On the Deactivation Analysis of IM-5 Zeolite in Pseudocumene Methylation with Methanol

Abstract

In the methylation of pseudocumene with methanol over IM-5 zeolite, the yield of durene can be enhanced. However, poorer stability of the catalytic activity was observed, especially at a higher methanol/pseudocumene ratio. In this paper, conventional characterization methods (XRD, XRF, TGA, SEM, physical adsorption, OH-IR, NH₃-TPD, and Py-IR) were used to characterize fresh and deactivated IM-5 zeolite and ZSM-5. FT-IR, XPS, TG-MS, GC-MS, FT-ICR MS, and NMR were employed to characterize deactivated IM-5 zeolite. It was found that the deactivation of IM-5 zeolite was mainly due to the severe coverage of acidic sites and pore channels by carbon deposits. The carbon deposits within the internal surface had a higher abundance, mainly in the form of linear unsaturated chain-like structures with a high degree of unsaturation. The carbon deposits on the external surface were mainly polycyclic aromatic hydrocarbons with alkyl side chains and a high degree of saturation, accompanied by unreacted methanol. Moreover, graphitized carbon existed on both the internal and external surfaces, which made the conventional coke-burning regeneration method unable to restore the activity of the post-reaction IM-5 zeolite. This work had certain reference significance for modulating the acidity and pore channels of zeolite catalysts, thus improving the activity and stability of the catalysts and extending their service life.

1. Introduction

The methylation of pseudocumene (1,2,4-trimethylbenzene) with methanol served as an important [1,2,3] and economic pathway [4] for producing durene (1,2,4,5-tetramethylbenzene), a key precursor in high-performance polyimide production [5,6]. Among all the catalysts, ZSM-5 zeolites (MFI code, 10-membered ring channels with the structure of pentasil) demonstrated molecular sieving capabilities by aligning the pore dimensions with the molecular sizes of trimethylbenzene and tetramethylbenzene isomers [7,8,9]. Nevertheless, ZSM-5’s moderate acid strength and restricted pore aperture (<5.5 Å) limited diffusion efficiency, leading to a low conversion of pseudocumene (10–30%), low yield of durene (5–15%), and the formation of by-products such as non-aromatics and BTX (benzene, toluene, and xylene) [10,11,12]. Moreover, it is regrettable that investigations on the catalysis of this reaction over ZSM-5 zeolites as well as other catalysts have exhibited a marked stagnation since the turn of the 21st century, with no further research outcomes having been publicly documented over the past two decades. These limitations underscore the necessity of selecting zeolites with larger pore sizes and stronger acid properties as catalysts.

Across 12-membered ring HY (FAU code) and H-mordenite (MOR code) zeolites with larger pore window sizes, the diffusion of reactants and products could be obviously improved, as well as the accessibility of active sites. Therefore, the conversion rate of pseudocumene (70–80%) was much higher than that of HZSM-5 (11%) [7,12]. However, the product distribution was close to the thermodynamic equilibrium (the proportion of durene in tetramethylbenzene was about 40%), which is much lower than that of ZSM-5 with shape selectivity [13,14]. Meanwhile, isomerization and further alkylation reactions are favored to produce more PLMS (products with large molecular sizes) [7,12].

IM-5 zeolite (IMF code), first synthesized by Benazzi et al. in 1998 [15], featured a unique 2D/3D pore system comprising intersecting 10-membered ring channels (0.48–0.55 nm), which had larger sizes than ZSM-5 zeolite and larger cavities (~2.5 nm) at the channel intersections [16,17,18,19,20]. It also possessed a high total acid amount, Brønsted acid amount, and proportion of strong acids [21,22,23]. These properties can enhance the reaction depth, shorten the mass transfer path, and reduce diffusion resistance, demonstrating significant application potential in the study of catalytic reactions [24,25,26], e.g., pseudocumene (1,2,4-trimethylbenzene) alkylation with methanol for durene (1,2,4,5-tetramethylbenzene). In our previous studies [27,28], compared with conventional ZSM-5 zeolite, IM-5 zeolite improved the conversion of pseudocumene and the yields of durene and tetramethylbenzene.

However, due to its large cavity size (2.5 nm) at the pore intersections [29], the IM-5 zeolite deactivated rapidly in many catalytic reactions [30,31,32], especially reactions with alcohols such as methanol and ethanol as feedstock [33,34,35,36,37]. The deactivation of zeolite catalysts mainly involved the coking (reversible) and structural changes (generally irreversible) occurring during catalytic reactions. Coking can be further categorized into acid site coverage and pore channel blockage, which led to a gradual loss in catalyst activity and shape selectivity [38]. Structural changes included poisoning and dealumination; the former primarily referred to the destruction of active sites by poisons such as bases, heavy metals, and sulfur-containing and halogen-containing substances, while the latter involved the removal of framework Al at high temperatures during the reaction (especially with H₂O involved) [39,40]. In this reaction, the deactivation of IM-5 zeolite catalyst can only be caused by coking and dealumination (H₂O originating from the dehydration of methanol). Extensive studies [33,35,36,37] have shown that in the catalytic conversion of methanol, the intense consecutive MTH (methanol-to-hydrocarbons) reactions within the pore channels of IM-5 zeolite led to the continuous accumulation of alkylaromatics (such as pentamethylbenzene). These precursors further underwent polymerization and dehygenation to form coke deposits. However, as documented in the study by Deependra Parmar et al. [37], the deactivation of zeolite catalysts in methanol-related reactions was inherently associated with the properties of coke deposits and the topological features of zeolite pore architectures, rather than the cumulative coke amount being the determinant factor. On the other hand, in steam catalytic cracking reactions, the participation of H₂O caused the dealumination of IM-5 zeolite, leading to structural changes and, thus, irreversible deactivation [30].

In summary, IM-5 zeolite was expected to significantly enhance the yield of the target product, durene, with decreased formation of smaller and larger moleculars, resulting in a product distribution dominated by tetramethylbenzenes. However, its rapid deactivation limited its potential long-term application. Therefore, in this work, the physicochemical properties—phase, morphology, textural properties, and acid properties—of fresh IM-5 zeolite and deactivated IM-5 zeolite after 60 h were analyzed using different characterization methods, such as XRD, XRF, SEM, ²⁹Si MAS NMR, ²⁷Al MAS NMR, OH-IR, Py-IR, NH₃-TPD, and thermal analysis. The coke properties and composition on the deactivated zeolites were investigated using FT-IR, XPS, and TG-MS. After pretreatment, the components of different types of carbon deposits were further determined in detail via FT-ICR MS and ¹³C (CR) MAS NMR. A comparative study was conducted with ZSM-5 zeolite throughout the process. This paper reported for the first time the deactivation analysis of the catalyst for pseudocumene alkylation with methanol, verified the deactivation behavior of IM-5 zeolite, and provided certain guiding significance for the pore structure modulation and acidity modulation of catalysts for the pseudocumene alkylation reaction with methanol.

2. Materials and Methods

2.1. Materials, Method of Preparation of Protonic-Form Zeolites, and Catalyst Testing

The raw materials used in the experiments, the method of preparation of protonic-form zeolites, and catalyst testing can be found in the Supplementary Materials.

2.2. The Characterization Methods of Deactivated and Fresh Zeolites

After separating the deactivated zeolites after 60 h of reaction from the inert quartz sand packed together in the reaction tube, the deactivated zeolites and fresh zeolites were subjected to characterization experiments of their physicochemical properties, including phase, morphology, textural properties, and acid properties. The powder X-ray diffraction (XRD) patterns of the samples were recorded on a D/MAX-IIIA X-ray diffractometer (Rigaku Corporation, Tokyo, Japan) using mono-chromatized Cu Kα radiation (35 kV, 35 mA). The samples were scanned from 5° to 50° at intervals of 0.013°, with a scanning duration of 30 s for each angle. The bulk chemical composition of the samples was characterized by X-ray fluorescence (XRF) measurement (Rigaku ZSX100E, Rigaku Corporation, Tokyo, Japan) running at 40 kV. An FEI Quanta 200F electron microscope (FEI Company, Hillsboro, OR, USA) operating at 20 kV was used to record the scanning electron microscopy (SEM) images. The temperature-programmed desorption curves of ammonia (NH₃-TPD) were recorded on an AutoChem II 2920 chemisorption analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA), where the desorption temperature was 600 °C. For the detailed schematic diagram of the NH₃-TPD, please refer to Figure S6. Thermal analysis experiments were carried out using a Netzsch STA 409CD analyzer (NETZSCH-Gerätebau GmbH, Bavaria, Germany) in a flow of air or argon (100 mL/min) at 10 °C/min from 30 to 1200 °C. It should be particularly noted that before conducting NH₃-TPD, Py-IR, and OH-IR, all deactivated zeolite samples must be calcined at 600 °C for 2 h to remove the partial carbon deposits, so as to prevent the partial volatilization of carbon deposits during the experiment heating process from blocking the instruments.

The ²⁹Si and ²⁷Al MAS NMR experiments were carried out on a Bruker AVANCE III 500 WB spectrometer equipped with a 7.0 mm probe at 11.7 T (Bruker BioSpin GmbH, Ettlingen, Germany) and a Bruker AVANCE III 600 WB spectrometer equipped with a 4.0 mm probe at 14.1 T (Bruker BioSpin GmbH, Ettlingen, Germany), respectively. The resonance frequencies for ²⁹Si and ²⁷Al were 99.2 and 156.4 MHz, respectively. The magic-angle spinning (MAS) speed for ²⁹Si and ²⁷Al was 5 and 8 kHz, respectively. The chemical shifts of ²⁹Si and ²⁷Al were referenced to kaolin and 1 M Al(NO₃) ₃ solutions, respectively. Both quantitative tests for the NMR were set with a 1 s recycle delay and 3000 scans. The N₂ adsorption–desorption isotherms and BJH desorption pore size distribution curves were recorded on a Micromeritics ASAP 2460 (Micromeritics Instrument Corporation, GA, USA). The total specific surface area of all the samples was calculated using the BET (Brunauer–Emmett–Teller) method, and the microporous area, external surface area, and microporous volume were calculated using the t-Plot method. The total pore volume was defined as the single-point adsorption total pore volume of pores with a diameter of less than 92.5480 nm at P/P₀ = 0.98, while the mesoporous volume was the difference between the total pore volume and the microporous volume.

2.3. Comprehensive Analytical Method for the Composition and Properties of Coke Deposition on Deactivated IM-5 Zeolite

Before describing the infrared characterization methods, a special clarification is needed here regarding the distinctions among FT-IR, Py-IR, and OH-IR. FT-IR (full spectrum) provided information on the overall chemical bonds and functional groups of the catalyst (such as the framework structure, coke deposits, and adsorbed species). OH-IR specifically detected the “acidic defects” of zeolites—the types of hydroxyl groups (particularly Si-OH-Al related to Brønsted acidity). Py-IR, on the other hand, used pyridine as a probe molecule to specifically distinguish and quantify the number and strength of Brønsted and Lewis acid sites.

All infrared (IR) characterizations were performed using a BIO-RAD FTS 3000 Fourier Transform Infrared spectrometer (Bio-Rad Laboratories Inc., Hercules, CA, USA).

• Sample Pretreatment: Proton-form zeolite samples, previously calcined at 550 °C to remove organic and ammonia molecules, and deactivated zeolites were irradiated with an infrared heating lamp (Shanghai Thermal Lighting Co., Ltd., Shanghai, China) for 3 min to desorb adsorbed moisture before pellet preparation. Particularly, before performing FT-IR, the zeolite powder should be thoroughly mixed with chromatographic-grade KBr. Subsequently, the samples were pressed into pellets using a 769YP-24B manual hydraulic press (Tianjin Keqi High-Tech Co., Ltd., Tianjin, China) at 20 MPa (12 tf) for 2 min. The pellet thickness (l) in the Lambert–Beer law was maintained constant at 0.5 mm by ensuring consistent mold dimensions and zeolite powder mass.
• Py-IR: Pellets were sealed in an in situ cell within the IR spectrometer, evacuated to 10⁻³ Pa at 400 °C for 60 min, then cooled to room temperature. A background spectrum was collected (32 scans, resolution 32 cm⁻¹, absorbance format, and wavenumber range 1700–1300 cm⁻¹ with automatic atmospheric background subtraction). After closing the evacuation valve, pyridine was introduced into the cell at 2.67 Pa for 3 min, followed by 17 min of equilibrium adsorption. The cell was then evacuated again to 1 × 10⁻³ Pa, heated to 200 °C for 90 min, cooled to room temperature, and scanned using the background file. Spectra were recorded at 350 °C (60 min hold time for higher temperatures). The pyridine adsorption quantities were calculated using the Lambert–Beer law as follows:

$$C={\displaystyle \frac{A}{\epsilon \times l\times m}}\times {10}^6$$

(1)

where C is the concentration of target acid sites (μmol/g), A is the integrated peak area after background subtraction (cm⁻¹·cm²), ε is the molar extinction coefficient (15,000 L·mol⁻¹·cm⁻¹ for Brønsted acid sites; 12,000 L·mol⁻¹·cm⁻¹ for Lewis acid sites), l is the sample thickness (cm), and m is the sample mass (g).

• FT-IR: Pellets were evacuated to 10⁻³ Pa at room temperature. Background spectra were collected (32 scans, resolution 32 cm⁻¹, absorbance format, and wavenumber range 4000–400 cm⁻¹ with automatic atmospheric background subtraction). The full spectrum was recorded under identical conditions using the background file.
• OH-IR: Pellets were evacuated to 10⁻³ Pa at 400 °C for 60 min, cooled to room temperature, and the background spectra were collected (32 scans, resolution 32 cm⁻¹, absorbance format, and wavenumber range 4000–3000 cm⁻¹ with automatic atmospheric background subtraction). The hydroxyl spectra were recorded using the same parameters with the background file.

The Agilent 7890 A-MSD 5975C mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) was employed for both the TG-MS and GC-MS coupled techniques. It featured a mass range of 10–1050 m/z, mass stability of ±0.10 μ/24 h, and a maximum scan rate of 20,000 μ/s.

2.4. The Analytic Methods of Deactivated IM-5 Zeolite’s External Surface, Internal Surface, and Insoluble Coke

The X-ray photoelectron spectroscopy (XPS) measurements were conducted on a Thermo Fisher VGESCALAB250 spectrometer (Thermo Fisher Scientific, Waltham, Massachusetts, USA) using a monochromatic Al Kα X-ray source operating at 150 W. Charge correction was applied by calibrating against the C 1s peak at 284.8 eV of contaminant carbon.

Prior to conducting GC-MS, FT-ICR, and ¹³C (CP) MAS NMR experiments, the deactivated zeolite samples must undergo pretreatment. The carbon deposits on the deactivated zeolite samples were categorized into three types: external surface organic adsorbates, internal surface (within the zeolite pores) organic adsorbates, and insoluble carbon deposits.

(1)

Extraction of External Surface Organic Adsorbates (Ultrasonic Extraction Method): The deactivated catalyst sample was immersed in toluene (extraction solvent) for 30 min, then transferred to a CREST 230D ultrasonic generator (Crest Ultrasonics, Ewing Township, NJ, USA) for 60 min of oscillatory extraction. The mixture of the extracted sample and solvent was filtered, and the filtrate was concentrated using an IKA rotary evaporator (IKA-Werke GmbH & Co. KG, Staufen im Breisgau, Germany);

(2)

Extraction of Internal Surface Organic Adsorbates (Within Zeolite Pores): A total of 10 g of the deactivated catalyst, pre-extracted for external surface adsorbates, was placed in a PTFE beaker. Then, 15–20 mL of hydrofluoric acid was added to dissolve the catalyst framework, releasing the carbon deposits within the pores. The acid solution was neutralized with sodium carbonate until gas evolution ceased. The resulting mixture was washed and filtered and the collected filter cake was dried at a low temperature. The dried cake was transferred to a conical flask, 20 mL of toluene was added, and ultrasonic extraction was performed. The extract was filtered to remove residues, and the filtrate was concentrated to 1–2 mL using the IKA rotary evaporator;

(3)

Insoluble Carbon Deposits: The filter cake remaining after internal surface extraction was dried at a low temperature (120 °C) to obtain the insoluble carbon deposits.

The carbon deposits were analyzed using a Bruker solariX XR Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS, Bruker Corporation, Bremen, Germany). The coke sample, dissolved and extracted in an organic solvent, was introduced into a metal-coated quartz capillary at 2000 V and nebulized at a flow rate of 360 μL/h. In the recorded mass spectra, mass-to-charge ratio (m/z) peaks with relative abundances exceeding 5.5 times the standard deviation of the root mean square (RMS) baseline noise (5.5 σ) were selected for molecular assignment. Chemical formulas (C_cH_hN_nO_o) were calculated from the m/z values with a relative error of <1 ppm. Double bond equivalent (DBE) values were computed according to Equation (2).

$$\mathrm{D}\mathrm{B}\mathrm{E}=c-{\displaystyle \frac{h}{2}}+{\displaystyle \frac{n}{2}}+1$$

(2)

where c, h, and n represent the number of carbon, hydrogen, and nitrogen atoms in the chemical formula, respectively.

The ¹³C (CP) MAS NMR experiments were performed on a Bruker AVANCE III 600 WB spectrometer (Bruker BioSpin GmbH, Ettlingen, Germany) fitted with a 4 mm probe at a magnetic field of 14.1 T. The resonance frequency was set at 150.9 MHz, with a magic-angle spinning (MAS) speed of 8 kHz. The acquisition parameters included an acquisition time of 0.02346 s; a pulse delay of 3 s; and 16,400 accumulated scans, and the chemical shifts were referenced to tetramethylsilane (TMS).

3. Results

3.1. Catalytic Test

When the molar ratio of n(1,2,4-TMB) to n(ME) was 2:1, within 60 h of reaction, the pseudocumene conversion of IM-5 zeolite was relatively high, 5–31% higher than that of ZSM-5 zeolite (Figure 1a). The selectivity of durene (Figure 1b) and tetramethylbenzene (Figure 1c) were also 5–40% and 15–60% higher than that of ZSM-5 zeolite. Overall, as shown in Figure 1d, the durene yield over IM-5 zeolite was 10% higher than that of ZSM-5 zeolite.

As shown in Figure 1e, when TOS = 30 h, in contrast to the products on ZSM-5 zeolite with non-aromatic hydrocarbons, BTX (benzene, toluene, and xylene), and trimethylbenzene isomers (84%) as main products, the products on IM-5 zeolite were mainly tetramethylbenzene (70%), and the distribution of C10+ aromatics on IM-5 zeolite was also higher (2.7% vs. 0.5%). This indicated that the reaction pathways of pseudocumene alkylation with methanol might be different for IM-5 and ZSM-5 zeolites. In our previous study [27,28], the IM-5 zeolite promoted the improvement in pseudocumene conversion and the generation of larger alkylaromatics due to its larger pore volume and Brønsted acid strength, therefore it was proposed that methanol was transformed through the methylation of pseudocumene. However, the ZSM-5 zeolite was mainly dominated by the separated MTH, disproportionation, and the isomerization of pseudocumene.

It was worth noting that the IM-5 zeolite deactivated rapidly. The conversion of pseudocumene (Figure 1a) and methanol (Figure 1f) decreased significantly with the progress, and the increasing rates of durene (Figure 1b) and tetramethylbenzene (Figure 1c) selectivity gradually decreased. Even the durene yield decreased after reaching the maximum (13%) at TOS = 43 h (Figure 1d). This was consistent with the research conclusion of Zhang et al., in which the IM-5 zeolite deactivated rapidly within 3 h in the MTP reaction [41].

With the n(1,2,4-TMB): n(ME) to 1:2, although the pseudocumene conversion (Figure 2a) and durene yield (Figure 2d) increased significantly, IM-5 zeolite exhibited rapid deactivation. The selectivity of durene (Figure 2b) and tetramethylbenzene (Figure 2c), durene yield (Figure 2d), and durene selectivity in tetramethylbenzene (Figure 2g) reached distinct peaks at TOS = 12–15 h, followed by a precipitous decline. Concurrently, the time required for complete methanol conversion was substantially reduced. For IM-5 zeolite, the product distribution shifted toward non-aromatics and C10+ aromatics, which not only indicated the progressive deactivation of IM-5, but also highlighted the extensive formation and accumulation of polyalkylaromatics. Notably, ZSM-5 zeolite maintained a relatively stable activity profile throughout the reaction.

A comprehensive analysis of the catalytic reaction results under different molar ratios of n(1,2,4-TMB) to n(ME) revealed that when the molar ratio of n(1,2,4-TMB) to n(ME) decreased from 2:1 to 1:2, the duration for IM-5 zeolite to maintain high activity and selectivity was shortened by 25 h, while the selectivity toward polyalkylbenzenes was increased by 8%. When the n(1,2,4-TMB): n(ME) = 1:2, the cmplete deactivation of IM-5 zeolite was observed at TOS = 30 h. In contrast, the conversion and selectivity of ZSM-5 zeolite remained low, but stable.

3.2. Chracterization of the Fresh and Deactivated Zeolites

To investigate the changes in the physicochemical properties of IM-5 and ZSM-5 zeolites before and after the reaction, using IM-5 zeolite and ZSM-5 zeolite before and after 60 h under the condition of a pseudocumene-to-methanol molar ratio of 2:1 as the research objects, the physicochemical properties of the zeolite samples before and after the reaction were investigated, including phase, elemental composition, morphology, textural properties, and acid properties.

3.2.1. Phase

The diffraction peaks observed at 2θ = 7.65°, 8.87°, 9.28°, 12.40°, 15.46°, 18.67°, 23.11°, 23.42°, 24.18°, 25.05°, 26.63°, 28.88°, and 31.10° of both the IM-5 and IM-5-60 h samples indicated a variety of typical IMF topological structures (Figure 3a) [42,43]. The XRD pattern of the ZSM-5 zeolites is shown in Figure 3b. The characteristic diffraction peaks at 2θ = 7.9°, 8.8°, 23.4°, 23.9°, and 24.4° indicated that the both ZSM-5 and ZSM-5-60 h were pure-phase zeolites with an MFI topological structure [44]. The intensity of the XRD peaks of the deactivated zeolites were all weakened, among which the IM-5-60 h sample had the lowest relative crystallinity (63.8%).

These observations suggest that carbonaceous species deposited within zeolitic pores, cages, or crystal surfaces absorbed and scattered incident X-rays. Furthermore, the infiltration of amorphous coke into the originally ordered crystalline pore networks diminished the periodic contrast of electron density distributions within the zeolite framework, thereby attenuating Bragg diffraction intensities. Notably, under elevated temperatures, water molecules generated during the MTH side-reaction facilitated zeolite dealumination, directly compromising the structural integrity and periodicity of the crystalline lattice to form amorphous phases, which collectively resulted in reduced overall crystallinity. Among the samples investigated, the IM-5 zeolite exhibited the most pronounced structural degradation.

3.2.2. Chemical Composition

The chemical element analyses using XRF of fresh and deactivated zeolites are shown in Table S1. The results indicated that there was no significant change in the overall silicon-to-aluminum ratio of the two zeolites before and after 60 h.

The weight loss of the catalyst can be divided into three stages [45,46]: low-temperature stage (<180 °C)—desorption of moisture; medium-temperature stage (200–400 °C)—weight loss corresponding to the combustion of low-molecular-weight carbon deposition precursors or relatively light carbon deposits; and the high-temperature stage (400–1000 °C)—combustion of high-molecular-weight carbon deposits (such as coke).

Thermogravimetric analysis revealed that the IM-5 zeolite had high-molecular-weight carbon deposits after 60 h (Figure 4).

3.2.3. Morphology

The SEM results showed that after 60 h, both zeolites exhibited agglomeration to varying degrees and irregular shapes, lacking the sharp-edged morphology, distinct from the fresh zeolites (Figure S1).

3.2.4. Textural Properties

As shown in Figure 5, after 60 h, the N₂ adsorption capacity of IM-5 zeolite decreased more significantly than that of ZSM-5 zeolite, indicating that the reduction in the total specific surface area of IM-5 zeolite was more pronounced than that of ZSM-5 zeolite. As shown in

Table 1. The textural properties of the samples.

Samples	SBET a/ (m2·g−1)	Smicro b/ (m2·g−1)	Sext b/ (m2·g−1)	Vtotal c/ (cm3·g−1)	Vmicro b/ (cm3·g−1)	Vmeso d/ (cm3·g−1)	D de e/ (nm)
IM-5	372	332	40	0.32	0.15	0.17	3.4
IM-5-60 h	77	15	62	0.21	0.005	0.205	10.75
ZSM-5	377	354	23	0.23	0.16	0.07	2.41
ZSM-5-60 h	250	234	16	0.16	0.11	0.05	2.63

^a. BET Surface Area; ^b. t-Plot; ^c. single-point adsorption total pore volume of pores less than 92.5480 nm diameter at P/Po = 0.98; ^d. V_total − V_micro; and ^e. adsorption average pore width (4V/A by BET).

, at 60 h, the total specific surface area of IM-5 zeolite decreased by 80%, the microporous area decreased by 95%, the total pore volume decreased by 34%, and the microporous volume decreased by 97%. Correspondingly, the decreases for ZSM-5 zeolite were 34, 35, 30, and 31%, respectively. Notably, the external surface area, mesoporous volume, and desorption average pore width of IM-5 zeolite increased significantly, which was mainly attributed to the fluffy carbon deposits covering the external surfaces of the IM-5 zeolite.

3.2.5. ²⁷Al MAS NMR and ²⁹Si MAS NMR

As shown in Figure 6a, after 60 h, the extra framework Al species in hexa-coordinated (with the chemical shift at −1 ppm) of IM-5 zeolite almost disappeared. The resonance peak at the chemical shift of 51 ppm (referring to framework Al species in tetra-coordinated) decreased in intensity and broadened. This observation indicated that the disordered distribution of framework aluminum in the zeolite became more pronounced, accompanied by an increase in structural defects [47]. For the IM-5-60 h sample after the reaction, the peak originally at 55.47 ppm shifted to 51.23 ppm, indicating that the shielding effect of the nuclei of some framework Al is weakened, i.e., the electron cloud density around the aluminum atoms decreased, thus leading to a change in the coordination environment. This usually occurred because the condensed-ring aromatics (precursors of carbon deposition) generated during the reaction were adsorbed on the acidic sites, affecting the local electric field of the adjacent framework aluminum through steric hindrance or electronic effects; or carbon deposition caused the pore channels to be blocked, resulting in local framework distortion and changing the Al-O-Si bond angles or bond lengths [48].

Figure 6b,c and

Table 2. The distribution of the chemical environment of silicon framework.

Samples	Si(4Si, 0Al)/%	Si(3Si, 1Al)/%	Si(2Si, 2Al)/%	Si(1Si, 3Al)/%
IM-5	72.36	12.57	14.10	0.97
IM-5-60 h	77.05	15.02	6.71	1.21

indicated that after 60 h of reaction, the Si(2Si, 2Al)—with the chemical shift of −103 ppm—decreased in IM-5 zeolite, while the Si(3Si, 1Al)—with the chemical shift of −106 ppm—increased, whereas the Si(4Si, 0Al)—with the chemical shift of − 112 ppm—and Si(1Si, 4Al)—with the chemical shift of − 97 ppm—showed little change. Calculations indicated that the SAR_FW of IM-5 zeolite increased slightly after the reaction (22.48 vs. 18.56).

During the methylation reaction of pseudocumene with methanol, the MTH (methanol—to—hydrocarbons) reaction first generated dimethyl ether through the dehydration of methanol, and then participated in the dual-cycle mechanism to generate olefins and aromatics. As mentioned earlier [30], the generated water caused the dealumination of the silica–alumina zeolite under high-temperature conditions, thus reducing the peak intensity of the framework aluminum with the coordination form of Si (2Si, 2Al). These observations were highly consistent with the XRD characterization results, further corroborating the structural insights derived from diffraction analysis.

3.2.6. Acidity

From the OH-IR spectra (Figure 7), it can be seen that the absorption peak intensities of both zeolites decreased after 60 h of reaction. After 60 h of reaction, the Si-OH-Al groups (with the wavenumber of 3610 cm⁻¹, representing Brønsted acid sites) and Al-OH extra-framework groups (with the wavenumber of 3666 cm⁻¹, representing Lewis acid sites) of the IM-5 zeolite almost disappeared, while the absorption peak of Si-OH groups on external surfaces (with the wavenumber of 3745 cm⁻¹) increased slightly, and further observation using FT-IR was needed. These phenomena were consistent with the NMR. In contrast, the degree of decrease in each absorption peak of the ZSM-5 zeolite was smaller.

As illustrated in Figure 8 and

Table 3. Acid properties of the samples from NH₃-TPD characterization.

Samples	Acid Amounts/(μmol/g)
	Total	Weak	Strong
IM-5	1462.6	793.5	669.1
IM-5-60 h	435.3	311.7	123.6
ZSM-5	721.8	409.0	312.8
ZSM-5-60 h	433.9	286.0	147.9

, after 60 h, both zeolites showed a significant decrease in strong acid sites. The acid amounts of the IM-5 zeolite decreased, with the total acid amounts decreasing by 70%, among which the weak acid amounts decreased by 61% and the strong acid amounts decreased by 81%. The total acid amounts of the ZSM-5 zeolite decreased by 40%, including a 30% decrease in weak acid amounts and a 53% decrease in strong acid amounts. Compared with the ZSM-5 zeolite, the desorption temperatures of the desorption peaks of the IM-5 zeolite decreased more significantly. The decrease in the desorption temperature correlated with the reduction in zeolite acid strength, with the IM-5 zeolite exhibiting a more pronounced decrease in acid strength.

As depicted in

Table 4. Acid properties of the samples from Py-IR characterization.

Samples	Brønsted Acid Amounts/(μmol/g)		Lewis Acid Amounts/ (μmol/g)		B/L
	200 °C	350 °C	200 °C	350 °C	200 °C	350 °C
IM-5	151.30	115.20	43.90	23.80	3.40	4.83
IM-5-60 h	25.78	15.34	19.25	8.34	1.34	1.84
ZSM-5	92.84	58.16	39.23	11.87	2.37	4.90
ZSM-5-60 h	28.67	13.86	19.05	6.66	1.51	2.08

and Figure S2, after 60 h, the acid amounts of both zeolites decreased significantly, with the decrease in Brønsted acid amounts being more pronounced. Specifically, the B/L ratio of the IM-5 zeolite decreased substantially. At desorption temperatures of 200 °C and 350 °C, the B/L ratios decreased by 61% and 62%, respectively. In contrast, the ZSM-5 zeolite showed only 36% and 58% decreases under the same conditions. This was consistent with the characterization results of OH-IR.

All the conventional characterizations indicated that both zeolites showed the phenomena of coverage on acidic sites and the blockage of pores after the reaction. Between them, the phenomenon was more severe in the IM-5 zeolite than in the ZSM-5 zeolite. Even a slight dealumination occurred in the IM-5 zeolite, leading to the irreversible deactivation of activity. Furthermore, the compositions and properties of the overall and classified (internal surface, external surface, and insoluble) carbon deposits of the IM-5 zeolite were analyzed to explore the intrinsic factors for catalyst deactivation.

3.3. Comprehensive Analysis of Coke Properties and Composition on Deactivated IM-5 Zeolite

3.3.1. FT-IR Analysis

As presented in Figure 9, the O-H stretching vibration peak at 3435 cm⁻¹ and the deformation vibration peak at 1624 cm⁻¹ (surface hydroxyl groups or adsorbed water) of the IM-5 zeolite showed weakened peak intensities and broadened peak shapes after the reaction. This indicated that the surface hydroxyl groups may participate in the alkylation reaction (such as acting as proton donors to catalyze electrophilic substitution), or that the adsorbed water decreased due to the competitive adsorption of organic substances. This was consistent with the characterization results of OH-IR.

After the reaction, new peaks appeared, the peak intensities increased at 2966 cm⁻¹ (methyl C-H stretching vibration peak) and 2920 cm⁻¹ (methylene C-H stretching vibration peak), and the peak intensity at 1517 cm⁻¹ (aromatic ring C=C stretching vibration peak) decreased and the peak shape at 1388 cm⁻¹ (methyl C-H deformation) distorted. This implied that pseudocumene (the aromatic ring parent) and its alkylation products (or intermediates) were adsorbed on the zeolite, reflecting the traces substrates/products.

The characteristic vibration peaks of the zeolite framework (such as the strong absorption peak near 1000 cm⁻¹) showed peak shape splitting and decreased intensity after the reaction, suggesting that the long-term reaction (60 h) may cause slight damage to the zeolite framework (such as the loss of acidic sites and pore blockage). This confirmed the results of all conventional characterizations [46,49].

In summary, the infrared spectrum revealed that the alkylation reaction occurred on the IM-5 zeolite, and the alkyl/aromatic ring products remained after the reaction. Moreover, the long-term reaction led to changes in the surface hydroxyl groups, framework structure, and adsorption state of the zeolite, indirectly reflecting the structure–performance relationship during the catalytic process.

3.3.2. TG-MS Analysis

As indicated in Figure 10, molecules with mass-to-charge ratios of 43/44/45/46 represented the acetyl fragment (CH₃CO⁺), carbon dioxide, the carboxyl fragment (COOH⁺), and the methyl formate fragment (HCOOCH₃⁺), respectively [50]. Among them, three fragments were derived from the oxidation of unreacted methanol and its dehydration product, dimethyl ether.

In the temperature range of 400–700 °C, the oxidation of unreacted methanol and its dehydration product dimethyl ether was relatively intense, generating large amounts of fragments such as acetyl and carboxyl groups, as well as carbon dioxide. This may be because key reaction steps occur on the catalyst within this temperature range, such as the oxidative dehydrogenation of methanol and dimethyl ether at the active sites of the catalyst, leading to the massive formation of related fragment ions.

This proved that methanol not only acted as an alkylation reagent, but may have also undergone dehydration, polymerization, or reacted with aromatic hydrocarbons to generate oxygen-containing intermediates. Eventually, these were deposited as oxygen-containing carbon deposits, covering the active sites of the zeolite and leading to catalyst deactivation. This also indicated that such carbon deposit species had moderate stability (not the “hard carbon” that decomposed only at high temperatures), reflecting that the carbon deposits were mainly composed of oxygen-containing organic species (related to the participation of methanol in the reaction, as methanol introduces oxygen elements). This had reference significance for catalyst regeneration, and the regeneration temperature can be designed around this interval (400–700 °C), which was consistent with the characterization results of TGA (Figure 4). However, the ion current intensity signal did not completely disappear after 800 °C in the figure, suggesting that a small number of stable species may remain. Therefore, for deep regeneration (the removal of stubborn carbon deposits), higher temperatures (>1200 °C) need to be considered.

3.4. The Analysis of Deactivated IM-5 Zeolite’s External Surface, Internal Surface, and Insoluble Coke

3.4.1. XPS Analysis of Coke on the External Surface

As demonstrated in Figure 11 and Figure S3, and Table S2, four binding energy values of energy bands were presented at 284.3 eV, 284.8 eV, 285.8 eV, and 288.8 eV, which correspond to graphitized carbon, amorphous carbon composed of aliphatic hydrocarbons, C-O bonds, and aromatic carbon, respectively, on the external surface of the IM-5-60 h sample [51,52].

It was concluded that the surface coke on IM-5 zeolite predominantly consisted of high-saturation amorphous carbon (i.e., sp³-hybridized alkyl carbon). This was followed by graphitized carbon—sp²-hybridized layered conjugated six-membered rings with delocalized π-electron systems, structurally analogous to aromatic carbon—which represented the ultimate form of coke evolution and resisted combustion below 1200 °C (consistent with TG-MS findings). Additionally, fractions of C-O bonded carbon and high-binding-energy species (e.g., C=O, -COOH, or surface carbonates) were present. Collectively, these results showed that the post-reaction IM-5 zeolite surface adsorbed diverse organic species—primarily alkylates, alongside aromatics and oxygenates—reflecting the reaction complexity arising from concurrent main and side pathways.

3.4.2. FT ICR MS Analysis of Coke on the External Surface and Internal Surface

For the pretreated IM-5-60 h sample, no peaks other than the solvent peak were detected via GC-MS for carbon deposits on either the internal or external surface. This may be due to the fact that the carbon deposition compounds had too large a molecular weight and too high a boiling point, as shown in Figures S4 and S5.

Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) was frequently used to analyze non-volatile and poorly distinguishable hydrocarbon compounds, as these compounds featured low abundance, high boiling points, and complex structures [53].

There were four main organic species in coke: oxygen-free aromatics (HC) and aromatics containing 1–3 oxygen atoms (O1, O2, and O3). As can be seen in Figure 12a, the carbon deposits on the external surface of the deactivated IM-5 zeolite were predominantly O2-type aromatics (containing two oxygen atoms), while the carbon deposits on the internal surface were predominantly oxygen-free aromatics (HC). Figure 12b,c showed that the mass-to-charge ratio (m/z) of carbon deposits on the external surface of the IM-5 zeolite was concentrated between 500 and 600, and that on the internal surface, it was concentrated between 400 and 600. Moreover, the carbon deposition on the internal surface caused more severe blockage.

Figure 13 depicted bubble charts of carbon number, DBE (double bond equivalent), and relative amount distributions for coke deposited on the external and inthernal surfaces of IM-5 zeolite. For the carbon deposits on the external surface of IM-5 zeolite, the carbon number was mainly concentrated in the range of 36–40, and the double bond equivalent (DBE) was approximately 12. The carbon number of oxygen-free aromatics in the pore channels of IM-5 zeolite mainly ranged from 33 to 40, with DBE between 22 and 27. Additionally, the DBE increased monotonically with the carbon number, meaning that the molecular formulas of the carbon deposits in the pore channels of IM-5 zeolite can be expressed as C_33–40H_24–28.

3.4.3. ¹³C (CP) MAS NMR Analysis of the Insoluble Carbon Deposits

As outlined in Figure 14, the chemical shifts in the ranges of 0–50 ppm, 50–90 ppm, and 90–150 ppm corresponded to species such as saturated carbon, C-O, and sp² carbon (aromatic carbon, olefinic carbon, or graphitized carbon), respectively [54]. The sidebands with chemical shifts of 183 ppm and 78 ppm are spinning sidebands. This is because the chemical shift of 130 ppm was between 183 ppm and 78 ppm, with an interval of approximately 53 ppm, which conformed to the interval law of spinning sidebands; and 130 ppm was the main peak, 183 ppm was its +1st order sideband (on the high-frequency side), and 78 ppm was its -1st order sideband (on the low-frequency side), which conformed to the characteristic of the relatively weak intensity of spinning sidebands [55].

The existence of these species not only reflected the occurrence of the main alkylation reaction (aromatic carbon + alkyl carbon), but also revealed the existence of side reactions (oxidation, dehydrogenation, and coke deposition) and side products (oxygen-containing combined carbon + carbonyl carbon), which was consistent with the characterization results of XPS, providing a structural basis for the subsequent optimization of reaction conditions (suppressing side reactions and improving selectivity).

After the reaction, the carbon deposits on IM-5 zeolite primarily consisted of amorphous carbon (saturated carbon), graphitized carbon (aromatic carbon), and oxygenates. The external surface exhibited a higher degree of saturation, whereas the internal pore channels showed lower saturation and a greater amount of coke deposition. The presence of these species inversely confirmed the occurrence of the deep methylation of pseudocumene by methanol and incomplete methanol conversion over IM-5 zeolite. However, the substantial formation of graphitized carbon rendered it extremely challenging to regenerate the zeolite’s activity via conventional coke-burning methods.

4. Discussion

In summary, the increase in methanol concentration in the feed promoted the deepening of alkylation reactions on the IM-5 zeolite. This not only enhanced the selectivity of the target products durene and tetramethylbenzene, but also accelerated the formation of large quantities of polyalkylbenzenes, thereby leading to the accelerated deactivation of the IM-5 zeolite.

The XRD results indicated that no other crystalline phases were observed in the two zeolite samples after 60 h of reaction, but their relative crystallinity decreased, among which the IM-5-60 h sample had the lowest relative crystallinity. The TGA results showed that the amount of carbon deposition on the IM-5 zeolite after 60 h was nearly three times that on the ZSM-5 zeolite. SEM and N₂ physisorption further confirmed that more carbon deposits covered the pores and acidic sites of IM-5 zeolite, and that these carbon deposits were relatively loose on the external surface. The NMR results concluded that slight dealumination occurred in the reacted IM-5 zeolite. Final acidic characterization revealed that compared with ZSM-5 zeolite, IM-5 zeolite after 60 h exhibited a more pronounced reduction in both acid quantity and strength, particularly for strong Brønsted acid sites.

Hence, under mild conditions with the n(1,2,4-TMB): n(ME) of 2:1, after 60 h of reaction, the IM-5 zeolites exhibited a more significant decrease in crystallinity, more substantial carbon deposition, more pronounced reductions in microporous area and volume, and a greater decrease in the amount of strong Brønsted acid. The reacted IM-5 zeolite was covered with loose carbon deposits on its external surface, and slight dealumination occurred in the IM-5 zeolite during the reaction.

FT-IR analysis demonstrated that the carbon deposits on IM-5 zeolite consisted of polycyclic aromatic hydrocarbons appended with alkyl side chains. TG-MS analysis showed that there were soft carbons that could be considered for removal by coking in relevant temperature (400–600 °C) conditions to regenerate the IM-5 zeolite. Nevertheless, hard carbons (graphitized carbon) present at temperatures above 800 °C could not be removed.

XPS analysis indicated that the carbon deposits on the external surface of the deactivated IM-5 zeolite were predominantly amorphous carbon (with high saturation), while the amounts of graphitized carbon and oxygenates were formed simultaneously. It can be concluded via FT-ICR MS that the organic compounds adsorbed on the external surface of the zeolite had higher saturation and oxygen content, while the carbon deposits in the pore channels exhibited higher unsaturation. The results of ¹³C (CP) MAS NMR analysis indicated that the insoluble coking components in the reacted IM-5 zeolite were predominantly composed of saturated carbon, while also containing significant amounts of C-O, sp² carbon (including aromatic, olefinic, or graphitized carbon) and COO species.

This distinct distribution can be explained from the perspective of diffusion limitation; the external surface was directly exposed to the reaction atmosphere, allowing reactants (such as olefins and aromatic precursors) to freely approach with a high concentration and no mass transfer resistance. The high reactant concentration promoted continuous hydrogen transfer reactions, leading to stepwise dehydrogenation and the cyclization of small molecules to form polycyclic aromatic hydrocarbons (e.g., anthracene and phenanthrene). In contrast, the steric hindrance from the narrow pore channels on the internal surface hindered multi-molecular synergistic reactions (e.g., aromatization requiring multiple molecules), preventing the formation of polycyclic aromatic hydrocarbons. The pore size was smaller than the planar dimension of polycyclic aromatic hydrocarbons (e.g., naphthalene: 0.72 nm), forcing the carbon deposits to exist as linear unsaturated chain structures (such as polyolefins or long-chain alkyl aromatics). And from the acid properties of the internal and external surfaces, the number of acid sites on the external surface was poor and dispersed, lacking precise shape-selective catalysis, making side reactions (such as excessive dehydrogenation and polymerization) more likely to occur. The density of Brønsted acid sites inside the pore channels was much higher than that on the external surface. Trapped olefins underwent repeated oligomerization reactions at these strong acid sites, forming long-chain compounds.

5. Conclusions

In conclusion, the primary cause of the deactivation of the IM-5 zeolite was the coverage of acidic sites and pores by carbon deposits (conventional characterization of pore channels and acidity), along with the slight removal of aluminum.

The carbon deposits on IM-5 zeolite had a high molecular weight, consisting of polyalkylbenzenes with alkyl side chains (GC-MS and FT-IR). Among them, the external surface was dominated by amorphous carbon with high saturation, featuring a carbon number range of 36–40, an unsaturation degree of only 12, and containing O elements derived from unreacted methanol (XPS, FT-ICR MS, TG-MS, and ¹³C (CP) MAS NMR). The internal surface was primarily composed of unsaturated oxygen-free aromatics with a carbon number range of 33–40 and an unsaturation degree of 22–27, exhibiting a higher abundance (FT-ICR MS). The phenomenon of different carbon deposition distributions between the internal and external surfaces may be attributed to the unique pore structure of the IM-5 zeolite, with the fundamental cause lying in the synergistic effects of diffusion limitation, spatial confinement, and acid site distribution. Furthermore, due to the presence of a high proportion of graphitized carbon, coke-burning regeneration under conventional conditions may be insufficient to restore the activity of IM-5 zeolite.

IM-5 zeolite, with its larger cavity at the pore intersections, provided more space for methanol molecules to further polymerize into long-chain olefins, aggravating the deactivation process. The abundant acidic sites on the external surface promoted continuous hydrogen transfer reactions of reactant molecules, gradually generating macromolecular polycyclic aromatic hydrocarbons that covered the pore channels and acidic sites. Therefore, during the modification of zeolite pore channels and acid properties, it was feasible to consider loading metals to fill part of the cavities while providing new acidic sites. Additionally, external surface passivation can be adopted to weaken the activity of side reactions, enabling reactant molecules to efficiently and selectively produce target products within pore channels under certain diffusion limitations. Furthermore, with methanol (CH₃OH) as an alkylation reagent, it needed to be dehydrated first to form active intermediates (such as surface methoxy groups or dimethyl ether), which had a low energy barrier, and the reaction activity of methanol-to-hydrocarbons (MTH) was more intense, easily leading to the formation of long-chain olefins through a dual-cycle mechanism [56]. Therefore, replacing the alkylation reagent should be considered. Dimethyl carbonate, as a mild alkylation reagent, participates in the reaction via a methyl carbene intermediate (:CH2) pathway, which should be prioritized in future research [57].