In situ upgrading of Shengli lignite pyrolysis vapors over metal-loaded HZSM-5 catalyst

In situ upgrading of Shengli lignite pyrolysis vapors over metal-loaded HZSM-5 catalyst

Fuel Processing Technology 160 (2017) 19–26 Contents lists available at ScienceDirect Fuel Processing Technology journal homepage: www.elsevier.com/...

1MB Sizes 5 Downloads 23 Views

Recommend Documents

No documents
Fuel Processing Technology 160 (2017) 19–26

Contents lists available at ScienceDirect

Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc

Research article

In situ upgrading of Shengli lignite pyrolysis vapors over metal-loaded HZSM-5 catalyst Tian-Long Liu, Jing-Pei Cao ⁎, Xiao-Yan Zhao, Jing-Xian Wang, Xue-Yu Ren, Xing Fan, Yun-Peng Zhao, Xian-Yong Wei Key Laboratory of Coal Processing and Efficient Utilization (Ministry of Education), China University of Mining & Technology, Xuzhou 221116, Jiangsu, China

a r t i c l e

i n f o

Article history: Received 4 January 2017 Received in revised form 11 February 2017 Accepted 13 February 2017 Available online xxxx Keywords: HZSM-5 Metal Catalyst Pyrolysis Aromatics

a b s t r a c t This work is aimed to study in situ upgrading of Shengli lignite pyrolysis vapors over different metal-loaded HZSM-5 in a drop tube reactor. Co/HZSM-5, Mo/HZSM-5 and Ni/HZSM-5 (5.0 wt%) were prepared by wet impregnation and characterized by N2 adsorption-desorption analyzer, X-ray diffraction, transmission electron microscope, Fourier transform infrared spectrometer and temperature programmed desorption of ammonia. The effects of temperature and catalyst on product yields and tar properties were investigated. The results show that the optimal temperature for liquid product was 600 °C and aromatics can be directly produced from solid lignite by catalytic fast pyrolysis over metal-loaded HZSM-5 under such mild condition. Due to the participation of metal and acid sites, the bifunctional metal-loaded HZSM-5 showed comparable catalytic activity for deoxygenation reaction in the valorization of oxygen content below 7.1%. The introduction of metal causes the increase of aromatics and the decrease of organic oxygen species in upgraded tar remarkably. Among the catalysts, Ni/ HZSM-5 exhibited the best performance for production of high quality tars with highest aromatics content of 94.2% (area%), which can be used as a potential candidate for catalytic upgrading of pyrolysis oil. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Pyrolysis, which provides mild conversion of coal into low-carbon fuels and chemicals appears to be a simple and sufficient thermochemical technique for clean and efficient utilization of coal. There has been considerable interest in lignite pyrolysis for liquid fuel production for about 20 years. However, the pyrolysis oil, especially derived from lignite cannot be directly used as transportation fuel because of its undesirable properties such as viscous and corrosive nature, thermal instability and low calorific values mainly resulted from different classes of organic oxygen species (OOSs) [1–3]. Thus the elimination of oxygen with a suitable upgrading method is imperative before large scale commercialization of pyrolysis oil become viable. Many upgrading process such as catalytic cracking and hydrodeoxygenation methods are currently in development [4,5]. Nonetheless, there exist more advantages to catalytic cracking for lignite conversion including (1) it does not require addition of hydrogen and could be operated at the atmospheric pressure, (2) it removes oxygen from pyrolysis oil in the form of water and carbon oxides while hydrotreatment removes the oxygen in the form of water [6], (3) the upgrading of pyrolysis oil can be achieved under simple and mild conditions by using various in-expensive catalysts. Furthermore, catalytic ⁎ Corresponding author. E-mail address: [email protected] (J.-P. Cao).

http://dx.doi.org/10.1016/j.fuproc.2017.02.012 0378-3820/© 2017 Elsevier B.V. All rights reserved.

cracking process produces products such as aromatics and olefins that already fit into existing infrastructure. In recent years, catalytic fast pyrolysis (CFP) has received substantial attention as a promising alternative process for thermochemical conversion of lignite and/or biomass to advanced liquid fuel [7–9]. This approach, which involves lignite pyrolysis and catalytic cracking of pyrolysis vapors in one continuous reaction system could avoid the problem of separating catalysts from residual chars, making it suitable for industrial application. The main problem for CFP is the selection of appropriate catalyst as the effectiveness of catalyst and conversion efficiency of this process still remain challenges for commercial production. Many studies focused on the use of microporous and mesoporous zeolites such as HZSM-5, USY, HY and SBA-15 for bio-oil upgrading. Especially, HZSM5 possesses well-defined ten-membered ring channel system with a pore size of 5.5–5.6 Å, making it difficult for large aromatic coke precursor to form inside the pores [10]. Besides, its high surface area and adjustable acidity allow to achieve a high catalytic activity through the reactions occurring inside the zeolite such as dehydration, decarbonylation, decarboxylation, and aromatization [11,12]. HZSM-5 zeolite has been tentatively proved as an outstanding catalyst for direct conversion of biomass into renewable aromatics [13–17]. Zhou et al. [15] investigated catalytic upgrading of fast pyrolysis lignin vapors over HZSM-5 in a fixed bed reactor and obtained an organic liquid product which has a low oxygen content of 4.0 wt% compared to the 23.4 wt% oxygen content without catalyst at 600 °C. Carlson et al. [16]

20

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26

investigated the influence of five zeolite structures including beta, Y, ZSM-5, silicalite and silica-alumina on the yields and chemical compositions of bio-oil produced from catalytic pyrolysis of biomass-derived oxygenates. They found that the aromatics selectivity is dependent on both the catalyst pore size and the nature of the active sites. It was observed that ZSM-5 had the highest aromatic yields (30% of carbon yield) and the least amount of coke. A distilled organic liquid product produced by non-catalytic cracking of soybean oil was successfully converted into a mixture containing high concentration of aromatic and cyclic hydrocarbons (58 and 13 wt%, respectively) using HZSM-5 as a catalyst [17]. Olazar et al. [18] studied CFP of sawdust with HZSM-5 catalyst at a gas residence time (GRT) of 50 ms and found that the oxygen content of the organic fraction is only 18.6%, which is less than the 38.5% of non-catalytic bio-oil. It is well known that coal originated from the accumulation, compaction and induration of variously altered plant remains original element compositions and chemical structure. Therefore, HZSM-5 could be likewise applied in lignite pyrolysis process for coal tar upgrading. In our previous work [19], we tested the effects of pyrolysis temperature (PT) and GRT on the distribution of OOSs and investigated the possible oxygen transformation routes during fast pyrolysis of lignite. The lignite-O was mainly converted into tar-O, water-O and gas-O during fast pyrolysis. In this study, we attempted to change oxygen transformation routes, reducing tar-O and simultaneously increasing light aromatics such as benzene (B), toluene (T), ethylbenzene (E), xylene (X) and naphthalene (N) in the tar. A commercial HZSM-5 was used as catalyst and the HZSM-5 was further modified by transition metals. The presence of transition metal was previously reported to change the surface properties of zeolite and affect the mode of oxygen rejection during CFP of biomass by producing more CO2 and less water, making more hydrogen available for incorporation into hydrocarbons [20–22]. Herein, this work is aimed to evaluate different metal-loaded HZSM-5 and study the effect of metal on product yields and chemical compositions of tar derived from Shengli lignite (SL). 2. Experiments

was slowly dropped into 20 mL of desired transparent metal solution with stirring and subsequently stored in vacuum for 24 h, followed by vacuum distillation at 40 °C, drying at 80 °C for 4 h and calcinating in air at 550 °C for 5 h. Modified HZSM-5 is denoted as Me/HZSM-5, where Me represents the loaded metal. The as-prepared catalysts were characterized by N2 adsorption-desorption analyzer, X-ray diffraction (XRD), transmission electron microscope (TEM), Fourier transform infrared (FTIR) spectrometer and temperature programmed desorption of ammonia (NH3-TPD). The detailed specification and suppliers are listed in Supplementary material. 2.2. Apparatus and procedure Fast pyrolysis of SL was performed at a prescribed temperature between 500 and 700 °C in a drop tube reactor (ID 1.1 cm, height 36 cm). This apparatus consists of a gas feed system, an electric furnace and temperature control system and a condenser unit as reported in our previous work [23]. In a typical CFP, the catalyst was placed on the quartz baffle with a height of 1.2 cm, located in the hot reactor zone (at 600 °C). Ar was used as the carrier gas with a flow rate of 100 mL min− 1. A typical GRT of the volatile phase in the catalyst bed is about 1.0 s. Me/HZSM-5 was firstly reduced by passing H2 at a flow rate of 100 mL min−1 for 1 h. After H2 was replaced by Ar completely, 3.0 g lignite sample was inserted from the top of the reactor with a rate of 0.1 g min−1. The produced pyrolysis volatiles were driven through the catalyst bed with the aid of Ar, while additional purging with Ar was performed for another 20 min to ensure a good material balance. Produced gas stream flew into a cooling system, traps with acetone submerged in a liquid bath (−20 °C). The condensable gases captured by the collector were converted into liquid oil. Meanwhile, non-condensable gases were collected with a gas bag and the solid char and catalyst were removed from the reactor for analysis after the experiment. In order to guarantee the reliability and suitability of results, each experiment was performed three times under the same conditions with the experimental error less than ± 3% and the average values were used for analysis.

2.1. Materials 2.3. Product analysis 2.1.1. Lignite sample SL was collected from SL coal field in Xilinhot, Inner Mongolia, China. It was pulverized to a particle size of 0.5–1.0 mm followed by drying at 107 °C for 24 h and then stored in an airtight container before use. The proximate and ultimate analyses are listed in Table 1. The ultimate analysis was conducted with an Elementar vario MACRO cube CHNS elemental determinator. Oxygen content was calculated by difference. 2.1.2. Catalysts HZSM-5 zeolite used here was purchased from the Catalyst Plant of Nankai University with a silicon-aluminum mole ratio of 25. It was pulverized to pass through a 40-mesh sieve followed by activating at 550 °C for 5 h in static air and then stored in an airtight container before use. Metal introduction (Co, Mo and Ni) was achieved by wet impregnation in aqueous solutions of corresponding nitrates or ammonium (Co(NO3)2·6H2O for Co, Ni(NO3)2·4H2O for Ni, (NH4)6Mo7O24·4H2O for Mo) in order to attain 5.0 wt% metal loading. Sieved HZSM-5 powder Table 1 Proximate and ultimate analyses of SL. Proximate analysis (wt%)

Ultimate analysis (wt%, daf) a

St,d

H/C

0.52

0.8760

a

Mar

Ad

VMd

FCd

C

H

N

O

12.76

18.72

43.64

37.64

67.47

4.96

1.27

N 25.77

A: ash; M: moisture; FC: fixed carbon; VM: volatile matter; ar: as received basis; d: dried basis; St,d: total sulfur in dried basis. a Calculated by difference.

The products derived from fast pyrolysis are classified into three categories: uncondensed gases, solid char and liquid oil (tar and water). The char, gas, and water yields were calculated according to the dryash-free (daf) basis of lignite. The tar yield was calculated by difference according to mass balance. The gases were analyzed with a Shimadzu GC-2014 gas chromatograph (GC) equipped with two detectors, a thermal conductivity detector (TCD) for analysis of H2, CO, CO2 and CH4 separated on PN and 13X packed columns, and a flame ionization detector (FID) for gaseous hydrocarbons separated on a alumina Bound/Na2SO4 capillary column. The tar was identified by Agilent 7890/5797 GC–MS and quantified by Shimadzu GC-2014 GC-FID. SH-Rtx-1 capillary column was used in GC-FID system. External standards were prepared for the major aromatics such as BTEXN, alkylbenzene (alkyl-B), methylnaphthalene (methyl-N), biphenyl (Biph) and fluorene (Fluor). A capillary column (60 m × 0.25 mm ID × 0.25 μm) coated with HP-5MS was used in GC– MS system. It was operated at a column flow of 1.0 mL min−1 with He (99.999%) as the carrier gas and the injector temperature was 250 °C and a split ratio of 20: 1 was employed. The MS was operated in EI model at 70 eV and the ion source temperature was set at 230 °C. The oven temperature was programmed from 50 °C to 300 °C at a heating rate of 5 °C min−1 and held for 10 min at 300 °C. The moisture content in the liquid oil was determined using a KEM MKV-710 Karl-Fischer titrator. The elemental composition of liquid oil were measured by the elemental analyzer after acetone was evaporated. The amount of coke deposited on catalyst was measured by

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26

thermogravimetric analysis using a Mettler-Toledo TGA/DSC1 analyzer. Samples were heated from 30 to 800 °C with a rate of 10 °C min−1 and maintained at the final temperature for 10 min under air atmosphere with a flow rate of 50 mL min−1. 3. Results and discussion 3.1. Catalyst characterization The metal oxides visible as dark grains are highly dispersed on HZSM-5 surface as Fig. 1 displays. However, no characteristic diffraction peaks of different metal oxides were observed in Fig. S1, which is probably due to their low loading content and weak crystallization, on the other hand, also implies a good dispersion of the very small metal oxides clusters on the zeolite external surface [24]. Besides, the similar XRD patterns and no significant band position shifts in Fig. S2 were observed between HZSM-5 and Me/HZSM-5, indicating that the structure of HZSM-5 remains intact and no isomorphous substitution in zeolite framework took place after loading different metals. The detailed description of Figs. S1 and S2 are listed in Supplementary material. The HZSM-5 and Me/HZSM-5 exhibited a combination of types I and IV according to the IUPAC classification as presented in Fig. S3, showing the characteristic of microporous and mesoporous structure. Table S1 summarizes their pore properties. In the case of parent HZSM-5, metal introduction leads to a decrease in the specific area and micropore volume. Similar observation was also reported by others, who described that the metal species could be agglomerated and block the micropores at high metal loading amount (5.0 wt%) [25–27]. On the other hand, data in Table 2 reveals that the participation of the metal species has a distinct effect on the acid properties. The presence of metal oxides reduced the number of total acid sites in all cases and a slight shift of the first peak (α1) to lower temperature were

21

observed in Fig. 2. This should be attributed to the zeolitic proton exchange by metal oxides during calcination process [20,28]. It should be noted that an additional adsorption peak (α3) is observed in Ni/ HZSM-5, indicating stronger adsorption sites as previously discovered by others [24,29]. This high temperature desorption peak could be due to the presence of aluminum with low coordination, which was formed by zeolite dehydroxylation during thermal treatment at temperature high than 500 °C [30]. These Lewis sites are important in many catalytic process, acting as an electron pair acceptor and thereby giving rise to charge transfer processes [31].

3.2. Fast pyrolysis of SL The product yields are plotted in Fig. 3, from which PT was found to be a vital parameter for SL fast pyrolysis. The range of char yield shows significant differences varying from 65.6% to 53.6% with PT increased from 500 to 700 °C. The tar yield also increased from 18.5% to 21.4% as PT increased from 500 to 600 °C and further decreased when PT was up to 700 °C. A continuous increase of uncondensed gas yield was also observed as PT increased. These tendencies indicated that higher PT could promote thermal cracking of lignite, due to further decomposition of lignite and secondary reactions of pyrolysis volatiles. Then the increased volatile matters converted into liquid and gaseous products. The effect of temperature on product distributions has been studied in our previous studies and similar results have been obtained [32,33]. However, when PT was further increased, secondary cracking of the produced volatiles dominated and therefore resulted in an increase of gas yield and a decrease of liquid yield. It should be noted that the yield of water generally generated from pyrolysis process either the remaining moisture of SL or from the decomposition of some compounds increased from 6.2% to 10.7% with PT increasing from 500 to 700 °C.

Fig. 1. TEM images of the catalysts.

22

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26

Table 2 Acidity of the catalysts.

α1 (mmol α2 (mmol α3 (mmol αT (mmol

NH3 g−1)a NH3 g−1)b NH3 g−1)c NH3 g−1)d

HZSM-5

Co/HZSM-5

Mo/HZSM-5

Ni/HZSM-5

2.69 3.10 – 5.79

2.15 2.43 – 4.58

2.12 2.59 – 4.71

1.97 3.03 0.27 5.28

– not detected. a mmol NH3 desorbed from the peak at 205 °C. b mmol NH3 desorbed from the peak at temperature ranging from 334 to 394 °C. c mmol NH3 desorbed from the peak at 571 °C. d Total NH3 desorbed.

Thereby, a moderate PT of 600 °C in the drop tube reactor should be chosen for maximum pyrolysis tar.

3.3. CFP over HZSM-5 Fig. 3. Effect of temperature on product yields.

3.3.1. Product yields The yields of product from SL fast pyrolysis at 600 °C and CFP of SL over HZSM-5 from 500 to 700 °C are presented in Fig. 4. Compared to non-catalytic experiment, there was a remarkable decrease in tar yield from 21.4% to 11.9% over HZSM-5, while the gas yield increased from 12.4% to 16.5%, indicating that pyrolysis vapors were further cracked by HZSM-5 into gases such as light hydrocarbons and carbon oxides. Similar to non-catalytic test, the yield of gas produced from CFP over HZSM-5 increased continuously with increasing PT from 500 to 700 °C. The upgraded tar yield reached a peak at 600 °C and then decreased at higher PT. One of the main challenges in catalytic cracking of pyrolysis vapors is undesired formation and retention of carbonaceous deposits, namely coke on catalyst. The formation of coke is unavoidable in acid catalyzed reactions over HZSM-5, especially at relatively low temperature [34,35]. HZSM-5 with unique 3-dimensional pore system [8] traps pyrolysis vapors in pores and these cracking products polymerize with each other and then dehydrogenate and aggregate to form coke on the acid sites of HZSM-5 [36]. There was a consequent decrease in coke yield with increasing catalytic PT. Therefore, a higher catalytic PT is necessary to avoid undesired decomposition reactions over the catalyst. A similar coke yield tendency was also reported by Williams and Nugranad [37]. More water was also detected in liquid oil than that of non-catalytic pyrolysis, indicating that the dehydration reactions become more intensive due to the Brønsted acid sites of HZSM-5 [15].

Fig. 2. NH3-TPD profiles of the catalysts.

3.3.2. Gas composition under different temperature The compositions of uncondensed gases, calculated by molar quantity on dry ash free are presented in Table 3, as deduced from which, HZSM-5 promoted the production of the gases (H2, CO2, CO, CH4 and C2–C4 hydrocarbons). Among the gas species, H2, CO2 and CO account for a significant portion and increase constantly as raising PT. During SL fast pyrolysis, H2, CO2 and CO are primarily formed through dehydrogenation reactions and the cracking of heterocyclic oxygen groups. The thermal decomposition of SL also includes a cracking of bridge carbons between ring systems and the formation of radicals. After recombination and hydrogenation reactions of aliphatic radicals and ring fragments, CH4 and C2–C4 hydrocarbons were produced [38]. In the presence of HZSM-5, the yields of H2 and CO increased to 2.8 and 2.9 mmol/g at 600 °C, respectively, while the yield of CO2 increased slightly from 1.0 to 1.3 mmol/g. Zhou et al. [15] discussed deoxygenation routes during HZSM-5 upgrading lignin derived bio-oil. The HZSM-5 mainly promotes decarbonylation reactions to form CO at high temperature (600 °C). As for the distribution of light hydrocarbons, olefins dominate and the selectivity of C2–C4 hydrocarbons increase towards ethene with increasing PT. The highest ethene yield was 0.4 mmol/g at 600 °C. Similar results were also obtained by Hong et al. [39]. HZSM-5 exhibits a high activity to promote the evolution of gases, a behavior attributed to its Brønsted acid sites that favor decarbonylation reactions to produce CO as well as cracking and hydrogen transfer reactions to form olefins [40–42].

Fig. 4. Effects of temperature and catalyst on product yields.

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26 Table 3 The composition of uncondensed gases (mmol/g lignite, daf). Catalyst

PT (°C)

H2

CO2

CO

CH4

C2H6

C2H4

C3–C4

CO2/CO

No Cat HZSM-5

600 500 550 600 650 700 600 600 600

1.5 0.5 1.4 2.8 4.4 7.0 3.4 4.8 10.4

1.0 0.8 1.0 1.3 1.4 1.5 2.7 2.4 3.6

2.2 1.4 2.3 2.9 3.2 3.9 2.8 2.3 1.1

0.6 0.4 0.6 0.7 0.8 1.2 0.8 0.7 1.3

0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1

0.1 0.3 0.3 0.4 0.3 0.2 0.4 0.2 0.1

b0.1 b0.1 b0.1 b0.1 b0.1 b0.1 b0.1 b0.1 b0.1

0.45 0.57 0.43 0.45 0.44 0.38 0.96 1.04 3.27

Co/HZSM-5 Mo/HZSM-5 Ni/HZSM-5

3.3.3. BTEXN release over HZSM-5 The effect of temperature on BTEXN release amount during CFP over HZSM-5 was investigated. As shown in Fig. 5, the amount of BTEXN increased from 7.7 mmol/g at 500 °C to 19.7 mmol/g at 700 °C. Higher PT enhanced the cleavage of bridge bounds in lignite and produced more olefins, alkanes, aromatics with side chains and phenols. These molecules can easily enter the HZSM-5 pores to yield more BTEXN. Compared with non-catalytic pyrolysis, the yield of BTEXN in the tar significantly increased 4.1 times after catalytic cracking over HZSM-5 at 600 °C. The amount of B and T increased 6.7 and 4.0 times, respectively, while the amount of B with a relatively long side chain such as E increased slightly. The side chains in the aromatic rings are activated by the protonic acids and the cracking reaction occurs in the catalyst bed via dealkylation of the side chains, resulting in a remarkable increase of B and T [43]. Furthermore, the amount of N increased 4.8 times after catalytic cracking. It should be noted that more m/p-X is produced than o-X, which may be ascribed to the fact that the steric hindrance of the two methyls of o-X is larger than that of the m/p-X and it trends to produce more stable m/p-X during catalytic reforming of pyrolysis vapors in HZSM-5 pores. The effect of HZSM-5 on the formation of light aromatics such as BTXEN is supposed to be related to its pore structure and acid sites. The molecular dimension of light aromatics is smaller than the pore diameter of HZSM-5, which makes it easier to enter into the porous HZSM-5 channels and to be desorbed out of the pores [10]. On the other hand, the acid sites also play critical roles as the catalytic reforming is apparently taking place on the acid sites within HZSM-5 pores [17]. Consequently, dealkylation of the side chains, dehydroxylation of phenolic compounds and aromatization of alkanes and olefins occur as the pyrolysis volatiles pass through HZSM-5 and thus light aromatics prevail in the reformate.

Fig. 5. Effect of temperature on BTEXN yields.

23

3.3.4. Reaction pathway The reaction pathway shown in Fig. 6 was proposed for CFP of SL over HZSM-5. In the first step, SL undergoes a rapid thermal decomposition and is converted into volatile organics and gases. Some of the heavy components, which mainly consisted of macromolecular weight OOSs are cracked to light organics. On the other hand, some heavy components are polymerized on the catalyst surface to form coke. The second step is the diffusion of the light organics into HZSM-5 pores. The light organics, which are “hydrocarbon pool” consists of various acids and esters, ketones, alcohols, ethers and phenols. The “hydrocarbon pool” mechanism proposed by Carslson et al. [44] stated that the light organics act as intermediate compounds, which is the precursor of final aromatics. These intermediate oxygenates undergo a series of reactions such as decarbonylation, decarboxylation, dehydration and oligomerization to form both monocyclic aromatics (MCA) and C2-C6 olefins [11,40]. The olefins then undergo a series of aromatization reactions to produce aromatics. The aromatization of olefins is called Diels-Alder reaction (Eq. (1)) which has studied by [45,46]. The polycyclic aromatics (PCA) such as N could be formed through a second series reaction where single-ring aromatics further react with another oxygenate. ð1Þ

3.4. CFP over Me/HZSM-5 3.4.1. Product yields CFP over Me/HZSM-5 was conducted under 600 °C to study the effects of Co, Mo and Ni on the catalytic performance. The product distribution is shown in Fig. 4 and the gas composition is presented in Table 3. It is clear that all the catalysts lead to the increase of gas yield and the decrease of total liquid oil yield in comparison to non-catalytic pyrolysis. The behavior is attributed to the enhancement of decarbonylation-decarboxylation reactions of lignite fragments and various hydrocarbon conversion reactions, such as cracking, dehydrogenation and cyclization-aromatization, which are catalyzed by the zeolitic Brønsted acid sites [47]. On the other hand, the gas yields increased from 16.5% over HZSM-5 to 21.1%, 19.9% and 22.2% by using Co/HZSM-5, Mo/HZSM-5

Fig. 6. Proposed reaction pathway for the conversion of SL into aromatics over HZSM-5.

24

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26

and Ni/HZSM-5, respectively, which should be attributed to the high activity of transition metal for steam gasification/reforming of some components in the tar [26,48,49]. When the reaction water contacted with pyrolysis vapors in the presence of transition metal, steam reforming reaction easily occurred and more gases such as H2 and CO can be produced. In addition, water formation was enhanced during the utilization of metal-loaded HZSM-5 due to the increase of OOSs dehydration reactions. The water yields in the liquid oil upgraded by Co/ HZSM-5 and Mo/HZSM-5 are lower than that of Ni/HZSM-5. This should be due to the decreased total acid sites for Co/HZSM-5 and Mo/HZSM-5 and the presence of strong Lewis acid sites for Ni/HZSM-5 proved by NH3-TPD analysis. Coke amounts deposited on Co/HZSM-5, Mo/HZSM5 and Ni/HZSM-5 increased from 3.6% (coke deposited on HZSM-5) to 4.0%, 4.1% and 5.0%, respectively. The coke can block the accessibility to the active sites and thus diminish the efficiency of the catalyst. The enhanced coke formation should be attributed to the higher activity of metal-loaded HZSM-5 related to the acid and metal sites, by which some oxygenated compounds with small molecule weight in pyrolysis tar can be easily cracked to coke on the catalyst [50].

3.4.2. Tar properties According to the ultimate analysis in Table 4, tar derived from fast pyrolysis of SL has an oxygen content of 21.3%. Such high oxygen content is the reason for almost all undesired properties, including low heating value and extreme corrosivity and instability. Thus the elimination of oxygenated components is one of the primary targets during upgrading of pyrolysis tar. The oxygen content decreased from 21.3% to 10.3%, 6.5%, 7.1% and 4.6% when used HZSM-5, Co/HZSM-5, Mo/ HZSM-5 and Ni/HZSM-5, respectively. Compared with HZSM-5, metalloaded HZSM-5 exhibited higher activity for removing oxygen and Ni/ HZSM-5 especially enabled the highest upgrading effect so that its generated tar has the lowest oxygen content. Besides, the upgraded tars with lower H/C molar ratio than that without catalyst suggest a strong aromatic character, indicating that the upgraded oils have higher qualities and might be used as transport fuel. In the presence of HZSM-5, oxygen is usually removed at the acid sites via three deoxygenation routes, including water by dehydration, CO2 by decarboxylation and CO by decarbonylation. As presented in Table 2, the acid sites decreased with the introduction of metal into HZSM-5, which is considered detrimental for its catalytic activity. However, the oxygen content in the tar was further reduced when Me/ HZSM-5 with less acid sites was used, indicating that the metal sites participated in the deoxygenation reactions in addition to the acid sites. As seen from Table 3, the total amount of CO2 and CO increased from 4.2 mmol/g (using HZSM-5) to 5.5, 4.7 and 4.7 mmol/g for Co/ HZSM-5, Mo/HZSM-5 and Ni/HZSM-5, respectively. Moreover, water formation was slightly enhanced by metal-loaded HZSM-5. The changes in CO2/CO ratio are indicative of the different decarbonylation-decarboxylation mechanism due to the difference in the acid and metal sites. As listed in Table 3, CO2 yield is 3.27 times higher than CO yield for Ni/HZSM-5, indicating that the oxygen is mainly removed via decarboxylation reaction that favored by Ni species. The increase of CO2 and CO amounts for both Co and Mo species compared with HZSM-5

catalyst suggests enhanced decarboxylation and decarbonylation reaction on Co and Mo species. Chemical compositions of tar produced from fast pyrolysis of SL and in situ upgrading of pyrolysis vapors are shown in Fig. 7. The representative organic compounds detected by GC–MS in the tars can be classified as aromatics, aliphatics, OOSs, organic nitrogen species (ONSs) and organic sulfur species (OSSs). It can be observed that the effect of catalyst on chemical compositions of the tar is noticeable. The major components of tar without catalyst were aromatics (56.7%) and considerable OOSs (27.5%) considered as undesired fraction, including phenols, acids, esters, aldehydes, alcohols and ketones. HZSM-5 leads to a remarkable increase in aromatics (86.2%) and a decrease in all other types of compounds. Higher aromatics content in tar is strongly beneficial in terms of its possible use as fuels [51]. The modified HZSM-5 by Co, Mo and Ni lead to the increase of aromatics content in the tar from 86.2% to 90.5%, 88.4% and 94.2%, respectively, indicating that more OOSs were converted to aromatics due to high activity of the catalysts. On the other hand, the combined function of the metal and acid sites can also provide ideal environment for typical aromatization reaction, leading to enhanced formation of aromatics [27,28]. Indeed, the highest percentage of aromatics (94.2%) was obtained using Ni/HZSM-5 as the catalyst. This may be ascribed to violent decarbonylation and dehydrogenation reactions of pyrolysis vapors on Ni species as suggested by the highest CO2/CO ratio and abundant H2 formation (see Table 3). It should be noted that the production of aliphatics in the presence of HZSM-5 is negligible, whereas metal-loaded HZSM-5 promoted the production of aliphatics, especially for Co/HZSM-5 (2.8%). 3.4.3. Aromatic selectivity The concentration on tar basis and the selectivity of main aromatics including BTEXN, alkyl-B, methyl-N, Biph and Fluor are shown in Fig. 8. The overall amount of the qualified aromatics accounts for a small part of 3.0% in the tar derived from fast pyrolysis of SL. The major components in this pyrolysis tar, also including tar from other coal pyrolysis technologies are heavy compounds with boiling points above 360 °C (pitch in fact), accounting for about 60–70 wt% of the total tar mass [52]. The concentration of the major aromatics in the tar catalyzed by HZSM-5, Co/HZSM-5, Mo/HZSM-5 and Ni/HZSM-5 increased significantly by 5.2, 11.4, 9.0 and 13.6 times, respectively than that without secondary upgrading. The catalytic upgrading caused the lower tar yield but higher fraction of light tar (the qualified aromatics). The highest light tar yield can be obtained over Ni/HZSM-5 catalyst, indicating that Ni/HZMS-5 has the strongest catalytic activity for cracking tar. As shown in Fig. 8, all the catalysts enhanced the selectivity towards BTN. The aromatic selectivity over HZSM-5 is B N T N N N methyl-

Table 4 Ultimate analysis of the tars. Catalyst

No Cat HZSM-5 Co/HZSM-5 Mo/HZSM-5 Ni/HZSM-5 b

Ultimate analysis (daf, wt%)

H/C

C

H

N

S

Ob

69.5 80.8 84.6 84.2 86.5

7.2 7.9 8.2 8.1 8.4

1.1 0.7 0.5 0.4 0.4

0.9 0.3 0.2 0.2 0.1

21.3 10.3 6.5 7.1 4.6

Calculated by difference.

1.2345 1.1651 1.1550 1.1463 1.1572 Fig. 7. Effect of catalyst on chemical compositions of pyrolysis oil.

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26

25

4. Conclusion The structure of modified HZSM-5 remains intact after loading different metals. The pore and acidic properties were affected as surface area, total pore volume and acid sites decreased slightly. PT was a vital parameter for product distributions and the optimal temperature for liquid products was 600 °C. The maximum tar yield of 11.9% was obtained over HZSM-5 at 600 °C. The gas yield increased while the coke deposited on HZSM-5 decreased when PT was increased from 500 to 700 °C. Compared to HZSM-5, Me/HZSM-5 showed comparable activity for deoxygenation reaction in the valorization of oxygen content below 7.1%. The oxygen species in the liquid oil was mainly converted to CO2, CO and H2O due to the participation of different metal and acid sites and the changes in CO2/CO ratio are indicative of different decarboxylation-decarbonylation mechanism. The use of bifunctional Me/HZSM5 is an effective method for enhanced selectivity towards aromatics. Ni/HZSM-5, can produce a tar with the aromatics content of 94.2% (area%), which seems to be a potential candidate in the catalytic upgrading of lignite pyrolysis oil. Acknowledgements This work was subsidized by the Fundamental Research Funds for the Central Universities (China University of Mining and Technology, Grant 2015QNA18), National Natural Science Foundation of China (Grant 21676292), Natural Science Foundation of Jiangsu Province (BK20151141), and the Priority Academic Program Development of Jiangsu Higher Education Institutions. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.fuproc.2017.02.012. References Fig. 8. Effect of catalyst on the selectivity (a) and the concentration (b) of main aromatics.

N N X N alkyl-B N E N Fluor N E N Biph. Compared to HZSM-5, the presence of Co increased the selectivity towards alkyl-B, X, N, Biph and Fluor. It was reported that aromatization-polymerization reactions were enhanced on Co species, leading to a further reaction of B with other oxygenated compounds to form N and alkylated B [28]. Mo/HZSM-5 exhibited the highest selectivity of BTEXN and alkyl-B. The synergistic effect between Mo species inside the HZSM-5 channels together with the Brønsted acid sites plays a crucial role in the aromatization of CH4 [53]. CH4 is activated to form CH3 and H free radicals (Eq. (2)) and the CH3 free radicals subsequently dimerized to ethane and ethylene (Eq. (3)) and then ethylene aromatizes to B with the aid of protons of HZSM-5 (Eq. (4)). Besides, generated CH3 and H free radicals easily combine with monocyclic and bicyclic aromatic radicals, which resulted in BTEXN formation. The presence of Ni/HZSM-5 enhanced B and N selectivity greatly. This is probably because that Ni species can not only promote dehydrogenation reactions that favor the production of aromatics but also exhibit an excellent performance of breaking side chains reported by David et al. [54]. MoOx

CH4 → CH3 ∙ þ H∙ MoOx

2CH3 ∙ → C2 H4 þ H2 Hþ

3C2 H4 → C6 H6 þ 3H2

ð2Þ

ð3Þ ð4Þ

[1] P.S. Rezaei, H. Shafaghat, W.M.A.W. Daud, Production of green aromatics and olefins by catalytic cracking of oxygenate compounds derived from biomass pyrolysis: a review, Appl. Catal. A Gen. 469 (2014) 490–511. [2] J.Z. Han, X.X. Liu, J.R. Yue, B. Xi, S. Gao, G. Xu, Catalytic upgrading of in situ coal pyrolysis tar over Ni-char catalyst with different additives, Energy Fuel 28 (2014) 4934–4941. [3] S. Czernik, A.V. Bridgwater, Overview of applications of biomass fast pyrolysis oil, Energy Fuel 18 (2004) 590–598. [4] Q. Zhang, J. Chang, T. Wang, Y. Xu, Review of biomass pyrolysis oil properties and upgrading research, Energy Convers. Manag. 48 (2007) 87–92. [5] S. Thangalazhy-Gopakumar, S. Adhikari, R.B. Gupta, Catalytic pyrolysis of biomass over H+ ZSM-5 under hydrogen pressure, Energy Fuel 26 (2012) 5300–5306. [6] H. Zhang, Y.T. Cheng, T.P. Vispute, R. Xiao, G.W. Huber, Catalytic conversion of biomass-derived feedstocks into olefins and aromatics with ZSM-5: the hydrogen to carbon effective ratio, Energy Environ. Sci. 4 (2011) 2297–2307. [7] R. French, S. Czernik, Catalytic pyrolysis of biomass for biofuels production, Fuel Process. Technol. 91 (2010) 25–32. [8] A. Aho, N. Kumar, K. Eränen, T. Salmi, M. Hupa, D.Y. Murzin, Catalytic pyrolysis of woody biomass in a fluidized bed: influence of the zeolite structure, Fuel 87 (2008) 2493–2501. [9] J. Han, X. Wang, J. Yue, S. Gao, G. Xu, Catalytic upgrading of coal pyrolysis tar over char-based catalysts, Fuel Process. Technol. 122 (2014) 98–106. [10] J. Jae, G.A. Tompsett, A.J. Foster, K.D. Hammond, S.M. Auerbach, R.F. Lobo, G.W. Huber, Investigation into the shape selectivity of zeolite catalysts for biomass conversion, J. Catal. 279 (2011) 257–268. [11] T.R. Carlson, Y.T. Cheng, J. Jae, G.W. Huber, Production of green aromatics and olefins by catalytic fast pyrolysis of wood sawdust, Energy Environ. Sci. 4 (2011) 145–161. [12] M. Bjørgen, S. Svelle, F. Joensen, J. Nerlov, S. Kolboe, F. Bonino, L. Palumbo, S. Bordiga, U. Olsbye, Conversion of methanol to hydrocarbons over zeolite H-ZSM-5: on the origin of the olefinic species, J. Catal. 249 (2007) 195–207. [13] A.J. Foster, J. Jae, Y.T. Cheng, G.W. Huber, R.F. Lobo, Optimizing the aromatic yield and distribution from catalytic fast pyrolysis of biomass over ZSM-5, Appl. Catal. A Gen. 423 (2012) 154–161. [14] H. Ben, A.J. Ragauskas, Influence of Si/Al ratio of ZSM-5 zeolite on the properties of lignin pyrolysis products, ACS Sustain. Chem. Eng. 1 (2013) 316–324. [15] G.F. Zhou, P.A. Jensen, D.M. Le, N.O. Knudsen, A.D. Jensen, Direct upgrading of fast pyrolysis lignin vapor over the HZSM-5 catalyst, Green Chem. 18 (2016) 1965–1975.

26

T.-L. Liu et al. / Fuel Processing Technology 160 (2017) 19–26

[16] T.R. Carlson, G.A. Tompsett, W.C. Conner, G.W. Huber, Aromatic production from catalytic fast pyrolysis of biomass-derived feedstocks, Top. Catal. 52 (2009) 241–252. [17] S. Fegade, B. Tande, A. Kubátová, W. Seames, E. Kozliak, Novel two-step process for the production of renewable aromatic hydrocarbons from triacylglycerides, Ind. Eng. Chem. Res. 54 (2015) 9657–9665. [18] M. Olazar, R. Aguado, J. Bilbao, Pyrolysis of sawdust in a conical spouted-bed reactor with a HZSM-5 catalyst, AICHE J. 46 (2000) 1025–1033. [19] X.B. Feng, J.P. Cao, X.Y. Zhao, C. Song, T.L. Liu, J.X. Wang, X. Fan, X.Y. Wei, Organic oxygen transformation during pyrolysis of Baiyinhua lignite, J. Anal. Appl. Pyrolysis 117 (2016) 106–115. [20] Y.T. Cheng, J. Jae, J. Shi, W. Fan, G.W. Huber, Production of renewable aromatic compounds by catalytic fast pyrolysis of lignocellulosic biomass with bifunctional Ga/ ZSM-5 catalysts, Angew. Chem. Int. Ed. 51 (2012) 1387–1390. [21] C. Zhao, J.A. Lercher, Upgrading pyrolysis oil over Ni/HZSM-5 by cascade reactions, Angew. Chem. Int. Ed. 51 (2012) 5935–5940. [22] B. Valle, A.G. Gayubo, A.T. Aguayo, M. Olazar, J. Bilbao, Selective production of aromatics by crude bio-oil valorization with a nickel-modified HZSM-5 zeolite catalyst, Energy Fuel 24 (2010) 2060–2070. [23] J.P. Cao, L.Y. Li, K. Morishita, X.B. Xiao, X.Y. Zhao, X.Y. Wei, T. Takarada, Nitrogen transformations during fast pyrolysis of sewage sludge, Fuel 104 (2013) 1–6. [24] A. Veses, B. Puértolas, M.S. Callén, T. García, Catalytic upgrading of biomass derived pyrolysis vapors over metal-loaded ZSM-5 zeolites: effect of different metal cations on the bio-oil final properties, Microporous Mesoporous Mater. 209 (2015) 189–196. [25] P.L. Tan, Y.L. Leung, S.Y. Lai, C.T. Au, The effect of calcination temperature on the catalytic performance of 2 wt.% Mo/HZSM-5 in methane aromatization, Appl. Catal. A Gen. 228 (2002) 115–125. [26] J. Rizkiana, G. Guan, W.B. Widayatno, J. Yang, X. Hao, K. Matsuoka, A. Abudula, Mgmodified ultra-stable Y type zeolite for the rapid catalytic co-pyrolysis of low-rank coal and biomass, RSC Adv. 6 (2016) 2096–2105. [27] E.F. Iliopouloua, S.D. Stefanidis, K.G. Kalogiannis, A. Delimitis, A.A. Lappas, K.S. Triantafyllidis, Catalytic upgrading of biomass pyrolysis vapors using transition metal-modified ZSM-5 zeolite, Appl. Catal. B Environ. 127 (2012) 281–290. [28] E.F. Iliopoulou, S. Stefanidis, K. Kalogiannis, A.C. Psarras, A. Delimitis, K.S. Triantafyllidis, A.A. Lappas, Pilot-scale validation of Co-ZSM-5 catalyst performance in the catalytic upgrading of biomass pyrolysis vapours, Green Chem. 16 (2014) 662–674. [29] Z.Y. Zakaria, J. Linnekoski, N.A.S. Amin, Catalyst screening for conversion of glycerol to light olefins, Chem. Eng. J. 207 (2012) 803–813. [30] S. Kotrel, M.P. Rosynek, J.H. Lunsford, Quantification of acid sites in H-ZSM-5, H-β, and H-Y zeolites, J. Catal. 182 (1999) 278–281. [31] M.H.W. Sonoemans, C. Den Heijer, M. Crocker, Studies on the acidity of mordenite and ZSM-5. 2. Loss of Brønsted acidity by dehydroxylation and dealumination, J. Phys. Chem. 97 (1993) 440–445. [32] X. Huang, J.P. Cao, P. Shi, X.Y. Zhao, X.B. Feng, Y.P. Zhao, X. Fan, X.Y. Wei, Influences of pyrolysis conditions on the production and chemical composition of the bio-oils from fast pyrolysis of sewage sludge, J. Anal. Appl. Pyrolysis 110 (2014) 353–362. [33] J.P. Cao, X. Huang, X.Y. Zhao, B.S. Wang, S. Meesuk, K. Sato, X.Y. Wei, Low-temperature catalytic gasification of sewage sludge-derived volatiles to produce clean H2rich syngas over a nickel loaded on lignite char, Int. J. Hydrog. Energy 39 (2014) 9193–9199. [34] M.J. Rice, A.K. Chakraborty, A.T. Bell, Al next nearest neighbor, ring occupation, and proximity statistics in ZSM-5, J. Catal. 186 (1999) 222–227.

[35] I. Torre, J.M. Arandes, M.J. Azkoiti, M. Olazar, J. Bilbao, Cracking of coker naphtha with gas-oil. Effect of HZSM-5 zeolite addition to the catalyst, Energy Fuel 21 (2007) 11–18. [36] N. Hosseinpour, A.A. Khodadadi, Y. Mortazavi, A. Bazyari, Nano-ceria-zirconia promoter effects on enhanced coke combustion and oxidation of CO formed in regeneration of silica-alumina coked during cracking of triisopropylbenzene, Appl. Catal. A Gen. 353 (2009) 271–281. [37] P.T. Williams, N. Nugranad, Comparison of products from the pyrolysis and catalytic pyrolysis of rice husks, Energy 25 (2000) 493–511. [38] M. Nguyen, C. Berndt, D. Reichel, S. Krzack, B. Meyera, Pyrolysis behaviour study of a tar- and sulphur-rich brown coal and GC-FID/MS analysis of its tar, J. Anal. Appl. Pyrolysis 115 (2015) 194–202. [39] C.G. Hong, F.Y. Gong, M.H. Fan, Q. Zhai, W.W. Huang, T.J. Wang, Q.X. Li, Selective production of green light olefins by catalytic conversion of bio-oil with Mg/HZSM-5 catalyst, J. Chem. Technol. Biotechnol. 88 (2013) 109–118. [40] A.G. Gayubo, B. Valle, A.T. Aguayo, M. Olazar, J. Bilbao, Olefin production by catalytic transformation of crude bio-oil in a two-step process, Ind. Eng. Chem. Res. 49 (2010) 123–131. [41] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Transformation of oxygenate components of biomass pyrolysis oil on a HZSM-5 zeolite. I. Alcohols and phenols, Ind. Eng. Chem. Res. 43 (2004) 2610–2618. [42] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, M. Olazar, J. Bilbao, Transformation of oxygenate components of biomass pyrolysis oil on a HZSM-5 zeolite. II. Aldehydes, ketones, and acids, Ind. Eng. Chem. Res. 43 (2004) 2619–2626. [43] L.J. Yan, X.J. Kong, R.F. Zhao, F. Li, K.C. Xie, Catalytic upgrading of gaseous tars over zeolite catalysts during coal pyrolysis, Fuel Process. Technol. 138 (2015) 424–429. [44] T.R. Carlson, J. Jae, G.W. Huber, Mechanistic insights from isotopic studies of glucose conversion to aromatics over ZSM-5, ChemCatChem 1 (2009) 107–110. [45] P.T. Williams, D.T. Taylor, Aromatization of tyre pyrolysis oil to yield polycyclic aromatic hydrocarbons, Fuel 72 (1993) 1469–1474. [46] R. Cypres, Aromatic hydrocarbons formation during coal pyrolysis, Fuel Process. Technol. 15 (1987) 1–15. [47] P.M. Mortensen, J.D. Grunwaldt, P.A. Jensen, K.G. Knudsen, A.D. Jensen, A review of catalytic upgrading of bio-oil to engine fuels, Appl. Catal. A Gen. 407 (2011) 1–19. [48] B.S. Wang, J.P. Cao, X.Y. Zhao, Y. Bian, C. Song, Y.P. Zhao, X. Fan, X.Y. Wei, T. Takarada, Preparation of nickel-loaded on lignite char for catalytic gasification of biomass, Fuel Process. Technol. 136 (2015) 17–24. [49] J.P. Cao, P. Shi, X.Y. Zhao, X.Y. Wei, T. Takarada, Catalytic reforming of volatiles and nitrogen compounds from sewage sludge pyrolysis to clean hydrogen and synthetic gas over a nickel catalyst, Fuel Process. Technol. 123 (2014) 34–40. [50] S. Vichaphund, D. Aht-Ong, V. Sricharoenchaikul, D. Atong, Catalytic upgrading pyrolysis vapors of Jatropha waste using metal promoted ZSM-5 catalysts: an analytical Py-GC/MS, Renew. Energy 65 (2014) 70–77. [51] I.V. Babich, M. van der Hulst, L. Lefferts, J.A. Moulijn, P. O'Connor, K. Seshan, Catalytic pyrolysis of microalgae to high-quality liquid bio-fuels, Biomass Bioenergy 35 (2011) 3199–3207. [52] X. Qu, P. Liang, Z.F. Wang, R. Zhang, D.K. Sun, X.K. Gong, Z.X. Gan, J.C. Bi, Pilot development of a polygeneration process of circulating fluidized bed combustion combined with coal pyrolysis, Chem. Eng. Technol. 34 (2011) 61–68. [53] L.Y. Chen, L.W. Lin, Z.S. Xu, X.S. Li, T. Zhang, Dehydro-oligomerization of methane to ethylene and aromatics over molybdenum/HZSM-5 catalyst, J. Catal. 157 (1995) 190–200. [54] D. Sutton, B. Kelleher, J.R.H. Ross, Review of literature on catalysts for biomass gasification, Fuel Process. Technol. 73 (2001) 155–173.