Initial optimization of the catalytic performance
Fig. 9 exhibited the catalytic performance of PMWW(31) and PMWW(74) in the liquid alkylation between benzene with 1-dodecene. As shown in Fig. 9a, PMWW(31) presented the comparable conversion of 1-dodecene compared with MZN4-14 since they had the similar concentration of external Brønsted acid sites (Table S2). While PMWW(74) showed the best activity in all of investigated MWW zeolites due to its largest concentration of external Brønsted acid sites (Table S2, Fig. S10), and the fraction of external Brønsted acid sites for PMWW(74) achieved 88.5 %, indicating that the optimized pillaring process was an effective method to expose the accessible Brønsted acid sites. Moreover, the 27Al multiquantum magic angle spinning (MQMAS) NMR spectroscopy was further conducted to reveal the potential differences in the spatial distribution of Al sites between PMWW(31) and PMWW(74) samples, as shown in Fig. 10.
The framework of MWW zeolite was composed of eight different T sites and three types of pores: 12-MR surface pockets, 10-MR windows accessing internal supercages and 10-MR sinusoidal channels, which was depicted in Fig. S11. It was noted from Fig. 10 that both PMWW(31) and PMWW(74) exhibited two peaks at around ~56 ppm and 0 ppm, which was assigned to the 4-coordinated framework (AlF) and 6-coordinated extra-framework aluminum (AlEF) species, respectively.33According to previous study,34 the broad peak around ~56 ppm can be resolved into three overlapped signals at 50, 56, and 61 ppm, which was attributed to T6+ T7 (δ = 50 ppm), T1+ T3 + T4 + T5 + T8 (δ = 56 ppm) and T2 (δ = 61 ppm), respectively. Following this way, the broad peak around ~56 ppm was deconvoluted, as presented in Table S3 and Fig. 10. It was observed that the proportion of AlF(50)species for PMWW(31) was about 49 % (the combination of δ = 50 and δ = 51), indicating the AlF species mainly located in the supercages and surface pockets,35 in which the alkylation occurred as a result of accessible Brønsted acid sites arised from the AlF(50) species. While the AlF(61) species for PMWW(31) was missing, which demonstrated that proportion of AlF(61) species in PMWW(31) was negligible. However, AlF(61) species were related to T2 sites that located on the exterior surface of MWW nanosheets close to the outer edge of the 12-MR pockets, and the T2 sites was the most accessible to bulky adsorbates.15 In this point of view, it was reasonable that PMWW(74) had an advantage over PMWW(31) in the alkylation due to the presence of considerable AlF(61) species (~13 %). In addition, PMWW(31) and PMWW(74) possessed the similar quantity of AlEF(0) species. Fig. 9b showed that PMWW(31) and PMWW(74) had comparable selectivity of 2-LAB regardless of the multilamellar arrangement of MWW zeolite nanosheets, because PMWW(31) and PMWW(74) shared the same 12-MR hemicavities structure on the upper and lower surface of MWW layers,36 which was in line with above computer calculation data (Fig. 5). These analyses corroborated that the fabrication of order multilamellar arrangement of MWW nanosheets and precise control of Al species distribution can greatly expose the accessible Brønsted acid sites.
Stability, acidity and catalytic performance of regularly spaced MWW zeolite nanosheets
The stability, acidity and catalytic performance of optimized PMWW(74) was further investigated in the alkylation between benzene with 1-dodecene. As shown in Fig. 11a, the conversion of 1-dodecene for PMWW(74) and MCM-22 was 85 % and 40 % in the first cycle, respectively. With the number of cycles increasing, the activity of PMWW(74) and MCM-22 all displayed the tendency declining. For MCM-22, the conversion dropped to 19 % after five cycles, while PMWW(74) still maintained 68 % conversion even after five recycling experiments, indicating that PMWW(74) possessed strong stability compared with MCM-22 due to the enhanced diffusion of PMWW(74) with open single-layered structure.21 As shown in Fig. S12 and Table S4, which clearly verified that the effective diffusional time constants (D eff/r2 ) of PMWW(74) was higher than MCM-22. Interestingly, the selectivity of 2-LAB for MWW zeolite catalysts was almost not influenced after cycle experiments (Fig. 11b), because MWW zeolite possessed special 12-MR pockets on the exterior surfaces that had the best matched steric configuration and charge force with 2-LAB isomer.
To examine the structure of zeolite catalysts before and after cycles, the corresponding XRD, N2 adsorption-desorption isotherm and acidity of fresh and spent catalysts after five cycles were performed, as given in Fig. 12, Fig. S13 and Fig. S14, respectively. It was observed from Fig. 12 that both PMWW(74) and MCM-22 exhibited the robust stability with similar reflection patterns before and after cycles. However, the coalescence of the (101) and (102) diffraction peaks into a broadband reflection for spent PMWW(74) was observed, indicating partial loss of vertical alignment order in the stackingc direction in the recycle experiments,37 but (001) and (002) reflections were still preserved, suggesting that multilamellar structure of PMWW(74) was also totally preserved.16 In addition, the fresh PMWW(74) and spent PMWW(74) showed the comparable N2 adsorption-desorption isotherms (Fig. S13) and textural parameters (Table S1), suggesting that multilamellar arrangement of MWW nanosheets over PMWW(74) was not destroyed, but the spent MCM-22 exhibited a decreased BET surface area (557 m2g-1 to 451 m2g-1, Table S1), indicating that the channels of MCM-22 was blocked on a certain degree due to diffusion limitation under the sole micropores. Moreover, the total Brønsted acid sites and external Brønsted acid sites of PMWW(74) and MCM-22 after five cycles were examined, as presented in Fig. S14 and Table S2. It was observed from Table S2 that the total Brønsted acid sites and external Brønsted acid sites of spent MCM-22 dropped by 83.2% and 34.6 %, respectively. While the total Brønsted acid sites and external Brønsted acid sites of spent PMWW(74) decreased by 32.3 % and 30.6 %, respectively. These data confirmed that multilamellar arrangement of MWW nanosheets for PMWW(74) exhibited the evident advantage of architecture compared with conventional 3D MCM-22, in which the sole micropores of MCM-22 suffered from serious diffusion limitation during the alkylation and their narrow micropore entrance was easily blocked by bulk alkylbenzene products,21 resulting in the concentration of total Brønsted acid sites drastically decreasing after cycles.