high yields of target products.
Some other reaction parameters, such as reaction temperature and catalyst content, are also investigated. We can see that improving the reaction temperature and catalyst content can increase the cyclohexanone conversion and total product selectivity. Under optimized conditions, the cyclohexanone conversion reaches about 60%, when the total selectivity is over 90% for the HTS zeolite catalyst. For the DHBEA zeolite catalyst, when the cyclohexanone conversion is 92%, the maximum 6-hydroxyhexanoic acid selectivity is over 95%.
According to these results, it is obvious that the cyclohexanone oxidation reactions catalyzed by the DHBEA and HTS zeolites can lead to high conversion and good target product selectivity in the absence of strong acid and organic solvents under mild condition. Thus, two environmentally friendly and controllable cyclohexanone transformation routes have been developed for producing different target products, such as 6-hydroxyhexanoic acid or adipic acid using clean and cheap 30% H2O2as the oxidant.
In order to investigate the different mechanisms of cyclohexanone oxidation catalyzed by two different catalysts, the molecular simulation methods were introduced to calculate the adsorption and activation of reagent molecules and determine their possible reaction pathways and their corresponding geometric structures and energy files. Herein, the trigonal Al(OSiH3)3and tetrahedral Ti(OSiH3)4species are chosen as the Lewis acid sites for the DHBEA and HTS zeolites, respectively. It is found that both cyclohexanone and H2O2molecules can be adsorbed on both Ti and Al clusters, which are attributed to the accepting-donating behavior of lone electron pairs between Lewis acid sites and the O atoms in cyclohexanone and H2O2molecules, as shown in Figure 4. When cyclohexanone and H2O2molecules adsorbed on the Lewis acid sites of trigonal Al species, the corresponding adsorption energy is -83.27 kJ/mol and -21.38 kJ/mol, respectively, as illustrated in Table 4. It means that cyclohexanone molecules are prone to be adsorbed on the Lewis acid sites of DHBEA zeolite, rather than H2O2molecules. Meanwhile, the adsorption energy of cyclohexanone and H2O2molecules on the tetrahedral Ti sites is -135.1 kJ/mol and -384.6 kJ/mol, respectively. It is indicated that H2O2molecules are much more easily to be adsorbed on the framework Ti species of HTS zeolite. Thus, it infers that the adsorption behavior of DHBEA zeolite is different from that of HTS zeolite.When the cyclohexanone and H2O2molecules are adsorbed on the trigonal Al species, the charges of the C atom of carbonyl group and the Obatom of H2O2are 0.50 and -0.39, while those in the original molecules are 0.40 and -0.43, respectively, as shown in Table 5. This fact suggests that the carbonyl groups in cyclohexanone molecules become more activated, but there is no impact on the activation of H2O2molecules after they are adsorbed on the Lewis acid sites of Al species. Furthermore, there are abundant Br?nsted acid (H+) sites which exist in the DHBEA zeolite. When H+species
react with the O atoms of carbonyl groups, the charge of C1atom in cyclohexanone is 0.519, which is greater than that in the original molecule (0.405). Thus, both the Br?nsted acid and the Lewis acid catalyzed BV oxidation mechanisms are proposed, as illustrated in Scheme 2.
Table 3 Catalytic performance of DHBEA and HTS1) zeolites in the Baeyer-Villiger oxidation of cyclohexanone and comparison to that of Sn-Beta zeoloite2)
Figure 4 Optimal adsorption models of cyclohexanone and H2O2 molecules on the Ti- and Al-active site clusters
Table 4 Adsorption energy of cyclohexanone and H2O2 on the Ti- and Al-clusters
Once the cyclohexanone molecules are activated by the Br?nsted and Lewis acid sites, the charges of C1atoms become larger, which means that they are much easier to be nucleophilically attacked by original H2O2molecules. After the attack initiated by H2O2molecules, the Criegee intermediates are formed. Then, the intramolecular rearrangement reaction takes place. Consequently, a Baeyer-Villiger oxidation process is completed, with the formation of ε-caprolactone molecules.
It is well known that the TS-1 zeolite is an effective catalyst to enhance the nucleophilic attack capability of H2O2molecules in many catalytic oxidation reactions[39-40].
Table 5 Charge and charge differences of cyclohexanone and H2O2 molecules on Al(OSiH3)3 and Ti(OSiH3)4 clusters
Scheme 2 Possible reaction mechanisms of cyclohexanone oxidation catalyzed by Br?nsted (a) or Lewis (b) acid sites on DHBEA zeolite
The tetrahedral Ti species can efficiently activate H2O2molecules, and then the BV oxidation mechanism via enhancing the nucleophilic attack capability is proposed, as shown in Scheme 3. At the initial stage of this route, the highly reactive Ti-OOH species are produced as a result of the reaction of Ti-OH groups with H2O2molecules, resulting in removal of H2O molecules[41]. Then, cyclohexanone molecules can be adsorbed on tetrahedral Ti(OSiH3)3OH sites via the hydrogen bonding interaction between the H atoms at Ti cluster and the O atoms in cyclohexanone molecules. In the next step, carbonyl groups can be nucleophilically attacked by Ti-OOH species, with the formation of the Criegee intermediate (CI) species. At the end, the ε-caprolactone molecules are generated through the intramolecular rearrangement reaction of CI species. After that, the whole BV oxidation reaction process is completed, giving birth to ε-caprolactone molecules.
Scheme 3 Possible reaction pathway of cyclohexanone B-V oxidation catalyzed by TS-1 zeolite involving the activation of H2O2 molecules
Above all, both DHBEA and HTS zeolites can serve as effective catalysts for the BV oxidation of cyclohexanone via different catalytic mechanisms. The DHBEA zeolite shows higher catalytic performance than the HTS zeolite thanks to its larger amount of Br?nsted and Lewis acid sites. However, the ε-caprolactone molecule is not stable with respect to the performance of H+species in the aqueous H2O2solution. It is calculated that the charge of C atoms of carbonyl groups in ε-caprolactone molecules is 0.69, which is much larger than that in the original molecules (0.59). It means that the carbonyl group can be easily subject to the nucleophilic attack of H2O molecules with the formation of the additive intermediate species. At last, 6-hydroxyhexanoic acid molecules are produced through an intramolecular rearrangement of the specific intermediate species, as shown in Scheme 4.
Scheme 4 One plausible mechanism of ring opening reaction of ε-caprolactone catalyzed by Br?nsted acid sites
As shown in Table 5, the Lewis acid species in Al atoms do not have the activation effect on enhancing the nucleophilic attack capability of H2O2molecules. Here, the adsorption models of 6-hydroxyhexanoic acid molecules on the trigonal Al sites are calculated, as illustrated in Figure 5. There are three O atoms in one 6-hydroxyhexanoic acid molecule, which can possibly be absorbed by trigonal Al species via the acceptor-donor interaction between Al sites of active sites and these O atoms. It is found that the Al…O bond distance in model (b) is shorter than that of other two models (a) and (c), which suggests that model (b) is the most possible one for further oxidation reaction. Table 6 gives the corresponding adsorption energy for these three models. When the adsorption energy of model (b) is defined as 0, then that of models (a) and (c) is 74.46 kJ/mol and 32.23 kJ/mol, respectively. Therefore, we can infer that the O atom of carbonyl group is apt to be adsorbed on trigonal Al sites. And in model (b), the charge of C atom of terminal C-OH group is -0.180, which is smaller than that in the original molecule (0.054). It is indicated that the terminal C-OH groups cannot be activated to accept the nucleophilic attack by H2O2molecules. Thus, according to the above-mentioned results, the reason why few adipic acid produced in DHBEA zeolite catalyzed system can be well explicated.
Being different from the trigonal Al species, the tetrahedral Ti species can act as the Lewis acid to make the H2O2molecules much more reactive (via forming Ti-OOH) during catalytic oxidation reactions. Hence, the terminal C-OH groups can be deeply oxidized to form carboxyl groups under the partcipation of Ti-OOH species, as shown in Scheme 5. Firstly, the 6-hydroxyhexanoic acid molecules can be adsorbed and oxidized by Ti-OOH species, with the formation of one 6-aldehyde caproic acid and one H2O molecule. However, 6-aldehyde caproic acid is highly reactive under these conditions, and it is adsorbed by Ti atom, and then the C atom of aldehyde group is nucleophilically attacked by Obatom of Ti-OOH species. Finally, one adipic acid and one H2O molecule are generated. Thus, the main products catalyzed by the HTS zeolite are ε-caprolactone, 6-hydroxyhexanoic acid and adipic acid.
Figure 5 Three models (a, b, and c) of 6-hydroxyhexanoic acid adsorbed on the Al(OSiH3)3 cluster. Some Al…O and O...H distances are given
Table 6 Energy difference among three hydroxyhexanoic acid adsorption models
Scheme 5 One plausible mechanism of deep oxidation reaction of 6-hydroxyhexanoic acid catalyzed by HTS zeolite
To investigate the impact of mass diffusion in the DHBEA and HTS zeolites catalyzed cyclohexanone oxidation, the diffusion energy barriers of cyclohexanone and H2O2inside the BEA and MFI topological structures are calculated by the Insight II software[42-44]. Here the straight channel along (001) of BEA zeolite, the straight channel along (100), and the zig-zag channel along (010) of the MFI zeolite are selected. As for the cyclohexanone molecule, the diffusion energy barrier (ED) in the straight channel of BEA zeolite is 18.22 kJ/mol, while that in the straight and the zig-zag channels is 105.67 kJ/mol and 137.77 kJ/mol, respectively, as shown in Table 7. It is inferred that cyclohexanone molecules can more easily diffuse inside the BEA zeolite than inside the MFI zeolite. Therefore, the DHBEA zeolite shows much higher catalytic activity and reaction speed than that of the HTS zeolite. On the other hand, the diffusion energy barriers of H2O2molecules in the BEA and MFI zeolites are very close (<10 kJ/mol), which are much lower than those for cyclohexanone molecules. As a result, we can confirm that the diffusion of cyclohexanone inside the MFI zeolite is a rate-dependent parameter for the cyclohexanone oxidation, which is helpful to the deep oxidation reactions, and agrees well with the experimental results.
Table 7 Diffusion energy barriers (ED) for cyclohexanone and H2O2 in the BEA and MFI zeolites
In this paper, two environmentally friendly cyclohexanone oxidation routes catalyzed by the DHBEA and HTS zeolites with 30% H2O2solution used as the oxidant have been developed, without using any organic solvents and toxic agents. The characterization results show that the BET surface area and the amount of acid sites in the DHBEA zeolite are much more than those of HTS zeolite, which are of great importance for enhancing their catalytic performance. Hence, the DHBEA zeolite shows higher cyclohexanone conversion and target product selectivity than the HTS zeolite under the same reaction conditions. By optimization of parameters, the cyclohexanone conversion and selectivity are about 60% and 90% for HTS zeolite, while both of them exceed over 95% for the DHBEA zeolite, respectively. Compared with a greater than 20% selectivity of adipic acid achieved by the HTS zeolite, there is almost no adipic acid detected in the DHBEA catalyzed system (with a 6-hydroxyhexanoic acid selectivity of over 90%). To investigate the catalytic performance difference between the DHBEA and HTS zeolites, molecular simulation methods were introduced to understand their chemical nature. It is demonstrated that cyclohexanone and H2O2molecules are preferentially to be adsorbed and activated on trigonal Al and tetrahedral Ti sites, respectively. Therefore different mechanisms for the BV oxidation of cyclohexanone are proposed via enhancing the reactivity of substrate and oxidant. The ring-opening of ε-caprolactone is considered to be accelerated by H+ions. Moreover, neither 6-hydroxyhexanoc acid nor H2O2molecules can become more reactive by the Lewis acid sites of Al over the DHBEA zeolite, while the formation of adipic acid is not detected. However, the 6-hydroxyhexanoic acid can be deeply oxidized to form adipic acid by Ti-OOH species. On the other hand, it is calculated that the DHBEA zeolite has greater diffusion performance of cyclohexanone than the HTS zeolite. Thus, the DHBEA zeolite demonstrates higher cyclohexanone conversion and 6-hydroxyhexanoic acid selectivity, while HTS zeolite is conducive to the formation of adipic acid.
Acknowledgement:This work was financially supported by the National Basic Research Program of China (973 Program, 2006CB202508), the Research Program of China Petrochemical Corporation (SINOPEC Group 20673054), and the National Key Research and Development Program of China (2017YFB0306800).