Intensified shape selectivity and alkylation reaction for the two-step conversion of methanol aromatization to p-xylene

Tingjun Fu*, Ran Wang, Kun Ren, Liangliang Zhang, Zhong Li*

State Key Laboratory of Clean and Efficient Coal Utilization, College of Chemical Engineering and Technology, Taiyuan University of Technology, Taiyuan 030024, China

Keywords:

ABSTRACT

1. Introduction

As the most important xylene isomer, p-xylene (PX) is mainly used to produce terephthalic acid (PTA) and dimethyl terephthalate (DMT), which can be further reacted with ethylene glycol(EG)to produce polyethylene terephthalate(PET)[1–4].The growing demand of those polyester products results in the continuous growth of PX consumption in recent years[2,5].However,the synthesis of PX in chemical industry is mainly based on the catalytic reformation of naphtha and catalytic cracking of gasoline [6,7]. In view of the inadequate supply and frequent price fluctuations of crude oil, it is necessary to explore the new petroleum-free route for the synthesis of PX[8–10].PX can be obtained via the catalytic conversion of methanol to aromatics(MTA),which is regarded as a novel technology for the synthesis of the PX with favorable development foreground [11–13]. This is because methanol can be obtained by the conversion of coal, biomass and natural gas[14,15]. The realization of this technology not only meets the demands of the PX market and reduces the dependence on petroleum route,but also promotes the efficient and clean utilization of the non-petroleum carbon-based resources.

For the formation of aromatics from methanol, methanol is firstly dehydrated to form dimethylether under the action of acid catalysts,and further convert to light olefin,and then the light olefin is further reacted to form aromatics and other by-product through polymerization, hydrogen transfer, cyclization or dehydrogenation [16–18]. As a kind of zeolites with typical melt flow index (MFI) topology, ZSM-5 zeolites, with plentiful acidic active sites, specific pore structure and high thermal stability, have been regarded as the most suitable catalyst for MTA[19].The specific 10 membered rings with straight channel (0.53–0.56 nm) and sinusoidal channel (0.51–0.55 nm) give aromatics the characteristic shape selectivity[20,21].Especially,since their pores sizes are similar to the size of PX(～0.53 nm),but smaller than that of m-xylene(～0.61 nm) and o-xylene (～0.66 nm), the high selectivity of PX could be achieved over ZSM-5 catalysts. However, the selectivity of PX in xylene(X)usually is only 23%-24%in the methanol to aromatics reaction [22]. The reason for this is that many acid sites existed on the external surface of ZSM-5 catalysts [23], which can promote the occurrence of the secondary reactions such as the alkylation, dealkylation, isomerization of X, due to the lack of shape selectivity and the diffusion constraint at pore-openings.Therefore, to improve the selectivity of PX via inhibiting the secondary reactions, numerous studies including silicon deposition by chemical vapor/liquid deposition [24], epitaxial growth of silicalite-1 [25], impregnation with P, Mg, or B, have been implemented to cover the external acidic sites of ZSM-5 [26,27].

Although the above-mentioned microporous structure gives ZSM-5 zeolites the shape selectivity, the presence of such typical micropores limits the diffusion of the large product molecules.These product molecules could be trapped in the channels, and eventually converted into the coke in micropores,which can cover the active sites and block the zeolite channels, and result in rapid deactivation of the catalyst [28]. So, the catalytic stability decreases significantly.In order to solve these diffusion limitations,a mass of works have been done in the synthesis of catalysts,including mesoporous construction via alkali treatment[29],morphology and size control [30,31]. These modification methods effectively shorten the diffusion path and the residence time of aromatic products in micropores,which can slow down the formation of coke and improve the catalytic stability.

It is worth noting that from the perspective of reaction process,MTA reaction can be regarded as two steps:methanol to light olefin and aromatization of light olefin to aromatics.Almost all of the methanol would be consumed in advance in the process of methanol to light olefin and the amount of methanol involved in the light olefin aromatization process is obviously reduced. Due to the fact that deep alkylation of aromatics for the coke formation is inhibited during the aromatization of light olefin to aromatics,the reaction stability for the second step can be significantly improved[32]. Based on this understanding, our previous work found that when the as-synthesized catalysts were loaded via the dual-bed mode with 1 g high Si/Al ratio (220) catalyst in the upstream and 1 g low Si/Al ratio (30) catalyst in the downstream in one reactor,the catalytic life was increased to 205 h, which was significantly higher than 22 h of 2 g aromatic catalyst packed individually in the reactor[33].It is easy to understand that passivating the external acid sites of the aromatization catalyst used in aromatization of light olefin can significantly improve the selectivity of PX and catalytic stability.However,although the whole reaction stability was obviously increased based on the two-step conversion of methanol to aromatics,methanol is also consumed excessively in the process of methanol to light olefin.This inhibits the alkylation of aromatics,especially benzene and toluene,which reduces the xylene selectivity for the final products. So, we can conclude that appropriate amount of methanol involved in the light olefin aromatization process is the key to obtain the high PX selectivity and catalytic stability.

In this work, firstly, silica coating was used to passivate the external acid sites of aromatization catalyst used in aromatization of light olefin by chemical liquid deposition (CLD) method to improve the PX selectivity in the two-step conversion of methanol to aromatics. The relationship between the deposition amount of silicon on the external acid sites and the catalytic performance,especially the PX selectivity and reaction stability, was investigated.In view of the fact that the aromatization process in the second step was partly inhibited as methanol was consumed in advance in the upper methanol-to-light hydrocarbons catalyst layer, part of methanol was directly introduced into the lower aromatization catalyst layer to promote the alkylation process of benzene and toluene for further improving the selectivity of PX.In addition,we also found that an appropriate external acid density was needed for aromatization catalyst to strengthen the alkylation process and improve the selectivity of xylene under the conditions of methanol introduction.The final results show a new insight into the methanol to aromatics reaction,and it may provide theoretical guidance for improving the selectivity of PX and catalytic stability in MTA process.

2. Materials and Methods

2.1. Materials

The following chemical materials were used for the synthesis of ZSM-5 catalyst and the evaluation of their catalytic performance.Tetraethylorthosilicate (TEOS, ≥96% (mass), Tianjin Kemeiou Chemical Reagent Co., Ltd.), tetrapropylammonium hydroxide(TPAOH, 25.0% (mass) in water, Zhejiang Kente Catalysts Inc.),sodium hydroxide(NaOH,≥96%(mass),Tianjin Kemeiou Chemical Reagent Co., Ltd.), aluminum sulfate (Al(SO4)3?18H2O, 99.0%(mass), Tianjin Kemeiou Chemical Reagent Co., Ltd.), aluminum nitrate (Al(NO3)3?9H2O, CP, Tianjin Kemeiou Chemical Reagent Co., Ltd.), deionized water (H2O), ethanol (EtOH, ≥95% (mass),Tianjin Kemeiou Chemical Reagent Co., Ltd.), silica sol (SiO2, 30%(mass), Dezhou Jinghuo Technology Glass Co., Ltd.), ammonium chloride (NH4Cl, ≥99% (mass), Tianjin Kemeiou Chemical Reagent Co., Ltd.), zinc nitrate (Zn(NO3)2, ≥99% (mass), Sinopharm Chemical Reagent Co., Ltd.), n-hexane (C6H14, ≥97% (mass), Tianjin Fuyu Fine Chemical Co., Ltd.), acetonitrile (C2H3N, ≥99% (mass), Shanghai Yi En Chemical Technology Co., Ltd.), tert-butylamine ((CH3)CNH2, ≥99% (mass), Shanghai Yi En Chemical Technology Co.,Ltd.), methanol (CH3OH, AR, Tianjin Kemeiou Chemical Reagent Co., Ltd.).

2.2. Experimental

2.2.1. Preparation of HZSM-5 with high Si/Al ratio (440)

A nano-HZSM-5 with high Si/Al ratio (440) was used for the conversion of methanol to light olefin. The specific preparation process was as follows:TEOS was added to TPAOH to form a mixed solution, and then stirred at 80 °C for 24 h. Next, the deionized water, Al(NO3)3?9H2O and NaOH were sequentially added dropwise to the solution. The mole ratio of the initial gel was 60SiO2:0.136Al2O3: 15TPAOH: 5Na2O: 500H2O. After mechanical stirring at 250 r?min-1for 5.5 h, the gel was crystallized in stainless steel autoclave at 170 °C for 24 h. Then, the mixture was centrifuged and washed repetitively by the deionized water until the supernatant liquid turned neutral. The solid product was dried at 100°C for 12 h, and then Na-ZSM-5 was prepared by calcining at 550°C for 6 h in a muffle furnace under air flow to remove the organic template.H-form high Si/Al ratio ZSM-5 was obtained by threefold ion-exchange with 0.8 mol?L-1NH4Cl aqueous solution(ml/ms=20)at 80 °C for 3 h and calcining at 550 °C for 6 h in static air. The obtained final H-form ZSM-5 was named as Z440.

2.2.2. Preparation of Zn modified HZSM-5 with low Si/Al ratio (75)

Zn modified HZSM-5 with low Si/Al ratio (75) was used for the aromatization of light olefin. Seeding method was employed to prepare the parental HZSM-5. Firstly, the preparation process of S-1 seed was as follows: TEOS was added to the mixed solution of TPAOH and NaOH to form a mixture with a molar ratio of TPAOH: Na2O: SiO2: H2O: EtOH = 4.4: 0.1: 25: 1048: 100 and stirred for 24 h at ambient temperature.The mixture liquid was transferred to the hydrothermal crystallization kettle, and the crystallization was rotated at 100 °C for 24 h. The obtained milky suspension was S-1 seed solution.Then,for the synthesis of parental HZSM-5 with low Si/Al ratio of 75, NaOH and EtOH was dissolved in deionized water, and then Al(SO4)3?18H2O and silica sol were added to form the precursor solution with molar ratio of SiO2: Na2O: Al2O3: EtOH: H2O = 1: 0.12: 1/75: 0.35: 15. After stirring for 30 minutes at ambient temperature, the neutral S-1 seed was added to the solution. Importantly, the addition quantity of seeds was 5% (mass) calculated by SiO2content in seed sol being divided by the total SiO2content in starting gel. Continued to stir for 3 h,and then the gel was crystallized in stainless steel autoclave at 140°C for 3 d.After cooled to ambient temperature,the mixture was centrifuged and washed repetitively by the deionized water until the supernatant liquid turned neutral. The solid product was dried at 100 °C for 12 h, and then Na-ZSM-5 was prepared by calcining at 550 °C for 6 h in air. The subsequent ionexchange process was the same as described in Section 2.2.1.

For the synthesis of Zn modified E75, 4 g E75 was immersed into 4 ml, 0.124 mol?L-1zinc nitrate solution and ultrasonicated for 30 min. After still standing at ambient temperature for 3 h,the product was dried at 100 °C for 12 h, and then calcined in a muffle furnace at 500 °C for 4 h. The acquired sample was named as ZnE75.

2.2.3. Preparation of silicon modified ZnE75 catalyst

To get the Si/Zn/ZSM-5 catalyst,ZnE75 was first dissolved in the n-hexane aqueous solution (ml/ms= 20). After stirring evenly at ambient temperature, TEOS was added to the mixed solution by droping, and stirred continuously for 4 h. Then, the n-hexane in the mixed solution was removed by the rotary evaporator. The sample was dried at 60°C for 3 h,and then calcined in a muffle furnace at 500 °C for 4 h. The obtained silicon deposited ZSM-5 was named as zSi/ZnE75 (z = 3, 7, 14, 28), and the ‘‘z” referred to the amount of silicon deposited on the external surface of ZnE75.

2.3. Catalyst characterization

The phase compositions and crystallinity of the synthesized ZSM-5 zeolites were analyzed using a Rigaku D/Max 2500 with Cu Kα radiation under 40 kV and 30 mA in the range of 2θ=5°–50°with a 2θ step size of 0.01°.A JEM-2100F transmission electron microscopy (TEM) was adopted to characterize the morphologies of seed and the synthesized catalysts.Nitrogen physisorption isotherms of all ZSM-5 catalysts were analyzed using a Beishide 3H-2000PS2 instrument at –196 °C. BET and T-plot methods were used to calculate specific surface area and micropore volume.

The strength and amount of acid sites of the synthesized ZSM-5 catalysts were tested by temperature-programmed desorption of ammonia (NH3-TPD) carried out on a Micromeritics Autochem II 2920 analyzer. The NH3-TPD curve was recorded at 120–650 °C with a rate of 10 °C?min-1. The amount of Br?nsted or Lewis acid sites of the ZSM-5 catalysts were analyzed by the pyridine-IR spectra using Bruker Tensor II FTIR spectrometer. The acid density on the external surface of the synthesized ZSM-5 zeolites was determined by the ZDJ-5 potentiometric titrator. Firstly, the catalysts were calcined at 500 °C for 2 h, and then 50 mg catalyst was dispersed in 40 ml acetonitrile solution and thoroughly stirred for 3 h. The obtained suspension was titrated with 0.01613 mol?L-1tert-butylamine acetonitrile solution.

Thermogravimetric analysis (TGA) of the used catalysts after MTA reaction were carried out on the TGSDTQ600 under the condition of a constant flow of air (30 ml?min-1) from 25 to 800 °C at a rate of 283 °C?min-1. GC-MS analysis of the used catalysts was carried out on Agilent 8890-5977B gas chromatograph-mass spectrometer equipped with a HP-5MS UI capillary column. The temperature was kept at 40 °C for 15 min firstly, and then heated to 300 °C at a rate of 5 °C?min-1and kept for 15 min.

2.4. Catalytic texts

The two-step reaction process of methanol to aromatics via light hydrocarbons was evaluated in a fixed-bed apparatus with a stainless steel tube reactor of 12 mm (i.d.). Both the olefin catalyst Z440 and aromatization catalyst zSi/ZnE75 (z = 3, 7, 14, 28)were pressed and sieved to 150–180 μm, and then the required amount of catalysts were mixed with quartz sand with a mass ratio of 1:2. The specific packing process of catalysts was as follows:0.5 g ZnE75 or corresponding silicon-coated catalysts were packed in the lower layer and 0.3 g Z440 was packed in the upper layer,and the two layers of catalyst were separated by 0.5 cm quartz wool.In order to optimize the depth of alkylation of the lower catalyst layer, a pipeline with a diameter of 3 mm was buried on quartz cotton between the two layers of catalyst to introduce part of methanol into the lower layer.For highlighting the catalytic performance of the two-step conversion of methanol to aromatics,the individual 0.8 g aromatization catalyst was packed in the reactor.Before the reaction, the catalysts were treated at 450 °C for 30 min in a N2flow of 35 ml?min-1. Then, the reaction was carried out at 430°C,atmospheric pressure and liquid weight hourly space velocity (WHSV) of 4 h-1. During the reaction, the liquid products were condensed in a cold trap and collected regularly.The oil samples in the liquid products were analyzed by GC-2014C gas chromatograph. Water samples in the gas and liquid products were analyzed by GC-7820.The formulas to calculate the methanol conversion,liquid hydrocarbon yield,product selectivity and BTX yield were presented as follows:

where, Y is the yield of liquid hydrocarbons, Cris the conversion of methanol, Sjis the selectivity of product j, mr,inis the quality of oil phase products obtained at a certain stage and mlhis the quality of inlet methanol at this stage,njis the number of carbons in the product j.

3. Results and Discussion

3.1. Texture property of the catalysts

Fig.1(a) exhibited the XRD patterns of ZSM-5 zeolites with different SiO2deposition amount. Each of the samples presented obvious characteristic diffraction peaks at 2θ = 8°–10° and 22.5°–25°,indicating that the samples exhibited the typical MFI topology even after liquid phase deposition of SiO2[34]. Defined the crystallinity of ZnE75 as 100%, and the relative crystallinity of SiO2deposition samples was calculated by peak area.When the amount of SiO2in the deposition solution increased from 3%to 28%,due to the increase of amorphous SiO2deposition on the outer surface,the characteristic diffraction peaks of zeolites decreased gradually and the relative crystallinity decreased from 97% to 86%.

N2adsorption-desorption was measured to determine the pore structures of ZSM-5 zeolites with different SiO2deposition amount.As exhibited in Fig.1(b),different samples still had high N2adsorption capacity at p/p0< 0.05 and hysteresis loops appeared at p/p0= 0.8–1.0. All the samples showed characteristic IV-type isotherms, indicating that the pores of the all samples were mainly micropores and mesoporous structures also existed[10].As shown in Table 1, the micropore surface area and volume of ZnE75 were 335.0 m2?g-1and 0.15 cm3?g-1, respectively. With increasing the amount of SiO2deposition,some silicon species caused pore blockage, resulting in the decrease of the micropore surface area and volume from 309 m2?g-1and 0.14 cm3?g-1of 3Si/ZnE75 to 179 m2?g-1and 0.08 cm3?g-1of 28Si/ZnE75.

3.2. The morphology of the catalysts

Fig. 1. (a) XRD patterns and (b) N2 adsorption-desorption isotherms of different ZSM-5 zeolites.

The TEM images of the all ZSM-5 samples were shown in Fig.2.In this work,the used ZSM-5 zeolites were synthesized by the seed induction method with S-1 as the seed and ethanol as the structure directing agent. S-1 seeds were spherical with dispersed particles and the uniform crystal diameter of about 350 nm (Fig. 2(a)).Fig. 2(b) showed the TEM image of ZnE75, which exhibited coffin-like morphology, with the particle size about 1.3–1.5 μm.No Zn particles were observed on the crystal surface, indicating the good dispersion of Zn species [35]. The samples after liquid phase deposition of SiO2still remained coffin-like. But with the increase of the amount of SiO2deposition,the SiO2shell thickened gradually, indicating that the thickness of silicon deposition coating was successfully controlled by changing the amount of SiO2in the deposition solution.

3.3. Acid properties of the catalysts

The acid properties of Zn/E75 and the samples after SiO2deposition were analyzed by NH3-TPD. As shown in Fig. 3(a), the synthesized samples had two ammonia desorption peaks, among which the peaks at 150–320 °C and 320–550 °C correspond to ammonia desorption at weak and strong acids, respectively. The desorption curve was further divided into three desorption peaks by Gaussian fitting, which corresponded to weak, medium and strong acid,respectively.The acid density distribution and quantitative calculation results were shown in Fig. 3 and Table 2. After deposition of SiO2, the total acid density of the samples gradually decreased. Especially, as the amount of SiO2deposition increased to 28%, the strong acid density decreased significantly.

The acidic type of the samples was determined by Py-IR. As exhibited in Fig. 4(a), all samples showed two peaks at 1540 cm-1and 1450 cm-1, which were related to the vibration of pyridine molecules bonded to BAS and LAS, respectively [36].Table 2 showed that with the increase of the amount of SiO2deposition,the density of BAS and LAS decreased gradually.In addition,the external surface acid density of the samples was further tested by the potential titration method, with Tert-butylamine as probe molecule, and the result was shown in Fig. 4(b). The external surface acid density of ZnE75 was 0.1 mmol?g-1. With the increase of the amount of SiO2deposition, the external surface acid density decreased from 0.1 mmol?g-1of ZnE75 to 0.03 mmol?g-1of 28Si/ZnEZ5,which indicated that silica species deposited on the surface of zeolites could interact with the silanol groups, resulting in the passivation of the external surface acid. The XPS characterization result indicated that the Al species content on the external surface after SiO2deposition decreased and the SiO2/Al2O3increased relatively.

3.4. Catalytic Tests: Methanol to aromatics reaction

3.4.1. Reaction stability

The prepared silicon-coated catalyst were packed on the below of Z440 to catalyze the aromatization reaction of the light olefin at P = 0.1 MPa, T = 430 °C and WHSV = 4 h-1. The catalytic performance of different aromatization catalyst samples in two-step conversion of methanol to aromatics was shown in Fig.5.Within 35 h of the reaction,the methanol conversion of all samples remained at about 99.0%,and then decreased gradually with the progress of the reaction. The liquid hydrocarbon yield of all samples reached the highest at 10 h, and then began to decline. The catalytic life and the highest liquid hydrocarbon yield of ZnE75 were 62 h and 15.9%, respectively. After SiO2deposition, the catalytic life was maintained at about 50 h, slightly lower than the parent catalyst.However, with the increase of the amount of SiO2deposition, the total acid density decreased from 0.62 mmol?g-1of ZnE75 to 0.44 mmol?g-1of 28Si/ZnE75. Especially, the strong acid density decreased from 0.25 mmol?g-1to 0.12 mmol?g-1, which resulted in the highest liquid hydrocarbon yield decreasing to 12.3% of 28Si/ZnE75.

In order to reflect the characteristics of silicon deposition catalyst in the two-step conversion of methanol to aromatics, ZnE75,3Si/ZnE75, and 28Si/ZnE75 were used for the direct methanol aromatization process, and the catalytic performance was shown in Fig.5(b).With the progress of the reaction,parent catalyst deactivated rapidly with the highest liquid hydrocarbon yield of 14.2%,which was significantly lower than that of in the two-step reaction.This was because that more methanol could be involved in the aromatization process,which promoted alkylation reaction of aromatics in the lower layer to product the heavy aromatics,and ultimately promoted the rapid deactivation of the catalyst.Due to the micropores were blocked after the SiO2deposition (Table 2), the catalysts were more prone to coke deposition and the catalytic life was further shortened to 10 h.

Table 1 Textual properties of different ZSM-5 zeolites

Fig. 2. TEM images of (a) S-1, (b) ZnE75, (c) 3Si/ZnE75, (d) 7Si/ZnE75, (e) 14Si/ZnE75, (f) 28Si/ZnE75.

Fig. 3. (a) NH3-TPD profiles and (b) distribution of acid density with different strength of different ZSM-5 zeolites.

Table 2 Acid properties of ZSM-5 with different SiO2/Al2O3

The conversion of methanol to aromatics via light olefin showed excellent catalytic stability, which can also be proved by the TG results of the spent catalysts. As shown in Fig. 6, when methanol directly participated in the aromatization process,the coke content and coke formation rate of ZnE75 were 0.07 g?g-1and 3.14×10-3g?g-1?h-1, respectively. In contrast, for the two-step conversion of methanol to aromatics, the consumption in advance of methanol by the upper olefin catalyst Z440 inhibited the deep alkylation of aromatics in the lower layer,which resulted in the decrease of coke content and formation rate of ZnE75 from 0.07 g?g-1and 3.14×10-3g?g-1?h-1to 0.09 g?g-1and 1.49×10-3g?g-1?h-1,respectively. In addition, the coke content and coke formation rate increased to 0.11 g?g-1, 2.10×10-3g?g-1?h-1of 28Si/ZnE75. This indicated that SiO2deposition blocked the pore channels of the zeolites, which limited the diffusion of macromolecular product and promoted the generation of coke species.

3.4.2. Distribution of products

Fig. 4. (a) Py-IR spectra and (b) density of the external surface sites of different ZSM-5.

Fig. 5. Comparison of the stability of (a) methanol to aromatics via light olefin and (b) methanol to aromatics directly.

Fig. 6. (a) TG curves and (b) the rate of coke formation of different ZSM-5 after deactivation.

Fig.7 exhibited the product selectivity when different aromatization catalysts were packed in the two-step conversion of methanol to aromatics.The C2=-4and aromatics were the main products of the MTA reaction. The total aromatics selectivity decreased from 34% of ZnE75 to 21% of 28Si/ZnE75, which indicated that as the amount of SiO2deposition increased, the decrease of total acid density inhibited the aromatization of light olefin. Meanwhile,the selectivity of C1-4light alkanes decreased from 23.8% of ZnE75 to 18.7%of 28Si/ZnE75,while the selectivity of C2=-4light olefin increased from 32.2% to 50.6%. The specific distribution of aromatics products as illustrated in Fig. 7(b), toluene and xylenes were the mainly aromatics products. As the amount of SiO2deposition increased, the selectivity of toluene and xylenes decreased from 12.2%,13.3%of ZnE75 to 6.7%,8.3%of 28Si/ZnZ5,respectively.Meanwhile, PX/X increased from 34.6% of ZnE75 to 60.0% of 28Si/ZnE75. These results indicated that the silicon deposition covered the external surface acid sites of zeolites, and inhibited secondary reaction such as alkylation and isomerization of xylenes.

Fig. 7. (a) Product distribution and (b) aromatic product distribution of different ZSM-5 over MTA via light olefin.

When methanol directly involved in the aromatization process,the variation of product selectivity with the amount of SiO2deposition was shown in Fig. 8. The selectivity of aromatics and BTX decreased from 36.4%, 27.1% of ZnE75 to 30.0%, 22.4% of 28Si/ZnE75.Especially,the aromatic products of methanol direct aromatization were mainly xylene, and its selectivity was 16.9% over 28Si/ZnE75, which was significantly higher than the 5.1% of toluene. In contrast, toluene and xylene were the main aromatic products of methanol to aromatics via two-step and the selectivity of them were 6.7%and 8.3%over 28Si/ZnE75,respectively.In addition,we found that when 28Si/ZnE75 catalyzed the direct aromatization reaction of methanol,the PX/X was 78.3%,which was much higher than the 60% in the two-step method. This indicated when packed individual aromatization catalyst, methanol directly involved in the alkylation process,which accelerated the formation of coke,thus causing significant effects on the acid sites of the zeolite and improving the selectivity of xylene.

To study the change rule of the two-step conversion process of methanol to aromatics and product selectivity with reaction, individual 0.3 g Z440 was packed to catalytic the conversion of methanol to aromatics. As shown in Fig. 9(a), the methanol conversion reached 99%at 10th h,and then started to drop quickly.The highest liquid hydrocarbon yield was only 5.6%. The product distribution was shown in Fig. 9(b). During the reaction, the highest aromatics selectivity was only 14.0%. In comparison, the highest selectivity of the C2=-4light olefin and C1-4light alkanes reached 74% and 17%, respectively. This was because olefin catalyst Z440 had low acid density, which promoted the ‘‘a(chǎn)lkene-based cycle”and suppressed the aromatization reaction. When 0.5 g 28Si/ZnE75 aromatization catalyst was packed in the lower layer, the products formed in the upper layer further involved in the aromatization reaction,which increased the liquid hydrocarbon yield and aromatics selectivity to 12.4% and 20.8%, respectively. Further analysis of products selectivity at different stages of the reaction,we found that with the reaction from 5 h to 52 h, coke blocked the pores of the zeolites gradually, which limited the diffusion of reactants and products and reduced the accessibility of the acid sites,resulting in the decrease of liquid hydrocarbon yield and aromatics selectivity from 12.4%and 18.0%to 3.7%and 16.0%,respectively.Fig.9(d)exhibited the aromatics selectivity of 28Si/ZnE75 in the two-step conversion of methanol to aromatics.The proportion of toluene in aromatics decreased from 28.1% to 21.9%, while the proportion of xylene and PX/X increased from 42.4% and 54.6% to 47.6%and 77.7%,respectively.It further indicated that the passivation of the external acid sites by coke inhibited secondary reactions, such as alkylation and isomerization.

3.4.3. Optimization of product selectivity

Fig. 8. (a) Product distribution and (b) aromatic product distribution of different ZSM-5 in direct conversion of methanol to aromatics.

Fig. 9. (a) Conversion of methanol and liquid hydrocarbon yield and (b) product distribution with time on stream on Z440, (c) total product distribution and (d) aromatic product distribution of 28Si/ZnE75 over MTA via light olefin with time on stream.

Fig. 10. (a) Reaction mode, (b) product distribution, (c) aromatic product distribution and (d) PX/X of methanol aromatization reaction under different catalyst packing modes.

Fig. 11. (a) TG and (b) the rate of coke formation of different ZSM-5 after reaction 10 h.

Fig. 12. GC-MS chromatograms of the organic extracts from the spent 3Si/ZnE75 catalyst in different modes. Reaction condition: 430 °C, 0.1 MPa, 10 h.

For the selectivity of the product,toluene accounted for a large proportion in aromatic products in the conversion of methanol to aromatics via light olefin (Fig. 10 Mode 1). This was because the pre-consumption of methanol in upper layer inhibited the alkylation of aromatics in the lower layer.In order to optimize the reaction depth,part of methanol was directly introduced into the lower layer(Mode 2),as shown in Fig.10.Comparing the aromatic products of the two modes,it was found that methanol introduced into the lower layer improved the selectivity of total aromatics, and part of methanol was alkylated with toluene,which resulted in the decrease of the proportion of benzene and toluene in aromatics and the increase of the proportion of xylene and PX/X.For example,the proportion of benzene and toluene over 3Si/ZnE75 decreased from 3.7%and 37.7%in Mode 1 to 0.8%and 14.3%in Mode 2,while the proportion of xylene and PX/X increased from 40.0%and 51.5%to 55.3%and 55.5%.In addition,the alkylation reaction of methanol and toluene in the lower layer was significantly affected by the external acid density of the catalyst. When 28Si/ZnE75 was used as aromatization catalyst,the obvious decrease of the external acid density inhibited the alkylation reaction of aromatics,so the selectivity of xylene was almost constant. In contrast, due to the suitable external surface acid density, 3Si/ZnE75 was more beneficial to promote the generation of xylene.

The introduction of methanol into the lower layer would affect the coke formation process. The used catalysts that the reaction time was 10 h in the two modes were analyzed by thermogravimetry, as shown in Fig. 11. The coke deposition rate of 3Si/ZnE75 in Mode 2 was 8.7 × 10-3g?g-1?h-1, while that was only 2.7 × 10-3g?g-1?h-1in Mode 1. This indicated the introduction of methanol into the lower layer promoted the coke deposition rate. Meanwhile, the reaction depth of Mode 2 was affected by the catalyst acidity.28Si/ZnE75 had a lower external acid density,which inhibited the secondary reaction, such as alkylation of aromatics. The deep alkylation reaction on the external surface of 3Si/ZnE75 promoted the formation of coke deposition and resulted in the higher coke deposition rate.

Fig. 13. (a) NH3-TPD profiles and (b) N2 adsorption-desorption isotherms of different catalysts after reaction in two reaction modes.

Table 3 Acid properties of different catalysts after reaction in two reaction modes

Table 4 Textual properties of different catalysts after reaction in two reaction modes

The introduction of methanol into the lower layer would affect the types of the organic deposits.As shown in Fig.12,the obtained chemical composition of the organic compounds occluded in the catalysts after two different reaction modes was analyzed by GCMS. By comparing the distribution and the intensities of the GCMS signals of the organic species in spent 3Si/ZnE75 catalyst in different modes, it can be found that the GC-MS signals of xylene,methylbenzenes (trimethylbenzene, tetramethylbenzenes and pentamethylbenzene) of the spent 3Si/ZnE75 in Mode 2 were much higher than that of in Mode 1. Especially, the GC-MS signals of naphthalene was even not detected of the spent 3Si/ZnE75 in Mode 1. These results indicated the direct introduction of part of methanol into the lower layer promoted methanol to be involved in the alkylation reaction of aromatics directly. This deepened the alkylation of aromatics and accelerated the formation of xylene and macromolecular species, such as polymethylbenzene and naphthalene, which was consistent with the TGA results.

The acidity and texture property of catalysts after reaction could also reflect the difference of reaction depth between the two modes. More methanol could be involved in the alkylation reaction of aromatics after introducing part of methanol into the lower layer, which promoted the generation of coke and decrease of acid density because of the passivation of acid sites by coke species(Fig.13).As shown in Table 2 and Table 3,after the same reaction time, the total acid density of 3Si/ZnE75 in Mode 1 decreased more than that in Mode 2.Meanwhile,the deep alkylation reaction also had a significant effect on the texture property of the catalysts after introducing part of methanol into the lower layer.The specific surface area and pore volume of 3Si/ZnE75 after reaction in Mode 1 also decreased more than that in Mode 2 (Table 4).

4. Conclusions

In this work,to improve the PX selectivity of the two-step conversion of methanol to aromatics with high catalytic stability,silica coating was firstly used to passivate external acid sites of ZSM-5 catalyst for the aromatization of light hydrocarbons by the chemical liquid deposition method.It was found that with the increase of SiO2deposition, the total acid density decreased from 0.62 mmol?g-1of ZnE75 to 0.44 mmol?g-1of 28Si/ZnE75, which inhibited the aromatization reaction, resulting in the decrease of liquid hydrocarbon yield and aromatics selectivity from 15.6%and 34.0% to 10.8% and 20.8%, respectively. Meanwhile, the selectivity of toluene and xylene decreased from 12.2% and 13.3% of ZnE75 to 5.5% and 8.3% of 28Si/ZnE75. While, the reduction of external surface acid amount inhibited the secondary reactions,such as alkylation and isomerization, and increased PX/X from 34.6 % of ZnE75 to 60.0 % of 28Si/ZnE75. In view of the fact that toluene and xylene are the main aromatics products in the twostep conversion of methanol to aromatics, in order to improve the selectivity of PX,part of methanol was directly introduced into the lower layer,which increased the selectivity of xylene and PX/X from 40.0%and 51.5%to 55.3%and 55.5%,respectively.In addition,the alkylation depth of aromatic products was significantly affected by the external acid density of catalyst. The dramatic decrease of the external acid density of 28Si/ZnE75 inhibited the alkylation of aromatics and the xylene selectivity only increased from 39.9% to 41.8%. In contrast, due to the suitable external surface acid density, 3Si/ZnE75 was more beneficial to promote the generation of xylene and the xylene selectivity increased from 40.0% to 55.3%.

Data Availability

The data that has been used is confidential.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We thank the financial support from the National Natural Science Foundation of China (21978191 and 22278292), Key Research and Development Project of Shanxi Province (InternationalScienceandTechnologyCooperationProgram)(201803D421011).

Nomenclature