- Email: [email protected]
Selective production of chemicals from biomass pyrolysis over metal chlorides supported on zeolite
Accepted Manuscript Selective production of chemicals from biomass pyrolysis over metal chlorides supported on zeolite Leng Shuai, Wang Xinde, Cai Qiuxia, Ma Fengyun, Liu Yue’e, Wang Jianguo PII: DOI: Reference:
S0960-8524(13)01528-9 http://dx.doi.org/10.1016/j.biortech.2013.09.096 BITE 12461
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
12 August 2013 19 September 2013 21 September 2013
Please cite this article as: Shuai, L., Xinde, W., Qiuxia, C., Fengyun, M., Yue’e, L., Jianguo, W., Selective production of chemicals from biomass pyrolysis over metal chlorides supported on zeolite, Bioresource Technology (2013), doi: http://dx.doi.org/10.1016/j.biortech.2013.09.096
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Selective Production of Chemicals from Biomass Pyrolysis over Metal Chlorides Supported on Zeolite Leng Shuai a, Wang Xinde a, Cai Qiuxia a, Ma Fengyun b, Liu Yue’e b, Wang Jianguo a* a College of Chemical Engineering and Materials Science, Zhejiang University of Technology, Hangzhou 310032,P.R. China b Key Laboratory of Oil and Gas Fine Chemicals of Ministry of Education, Xinjiang University, Urumuqi 830046, P.R. China * Corresponding author: Tel/fax: +86 571 88871037, E-mail: [email protected] Abstract Direct biomass conversion into chemicals remains a great challenge because of the complexity of the compounds; hence, this process has attracted less attention than conversion into fuel. In this study, we propose a simple one-step method for converting bagasse into furfural (FF) and acetic acid (AC). In this method, bagasse pyrolysis over ZnCl2/HZSM-5 achieved a high FF and AC yield (58.10%) and a 1.01 FF/AC ratio, but a very low yield of medium–boiling point components. However, bagasse pyrolysis using HZSM-5 alone or ZnCl2 alone still remained large amounts of medium–boiling point components or high–boiling point components. The synergistic effect of HZSM-5 and ZnCl2, which combines pyrolysis, zeolite cracking, and Lewis acid–selective catalysis results in highly efficient bagasse conversion into FF and AC. Therefore, our study provides a novel, simple method for directly converting biomass into high-yield useful chemical.
1
Key Words: Biomass Conversion; Chemical; Metal Chlorides; Pyrolysis; Synergistic Effect
1. Introduction Fossil resources are sources of fuel as well as the backbone of the contemporary chemical production. However, fossil reserves are gradually diminishing (Dietenberger and Anderson, 2007); hence, finding a renewable source for the production of chemicals and fuels is urgently necessary. Therefore, the conversion of biomass into fuel and energy has attracted increasing attention in recent years (Douglas et al., 2009; Teerawit et al., 2013; Zhang et al., 2005). Given their extremely complex biomass composition, obtaining one or several chemicals with high yields compared with pyrolysis bio-oil (the mixture of different compounds) is difficult and challenging. At present, most biomass conversion studies have focused on energy rather than chemicals (Alonso et al., 2013; Dickerson and Soria, 2013; Zhang et al., 2005). Several
methods
for
producing
chemicals
are
available,
such
as
5-hydroxymethylfurfural and furfural (FF), from the simple model substances and/or biomass. Zhao (Zhao et al., 2007) et al. first obtained 5-hydroxymethylfurfural from the conversion of sugars using ionic liquid as liquid phase environment and metal halides as catalysts, in which, chromium (II) chloride in 1-alkyl-3-methylimidazolium was found to be a uniquely effective catalyst. This pioneering work stimulated many studies for 5-hydroxymethylfurfural production using different substrates under different ionic liquid environments and metal chloride catalysts (Zhang et al., 2011). Paired
2
CuCl2/PdCl2 catalyst in 1-ethyl-3-methylimidazoliun chloride accelerated cellulose depolymerization and promoted the formation of 5-hydroxymethylfurfural (Su et al., 2011). The conversion of cellulose into 1-(4-sulfonic acid) butyl-3-methylimidazolium hydrogen sulfate by MnCl2 achieved high yields of 5-hydroxymethylfurfural and FF (Tao et al., 2011). However, the expensive ionic liquid (Weerachanchai et al., 2012) and the simple model substances rather than biomass (Liu et al., 2012; Zhang et al., 2013) were two major limits of this method. To overcome these limits, Yang (Yang et al., 2012) et al. obtained 5-hydroxymethylfurfural and 5-ethoxymethylfurfural from the catalytic conversion of glucose by AlCl3 in an ethanol–water solvent system. The aforementioned studies had been conducted under solvent environment. Recently, the pyrolysis of corncob (Lu et al., 2011) impregnated with acids, chlorides, and sulfates, under nitrogen atmosphere, resulted in the production of FF and acetic acid (AC). Moreover, they found that the addition of ZnCl2 improved the yield for FF and AC. Corncob pyrolysis over several metal salts was carried out in a packed-bed at 800 K, and the results indicated that H2SO4 and ZnCl2 treatment exhibited the best FF yield (Branca et al., 2012). However, the loss of metal is inevitable and some components with high boiling points are still hard to utilize regardless of whether ZnCl2 is mixed with biomass or impregnated in the pretreatment. The industrial preparation of FF includes acidolysis, stripping, condensation, distillation, and vacuum refining to obtain pure FF. However, the rigorous conditions impede its development and the enterprise scales are mainly SMEs. Therefore, efficiently and selectively transforming biomass into valuable chemicals, such as 5-hydroxymethylfurfural and FF, has become an urgent issue.
3
In this study, a novel, highly efficient, one-step method has been developed for selectively producing FF and AC from bagasse pyrolysis, over a series of metal chlorides on different supports. The results show that ZnCl2/HZSM-5 has the best catalytic performance because of the synergistic effect of HZSM-5 and ZnCl2, which combines pyrolysis, zeolite cracking, and Lewis acid–selective catalysis.
2. Materials and Methods 2.1. Catalyst synthesis and characterization The catalyst was prepared via incipient wetness coimpregnation using an aqueous solution (deionized water) containing metal precursors [a series of metal chlorides (AR, Aladdin)]. The aqueous solution was added to the supports [HZSM-5, SBA-15, and MCM-41 (Nanjing XFNANO Materials TECH Co., Ltd.); TiO2, C, and γ-Al2O3 (Aladdin)], fully stirred, and the mixture was set aside for 24 h. After impregnation, the catalyst was dried for 6 h at 80 °C and then kept at 110 °C for another 10 h. X-ray powder diffraction pattern was performed on an X’Pert PRO X-ray Diffractometer (PANalytical), using Cu Ka radiation generated at 40 kV and 40 mA. The scans ranged from 10° to 80°. The results of the GC-MS were analyzed on a ThermoFisher DSQ. The gas chromatography conditions were as follows: HP-5 MS elastic quartz capillary column (30 m × 0.25 mm × 0.25 μm); carrier gas: highly pure He, 1 mL min
-1
; column
temperature, 50 °C to 300 °C, 10 °C/min. The mass spectrum conditions were as follows: ET source, 70 eV; filament current: 100 μA; voltage multiplier: 1200 V; full
4
scan. The simulated distillations were carried on Agilent SimDis 6890N. The gas chromatograph conditions were as follows: column: DB-2887 SimDis (10 m × 0.53 mm × 3.00 μm); carrier gas: high purity He, 16 mL min -1; column temperature: 35 °C to 350 °C, 10 °C/min; FID temperature: 350 °C; H2 speed: 35 mL min
-1
; air speed:
350 mL min-1; PTV injection port, 100 °C to 350 °C, 15 °C/min; injection volume, 1 µL; split ratio, 2:1.
2.2 Catalytic activity measurements The pyrolysis experiments over different catalysts were evaluated in a tubular quartz packed-bed reactor that was heated by electricity. In each run, a well-mixed mixture (Gopakumar et al., 2011) of catalyst sample and bagasse powder (3g) (to a ratio) were placed at the center of the reactor quartz tube (diameter = 25 mm, length = 400 mm) above the layer of quartz wool. Before the reaction, a N2 flow (7800 h-1, 99.999%, Shanghai) was bubbled into the reactor to eliminate the air. The reactor was heated to 500 °C at a rate of 50 °C/min, held for 5 min to collect splitting products, and then cooled down to room temperature. The reaction products were collected in a conical flask in an ice-water bath and analyzed via GC out-line (Zhejiang Fuli Analytical Instrument Co. Ltd., 9790), using an AT SE-5 capillary column and a FID detector. The yields of FF and AC were calculated base on area percentage from GC. Principal component analysis of the liquid products was determined using GC-MS. The tail gas was excluded from the air.
5
3. Results and Discussion 3.1. Screening of metal chlorides We first conducted the blank bagasse pyrolysis experiments with and without HZSM-5. Without HZSM-5, the liquid products were very complex and uniformly distributed in the boiling range, with maximum yield for AC, but only 5.56% through GC-MS analysis. With HZSM-5, which had both Lewis acid sites and Bronsted acid sites, the yield for FF and AC increased but remained low (See Fig.1), and a yield distribution drift of higher boiling point components to lower boiling point components was detected due to the zeolite cracking. But it didn’t show that which acid sites played the major role. So a preliminary work of the role of Bronsted acid was also performed, which was impregnated by H2SO4 in advance, but the results found that the yield of liquid was very low (only about 13%). That’s to say that Bronsted acid sites were so strong that making the larger molecules crake into gas molecules. Therefore, to improve the yield to FF and AC, a series of metal chlorides supported on HZSM-5 to enhance the Lewis acid were evaluated (See Fig.1). Surprisingly, the yields of FF and AC were 2 to 3 times that with HZSM-5, and 4 to 6 times more than without HZSM-5. Meanwhile, the FF/AC ratio also significantly increased. Among these metal chlorides, ZnCl2 showed the best performance for FF production (Branca et al., 2012). The FF and AC yield was 58.10% and the FF/AC ratio was 1.01, which were much higher than those without catalysts (9.17% and 0.23). The FF/AC ratio over SnCl2/HZSM-5 was also high, up to 1.01, but the FF and AC yield decreased by about 11.7%. CuCl2/HZSM-5 showed worse effects than ZnCl2/HZSM-5 (FF and AC yield of 50.92%, FF/AC ratio of 0.91).
6
The other four catalysts (MnCl2/HZSM-5, NiCl2/HZSM-5, CaCl2/HZSM-5, and CoCl2/HZSM-5) exhibited nearly the same FF/AC ratio (about 0.55) and FF and AC yield (about 37.5 %), except for CaCl2/HZSM-5 (43%, slightly higher). ZnCl2/ HZSM-5 showed very good performance for the production of FF and AC, which indicates that the intensity of the Lewis acid has a vital role in the selectively conversion of bagasse into FF (Lu et al., 2011). In this work, the product distribution showed little difference, which was determined by heating rate and final temperature. The distribution had a range of 20-25% (solid), 25-30% (gas) and 45-55% (liquid). The supported catalyst had another superiority that the fix of chlorides, which could be identified by ICP-MS (ELAN DRC-e, Perkin-Elmer, Canada) towards to fresh (13.7%) and used catalyst (13.7%).
3.2. Optimal conditions for bagasse pyrolysis over ZnCl2/HZSM-5 ZnCl2 was the best catalyst supported on HZSM-5 for FF production from bagasse pyrolysis. The ratio of bagasse/catalyst and ZnCl2 loading was further optimized, as shown in Figs. 2 a and 2 b. The optimal loading of ZnCl2 supported on HZSM-5 was determined by changing the mass ratio of ZnCl2/HZSM-5 (Figs. 2 a). The catalyst with a mass ratio of 0.4 achieved the best FF and AC yield (58.10 %) yield and an FF/AC ratio (1.01). The lower or higher loadings had worse performances because the low loadings might not provide enough active sites, whereas the high loadings may deactivate aggregation to certain degree. A series of experiments of catalyst/bagasse mass ratio were also carried out (Figs. 2 b). When the ratio bagasse/catalysts was 1, the
7
FF and AC yield was slightly higher than that at the higher ratios (3, 5, 7), but the FF yield was extremely low (FF/AC ratio was below 0.4). The FF and AC yield decreased with increasing bagasse/catalysts ratio (7). Therefore, the bagasse/catalysts mass ratio was optimized to 5, at which the FF and AC yield was 58.10 % and the FF/AC ratio was about 1.01. The results of the simulated distillation liquid products were shown in Fig. 3. The curves could be divided into four parts based on the boiling temperature. These liquid products could be classified as low–boiling point (LBPC) if their boiling temperatures were less than 130 °C, medium–boiling point (MBPC) from 130 °C to 215 °C, and high–boiling point (HBPC) from 215 °C to 330 °C, which corresponded to the first three parts in the curve. The yields of these products at boiling temperatures above 330 °C without any catalyst and with ZnCl2/HZSM-5 were nearly the same, which might have mainly formed through polymerization reactions when the temperature increased because of the thermal instability of the liquid product. The liquid products without any catalyst showed a uniform distribution at different boiling points. Over ZnCl2/HZSM-5, the LBPC yields dramatically increased, whereas those of MBPCs was slightly decreased and those of the HPBCs was sharply decreased, which indicated that both HBPCs and MBPCs efficiently converted into LBPCs by ZnCl2/HZSM-5.
3.3. Influence of the state of ZnCl2 on FF and AC yield In traditional catalyst preparation, calcination is necessary to stabilize the catalyst. However, the growth of crystal grains is inevitable. Therefore, based on the effects of
8
thermal treatment, catalysts calcined at 500 °C for 3 h was also evaluated for comparison. A lower FF and AC yield of 29.29% was achieved using the calcined catalyst than using fresh catalyst (58.10%, Figs. S1 a). The TEM micrographs (Figs. S1 c and d) showed that ZnCl2 particles could not be observed in the fresh catalyst (Figs. S1 c), but were easily observed in the calcined catalyst (Figs. S1 d), which indicated that the ZnCl2 species were uniformly distributed or aggregated in the uncalcined and the calcined HZSM-5.This result indicated that the state of ZnCl2 significantly influenced the catalytic performance of the pyrolysis of bagasse over the ZnCl2/HZSM-5 catalyst. However, ZnCl2 clusters were still well dispersed even on calcined catalyst (Figs. S1 b), which meant that the calcined catalyst showed well physical property and only the state of ZnCl2 played a vital role. After reaction, the catalyst was surrounded by bio-char and was hard to be eliminated clearly. What’s more, the used catalyst suffered thermal treatment and the state of ZnCl2 must have some degree of agglomeration. Therefore, the recycled catalyst should have worse performance than fresh one.
3.4. Synergistic effect of ZnCl2 and HZSM-5 A series of supports (C, Al2O3, TiO2, HZSM-5, MCM-41, and SBA-15) was also investigated, as shown in Fig. 4. HZSM-5 showed the lowest FF and AC yield (19.49%) among the supports, whereas the traditional supports (TiO2, Al2O3, and C) showed better performance (28.24%, 29.68%, and 24.75%, respectively). MCM-41 and SBA-15, which had regular pore sizes, showed similar or slightly better yield for FF and AC
9
yield (21.91% and 27.98%) than the traditional supports. However, the FF and AC yield over ZnCl2 on these supports were different from those over supports without ZnCl2. The FF and AC yields of ZnCl2 supported on HZSM-5, MCM-41, and SBA-15 were 58.09%, 55.80%, and 63.58%, respectively. The performances over ZnCl2/zeolites were much higher than these on ZnCl2/traditional supports. This finding indicated that the zeolite and the ZnCl2 had different roles in bagasse pyrolysis. Zeolite, especially HZSM-5, had excellent catalytic cracking properties: effectively transforming HBPCs into MBPCs and small amounts of LBPCs. By contrasts, traditional supports partly converted MBPCs into LBPCs. ZnCl2 supported on zeolites produced high FF and AC yields, which indicated that ZnCl2 effective catalyzed the conversion of MBPCs into LBPCs (such as FF). Therefore, the excellent catalytic properties of ZnCl2/HZSM-5 catalysts synergistically improved cracking (zeolite) and Lewis acid–selective catalysis (ZnCl2). The uncatalyzed bagasse pyrolysis showed very little yield for the specific chemicals, in which many products were obtained in a wide boiling range (Fig. 5). The cracking over zeolite partly upgraded these products, but a large proportion of the products were MBPCs. Bagasse catalysis over ZnCl2 dramatically promoted LBPCs yield, but still had a certain extent of unconverted HBPCs. The ZnCl2 /HZSM-5 catalyst exhibited excellent LBPCs yield by combining pyrolysis, cracking, and selective catalysis into one process. Furthermore, the nature of ZnCl2 catalysis predominantly converted MBPCs, but the ability to transform HBPCs was weak, and this task was satisfactorily performed by HZSM-5. The role of HZSM-5 was primarily degrading HBPCs into MBPCs and small
10
amounts of LBPCs. The formation of LBPCs was mainly attributed to the cracked fragments.
4. Conclusions The direct conversion of bagasse, a typical biomass, into FF and AC was achieved through bagasse pyrolysis over ZnCl2/HZSM-5. The maximum FF and AC yield (58.10%) and a 1.01 FF/AC ratio was obtained. However, bagasse pyrolysis using HZSM-5 alone or ZnCl2 alone obtained large amounts of MBPCs or HBPCs. This simple one-step method that combined pyrolysis, zeolite cracking, and Lewis acid–selective catalysis resulted in highly efficient bagasse conversion into FF and AC because of the synergism between HZSM-5 and ZnCl2. Therefore, our study provided a novel, simple method for the direct conversion of biomass into high-yield useful chemicals.
Acknowledgements This work was supported by National Basic Research Program of China (973 Program) (2013CB733501), the National Natural Science Foundation of China (No. 21176221, 21136001, 21101137 and 21306169), Zhejiang Provincial Natural Science Foundation of China (No. R4110345), the New Century Excellent Talents in University Program (NCET-10-0979) and the Open Funding (No. XJDX0908-2012-7).
References
11
1) Alonso, D.M., Wettstein, S.G., Dumesic, J.A., 2013. Gamma - valerolactone, a sustainable platform molecule derived from lignocellulosic biomass. Green Chem. 15, 584–595. 2) Branca, C., Blasi, C.D., Galgano, A., 2012. Catalyst Screening for the Production of Furfural from Corncob Pyrolysis. Energ. Fuel. 26, 1520–1530. 3) Dickerson, T., Soria, J., 2013. Catalytic Fast Pyrolysis: A Review. Energies 6, 514– 538. 4) Dietenberger, M.A., Anderson, M., 2007. Vision of the US biofuel future: A case for hydrogen-enriched biomass gasification. Ind. Eng. Chem. Res. 46, 8863–8874. 5) Douglas, C.E., Todd, R.H., Gary, G.N., Leslie, J.R., Alan, H.Z., 2009. Catalytic Hydroprocessing of Biomass Fast Pyrolysis Bio-oil to Produce Hydrocarbon Products. Environ. Prog. Sustain. 28, 441–449. 6) Gopakumar, S.T., Adhikari, S., Gupta, R.B., Tu, M.B., Taylor, S., 2011. Production of hydrocarbon fuels from biomass using catalytic pyrolysis under helium and hydrogen environments. Bioresour. Technol. 102, 6742–6749. 7) Liu, D.J., Chen, E.Y.X., 2012. Ubiquitous aluminum alkyls and alkoxides as effective catalysts for glucose to HMF conversion in ionic liquids. Appl. Catal. A–Gen. 435 –436, 78–85. 8) Lu, Q., Dong, C.Q., Zhang, X.M., Tian, H.Y., Yang, Y.P., Zhu, X.F., 2011. Selective fast pyrolysis of biomass impregnated with ZnCl2 to produce furfural: Analytical PY-GC/MS study. J. Anal. Appl. Pyrolysis 90, 204–212. 9) Lu, Q., Wang, Z., Dong, C.Q., Zhang, Z.F., Zhang, Y., Yang, Y.P., Zhu, X.F., 2011.
12
Selective fast pyrolysis of biomass impregnated with ZnCl2 : Furfural production together with acetic acid and activated carbon as by-products. J. Anal. Appl. Pyrolysis 91, 273–279. 10) Su, Y., Brown, H.M., Li, G.S., Zhou, Z.D., Amonette, J.E., Fulton, J.L., Camaioni, D.M., Zhang, Z.C., 2011. Accelerated cellulose depolymerization catalyzed by paired chlorides in ionic liquid solvent. Appl. Catal. A - Gen. 391, 436–442. 11) Tao, F.R., Song, H.L., Yang, J., Chou, L.J., 2011. Catalytic hydrolysis of cellulose into furans in MnCl2-ionic liquid system. Carbohyd. Polym. 85, 363–368. 12) Teerawit, P., Tarit, N., Yuriy, R.L., 2013. Effective hydrodeoxygenation of biomass-derived oxygenates into unsaturated hydrocarbons by MoO3 using low H2 pressurest. Energ. Environ. Sci. 6, 1732－1738. 13) Yang, Y., Hu, C.W., Omar, M.M.A., 2012. Conversion of glucose into furans in the presence of AlCl3 in an ethanol-water solvent system. Bioresour. Technol. 116, 190 –194. 14) Weerachanchai, P., Leong, S.S.J., Chang, M.W., Ching, C.B., Lee, J.M., 2012. Improvement of biomass properties by pretreatment with ionic liquids for bioconversion process. Bioresour. Technol. 111, 453–459. 15) Zhang, L.X., Yu, H.B., Wang, P., Dong, H., Peng, X.H., 2013. Conversion of xylan, D-xylose and lignocellulosic biomass into furfural using AlCl3 as catalyst in ionic liquid. Bioresour. Technol. 130, 110–116. 16) Zhang, S.P., Yan, Y.J., Li, T.C., Ren, Z.W., 2005. Upgrading of liquid fuel from the pyrolysis of biomass. Bioresour. Technol. 96, 545–550.
13
17) Zhang, Z.H., Wang, Q., Xie, H.B., Liu, W.J., Zhao, Z.B., 2011. Catalytic Conversion of Carbohydrates into 5-Hydroxymethylfurfural by Germanium (IV) Chloride in Ionic Liquids. ChemSusChem 4, 131–138. 18) Zhao, H.B., Holladay, J.E., Brown, H., Zhang, Z.C., 2007. Metal Chlorides in Ionic Liquid Solvents Convert Sugars to 5-Hydroxymethylfurfural. Science 316, 1597– 1600.
14
Figure captions
Fig. 1 Yield of FF and AC and their ratios from the bagasse pyrolysis under different metal chlorides supported on HZSM-5 Fig. 2 Optimal conditions for bagasse pyrolysis over ZnCl2/HZSM-5. Mass ratio of (a) ZnCl2 to HZSM-5 and (b) bagasse to ZnCl2/HZSM-5. Fig. 3 Results of the simulated distillation liquid products from bagasse pyrolysis with and without ZnCl2/HZSM-5. Fig. 4 Synergistic effect of ZnCl2 and support on the production of FF and AC from bagasse pyrolysis Fig. 5 Reaction scheme and roles of different catalysts for bagasse pyrolysis
15
Fig. 1 Yield of FF and AC and their ratios from the bagasse pyrolysis under different metal chlorides supported on HZSM-5
16
Fig. 2 Optimal conditions for bagasse pyrolysis over ZnCl2/HZSM-5. Mass ratio of (a) ZnCl2 to HZSM-5 and (b) bagasse to ZnCl2/HZSM-5.
17
Fig. 3 Results of the simulated distillation liquid products from bagasse pyrolysis with and without ZnCl2/HZSM-5.
18
Fig. 4 Synergistic effect of ZnCl2 and support on the production of FF and AC from bagasse pyrolysis
19
Fig. 5 Reaction scheme and roles of different catalysts for bagasse pyrolysis
20
Highlights:
A simple one step method was proposed to the conversion of biomass into chemical
The pyrolysis of bagasse with HZSM-5 or ZnCl2 obtains large amount of MBPC or HBPC
The pyrolysis of bagasse over ZnCl2/HZSM-5 obtains high yield of furfural
The synergistic effect makes ZnCl2/HZSM-5 high selectivity to chemicals.
S0960-8524(13)01528-9 http://dx.doi.org/10.1016/j.biortech.2013.09.096 BITE 12461
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
12 August 2013 19 September 2013 21 September 2013
Please cite this article as: Shuai, L., Xinde, W., Qiuxia, C., Fengyun, M., Yue’e, L., Jianguo, W., Selective production of chemicals from biomass pyrolysis over metal chlorides supported on zeolite, Bioresource Technology (2013), doi: http://dx.doi.org/10.1016/j.biortech.2013.09.096
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Selective Production of Chemicals from Biomass Pyrolysis over Metal Chlorides Supported on Zeolite Leng Shuai a, Wang Xinde a, Cai Qiuxia a, Ma Fengyun b, Liu Yue’e b, Wang Jianguo a* a College of Chemical Engineering and Materials Science, Zhejiang University of Technology, Hangzhou 310032,P.R. China b Key Laboratory of Oil and Gas Fine Chemicals of Ministry of Education, Xinjiang University, Urumuqi 830046, P.R. China * Corresponding author: Tel/fax: +86 571 88871037, E-mail: [email protected] Abstract Direct biomass conversion into chemicals remains a great challenge because of the complexity of the compounds; hence, this process has attracted less attention than conversion into fuel. In this study, we propose a simple one-step method for converting bagasse into furfural (FF) and acetic acid (AC). In this method, bagasse pyrolysis over ZnCl2/HZSM-5 achieved a high FF and AC yield (58.10%) and a 1.01 FF/AC ratio, but a very low yield of medium–boiling point components. However, bagasse pyrolysis using HZSM-5 alone or ZnCl2 alone still remained large amounts of medium–boiling point components or high–boiling point components. The synergistic effect of HZSM-5 and ZnCl2, which combines pyrolysis, zeolite cracking, and Lewis acid–selective catalysis results in highly efficient bagasse conversion into FF and AC. Therefore, our study provides a novel, simple method for directly converting biomass into high-yield useful chemical.
1
Key Words: Biomass Conversion; Chemical; Metal Chlorides; Pyrolysis; Synergistic Effect
1. Introduction Fossil resources are sources of fuel as well as the backbone of the contemporary chemical production. However, fossil reserves are gradually diminishing (Dietenberger and Anderson, 2007); hence, finding a renewable source for the production of chemicals and fuels is urgently necessary. Therefore, the conversion of biomass into fuel and energy has attracted increasing attention in recent years (Douglas et al., 2009; Teerawit et al., 2013; Zhang et al., 2005). Given their extremely complex biomass composition, obtaining one or several chemicals with high yields compared with pyrolysis bio-oil (the mixture of different compounds) is difficult and challenging. At present, most biomass conversion studies have focused on energy rather than chemicals (Alonso et al., 2013; Dickerson and Soria, 2013; Zhang et al., 2005). Several
methods
for
producing
chemicals
are
available,
such
as
5-hydroxymethylfurfural and furfural (FF), from the simple model substances and/or biomass. Zhao (Zhao et al., 2007) et al. first obtained 5-hydroxymethylfurfural from the conversion of sugars using ionic liquid as liquid phase environment and metal halides as catalysts, in which, chromium (II) chloride in 1-alkyl-3-methylimidazolium was found to be a uniquely effective catalyst. This pioneering work stimulated many studies for 5-hydroxymethylfurfural production using different substrates under different ionic liquid environments and metal chloride catalysts (Zhang et al., 2011). Paired
2
CuCl2/PdCl2 catalyst in 1-ethyl-3-methylimidazoliun chloride accelerated cellulose depolymerization and promoted the formation of 5-hydroxymethylfurfural (Su et al., 2011). The conversion of cellulose into 1-(4-sulfonic acid) butyl-3-methylimidazolium hydrogen sulfate by MnCl2 achieved high yields of 5-hydroxymethylfurfural and FF (Tao et al., 2011). However, the expensive ionic liquid (Weerachanchai et al., 2012) and the simple model substances rather than biomass (Liu et al., 2012; Zhang et al., 2013) were two major limits of this method. To overcome these limits, Yang (Yang et al., 2012) et al. obtained 5-hydroxymethylfurfural and 5-ethoxymethylfurfural from the catalytic conversion of glucose by AlCl3 in an ethanol–water solvent system. The aforementioned studies had been conducted under solvent environment. Recently, the pyrolysis of corncob (Lu et al., 2011) impregnated with acids, chlorides, and sulfates, under nitrogen atmosphere, resulted in the production of FF and acetic acid (AC). Moreover, they found that the addition of ZnCl2 improved the yield for FF and AC. Corncob pyrolysis over several metal salts was carried out in a packed-bed at 800 K, and the results indicated that H2SO4 and ZnCl2 treatment exhibited the best FF yield (Branca et al., 2012). However, the loss of metal is inevitable and some components with high boiling points are still hard to utilize regardless of whether ZnCl2 is mixed with biomass or impregnated in the pretreatment. The industrial preparation of FF includes acidolysis, stripping, condensation, distillation, and vacuum refining to obtain pure FF. However, the rigorous conditions impede its development and the enterprise scales are mainly SMEs. Therefore, efficiently and selectively transforming biomass into valuable chemicals, such as 5-hydroxymethylfurfural and FF, has become an urgent issue.
3
In this study, a novel, highly efficient, one-step method has been developed for selectively producing FF and AC from bagasse pyrolysis, over a series of metal chlorides on different supports. The results show that ZnCl2/HZSM-5 has the best catalytic performance because of the synergistic effect of HZSM-5 and ZnCl2, which combines pyrolysis, zeolite cracking, and Lewis acid–selective catalysis.
2. Materials and Methods 2.1. Catalyst synthesis and characterization The catalyst was prepared via incipient wetness coimpregnation using an aqueous solution (deionized water) containing metal precursors [a series of metal chlorides (AR, Aladdin)]. The aqueous solution was added to the supports [HZSM-5, SBA-15, and MCM-41 (Nanjing XFNANO Materials TECH Co., Ltd.); TiO2, C, and γ-Al2O3 (Aladdin)], fully stirred, and the mixture was set aside for 24 h. After impregnation, the catalyst was dried for 6 h at 80 °C and then kept at 110 °C for another 10 h. X-ray powder diffraction pattern was performed on an X’Pert PRO X-ray Diffractometer (PANalytical), using Cu Ka radiation generated at 40 kV and 40 mA. The scans ranged from 10° to 80°. The results of the GC-MS were analyzed on a ThermoFisher DSQ. The gas chromatography conditions were as follows: HP-5 MS elastic quartz capillary column (30 m × 0.25 mm × 0.25 μm); carrier gas: highly pure He, 1 mL min
-1
; column
temperature, 50 °C to 300 °C, 10 °C/min. The mass spectrum conditions were as follows: ET source, 70 eV; filament current: 100 μA; voltage multiplier: 1200 V; full
4
scan. The simulated distillations were carried on Agilent SimDis 6890N. The gas chromatograph conditions were as follows: column: DB-2887 SimDis (10 m × 0.53 mm × 3.00 μm); carrier gas: high purity He, 16 mL min -1; column temperature: 35 °C to 350 °C, 10 °C/min; FID temperature: 350 °C; H2 speed: 35 mL min
-1
; air speed:
350 mL min-1; PTV injection port, 100 °C to 350 °C, 15 °C/min; injection volume, 1 µL; split ratio, 2:1.
2.2 Catalytic activity measurements The pyrolysis experiments over different catalysts were evaluated in a tubular quartz packed-bed reactor that was heated by electricity. In each run, a well-mixed mixture (Gopakumar et al., 2011) of catalyst sample and bagasse powder (3g) (to a ratio) were placed at the center of the reactor quartz tube (diameter = 25 mm, length = 400 mm) above the layer of quartz wool. Before the reaction, a N2 flow (7800 h-1, 99.999%, Shanghai) was bubbled into the reactor to eliminate the air. The reactor was heated to 500 °C at a rate of 50 °C/min, held for 5 min to collect splitting products, and then cooled down to room temperature. The reaction products were collected in a conical flask in an ice-water bath and analyzed via GC out-line (Zhejiang Fuli Analytical Instrument Co. Ltd., 9790), using an AT SE-5 capillary column and a FID detector. The yields of FF and AC were calculated base on area percentage from GC. Principal component analysis of the liquid products was determined using GC-MS. The tail gas was excluded from the air.
5
3. Results and Discussion 3.1. Screening of metal chlorides We first conducted the blank bagasse pyrolysis experiments with and without HZSM-5. Without HZSM-5, the liquid products were very complex and uniformly distributed in the boiling range, with maximum yield for AC, but only 5.56% through GC-MS analysis. With HZSM-5, which had both Lewis acid sites and Bronsted acid sites, the yield for FF and AC increased but remained low (See Fig.1), and a yield distribution drift of higher boiling point components to lower boiling point components was detected due to the zeolite cracking. But it didn’t show that which acid sites played the major role. So a preliminary work of the role of Bronsted acid was also performed, which was impregnated by H2SO4 in advance, but the results found that the yield of liquid was very low (only about 13%). That’s to say that Bronsted acid sites were so strong that making the larger molecules crake into gas molecules. Therefore, to improve the yield to FF and AC, a series of metal chlorides supported on HZSM-5 to enhance the Lewis acid were evaluated (See Fig.1). Surprisingly, the yields of FF and AC were 2 to 3 times that with HZSM-5, and 4 to 6 times more than without HZSM-5. Meanwhile, the FF/AC ratio also significantly increased. Among these metal chlorides, ZnCl2 showed the best performance for FF production (Branca et al., 2012). The FF and AC yield was 58.10% and the FF/AC ratio was 1.01, which were much higher than those without catalysts (9.17% and 0.23). The FF/AC ratio over SnCl2/HZSM-5 was also high, up to 1.01, but the FF and AC yield decreased by about 11.7%. CuCl2/HZSM-5 showed worse effects than ZnCl2/HZSM-5 (FF and AC yield of 50.92%, FF/AC ratio of 0.91).
6
The other four catalysts (MnCl2/HZSM-5, NiCl2/HZSM-5, CaCl2/HZSM-5, and CoCl2/HZSM-5) exhibited nearly the same FF/AC ratio (about 0.55) and FF and AC yield (about 37.5 %), except for CaCl2/HZSM-5 (43%, slightly higher). ZnCl2/ HZSM-5 showed very good performance for the production of FF and AC, which indicates that the intensity of the Lewis acid has a vital role in the selectively conversion of bagasse into FF (Lu et al., 2011). In this work, the product distribution showed little difference, which was determined by heating rate and final temperature. The distribution had a range of 20-25% (solid), 25-30% (gas) and 45-55% (liquid). The supported catalyst had another superiority that the fix of chlorides, which could be identified by ICP-MS (ELAN DRC-e, Perkin-Elmer, Canada) towards to fresh (13.7%) and used catalyst (13.7%).
3.2. Optimal conditions for bagasse pyrolysis over ZnCl2/HZSM-5 ZnCl2 was the best catalyst supported on HZSM-5 for FF production from bagasse pyrolysis. The ratio of bagasse/catalyst and ZnCl2 loading was further optimized, as shown in Figs. 2 a and 2 b. The optimal loading of ZnCl2 supported on HZSM-5 was determined by changing the mass ratio of ZnCl2/HZSM-5 (Figs. 2 a). The catalyst with a mass ratio of 0.4 achieved the best FF and AC yield (58.10 %) yield and an FF/AC ratio (1.01). The lower or higher loadings had worse performances because the low loadings might not provide enough active sites, whereas the high loadings may deactivate aggregation to certain degree. A series of experiments of catalyst/bagasse mass ratio were also carried out (Figs. 2 b). When the ratio bagasse/catalysts was 1, the
7
FF and AC yield was slightly higher than that at the higher ratios (3, 5, 7), but the FF yield was extremely low (FF/AC ratio was below 0.4). The FF and AC yield decreased with increasing bagasse/catalysts ratio (7). Therefore, the bagasse/catalysts mass ratio was optimized to 5, at which the FF and AC yield was 58.10 % and the FF/AC ratio was about 1.01. The results of the simulated distillation liquid products were shown in Fig. 3. The curves could be divided into four parts based on the boiling temperature. These liquid products could be classified as low–boiling point (LBPC) if their boiling temperatures were less than 130 °C, medium–boiling point (MBPC) from 130 °C to 215 °C, and high–boiling point (HBPC) from 215 °C to 330 °C, which corresponded to the first three parts in the curve. The yields of these products at boiling temperatures above 330 °C without any catalyst and with ZnCl2/HZSM-5 were nearly the same, which might have mainly formed through polymerization reactions when the temperature increased because of the thermal instability of the liquid product. The liquid products without any catalyst showed a uniform distribution at different boiling points. Over ZnCl2/HZSM-5, the LBPC yields dramatically increased, whereas those of MBPCs was slightly decreased and those of the HPBCs was sharply decreased, which indicated that both HBPCs and MBPCs efficiently converted into LBPCs by ZnCl2/HZSM-5.
3.3. Influence of the state of ZnCl2 on FF and AC yield In traditional catalyst preparation, calcination is necessary to stabilize the catalyst. However, the growth of crystal grains is inevitable. Therefore, based on the effects of
8
thermal treatment, catalysts calcined at 500 °C for 3 h was also evaluated for comparison. A lower FF and AC yield of 29.29% was achieved using the calcined catalyst than using fresh catalyst (58.10%, Figs. S1 a). The TEM micrographs (Figs. S1 c and d) showed that ZnCl2 particles could not be observed in the fresh catalyst (Figs. S1 c), but were easily observed in the calcined catalyst (Figs. S1 d), which indicated that the ZnCl2 species were uniformly distributed or aggregated in the uncalcined and the calcined HZSM-5.This result indicated that the state of ZnCl2 significantly influenced the catalytic performance of the pyrolysis of bagasse over the ZnCl2/HZSM-5 catalyst. However, ZnCl2 clusters were still well dispersed even on calcined catalyst (Figs. S1 b), which meant that the calcined catalyst showed well physical property and only the state of ZnCl2 played a vital role. After reaction, the catalyst was surrounded by bio-char and was hard to be eliminated clearly. What’s more, the used catalyst suffered thermal treatment and the state of ZnCl2 must have some degree of agglomeration. Therefore, the recycled catalyst should have worse performance than fresh one.
3.4. Synergistic effect of ZnCl2 and HZSM-5 A series of supports (C, Al2O3, TiO2, HZSM-5, MCM-41, and SBA-15) was also investigated, as shown in Fig. 4. HZSM-5 showed the lowest FF and AC yield (19.49%) among the supports, whereas the traditional supports (TiO2, Al2O3, and C) showed better performance (28.24%, 29.68%, and 24.75%, respectively). MCM-41 and SBA-15, which had regular pore sizes, showed similar or slightly better yield for FF and AC
9
yield (21.91% and 27.98%) than the traditional supports. However, the FF and AC yield over ZnCl2 on these supports were different from those over supports without ZnCl2. The FF and AC yields of ZnCl2 supported on HZSM-5, MCM-41, and SBA-15 were 58.09%, 55.80%, and 63.58%, respectively. The performances over ZnCl2/zeolites were much higher than these on ZnCl2/traditional supports. This finding indicated that the zeolite and the ZnCl2 had different roles in bagasse pyrolysis. Zeolite, especially HZSM-5, had excellent catalytic cracking properties: effectively transforming HBPCs into MBPCs and small amounts of LBPCs. By contrasts, traditional supports partly converted MBPCs into LBPCs. ZnCl2 supported on zeolites produced high FF and AC yields, which indicated that ZnCl2 effective catalyzed the conversion of MBPCs into LBPCs (such as FF). Therefore, the excellent catalytic properties of ZnCl2/HZSM-5 catalysts synergistically improved cracking (zeolite) and Lewis acid–selective catalysis (ZnCl2). The uncatalyzed bagasse pyrolysis showed very little yield for the specific chemicals, in which many products were obtained in a wide boiling range (Fig. 5). The cracking over zeolite partly upgraded these products, but a large proportion of the products were MBPCs. Bagasse catalysis over ZnCl2 dramatically promoted LBPCs yield, but still had a certain extent of unconverted HBPCs. The ZnCl2 /HZSM-5 catalyst exhibited excellent LBPCs yield by combining pyrolysis, cracking, and selective catalysis into one process. Furthermore, the nature of ZnCl2 catalysis predominantly converted MBPCs, but the ability to transform HBPCs was weak, and this task was satisfactorily performed by HZSM-5. The role of HZSM-5 was primarily degrading HBPCs into MBPCs and small
10
amounts of LBPCs. The formation of LBPCs was mainly attributed to the cracked fragments.
4. Conclusions The direct conversion of bagasse, a typical biomass, into FF and AC was achieved through bagasse pyrolysis over ZnCl2/HZSM-5. The maximum FF and AC yield (58.10%) and a 1.01 FF/AC ratio was obtained. However, bagasse pyrolysis using HZSM-5 alone or ZnCl2 alone obtained large amounts of MBPCs or HBPCs. This simple one-step method that combined pyrolysis, zeolite cracking, and Lewis acid–selective catalysis resulted in highly efficient bagasse conversion into FF and AC because of the synergism between HZSM-5 and ZnCl2. Therefore, our study provided a novel, simple method for the direct conversion of biomass into high-yield useful chemicals.
Acknowledgements This work was supported by National Basic Research Program of China (973 Program) (2013CB733501), the National Natural Science Foundation of China (No. 21176221, 21136001, 21101137 and 21306169), Zhejiang Provincial Natural Science Foundation of China (No. R4110345), the New Century Excellent Talents in University Program (NCET-10-0979) and the Open Funding (No. XJDX0908-2012-7).
References
11
1) Alonso, D.M., Wettstein, S.G., Dumesic, J.A., 2013. Gamma - valerolactone, a sustainable platform molecule derived from lignocellulosic biomass. Green Chem. 15, 584–595. 2) Branca, C., Blasi, C.D., Galgano, A., 2012. Catalyst Screening for the Production of Furfural from Corncob Pyrolysis. Energ. Fuel. 26, 1520–1530. 3) Dickerson, T., Soria, J., 2013. Catalytic Fast Pyrolysis: A Review. Energies 6, 514– 538. 4) Dietenberger, M.A., Anderson, M., 2007. Vision of the US biofuel future: A case for hydrogen-enriched biomass gasification. Ind. Eng. Chem. Res. 46, 8863–8874. 5) Douglas, C.E., Todd, R.H., Gary, G.N., Leslie, J.R., Alan, H.Z., 2009. Catalytic Hydroprocessing of Biomass Fast Pyrolysis Bio-oil to Produce Hydrocarbon Products. Environ. Prog. Sustain. 28, 441–449. 6) Gopakumar, S.T., Adhikari, S., Gupta, R.B., Tu, M.B., Taylor, S., 2011. Production of hydrocarbon fuels from biomass using catalytic pyrolysis under helium and hydrogen environments. Bioresour. Technol. 102, 6742–6749. 7) Liu, D.J., Chen, E.Y.X., 2012. Ubiquitous aluminum alkyls and alkoxides as effective catalysts for glucose to HMF conversion in ionic liquids. Appl. Catal. A–Gen. 435 –436, 78–85. 8) Lu, Q., Dong, C.Q., Zhang, X.M., Tian, H.Y., Yang, Y.P., Zhu, X.F., 2011. Selective fast pyrolysis of biomass impregnated with ZnCl2 to produce furfural: Analytical PY-GC/MS study. J. Anal. Appl. Pyrolysis 90, 204–212. 9) Lu, Q., Wang, Z., Dong, C.Q., Zhang, Z.F., Zhang, Y., Yang, Y.P., Zhu, X.F., 2011.
12
Selective fast pyrolysis of biomass impregnated with ZnCl2 : Furfural production together with acetic acid and activated carbon as by-products. J. Anal. Appl. Pyrolysis 91, 273–279. 10) Su, Y., Brown, H.M., Li, G.S., Zhou, Z.D., Amonette, J.E., Fulton, J.L., Camaioni, D.M., Zhang, Z.C., 2011. Accelerated cellulose depolymerization catalyzed by paired chlorides in ionic liquid solvent. Appl. Catal. A - Gen. 391, 436–442. 11) Tao, F.R., Song, H.L., Yang, J., Chou, L.J., 2011. Catalytic hydrolysis of cellulose into furans in MnCl2-ionic liquid system. Carbohyd. Polym. 85, 363–368. 12) Teerawit, P., Tarit, N., Yuriy, R.L., 2013. Effective hydrodeoxygenation of biomass-derived oxygenates into unsaturated hydrocarbons by MoO3 using low H2 pressurest. Energ. Environ. Sci. 6, 1732－1738. 13) Yang, Y., Hu, C.W., Omar, M.M.A., 2012. Conversion of glucose into furans in the presence of AlCl3 in an ethanol-water solvent system. Bioresour. Technol. 116, 190 –194. 14) Weerachanchai, P., Leong, S.S.J., Chang, M.W., Ching, C.B., Lee, J.M., 2012. Improvement of biomass properties by pretreatment with ionic liquids for bioconversion process. Bioresour. Technol. 111, 453–459. 15) Zhang, L.X., Yu, H.B., Wang, P., Dong, H., Peng, X.H., 2013. Conversion of xylan, D-xylose and lignocellulosic biomass into furfural using AlCl3 as catalyst in ionic liquid. Bioresour. Technol. 130, 110–116. 16) Zhang, S.P., Yan, Y.J., Li, T.C., Ren, Z.W., 2005. Upgrading of liquid fuel from the pyrolysis of biomass. Bioresour. Technol. 96, 545–550.
13
17) Zhang, Z.H., Wang, Q., Xie, H.B., Liu, W.J., Zhao, Z.B., 2011. Catalytic Conversion of Carbohydrates into 5-Hydroxymethylfurfural by Germanium (IV) Chloride in Ionic Liquids. ChemSusChem 4, 131–138. 18) Zhao, H.B., Holladay, J.E., Brown, H., Zhang, Z.C., 2007. Metal Chlorides in Ionic Liquid Solvents Convert Sugars to 5-Hydroxymethylfurfural. Science 316, 1597– 1600.
14
Figure captions
Fig. 1 Yield of FF and AC and their ratios from the bagasse pyrolysis under different metal chlorides supported on HZSM-5 Fig. 2 Optimal conditions for bagasse pyrolysis over ZnCl2/HZSM-5. Mass ratio of (a) ZnCl2 to HZSM-5 and (b) bagasse to ZnCl2/HZSM-5. Fig. 3 Results of the simulated distillation liquid products from bagasse pyrolysis with and without ZnCl2/HZSM-5. Fig. 4 Synergistic effect of ZnCl2 and support on the production of FF and AC from bagasse pyrolysis Fig. 5 Reaction scheme and roles of different catalysts for bagasse pyrolysis
15
Fig. 1 Yield of FF and AC and their ratios from the bagasse pyrolysis under different metal chlorides supported on HZSM-5
16
Fig. 2 Optimal conditions for bagasse pyrolysis over ZnCl2/HZSM-5. Mass ratio of (a) ZnCl2 to HZSM-5 and (b) bagasse to ZnCl2/HZSM-5.
17
Fig. 3 Results of the simulated distillation liquid products from bagasse pyrolysis with and without ZnCl2/HZSM-5.
18
Fig. 4 Synergistic effect of ZnCl2 and support on the production of FF and AC from bagasse pyrolysis
19
Fig. 5 Reaction scheme and roles of different catalysts for bagasse pyrolysis
20
Highlights:
A simple one step method was proposed to the conversion of biomass into chemical
The pyrolysis of bagasse with HZSM-5 or ZnCl2 obtains large amount of MBPC or HBPC
The pyrolysis of bagasse over ZnCl2/HZSM-5 obtains high yield of furfural
The synergistic effect makes ZnCl2/HZSM-5 high selectivity to chemicals.