Highly selective BTX from catalytic fast pyrolysis of

International Journal of Biological Macromolecules 91 (2016) 278–293

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac

Highly selective BTX from catalytic fast pyrolysis of lignin over supported mesoporous silica A.M. Elfadly a , I.F. Zeid b , F.Z. Yehia a , A.M. Rabie a,c,∗ , M.M. aboualala a , Sang-Eon Park c,∗ a b c

Egyptian Petroleum Research Institute, Petrochemical department, Nasr City, Cairo 11727, Egypt Chemistry Department, Faculty of Science, Menofia University, Shibein El-Kom, Egypt Laboratory of Nano-Green Catalysis, Department of Chemistry, Inha University, Incheon, Republic of South Korea

a r t i c l e

i n f o

Article history: Received 24 March 2016 Received in revised form 11 May 2016 Accepted 14 May 2016 Available online 16 May 2016 Keywords: Lignin Fast pyrolysis BTX MCM-48

a b s t r a c t The post synthesis of Al3+ or Zr4+ substituted MCM-48 framework with controlled acidity is challenging because the functional groups exhibiting acidity often jeopardize the framework integrity. Herein, we report the post-synthesis of two hierarchically porous MCM-48 composed of either aluminum (Al3+ ) or zirconium (Zr4+ ) clusters with high throughput. All prepared catalysts have been characterized by HR-TEM, XRD, IR, N2 -adsorption, NH3 -TPD, TGA and MAS NMR. They exhibit BET surface areas of 597 and 1112 m2 g−1 for 8.4% Al/MCM-48 and 2.9% Zr/MCM-48, respectively. XRD analysis reveals that the hierarchical porosity of parental MCM-48 is reserved even after incorporation of Al3+ or Zr4+ . Zr/MCM-48 catalysts are demonstrate a superior performance versus that of Al/MCM-48 and MCM-48 because of the mild (ZrO2 ) or nil (SiO2 ) Lewis acidity contributed from Zr-␮2-O group as well as smaller pore sizes suitable for the restriction of unwanted side reactions. The reaction conditions which were affecting the catalytic pyrolysis and final products were gas flow rate, pyrolysis temperature, and catalyst to lignin ratio. A total of 49% of BTX product were obtained over 2.9% Zr/MCM-48 at 600 ◦ C. The Lewis acid character was the governing factor which helps in pyrolysis and directly affects the BTX formation. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The renewable sources for energy generation are considered to be one of most crucial task for scientists and engineers [1,2]. Studies on other resources like solar, tidal and wind for energy generation are still under infant states [3]. Biomass is the potential abundant candidate of renewable energy [4–6]. Amorphous branched and cross-linked lignin structure having biomass weight fraction of 15–33 wt% is the second most plentiful component in biomass [7] and account ≈40% of the biomass energy content. In recent, one of the main sources of lignin is the production of pulp and paper industry (50 million tons/year) and as a waste of lignocelluloses based on bioethanol industry. The challenge is about 2% of the lignin residue is used for manufacturing 106 tons/year of lignosulfonates and 106 tons/year of kraft lignin [8]. Consequently,

∗ Corresponding authors at: Egyptian Petroleum Research Institute, Petrochemical department, Nasr City, Cairo 11727, Egypt. E-mail addresses: [email protected], [email protected] (A.M. Rabie), [email protected] (S.-E. Park). http://dx.doi.org/10.1016/j.ijbiomac.2016.05.053 0141-8130/© 2016 Elsevier B.V. All rights reserved.

lignin considered as a potential source for many useful products likes catalysts and bio-oil. There are two approaches of pyrolysis, one is slow pyrolysis and the other is fast pyrolysis. The slow pyrolysis has been applied with low heating rate to enhance char production. However, slow pyrolysis has some disadvantages like taking long decomposition time, very low heat transfer and subsequently requires high energy consumption [9,10]. Hence, these make it less suitable for getting high quality of bio-oil production. On the other hand, in the fast pyrolysis process biomass is rapidly exposed to a high temperature reactor under inert condition. The major advantages of the fast pyrolysis process are; (1) rapid heat transfer (2) high heating rate (3) short decomposition time, and (4) fast cooling of vapors to furnish the high yield on bio-oil with good quality control at the end products [11]. Moreover, fast pyrolysis process needs a relatively low capital investments with high energy efficiencies compared to other processes, especially on a small scale [12]. Fast pyrolysis of lignin is a promising process for lignin utilization. The latest strategies of fast pyrolysis of lignin include; (1) catalytic fast pyrolysis using acidic silica-alumina zeoletic materials to produce high value aromatics [13,14] (2) hydrogen donor materials likes calcium formate/formic acid to produce oxygen free

A.M. Elfadly et al. / International Journal of Biological Macromolecules 91 (2016) 278–293

279

Fig. 2. N2 adsorption–desorption isotherms and pore size distribution curves of Zr/MCM-48 catalysts. Fig. 1. N2 adsorption–desorption isotherms and pore size distribution curves of Al/MCM-48 catalysts.

chemicals [15]. However, the economic challenge and deactivation of the catalysts is high barrier to commercialize the previous strategies. In addition, one of the drawbacks and challenge of lignin pyrolysis is the generation of unwanted oxygenates as by-products such as (phenol,2-methoxyphenol, etc.). The current process for benzene, toluene, and xylenes (BTX) production is mainly rely the petrochemical industry suffers from depletion of resources and some environmental problems such as high energy consumption and by-products. In this aspect, lignin could be one of the abundant, cheap resources for the sustainable production of BTX [16–18]. BTX is used as basic petrochemical feed stocks for the downstream applications and can be hydrogenated to alkanes or used in the preparation of other specialty chemicals [19]. Pyrolysis is used as one of direct and effective methods for biomass valorization [20]. Unlike I-D- or 2-D M41S channels such as HMS, MCM-41 and SBA-15, 3D mesoporous silica MCM-48 having a bicontinuous cubic ¯ meso-structure with mesopores and thick pore walls as well as Ia3d complementary pores linked two main channel systems has been known as good catalyst or support in catalysis [21–23]. Mochizuki et al. [24] reported that the catalytic fast pyrolysis is more efficient upon using medium porosity silica materials due to reducing the oxygenated compounds contents. According to Adam et al. [25] the use of aluminated MCM41 material eliminated formation of levoglucosan and decreased phenols production with increasing furans, acetic acid and hydrocarbons. Pattiya et al. [26] reported that the utilization of

aluminated MCM-41 and MSU-F materials increases the aromatic hydrocarbon while lignin-derived products decreased. Lu et al. [27] states that aluminated SBA-15 materials showed a similar catalytic to AlMCM-41, however the decreasing high aluminated material exhibited high aromatic yield. We focused on the optimization of metal loading on the MCM-48 support to improve catalyst stability towards coking and improve selectivity toward BTX materials. This approach was successfully applied by incorporation of Zr4+ and Al3+ into MCM-48 materials for the advancement of the biomass fast pyrolysis products. The effect of acidity and pore size of Al/MCM-48 and Zr/MCM-48 on the product distribution, especially BTX, of catalytic fast pyrolysis of lignin was investigated for comparison. In addition, extensive characterization by means of various techniques of both the fresh and spent samples allowed comparing the main catalysts physicochemical properties and to comment activity, stability and selectivity data.

2. Experimental Alkaline lignin, tetraethyl orthosilicate (TEOS), hexadecyltrimethylammonium bromide (CTAB), aluminum chloride, zirconium chloride were purchased from Sigma-Aldrich and used without further purification. The elemental analysis of lignin indicates the weight composition of lignin was C4 :H3 :O3 ratio. This lignin contained about 4 wt.% of ash.

280


Fig. 3. FTIR spectra of AL&Zr/MCM-48 catalysts.

2.1. Catalysts 2.1.1. Preparation of MCM-48 Mesoporous silica MCM-48 material was synthesized as described elsewhere [28] using TEOS and CTAB as a silica source and structure directing agent with slight modification. The preparation was carried out in round 500 cc flask following conventional procedures using the following molar composition: 0.34 TEOS: 0.2 CTAB: 0.167 NaOH: 19H2 O. TEOS was added under vigorous stirring to a solution of NaOH and CTAB at 40 ◦ C for 4 h. The milky solution was

transferred to a Teflon-lined stainless steel autoclave, and heated at 110 ◦ C for 48 h the white solid materials were obtained by filtration and dried under vacuum at 70 ◦ C, followed by calcination at 600 ◦ C for 5 h to remove organic remaining material. 2.1.2. Preparation of Al/MCM-48 and ZrMCM-48 The metallic (Al3+ or Zr4+ ) MCM-48 catalysts was prepared by solution impregnation method. A definite weight of aqueous metal chloride corresponding to 5–20-wt% AlCl3 and 5–10-wt% ZrCl4 to MCM-48 support. After stirring for 3 h, an amount of ammonia solu-


281

Fig. 4. XRD high angel pattern of the prepared catalysts.

tion corresponding half molar ratio of metal chloride was added and the solution kept at rest for additional 3 h. After, filtration and vacuum drying, the composite is air calcined at 550 ◦ C for Al/MCM48 and 480 ◦ C for Zr/MCM-48. XRF indicated that the final of Al3+ in Al/MCM-48 was 2.7, 5.9 and 8.4%, while Zr4+ was of 1.5, 2.9 and 5.5%. Consequently, the sample donated as Al/MCM-48 (2.7), Al/MCM48 (5.9), Al/MCM-48 (8.4), Zr/MCM-48 (1.5), Zr/MCM-48 (2.9) and Zr/MCM-48 (5.5). 2.1.3. Catalytic test Catalytic fast pyrolysis was carried out using fixed −bed quartz reactor system. The reactor dimensions are height 25 cm with 3 cm internal diameter. For each run, 4.0 g of premixed lignin-catalyst mixture using a mortar diluted with 3.0 g of glass beads was loaded into the reactor. The reactor was pretreated with nitrogen flow for

1 h at room temperature to remove the extraneous air from the reactor. The reactor was then introduced into the oven directly in order to achieve rapid pyrolysis at desired reaction temperature. A constant stream of N2 (oxygen free) was fed from the top of the reactor for the continuous withdrawal products and maintenance of the inert atmosphere during pyrolysis. At the end of the experiment, the reactor was cooled and purged for 10 min with N2 . The liquid products were collected using dry ice trap. The yielded products (gaseous and liquid) were analyzed by the gas chromatograph (Agilent 7890A). The liquid products were identified using Agilent GC/MS. An estimate of the products distribution in the liquid product was done based on the peak area of the GC/MS chromatogram after calibration with the individual component to ignore the detector response factor. This estimate is useful for the

282


Fig. 5. XRD small angel pattern of the prepared catalysts.

relative comparison of the compounds produced from the different catalysts and reaction condition. 2.1.4. Characterization of catalysts 2.1.4.1. XRD. Low angle X-ray diffraction (XRD) was recorded on a Bruker D5005 XPERT X-ray diffraction using CuK␣ radiation (␭ = 0.1542 nm) operated at 40 mA and 45 kV in the angular range from 0.5 to 10◦ (2) with a scanning rate of 0.002◦ (2)/s. Wideangle X-ray diffraction patterns were recorded on a XPERT X-ray diffraction in the angular range from 10 to 80◦ (2).

2.1.4.2. XRF. Si/Al and Si/Zr ratio were obtained by X-ray fluorescence (XRF) using a Bruker (S4 EXPLORER) operated at 20 mA and 50 kV. 2.1.4.3. HR-TEM. The TEM Images were obtained using a JEOL JEM2100 (Japan) operating at 200 kV. The powdered material was dispersed in ethanol solution and then dispersed by sonication for 30 min. This solution was dropped on a TEM copper grid and dried in air at room temperature for remove the ethanol from the TEM grid.


283

Fig. 6. NH3 -TPD patterns of Al&Zr/MCM-48.

2.1.4.4. BET. The Brunauer, Emmett and Teller (BET) surface area of different samples was measured from the adsorption of the nitrogen gas at −195.8 ◦ C using a volumetric apparatus of the conventional type using a Quantachrome. The catalysts were perfectly degassed at 200 ◦ C for 6 h before experiments. The SBET was determined from the adsorption curve. The pore size was calculated from the BJH desorption branch. FTIR analyses were performed using Nicolet IS-10 FTIR in the range wave number of 4000–400 cm−1 .

2.1.4.5. NH3 -TPD. Studies were obtained using an adsorption unit Micromeritics, (Chemisorb-2705) equipped with TCD detector. 50.0 mg catalyst sample was pretreated in under helium flow at 300 ◦ C for 30 min and allowed it to cool up to 100 ◦ C, and at this temperature the gas was switched to 5% NH3 in helium with a flow rate of 20 ml/min for 30 min and subsequently purged with helium gas at 100 ◦ C for 1 h to remove the physisorbed NH3 . NH3 TPD curves were obtained from 100 ◦ C to 800 ◦ C with ramping rate of 10 ◦ C/min for 30 min.

284


Fig. 7. TEM images of the prepared catalysts.

2.1.4.6. TGA. The thermogravimetric measurement was carried out using TGA Bruker 2010SA at a heating rate of 10 ◦ C/min to 800 ◦ C in air.

2.1.4.7. Solid NMR. Solid state 29 Si-NMR experiments are acquired using Bruker AVANCE II+ 400 MHz NMR system. The data were obtained at radio frequency (RF) 79.48 MHz with a delay time 100 s and spinning rate 12 KHz using calibration TMS; 0 ppm. In the case


285

Fig. 8. (a) 29 Si MAS NMR of catalysts MCM48, Al/MCM-48(8.4%) and Zr/MCM-48(2.9%) (b) 27 Al MAS NMR of 20% Al/MCM-48 catalyst.

of 27 Al-NMR, the data was obtained at RF 104.3 with delay time 3 s and spinning rate 14 KHz using calibration AlCl3 ; 0 ppm. 3. Results and discussion 3.1. Characterization and analysis 3.1.1. Nitrogen adsorption–desorption analysis The N2 full isotherms of Al/MCM-48(2.7–8.4) and Zr/MCM48(1.5–5.5) are shown in Figs. 1 and 2 and the SBET and pore diameter data are summarized in Table 1. All samples showed Type IV with a strong capillary condensation step, indicating tunable mesoporous framework [29]. The obtained data implies that the tunable mesoporosity remains after impregnation of AlCl3 and ZrCl4 with different ratios. However, the gradual decline in surface area was high in case of Al3+ , indicating diffusion of Al3+ in MCM-48

frame work. In addition, the increase of surface upon increasing Zr4+ may be come from the formation of external zirconia frame work due to the difference of electronegativity that can favor oxo-bridged species over their hydroxo-bridged counterparts, indicating a tendency toward oxolation over olation condensation. ZrIV is not consistent with the trends seen for the 5f series; it has preponderance for hydroxo-bridged oligomers in solution [30]. The steady decrement in the pore volume was observed with increasing the loading of Al3+ or Zr4+ , indicating portion of Al3+ or Zr4+ occupied internal pores. A sharp increment in nitrogen uptake at the relative pressure of P/P0 = 0.22–0.35 was observed to be reversible without exhibiting hysteresis loop, indicating short pore length. 3.1.2. The FTIR FTIR spectra of MCM-48, Al/MCM-48, and Zr/MCM-48 are shown in Fig. 3. The vibration bands around 3500 and 1619 cm−1 are

286


Fig. 9. Effect of carrier gas flows on the catalytic pyrolysis of lignin over MCM-48 at Temp. = 600 ◦ C and Catalyst/lignin ratio was 3:1.

Table 1 Specific surface area (SBET ), total pore volume and pore diameter of all prepared catalysts. Support/Catalyst

BETa m2 /g

Pore volumeb cm3 /g

Pore diameterc (nm)

MCM-48 Al/MCM-48(2.7%) Al/MCM-48(5.9%) Al/MCM-48(8.4%) Zr/MCM-48(1.5%) Zr/MCM-48(2.9%) Zr/MCM-48(5.5%)

1302 1132 827 597 1024 1112 1212

0.86 0.63 0.46 0.33 0.57 0.59 0.64

2.34 2.20 2.34 2.35 2.23 2.28 2.18

a Surface area was calculated by Brunauer–Emmett–Teller (BET) method in the relative pressure range of (P/Po ) range = 0.040–.25. b Total pore volume was derived at P/Po = 0.98. c Pore diameter is analyzed by the desorption branch of the isotherms by the Barrett–Joyner–Halenda (BJH) method.

due to stretching and bending vibration mode of OH groups. The symmetric and asymmetric stretching vibrations of SiO4 species appears at 1105, 811 and 471 cm−1 , while Si OH vibration noticed at 985 cm−1 . A similar spectrum was observed for Al/MCM-48 and Zr/MCM-48 except that the intensity of band at 985 cm−1 becomes low is shifted to 995 cm−1 , suggesting the formation of Si-O-Al or Si-O-Zr species [31]. 3.1.3. XRD The low and high angles XRD patterns of MCM-48, Al/MCM48 and Zr/MCM-48 are presented in Figs. 4 and 5. The sharp and high diffraction peaks indicated the well-ordered mesoporosity [32]. Fig. 5. exhibited a steady decrease in the diffraction peaks as the increase of Al3+ loadings, indicating partial destruction of the MCM-48 framework. On the other hand, the diffraction peaks of Zr/MCM-48 increased with Zr4+ loadings up to 5.5%, suggesting formation of zirconium extra framework. Furthermore, it was found that, besides the peak of MCM-48 (2 ≈ 22.8◦ ), there were no other peaks in the high-angle region of both Zr/MCM-48 or Al/MCM-48, indicating that Zr4+ and Al3+ species were highly dispersed over MCM-48. 3.1.4. TPD analysis NH3 -TPD analysis was carried out to evaluate the distribution of acid sites in MCM-48, Al/MCM-48 (2.7%), Al/MCM-48

(5.9%), Al/MCM-48 (8.4%), Zr/MCM-48 (1.5%), Zr/MCM-48 (2.9%) and Zr/MCM-48 (5.5%) catalysts. Fig. 6 shows the amount of ammonia desorbed, which is used as a measure of the total acidity as well as its distribution over the catalyst surface. Based on the desorption spectra, the acidity was classified as weak, moderate and strong sites [33]. MCM-48 showed two broad desorption peaks, which corresponds to the existence of weak and moderate acid sites. The Al/MCM-48 showed three broad desorption maxima at different temperatures, indicating a broad distribution of acid site (Weak, moderate and strong). The strong acidity at high temperature was observed to increase with increasing the aluminum contents. The total acidity of the catalyst is summarized in Table 2, which shows the catalyst with higher aluminum contents had a greater number of total acids sites [34]. However, Zr/MCM-48(1.5–5.5) showed two peaks centered at (150 ◦ C and 700 ◦ C) due to the existence of weak and very strong Brønsted acid sites. A notable increment in the acid sites can be observed in the Al/MCM-48 or Zr/MCM-48 compared with MCM-48. 3.1.5. TEM analysis Fig. 7 shows some representative HRTEM images of MCM-48 and Al/MCM-48 or Zr/MCM-48 with different loadings. The silica channels are clearly observed with no agglomeration of Al3+ and Zr4+ species on the catalyst surface, which are in agreement with XRD results (Fig. 4). All samples showed the well-ordered mesoporous structure of MCM-48. 3.1.6. MAS NMR The 29 Si MAS NMR of MCM-48, Al/MCM-48(8.4) and Zr/MCM48(5.5) were analyzed to investigate the incorporation of Al3+ and Zr4+ on the MCM-48 matrix as shown in Fig. 8a. The parent support exhibits a resonance peak with high intensity at 109 ppm which is related to the OH/Si ratio of MCM-48. On successive impregnation of Al3+ or Zr4+ , a sharp decrement in the intensity was observed, which is attributed to the change in condensation degree of silicates [32]. On the other hand, the 27 Al MAS NMR (Fig. 8b) of Al/MCM48(8.4) catalyst showed a resonance peak around 6, 21 and 52 ppm. The peaks corresponding to 6 and 21 ppm are assigned to pentaand hexacoordinated aluminum, indicating leaching of Al3+ during free calcination, while the peak at 53 is attributed to aluminum


Fig. 10. Effect of temperature on the catalytic pyrolysis of lignin (a) MCM-48 (b) Al/MCM-48 (8.4%) (c) Zr/MCM-48 (2.9%) at low = 25.

287

288


Fig. 11. Effect of Catalyst/Lignin ratio on the catalytic pyrolysis of lignin over (a) MCM-48 (b) Al/MCM-48 (8.4%) (c) Zr/MCM-48 (2.9%) at flow = 25 and Temp. = 600 ◦ C..


289

Table 2 NH3 -TPD for determination of Weak and strength of acid sites. Catalyst

Weak Lewis acid site

Moderate acid site

Strong bronosted acid site

Total acid sites (mmol/g)

MCM-48 Zr/MCM-48(1.5%) Zr/MCM-48(2.9%) Zr/MCM-48(5.5%) Al/MCM-48(2.7%) Al/MCM-48(5.9%) Al/MCM-48(8.4%)

0.027 0.240 0.284 0.372 0.735 0.109 0.348

0.253 absence absence absence 0.011 0.012 0.281

absence 0.346 0.449 0.393 absence 0.813 0.389

0.28 0.586 0.733 0.765 0.746 0.934 1.018

Fig. 12. The yield of total BTX as a function of catalyst acidity; (A) Al/MCM-48 catalysts, (B) Zr/MCM-48 catalysts at temp. = 600 ◦ C, flow = 25 ml/min.

in tetrahedral coordination sphere [33], and this result is in good agreement with NH3 -TPD results. 3.2. Catalytic activity 3.2.1. The effect of carrier gas flows Effect of carrier gas flow rate on product distributions in catalytic pyrolysis of lignin was studied (Fig. 9). The yield of BTX was observed to be highest as ≈17% at the flow rate of 25 ml min−1 , and sharply declined as the carrier gas flow reaches ml.min−1 , due to decreasing the contact time while appended increase over 25 ml min−1 may result from the dilution effect that enables some bulky figments entrance catalyst pores. Further parameter studies were carried out using the optimum carrier flow conditions of 25 ml min−1 . 3.2.2. The effect of reaction temperature Fig. 10A–C shows the plots for the effect of temperature on the yield of liquid products over MCM-48, Al/MCM-48 (8.4), and Zr/MCM-48 (5.5) catalysts. It was observed that temperature plays an important role as modulating the BTX yields. As shown in Fig. 10A. It was observed that the aromatic yield reached 1.3%,

18.0% and 17.0% at 500, 550 and 600 ◦ C, respectively. However, further temperature increasing decline formation of BTX yields, which reaches 6.8% at 650 ◦ C. This could be attributed to cock formation. In the case of Al/MCM-48(8.4) in Fig. 10B, the optimum temperature was 600 ◦ C, at which 32.5% of BTX yield is obtained. In Fig. 10C, it was found the Zr/MCM-48(2.9) catalyst gave the highest yield on BTX (49.4%) at 600 ◦ C. 3.2.3. The effect of catalyst/lignin ratios To further promote the catalytic activity, the studies on the effect of catalyst/lignin ratios (C/L) were studied. Fig. 11A shows the effect of C/L ratio on the BTX yield over MCM-48, Al/MCM-48(8.4) and Zr/MCM-48(2.9) catalysts. The yield of BTX was improved from 12.4 to 17.0% by increasing the C/L ratio from 1:1 to 3:1, and further increase leads to a slight decreased in BTX yield. Mullen et al. [34] attributed this trend to the nature of catalyst that could be resisted or help heat transfer rates to lignin materials resulting from the heat capacity of the additional catalyst mass. Fig. 11B shows the C/L ratio of 3:1 gave the higher BTX yield (32.5%) than that of the parent MCM-48 catalyst (17.0%). Nevertheless, the Zr/MCM-48(2.9) showed the higher yield of BTX (49.4%) at C/L ratio of 2:1 (Fig. 11C). Betiha et al. [22] stated that the accessibility of acidic site enables

290


Fig. 13. The maximum BTX yield% at different loadings for each catalyst.

Fig. 14. The maximum BTX yield% at different temperatures for each catalyst.

Fig. 15. Effect of the regeneration of catalysts as a function of the yield% of BTX at the optimum parameter of each catalyst.


291

Fig. 16. XRD of the reused catalysts (a) MCM-48 (b) Al/MCM-48(8.4%) and (c) Zr/MCM-48(2.9%).

more molecules to adsorb and as the cracking reaction proceeds on catalysts, the formation of smaller fragments with an easier access to the internal active sites increased and facilitates the second stage where complete degradation of fragments into smaller and volatile species can take place. Consequently, the high catalytic cracking of Zr/MCM-48(2.9) is attributed to more pre-cracking at the surface and/or near surface acid sites of catalysts that is responsible for initiating the carbo-cationic mechanism, leading to some products diffusing into openings and further cracking to lighter products because of the Zr/MCM-48(2.9) catalyst contains surface area of 1112 m2 /g and total acidity of 0.733 mmol/g, making this a more attractive catalyst for feeds with bulky substituents. 3.2.4. The effect of catalyst acidity The yield of BTX (Fig. 12A) was observed to be closely related to the total acidity and the type of acid sites (Table 2 and Fig. 6)

of the Al/MCM-48 and Zr/MCM-48 catalysts. The total acidity of catalyst increased with Al3+ loadings and was observed to be MCM48 < Al/MCM-48 (2.7) < Al/MCM-48 (5.9) < Al/MCM-48 (8.4). In the case of Zr/MCM-48 (Table 2, Fig. 12B), the increment in BTX yield was observed as the catalyst acidity increased on Zr/MCM-48 (2.9). Further increment in the total acidity had rather negative effect on the overall BTX yield (Table 2). From the distribution of the acid sites Table 2. Pristine MCM-48 catalyst exhibited weak and moderate acid sites. While the Al/MCM-48 catalysts had three types of acid sites (weak, moderate and strong) depending on the load of impregnated Al3+ species. Zr/MCM-48 catalysts had two types of acidity (weak and very strong) and gave the highest activity than Al/MCM48. This tells that the lignin pyrolysis to BTX depends mainly on the weak and strong acid sites. The surface area of the Zr/MCM-48 catalysts was observed to increase with Zr4+ loading which might

292


Fig. 17. TEM of the reused catalysts MCM-48, Al/MCM-48(8.4%) and Zr/MCM-48(2.9%). Table 3 The yield of BTX based on the catalyst regeneration. Yield% of BTX

MCM-48 Al/MCM-48(8.4%) Zr/MCM-48(2.9%)

Before regeneration

First regeneration

Second regeneration

17.01 32.45 49.39

11.48 30.3 16.46

6.6 8.05 13.45

also assist to get higher catalytic activity due less effective of carbon deposits. 3.2.5. The effect of loadings of Al or Zr and temperature Fig. 13 shows the effect of Al3+ loadings on the yield of BTX and the results revealed that the yield of BTX increases with increasing of Al3+ loadings. But in the case of Zr4+ loadings, the yield of BTX increased with the loading up to 2.9% and then decreased of even higher Zr4+ loading 5.5%. This may be attributed to the high content of very strong acid sites as the loading of zirconia increase beyond 2.9%. The effect of temperature on BTX yield was depicted in Fig. 14. The yield on BTX was observed to increase as increasing temperatures up to 600 ◦ C and decreased at 650 ◦ C. 3.2.6. The effect of catalyst regeneration Because of the stability and easy regeneration of the tested catalyst is one of the important factors, recycled catalyst tests were conducted. The spent catalysts MCM-48, Al/MCM-48(8.4) and Zr/MCM-48(2.9) were regenerated by calcination at 600 ◦ C in air atmosphere for 5 h to remove the surface coke deposits and reused under appropriate optimum pyrolysis condition. The catalytic activity was observed to decline after regeneration in all the cases, (illustrated in Fig. 15 and Table 3). The yield of BTX dropped into 11.5%, 30.3%, and 16.5%, respectively, after 1st regeneration and further decline in BTX yield was obtained the 2nd regeneration. In order to evaluate the catalyst deactivation, XRD and HRTEM analyses were carried (Figs. 16 and 17). As shown in Fig. 16, the XRD diffraction of the fresh and reused catalysts after first and second regeneration cycles. The fresh catalysts of MCM-48, Al/MCM48(8.4) and Zr/MCM-48(2.9) exhibited characteristic broad peaks

at 2 = 24◦ . After the first regeneration, catalysts did not show any sintering of agglomeration of Al3+ , or Zr4+ moieties, however after the 2nd of Al/MCM-48(8.4), XRD pattern showed formation of sillimanite phase (Al2 Si O5 , JCPDS card no. 22-18) or formation of Al2 O3 as mentioned by Kim et al. [35]. On the other hand, it well known that the highly dispersed nature of the incorporated ZrO2 was verified by the absence of monoclinic or tetragonal phase diffraction features in the XRD region, indicating dimensions below the instrumental detection limit of ∼2 nm, however the XRD pattern indicates presence of polymorphs of ZrO2 (monoclinic phase; PDF# 37-1484). The TEM analysis of MCM-48, Al/MCM-48(8.4) and Zr/MCM48(2.9) revealed structural deformation with heavy coke deposition (Fig. 17), which in turn could restrict the molecular diffusion inside pores due to the blocking of pore. To investigate the carbonaceous deposits percent, thermogravimetric analysis was carried out for the used samples MCM-48, Al/MCM-48(8.4) and Zr/MCM-48(2.9) (Fig. 18). The amount of coke formation was of 16.1%, 15.55% and 14.71% for MCM-48, Al/MCM48(8.4), and Zr/MCM-48(2.9), respectively. It can be inferred that Zr/MCM-48(2.9) catalyst showed less coke formation and thus in turn exhibits highest catalytic activity. Based on the XRD, TEM and TGA analysis of the spent catalyst, partial pore blocking, heavy coke deposition and structural deformation is the main reason for the loss of catalytic activity after recycling. 4. Conclusions The lignin was pyrolyzed over MCM-48 support loaded with different amount of Al3+ or Zr4+ metals for the production of aromatic hydrocarbon, especially BTX. The yield of BTX increased from 17.0% in case of MCM-48 to 32.5% in case of Al/MCM-48 (8.4%) and 49.4% in case of Zr/MCM-48 (2.9%) due to the enhanced acidity of the catalysts. Other experimental parameters such as carrier flow rate, temperature, C/L ratios and acidity of different catalysts gave positive effects on increasing the BTX yield. Among the tested catalysts, Zr4+ loaded onto MCM-48 was considered promising catalyst for the pyrolysis of lignin with a good activity as well as selectivity to BTX yield.


293

Acknowledgements This work was supported by the Egyptian Petroleum Research Institute. We thank Laboratory of Nano-Green Catalysis, Department of Chemistry, Inha University; Incheon which is supported in the catalyst characterization. We thank KBSI Seoul Western Center for Solid-state NMR analysis. References [1] R. Luque, J.A. Menéndez, A. Arenillas, J. Cot, Energy Environ. Sci. 5 (2012) 5481–5488. [2] M. Pudukudy, Z. Yaakob, M. Mohammad, B. Narayanan, K. Sopian, Renew. Sustain. Energy Rev. 30 (2014) 743–757. [3] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Ind. Eng. Chem. 43 (2004) 2610–2618. [4] R. Ma, W. Hao, X. Ma, Y. Tian, Y. Li, Angew. Chem. Int. Ed. 53 (2014) 7310–7315. [5] S. Du, J.A. Valla, G.M. Bollas, Green Chem. 15 (2013) 3214–3229. [6] J.M. Nichols, L.M. Bishop, R.G. Bergman, J.A. Ellman, J. Am. Chem. Soc. 132 (2010) 12554–12555. [7] D. Mohan, J. Pittman, D. Mohan, P.H. Steele, Energy Fuels 20 (2006) 848–889. [8] T.N. Trinh, P.A. Jensen, Z. Sarossy, K. Dam-Johansen, N.O. Knudsen, H.R. Sørensen, H. Egsgaard, Energy Fuels 27 (2013) 3802. [9] D.S. Xiao, S. GU, K. Luo, RSC Adv. 1 (2011) 1641–1660. [10] T.R. Carlson, Y.-T. Cheng, J. Jae, G.W. Huber, Energy Environ. Sci. 4 (2011) 145–161. [11] N. Tippayawong, J. Kinorn, S. Thavornun, Energy Source Part A 30 (2008) 1572–1578. [12] M.P. Pandey, C.S. Kim, Chem. Eng. Technol. 34 (2011) 29–41. [13] J.S. Shabtai, W.W. Zmierczak, E. Chornet, D.U.S. Johnson, Patent No. US20030115792, 2003. [14] D. Goheen, Adv. Chem. 59 (1966) 205–225. [15] A. Effendi, H. Gerhauser, A.V. Bridgwater, Renew. Sustain. Energy Rev. 12 (2008) 2092–2116. [16] Q. Zhang, J. Chang, T.J. Wang, Y. Xu, Energy Convers. Manag. 48 (2007) 87–92. [17] R.N. Olcese, J. Francois, M.M. Bettahar, D. Petitjean, A. Dufour, Energy Fuels 27 (2013) 975–984. [18] H.W. Lee, T.H. Kim, S.H. Park, J. Jeon, D.J. Suh, Y. Park, J. Nanosci. Nanotechnol. 13 (2013) 2640–2646. [19] J. Zakzeski, P.C.A. Bruijnincx, A.L. Jongerius, B.M. Weckhuysen, Chem. Rev. 110 (2010) 3552–3599. [20] C.Q. Dong, Z.F. Zhang, Q. Lu, Y.P. Yang, Energy Convers. Manag. 57 (2012) 49–59. [21] H.S. Kibombo, V. Balasanthiran, C. Wu, R. Peng, R.T. Koodali, Microporous Mesoporous Mater. 198 (2014) 1–8. [22] M.A. Betiha, M.F. Menoufy, A.M. Al-Sabagh, H.M.A. Hassan, S.A. Mahmoud, Microporous Mesoporous Mater. 204 (2015) 15–24. [23] H.M.A. Hassan, E.M. Saad, M.S. Soltan, M.A. Betiha, I.S. Butler, S.I. Mostafa, Appl. Catal. A Gen. 488 (2014) 148–159. [24] T. Mochizuki, D. Atong, S.Y. Chen, M. Toba, Y. Yoshimura, Catal. Commun. 36 (2013) 1–4. [25] J. Adam, M. Blazsó, E. Mészáros, M. Stöcker, M.H. Nilsen, A. Bouzga, J.E. Hustad, M. Grønli, G. Øye, Fuel 84 (2005) 1494–1502. [26] A. Pattiya, J.O. Titiloye, A.V. Bridgwater, J. Anal. Appl. Pyrolysis 81 (2008) 72–79. [27] Q. Lu, W.Z. Li, D. Zhang, X.F. Zhu, J. Anal. Appl. Pyrolysis 84 (2009) 131–138. [28] K. Wang, Y. Lin, M.A. Morris, J.D. Holmes, J. Mater. Chem. 16 (2006) 4051–4057. [29] M.A. Betiha, H.M.A. Hassan, A.M. Al-Sabagh, A.S. Khder, E.A. Ahmed, J. Mater. Chem. 22 (2012) 17551–17559. [30] Y.-J. Hu, K.E. Knope, S. Skanthakumar, M.G. Kanatzidis, J.F. Mitchell, L.J. Soderholm, Am. Chem. Soc. 135 (2013) 14240–14248. [31] M.A. Betiha, A.M. Rabie, A.M. Elfadly, F.Z. Yehia, Microporous Mesoporous Mater. 222 (2016) 44–54. [32] L. Wang, J. Zhang, F. Chen, Microporous Mesoporous Mater. 122 (2009) 229–233. [33] M. Janicke, D. Kumar, G.D. Stucky, B.F. Chemlka, Stud. Surf. Sci. Catal. 84 (1994) 243–250. [34] C.A. Mullen, A.A. Boateng, Fuel Process. Technol. 91 (2010) 1446–1458. [35] S. Kim, E. Sasmaz, J. Lauterbach, Appl. Catal. B: Environ. 168–169 (2015) 212–219. Fig. 18. TGA of used samples (a) MCM-48 (b) Al/MCM-48(8.4%) and (c) Zr/MCM-48 (2.9).