Anchoring Effect of Organosilanes on Hierarchical ZSM-5 Zeolite for Catalytic Fast Pyrolysis of Cellulose to Aromatics.

As an essential chemical feedstock, aromatics can be obtained from biomass by catalytic fast pyrolysis (CFP) technology, in which diffusion limitation is still a problem. In this study, several ZSM-5 zeolites with intercrystal stacking macropores were synthesized by adding organosilanes (OSAs) with different alkyl chain groups. Due to the structure-directing effect of the OSA, the prepared ZSM-5 zeolites possess a larger external surface area and pore volume than Blank-Z5. Moreover, the pore size is related to the extent of anchoring of the OSA and silicon-aluminum species in the zeolite precursor. Pyridine Fourier transform infrared (Py-FTIR) and NH3-temperature-programmed desorption (TPD) analyses show that the obtained ZSM-5 zeolites have a higher Brønsted acidity and total number of acid sites. In addition, excessive addition of OSA is not conducive to the growth of ZSM-5 zeolites. The catalytic performance of the synthesized ZSM-5 zeolites was evaluated by Py-GC/MS. The larger external surface area and pore volume improve the accessibility of the acid sites and thus promote the conversion of biomass into aromatics.

Introduction

With the shortage of fossil resources and increasing carbon emissions, biomass energy, as the only renewable carbon source, has gradually attracted attention to replace traditional fossil resources.[1−5] Catalytic fast pyrolysis (CFP) is a promising technology for producing aromatics directly from biomass that integrates fast pyrolysis and catalytic cracking technology.[6,7] In recent years, extensive research into the reaction mechanism[8−11] and catalysts[12−15] for CFP has been reported. Aromatization is the main pathway for small-molecule oxygenates and olefins to form aromatics through the “hydrocarbon pool mechanism,” activated by the shape-selective effect of ZSM-5 zeolite channels.[16]

The average pore dimensions of ZSM-5 are close to the kinetic diameter of aromatic products, and ZSM-5 zeolite exhibits favorable aromatization capacity, shape selectivity, and tunable acidity.[17] Therefore, ZSM-5 zeolite is considered a suitable catalytic pyrolysis catalyst for the CFP process. However, conventional ZSM-5 zeolite microporous channels can only accommodate the transfer of small-molecule organics. Primary pyrolysis oxygen compounds are prone to occur polymerization and condensation reactions on the outer surface of the zeolite and inside the micropores to form coke.[18] Thus, it is necessary to improve the diffusion of macromolecule substances and increase the accessibility of acid sites in the pores of the zeolite.[19,20]

Hierarchical ZSM-5 zeolite has been verified to be an effective material for solving the above problem.[21] Qiao et al. reported a 41.8% aromatic yield for CFP of cellulose with hierarchical ZSM-5 zeolite.[22] There have been some reports on the preparation of hierarchical ZSM-5 zeolites by alkaline treatment. Appropriate treatment conditions can effectively introduce interconnected mesopores in parent zeolites.[23,24] In addition, mesopores or macropores can be introduced by adding organosilanes (OSAs) to the zeolite precursor. However, the OSA might prolong the crystallization time and reduce the crystallinity and acidity, such as for bipedal OSAs with the structure of (OR2)3SiR1Si(OR2)3, which can strongly impede the aggregation of nanounits, leading to a partially amorphous material.[25] Nonetheless, the mesopore diameter can be tailored by altering the OSA group or adding a hydrophobic swelling agent to modulate the interaction of OSA with silicon–aluminum species.[22,26−28] For instance, a new structure-directing agent was used to prepare single-walled zeolitic nanotubes, presumably a long-chain agent containing π-stacking interactions that forms a template for nanotube zeolites.[29] The presence of an amine moiety in OSA increases the extent of anchoring of OSA to the primary zeolite nanounits in the gel, resulting in more mesopores.[30,31] Therefore, the effect of the OSA structure on the physicochemical properties of zeolites should be considered for designed reactions.

In this research, three OSAs with different alkyl chain groups were used to prepare hierarchical ZSM-5 zeolites to investigate the anchoring effect of OSAs on hierarchical ZSM-5 zeolite. The obtained ZSM-5 zeolites form a special morphology and possess a large pore volume and relatively high crystallinity. The prepared zeolites were used to produce aromatics through CFP of cellulose. The results indicate that a better balance between acidity and the pore structure can be achieved by selecting an appropriate OSA to prepare hierarchical ZSM-5 zeolites. The ZSM-5 zeolite with the highest acidity and moderate pore volume shows the highest aromatic yield (42.2%) and a lower coke yield (29.3%).

Results and Discussion

Characterization

The X-ray diffraction (XRD) patterns for Blank-Z5 and catalysts with different OSAs added are shown in Figure . Figure S1 shows the XRD characterization of ZSM-5 zeolites with a varying amount of ODDMMS addition. All synthesized ZSM-5 zeolites show an MFI topological structure (JCPDS: 01-087-1527).[22] The intensity of the peak varies with the type and amount of OSA, indicating that the type and amount of OSA affect the ZSM-5 zeolite crystallinity. The calculation of crystallinity is based on the sum of the diffraction peak areas in the range of 2θ = 22–25°. The relative crystallinity based on Blank-Z5 is displayed in Table S1. ODMDES(0.05)-Z5 shows the lowest relative crystallinity, while the relative crystallinity of ODTMS(0.05)-Z5 and ODDMMS(0.05)-Z5 is only slightly lower than that of Blank-Z5. This result suggests that the addition of ODTMS and ODDMMS can overcome the problem of significantly decreased crystallinity for hierarchical ZSM-5 zeolite.[22,32] However, the relative crystallinity of ZSM-5 zeolite distinctly decreases with increasing addition of ODDMMS. This proves that excessive addition of ODDMMS can suppress the crystallization process of zeolites.

Figure 1

XRD patterns of synthesized ZSM-5 zeolites.

27Al MAS NMR spectra characterizing the environment of Al atoms in zeolites are shown in Figures and S2. The chemical shifts observed at 55 and 0 ppm represent the tetrahedral Al in the zeolite framework (FAl) and extraframework octahedral Al (EFAl) atoms, respectively.[33] It can be observed that the proportion of EFAl in all prepared ZSM-5 zeolites is close to that in the standard ZSM-5 zeolite (in the range 5–10%).[34] The FAl content in the three zeolites synthesized with OSA is slightly lower than that in Blank-Z5, indicating that Al atoms tend to form EFAl when preparing zeolite using OSA. When ODDMMS/SiO2 = 0.15, the EFAl content reaches up to 10.4%. This may be related to the significant decrease in the relative crystallinity of ODDMMS(0.15)-Z5.

Figure 2

27Al MAS NMR spectra of the ZSM-5 zeolites.

Figure displays the scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of ODDMMS(0.05)-Z5 and Blank-Z5 zeolites. As clearly shown in Figure D, Blank-Z5 is composed of aggregates of loose small particles. ODDMMS(0.05)-Z5 shows a cactus-like morphology. Many small particles are attached to the surface of large particles of approximately 300 nm in size. As shown in Figure A, the small particle size of ODDMMS(0.05)-Z5 mainly ranges from 30 to 50 nm. Based on the TEM images and our previous work, the possible mechanism of OSA affecting the morphology of ZSM-5 can be explained as follows. First, protozeolitic nanounits formed in the precrystallization process in the presence of the microporous template tetrapropyl ammonium hydroxide (TPAOH). Then, the RO–Si– moieties of OSA hydrolyzed to produce the −Si–OH structure and formed −Si–O–Al– and −Si–O–Si– linkages that anchored the OSA in the zeolite framework. ZSM-5 zeolites prepared with OSA show a similar morphology, which is different from Blank-Z5, indicating that the long carbon chain group in OSA plays a major role in the morphology of ZSM-5 zeolites. The long carbon chain adsorbed more silicon–aluminum species and restrained the accumulation of nanoparticles. Therefore, molecular addition became the predominant mechanism involving two-dimensional (2D) layer nucleation and stepwise advancement of layers, resulting in the formation of large particles. Later, silicon–aluminum species adsorbed by the carbon chain of OSA gave another 2D layer and small particles continued to grow one by one by a second 2D layer growth pathway. Finally, the morphology of small particles attached to the surface of large particles was formed.[25,26,30,32] The formation of large particles may be the reason why ZSM-5 zeolites prepared with OSA still show high crystallinity. TEM images of ODDMMS(0.10)-Z5 and ODDMMS(0.15)-Z5 are shown in Figure S3C,D. As the addition of ODDMMS increases, the small particles attached to the surface of the large particles gradually become loose and even detach from the surface of the large particles. This may be because excessive ODDMMS hinders the aggregation of small particles. The morphology of ODTMS(0.05)-Z5 and ODMDES(0.05)-Z5 shown in Figure S3A,B shows similar characteristics, which prove the above assumption.

Figure 3

SEM and TEM images of the ZSM-5 zeolites. (A) SEM of ODDMMS(0.05)-Z5, (B,C) TEM of ODDMMS(0.05)-Z5, and (D) TEM of Blank-Z5.

The textural properties including the Brunauer–Emmett–Teller (BET) surface area and pore volume for different zeolites were calculated by N2 adsorption–desorption. The isotherms and the pore size distribution are shown in Figures and 5, respectively. As shown in Figure , the ZSM-5 zeolites contain both microporous and macroporous structures, which explains the rapid increase in the isotherms observed when P/P0 is 0 and 0.9–1.0, respectively. Mesopores and macropores with sizes larger than 10 nm are mainly generated by the accumulation of ZSM-5 zeolite particles.[26,35]

Figure 4

N2 adsorption–desorption isotherms of the prepared ZSM-5 zeolites.

Figure 5

Pore size distributions of the prepared ZSM-5 zeolites.

The BET surface area and pore volume are displayed in Table . Compared with Blank-Z5, the zeolites prepared with OSA show a smaller micropore surface area and micropore volume but a larger external surface area and mesopore volume, showing an increase of at least 27% from 95 m2/g (Blank-Z5) to 120 m2/g (ODMDES(0.05)-Z5), indicating that the OSA influences the crystal growth process. According to the TEM characterization (Figure ), there are many small particles with a size of tens of nanometers on the surface of Blank-Z5 and ZSM-5 zeolites prepared with OSA, indicating that the increase in the external surface area is mainly due to the intercrystal stacking of mesopores and macropores rather than a decrease in crystal size. By increasing the external surface area, more reactive sites can be exposed, and the catalytic reaction at these active sites has a smaller steric hindrance than that in the micropores. During crystallization of zeolites, the RO–Si– moieties of OSA hydrolyzed to produce the −Si–OH structures and the OSA anchored to protozeolitic nanounits through covalent bonds. Then, OSA affected the crystallization process of zeolites, and the external surface area and pore volume increased finally.[25,30,32] Interestingly, ODTMS contains three RO–Si groups, resulting in more protozeolitic nanounits anchored by ODTMS. Thus, ODTMS(0.05)-Z5 shows the largest BET surface area and mesopore volume but the smallest micropore surface area and micropore volume. This can lead to a lower acidity, which then reduces the catalytic pyrolysis activity in the CFP process.[31] However, adding ODDMMS(0.05) to the precursor not only increases the external surface area but also has a small effect on the micropore structure of the zeolite. With increasing addition of ODDMMS, the micropore surface area and micropore volume of ZSM-5 zeolites gradually decrease, indicating that the addition of ODDMMS hinders the formation of the microporous structure to a certain extent.[22] The intercrystal stacking macroporous volume decreases, in good agreement with the gradual loosening of small particles as the addition of ODDMMS increases.

Table 1

Textural Properties of the ZSM-5 Zeolites

catalysts	SBET (m2/g)a	Smic (m2/g)	Sext (m2/g)	Vtol (cm3/g)b	Vmic (cm3/g)	Vmes (cm3/g)
Blank-Z5	371	276	95	0.311	0.163	0.148
ODTMS(0.05)-Z5	385	250	135	0.349	0.136	0.213
ODMDES(0.05)-Z5	384	264	120	0.321	0.152	0.169
ODDMMS(0.05)-Z5	384	261	123	0.329	0.152	0.177
ODDMMS(0.10)-Z5	375	256	119	0.278	0.156	0.122
ODDMMS(0.15)-Z5	377	252	125	0.305	0.145	0.160

BET method.

Volume adsorbed at P/P0 = 0.99.

The Py-FTIR results are shown in Figure . The peak located at ∼1460 cm–1 corresponds to Lewis acid sites. The peak at ∼1545 cm–1 is attributed to Brønsted acid sites. The peak at ∼1490 cm–1 corresponds to the interaction of Brønsted acid and Lewis acid sites.[36,37] The results obtained from a quantitative analysis of acidity are displayed in Table . Table shows that ODDMMS(0.05)-Z5 has the highest concentration of Brønsted acid sites. ODTMS(0.05)-Z5 shows the lowest concentration of Brønsted acid sites compared with other zeolites with the addition of OSA, which may be related to the low microporosity of the structure. The NH3-TPD results for different zeolites are shown in Figure . The two main characteristic peaks are located at 150–300 and 300–500 °C, representing the weak acid sites and strong acid sites of zeolites, respectively.[38] The total acid amounts for ODTMS(0.05)-Z5, ODMDES(0.05)-Z5, and ODDMMS(0.05)-Z5 are all increased compared with Blank-Z5, and ODDMMS(0.05)-Z5 shows the highest total acid amount among all samples. This result is consistent with the Py-FTIR results.

Figure 6

Py-FTIR profiles of the synthesized ZSM-5 zeolites.

Table 2

NH3-TPD Data and Acid Amount of the Prepared ZSM-5 Zeolites

	tpeak (°C)		acid amount (μmol/g)
catalysts	LT peak	HT peak	total acidity	weak acidity	strong acidity	Ba	La
Blank-Z5	207.0	390.3	921	406	515	902	258
ODTMS(0.05)-Z5	209.1	402.4	1002	508	494	966	293
ODMDES(0.05)-Z5	204.7	389.1	1015	503	512	1047	312
ODDMMS(0.05)-Z5	212.7	416.7	1262	621	641	1140	236
ODDMMS(0.10)-Z5	202.2	383.5	1048	515	533	1005	259
ODDMMS(0.15)-Z5	202.8	385.9	1028	496	532	957	254

Measured at 150 °C.

Figure 7

NH3-TPD profiles of the synthesized ZSM-5 zeolites.

The Py-FTIR and NH3-TPD profiles of ZSM-5 zeolites prepared with different ODDMMS addition amounts are shown in Figures S4 and S5. With increasing ODDMMS addition, the Brønsted acid amount and the total acid amount of the zeolite decrease, which may be related to the decrease in micropore surface area and crystallinity. Compared with the lower Brønsted acid sites of the hierarchical ZSM-5 zeolite prepared by alkaline treatment, the addition of OSA can reduce the damage to the structure and acid amount of the ZSM-5 zeolite.[38−40] Many studies have shown that Brønsted acid sites and pore structure are critical for the CFP process.[11,31,41,42] Thus, it is practical to prepare hierarchical ZSM-5 zeolite by adding a suitable amount of ODDMMS to the precursor gel.

29Si MAS NMR spectra characterizing the environment of Si atoms in zeolites are shown in Figure . The chemical shifts at ∼−102, ∼−107, and >−110 ppm can be assigned to SiOH, Si(1Al, 3Si), and Si(4Si), respectively. It is generally accepted that the SiOHAl groups (corresponding to Si(1Al,3Si)) can offer Brønsted acid sites and strong acid sites for zeolites and SiOH can offer weak acid sites. It is interesting to note that zeolites prepared with OSA show a higher content of Si(1Al,3Si) and SiOH sites (Figure S1) compared with Blank-Z5, leading to the increase of acid amount.[33,43−45] According to 29Si MAS NMR spectra, we can conclude that OSA can modulate the coordination environment of Si in zeolites and thus promote the formation of acid sites.

Figure 8

29Si MAS NMR spectra of prepared zeolites. (A) Blank-Z5, (B) ODTMS(0.05)-Z5, (C)ODMDES(0.05)-Z5, and (D) ODDMMS(0.05)-Z5.

Catalytic Performance

A microreactor directly connected to a gas chromatography/mass spectrometry (GC/MS) system was used for the qualitative and quantitative analyses of the pyrolysis products for the catalytic pyrolysis of cellulose, and the yields of the main products are shown in Figure A. A 37.0% yield of aromatics, a 14.4% yield of CO, an 8.4% yield of CO2, a 2.8% yield of short-chain alkanes and olefins, and a 36.7% coke yield were obtained when catalyzed by Blank-Z5, which is consistent with the reported results.[22,46] For CFP of cellulose over ODDMMS(0.05)-Z5, the yields of aromatics and coke are 42.2 and 29.3%, respectively. The formation of CO, CO2, and H2O is mainly derived from the primary pyrolysis of oxygenated compounds. Decarbonylation, decarboxylation, and dehydration concurrently occur on the acid sites of ZSM-5 zeolites. The yield of CO is much higher than that of CO2, indicating that the oxygen atom in the primary pyrolysis products tends to be more easily removed by decarbonylation.[47]

Figure 9

Product yield and aromatic distributions obtained for the catalytic fast pyrolysis of cellulose over the synthesized ZSM-5 zeolites. (A) Yield of products for CFP of cellulose over ZSM-5 zeolites prepared with different OSAs; (B) aromatic selectivity for CFP of cellulose over ZSM-5 zeolites prepared with different OSAs; (C) yield of products for CFP of cellulose over ZSM-5 zeolites prepared with different amounts of ODDMMS; and (D) aromatic selectivity for CFP of cellulose over ZSM-5 zeolites prepared with different amounts of ODDMMS.

Many studies have proven that Brønsted acid sites can promote the conversion of primary pyrolysis products of biomass into final aromatics. Meanwhile, Lewis acid sites can inhibit side reactions in the Diels–Alder reaction.[20,40,48] ODMDES(0.05)-Z5 (aromatic yield of 40.2%) and ODDMMS(0.05)-Z5 (aromatic yield of 42.2%) possess similar porous parameters while the latter shows more Brønsted acid sites and therefore higher yield for aromatics during the CFP process. This result proves the importance of Brønsted acid sites in the CFP process.[31] In addition, we should note that ODTMS(0.05)-Z5 (aromatic yield of 40.6%) shows fewer Brønsted acid sites and a larger external surface area than ODMDES(0.05)-Z5 (aromatic yield of 40.2%). The two zeolites show a similar yield of aromatics, which can prove that an increase in the external surface area can improve the accessibility of acid sites and make up for a slight lack of acidity.[22,49] A comprehensive consideration of the Brønsted acid sites and pore structure can explain why zeolites prepared with OSA reflect a higher yield for aromatics and the production of less coke compared with Blank-Z5. Interestingly, although ODTMS(0.05)-Z5 shows a larger external surface area and mesoporous pore volume, it possesses the smallest micropore area, resulting in a lower aromatic yield than ODDMMS(0.05)-Z5. This proves the importance of the micropore structure and acid sites in the CFP process. Thus, a balance between the proportion of micropores and macropores is essential to obtain a higher aromatic yield.

The catalytic performance of ODDMMS(0.10)-Z5 and ODDMMS(0.15)-Z5 is shown in Figure C. With increasing ODDMMS addition, the yield of aromatics is reduced, which shows the same tendency as the acid amount for the zeolites. This proves the above conclusion. In addition, the yield of aromatics decreases to its lowest value of 38.1% when ODDMMS/SiO2 = 0.15. This may be related to the obvious decrease in crystallinity and micropore surface area. The external framework silicon–aluminum species are easily dissolved under biomass reaction conditions. These dissolved species are catalytically active and can lead to undesirable side reactions, thereby resulting in loss of aromatic selectivity and yield.[50]

The distribution of components in the aromatic products is shown in Figure B–D. Based on the carbon number in aromatics, the products are divided into benzene (B), toluene (T), xylene (X), trimethylbenzene (C9A), C10 aromatics (C10A), and C11 and C11+ aromatics (C11A+). The distribution of products under different zeolite-catalyzed CFP processes is similar. Compared with Blank-Z5, the BTX selectivity in other ZSM-5-catalyzed cellulose pyrolysis processes slightly decreases. This might be due to the larger external surface area and pore volume making it easier to generate polycyclic aromatics and the undesirable side reactions caused by the increase in EFAl.[19,50]

Conclusions

Three organosilanes with different alkyl chain groups were used to prepare hierarchical ZSM-5 zeolites. ODTMS with three RO–Si groups can be more easily anchored to protozeolitic nanounits, resulting in a larger external surface area and pore volume while also reducing the micropore area of ZSM-5 zeolites. However, there are three alkyl chain groups in ODDMMS that act as a hydrophobic swelling agent, which increases the external surface area and pore volume while reducing the impact on the micropore structure and showing the highest acidity. In addition, the influence of the OSA addition amount on the physicochemical properties of ZSM-5 zeolites was studied. Excessive addition of OSA is not beneficial to the growth of zeolites. As the addition of ODDMMS is increased, the FAl content, acid amount, and aromatic yield decrease. Compared with the catalytic activities of different ZSM-5 zeolites on CFP of cellulose to aromatics, sufficient acid sites and suitable pore structures are conducive to the formation of aromatics. The ODDMMS (0.05)-Z5 zeolite shows the largest acid amount and a larger external surface area and total pore volume, thereby showing the highest aromatic yield of 42.2% for CFP of cellulose to aromatics.

Experimental Section

Materials

Sodium aluminate (NaAlO2), octadecyl trimethoxysilane (ODTMS, >90%), benzene (>99%, GC), xylene (>99%, GC), 1,3,5-trimethylbenzene (>99%, GC), and naphthalene (>99.7%, GC) were purchased from Aladdin Chemicals, Shanghai, China. Tetraethyl orthosilicate (TEOS, AR), tetrapropyl ammonium hydroxide (TPAOH, 25 wt % in water), methylbenzene (>99.5%, GC), 1-methyl naphthalene (>99%, GC), ammonium chloride (NH4Cl, AR), and acetone (AR) were obtained from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China. Octadecyl dimethylmethoxysilane (ODDMMS, >90%) was obtained from TCI, Japan. Octadecylmethyl diethoxylsilane (ODMDES, >97%) was obtained from Acme, China. Microcrystalline cellulose was obtained from Sigma-Aldrich.

Catalyst Synthesis

The catalyst was synthesized by a hydrothermal synthesis method (Scheme ), using NaAlO2 as the aluminum source, TEOS as the main silicon source, TPAOH as the microporous template, and a certain amount of OSA. Typically, a known quantity of NaAlO2 was dissolved in deionized water and stirred until the solution was clear. Then, a certain amount of TPAOH was added with continuous stirring. Following this, TEOS was added drop by drop until it was totally hydrolyzed. The precrystallization was conducted at 90 °C with stirring for 8 h to obtain the precursor under reflux condensation conditions. The molar ratio in the obtained precursor was Al2O3/SiO2/TPAOH/H2O/Na2O = 1:35:8:1000:1.3. Then, a certain quantity of OSA was added and stirred for 8 h. During this procedure, we used ODTMS, ODMDES, and ODDMMS as three different kinds of OSA. Afterward, the precursor was transferred into a 100 mL Teflon-lined autoclave for crystallization at 170 °C for 72 h to obtain the solid. The solid was separated, washed with deionized water, and dried thoroughly. Then, the solid was calcined at 550 °C for 5 h at a heating rate of 2 °C/min. Next, the calcined solid was subjected to ion exchange with 1 mol/L NH4Cl solution at 80 °C for 8 h. The ion exchange was repeated three times. After each ion exchange, the catalyst was filtered and thoroughly washed. Finally, the catalysts were calcined again at 550 °C for 5 h at a heating rate of 2 °C/min. The catalyst without any OSA was denoted as Blank-Z5. The catalysts were named according to the molar ratio for the different amounts of OSA added. For instance, when adding ODDMMS as OSA with OSA/SiO2 = 0.05, the catalyst was named ODDMMS(0.05)-Z5.

Scheme 1

Preparation of Hierarchical ZSM-5 Zeolite with OSA

The powder X-ray diffraction (XRD) patterns were collected using a Shimadzu XRD-6000 operated at a voltage and current of 40 kV and 30 mA, respectively, with Cu Kα radiation at a scanning rate of 4°/min from 5 to 50°.

27Al MAS NMR spectroscopy was performed using a Bruker Avance III 500 spectrometer with a 4 mm ZrO2 rotor. The 27Al MAS NMR spectrum was recorded using single-pulse sampling with a pulse width of 0.22 μs (π/12), and the number of revolutions of the MAS was 12 kHz.

N2 adsorption–desorption isotherms were measured by a Micromeritics 3Flex type full microporous physical adsorption–desorption instrument. All samples were degassed at 350 °C for 4 h before testing. The surface area was determined using the Brunauer–Emmett–Teller (BET) method. The external surface area was calculated by the t-plot method, and the micropore volume was calculated according to the refs (51, 52). The total pore volume was derived from the amount of N2 adsorbed at P/P0 = 0.99.

Scanning electron microscopy (SEM) images were obtained using an S-4800 cold-field emission electron microscope (Hitachi, Japan) to observe the surface morphology and crystal size of the catalysts. The samples were evenly dispersed on the conductive adhesive for testing.

The transmission electron microscopy (TEM) images were measured on a JEM-2100, 200 kV high-resolution transmission electron microscope (JEOL Co., Ltd).

NH3-temperature-programmed desorption (NH3-TPD) was conducted using an AMI-3300 automatic temperature program chemical adsorption instrument (ALTAMIRA Company). The sample was first heated to 550 °C at a heating rate of 10 °C/min under an argon atmosphere and held for 2 h to remove impurities. After the temperature was reduced to 100 °C, 10% NH3–He gas flow was introduced until the adsorption became saturated. After the excess NH3 was blown by high-purity He, the sample was heated to 600 °C at a rate of 10 °C/min. The signal value recorded by the thermal conductivity detector (TCD) was used to calculate the amount of acid based on the pulse integral area.

Brønsted acid sites and Lewis acid sites were measured using an FTIR spectrometer (Py-FTIR, frontier FTIR spectrometer from PerkinElmer) with the adsorption of pyridine. The sample was kept at 400 °C for 2 h to remove impurities. Then, pyridine was introduced into the sample until the absorption was saturated. The infrared spectrum was recorded when the temperature dropped to 150 °C. After subtracting the background peak, the amount of Brønsted acid and Lewis acid was calculated quantitatively according to the area, and the extinction coefficients for Brønsted acid sites and Lewis acid sites are 1.67 and 2.22 cm/μmol, respectively.[53]

29Si MAS NMR spectroscopy was performed using a Bruker Avance III HD spectrometer. 29Si MAS NMR spectra were recorded using single-pulse sampling with 10 s recycle delay and 600 scans.

Catalytic Performance Evaluation

The catalytic pyrolysis of cellulose to produce aromatics was carried out in a tandem microreaction system (Rx-3050 TR, Frontier Laboratories, Japan). This reaction system was composed of upper and lower reactors with heater temperatures ranging from 40 to 900 °C. The reaction was conducted in the upper reactor at 600 °C. The system was directly connected to a GC-MS system (Agilent GC7890B/MS5977A) equipped with a mass spectrometer detector (MSD), a hydrogen flame ionization detector (FID), and a thermal conductivity detector (TCD).

In a typical CFP reaction experiment, 4 mg of cellulose was added into a weighing bottle (30 mm × 60 mm) with a certain amount of zeolite screened less than 400 meshes. The mass ratio of catalyst to cellulose was 20:1, which was selected from our earlier research.[13] After the mixture in the bottle was vibrated and mixed thoroughly, 4 mg of the mixture was charged into a stainless steel sample cup. Then, the cup was put into the reactor once the temperature reached the set value. The reaction was carried out under high-purity helium. The pyrolysis products passed through the reactor and the cold trap and finally entered the chromatographic column Ultra Alloy-5 (30 m × 0.25 mm × 2 μm). The qualitative and quantitative analyses for the product were completed using the MSD, FID, and TCD. The heating procedure is described as follows: hold at 35 °C for 5 min, ramp to 280 °C at 10 °C/min, and then hold for 10 min. For the gas products, CO and CO2 were analyzed by TCD, and C2H6, C2H4, C3H8, C3H6, C4H8, and aromatics were analyzed by FID. The coke was quantitatively determined by elemental analysis (Elemental Vario Micro cube, Germany). All product yields were finally calculated based on the carbon yield (the molar ratio of carbon in the product to that in the reactant). The aromatic selectivity was defined as the ratio of the number of moles of carbon in a particular aromatic component to the total number of moles of carbon in all aromatics.