Modeling and Kinetic Study of Degradative Solvent Extraction of

Apr 17, 2017 - *Telephone/Fax: +86-27-87545526. E-mail: [email protected]., *Telephone/Fax: +86-27-87545526. E-mail: [email protected]...
0 downloads 0 Views 933KB Size
Download PDF
Article pubs.acs.org/EF

Modeling and Kinetic Study of Degradative Solvent Extraction of Biomass Wastes Xianqing Zhu,† Jian Tang,† Xian Li,*,† Wei Lan,† Kai Xu,† Yuan Fang,† Ryuichi Ashida,‡ Kouichi Miura,§ Guangqian Luo,† and Hong Yao*,† †

State Key Laboratory of Coal Combustion, School of Energy and Power Engineering, Huazhong University of Science and Technology, Wuhan, Hubei 430074, People’s Republic of China ‡ Department of Chemical Engineering and §Institute of Advanced Energy, Kyoto University, Kyoto 615-8510, Japan ABSTRACT: The degradative solvent extraction (DSE) has been found to be an effective method to convert and upgrade biomass wastes into high-quality extracts with various applications. However, the reaction pathways and kinetics of this DSE process are still not clear and need to be clarified. Hence, in this study, the biomass was treated in 1-methylnaphthalene at various conditions (300−350 °C and 0−90 min) by the DSE method and separated into five products: residue (unreacted biomass), two extracts (deposit and soluble), liquid, and gas. A lumped reaction model was proposed to simulate this biomass DSE process. The kinetic parameters were estimated by the least squares fit based on the experimental data and the model, which was optimized by the MATLAB optimization program. The results showed that the proposed model was valid and well-capable of describing the biomass DSE process. It can be concluded that different conversion pathways existed at 300 and 350 °C. At 300 °C, the dominant reactions were the conversion of residue (unreacted biomass) to the extraction products [deposit (k = 0.0104 min−1), soluble (k = 0.0042 min−1), and liquid (k = 0.0044 min−1)]. The deoxygenation reactions mainly occurred at relatively mild conditions. While at 350 °C, the rate-controlled process was the conversion of high-molecular-weight extract (deposit) to lowmolecular-weight extract (soluble) (k = 0.0155 min−1). A comprehensive understanding of the reaction pathways of the biomass DSE conversion process was provided.

1. INTRODUCTION As a result of the concerns of fossil fuel exhaustion and climate change, renewable alternative resources are receiving more and more interest and growingly considered as potential substitutes for fossil energies. Among them, biomass is one of the most important and abundant renewable energy resources, which accounts for around 10−13% of the world primary energy consumption.1−3 In addition, biomass has less negative impacts on the environment compared to fossil fuels, owing to its low sulfur and nitrogen contents as well as CO2-neutral characteristic.4−6 Therefore, biomass is expected to play a more vital role in energy production in the near future. The existing biomass utilization technologies mainly include thermochemical conversion technologies (such as pyrolysis, gasification, and combustion)7−9 and biochemical conversion technologies (such as digestion and fermentation).10 Generally, thermochemical processes have higher utilization efficiencies than biochemical processes in terms of the much lower reaction time and the superior capability to convert most of the organic compounds in biomass, especially for lignin.11 The thermochemical processes are the most common methods up to date. However, the inherently disadvantageous features of biomass, such as high moisture and oxygen contents and low energy density, lead to its low economic efficiency and limit its practical application for conventional thermochemical technologies.12 Hence, to realize the efficient utilization of biomass feedstocks, it is necessary to achieve the dewatering and deoxygenation of them before introducing them into traditional thermochemical processes. This process is referred to as biomass upgrading. © XXXX American Chemical Society

Currently, two main biomass upgrading technologies, slow pyrolysis (torrefaction)13−16 and hydrothermal carbonization,17−20 have been employed to upgrade and convert biomass into carbonaceous solid products (biochar and hydrochar). Slow pyrolysis of biomass is typically performed between 200 and 300 °C with a low heating rate in the absence of oxygen. Hydrothermal carbonization is the conversion of biomass to hydrochar using subcritical water at moderate temperatures (180−300 °C). In general, these two methods could increase the carbon content and reduce the oxygen content of the biomass feedstocks to some extent and narrow the difference between the diversities of biomasses. However, the carbon contents of biochars produced from slow pyrolysis are generally lower than 65%, and the oxygen contents are still as considerably high as 25−35%.21 As for hydrothermal carbonization, it is commonly carried out under harsh subcritical conditions with the pressure higher than 16 MPa, which requires a complex and expensive reaction system. Moreover, the ash contents of both biochar and hydrochar are higher than those of the corresponding raw biomasses, which can result in ash-related problems during further utilization. Apart from the traditional technologies mentioned above for biomass upgrading, an alternatively novel method (termed degradative solvent extraction, abbreviated to DSE) has been recently proposed to upgrade and convert a wide variety of biomass wastes into several high-quality solid fractions in our Received: December 23, 2016 Revised: April 7, 2017 Published: April 17, 2017 A

DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels Table 1. Ultimate and Proximate Analyses of SD ultimate analysis (wt %, daf)

a

proximate analysis (wt %, db)

atomic ratio

sample

C

H

N

Oa

VM

A

FC

H/C

O/C

SD

53.1

5.8

0.1

41.0

84.4

1.2

14.4

1.3

0.6

By difference.

previous works.22−24 In this method, the biomass feedstocks are treated in a nonpolar solvent below 350 °C under an inert atmosphere by a specially designed autoclave. The obtained solid products are composed of a non-extractable fraction (termed residue), high-molecular-weight extract (deposit), which precipitates from the solvent at room temperature, and low-molecular-weight extract (soluble), which is maintained soluble in the solvent at room temperature. It was found that the physicochemical properties of the solubles or deposits obtained from different kinds of biomass feedstocks resemble each other, which means that the DSE method narrows the differences between different types of biomasses. In comparison to raw biomasses, the solubles and deposits have much higher carbon contents (as high as 85%) and lower oxygen contents (as low as 7.3%). Furthermore, solubles and deposits contain almost no ash and exhibit unique thermoplastic properties. In our previous works, the high feasibilities of using the soluble and deposit for several purposes have already been found. The soluble and deposit (especially deposit) are effective additives for coke making, which can markedly improve the quality of coke.25 Benefiting from the thermoplastic properties and high carbon content, the soluble is a good precursor for carbon fiber production.26 In addition, as a result of the low oxygen and ash contents, the soluble and deposit can be further liquefied or pyrolyzed for high-quality liquid fuel or chemicals production.27,28 Our previous works mainly focused on the characterizations and utilizations of the products (soluble and deposit) obtained by biomass DSE. The reaction mechanisms of the biomass DSE process are complex and still remain unclear. The determination of proper reaction pathways describing the mechanism or interconversion of the product fractions is essential and meaningful in providing basic information for the process optimization or reactor design for a future development of the biomass DSE method. However, it has been rarely reported on the reaction pathways and kinetics of the DSE method. Therefore, in this study, a reaction kinetic model was proposed to simulate the biomass DSE process. The kinetic parameters were estimated by the least squares fit based on the experimental and model results, which were optimized by the MATLAB optimization program (fourth-order Runge−Kutta numerical integration).29 The main reaction pathways as well as the rate-dominating reaction during the biomass DSE process were established according to the obtained kinetic parameters. This study is supposed to provide a comprehensive understanding of the reaction pathways of the biomass DSE process.

Figure 1. Diagram of the DSE apparatus. The detailed experimental steps had already been described in our previous studies.25 Briefly, around 20 g of SD (air-dried basis) and 300 mL of 1-MN were mixed and placed into the autoclave extractor. After purged adequately by N2, the extractor was sealed with pure N2 (99.999%) of an initial pressure of 0.2 MPa, heated to extraction temperature (300 or 350 °C) at the heating rate of 5 °C/min, and maintained at the extraction temperature for the designed residence time (0, 15, 30, 45, 60, and 90 min). When time reached the designed residence time, the valve below the filter was quickly opened to allow the mixture of the extracts and solvent to be transferred to the reservoir, implementing the thermal in situ separation of the extracts from the non-extractable fraction (residue). The residue was considered as the unreacted biomass during the kinetic model calculation in this work. The reservoir was cooled to ambient temperature by circulating cooling water. A part of the extracts, which was extractable at the extraction temperature but precipitated as a solid in the reservoir at ambient temperature, was obtained by filtration and termed as the deposit. Another part of the solid extracts, which was still dissolved in the solvent at ambient temperature, was obtained as a solid through removing the solvent by a vacuum rotary evaporator and termed as soluble. The schematic diagram of the separation procedure of biomass DSE products is presented in Figure 2. Subsequently, the solid residue, deposit, and soluble were further dried at 150 °C under vacuum to remove the residual solvent sufficiently. The gaseous products (gas) were collected in a gas bag and analyzed quantitatively by a gas chromatograph (Agilent MicroGC3000). The yields of residue, deposit, soluble, and gas were calculated by their weights. The yield of liquid (mostly H2O) was then determined by mass balance. The extraction experiments were repeated, with the average values of the yields obtained and used for the kinetic calculation, and the experimental errors of the yields of the DSE products were all less than 6%.

2. MATERIALS AND METHODS 2.1. Biomass and Solvent Used. A representative biomass waste, fir sawdust (SD), was used (from Wuhan, Hubei, China) for the experimental material. The proximate and elemental analyses of the raw SD are listed in Table 1. 1-Methylnaphthalene (1-MN), a nonpolar solvent (analytically pure), was selected for the DSE experiments. 2.2. DSE Procedure. The biomass DSE experiments were carried out in a specially self-designed batch autoclave, as shown in Figure 1.

3. RESULTS AND DISCUSSION 3.1. Product Yields of DSE of SD. In our previous study, it was found that the temperature higher than 300 °C was proper for the biomass DSE process.25 Consequently, in this study, B

DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

and 350 °C, respectively. It can be seen that the residue was the major product at the beginning of the DSE process at 300 °C. With the residence time increasing, the yield of residue decreased rapidly to 15.3% at the residence time of 90 min, while the yields of deposit, soluble, liquid, and gas all increased continuously with the residence time in the range of 0−90 min. According to the yield change trends mentioned above, it can be concluded that the main reaction during the SD DSE at 300 °C should be the conversion of residue to other products (such as deposit, soluble, liquid, and gas). On the contrary, the variation trends of the product yields at 350 °C were totally different from those at 300 °C, as shown in Table 3. Deposit, instead of residue, was the main solid product at 350 °C at the residence time of 0 min. The yields of residue were less than 5.3%, indicating the high conversion rate of biomass at 350 °C. The yield of deposit decreased unceasingly from 28.7% at 0 min to 8.8% at 90 min. The soluble yield first increased in the range of 0−45 min and was then kept steady. In the meantime, the yields of residue, gas, and liquid maintained only a minor change. According to the yield change trends above, it can be deduced that the main reaction during the SD DSE process at 350 °C may be the conversion of deposit to soluble. The different variation tendencies of the product yields suggest that the DSEs of biomass at 300 and 350 °C belong to different reaction stages of this process. Table 4 displays the elemental composition and proximate analyses of the biomass DSE solid products. It is obvious that the carbon contents of the solid products (soluble, deposit, and residue) increased significantly compared to raw SD, while the oxygen contents of the solid products were much lower than that of raw SD. The ash contents of the two extracts, solubles and deposits, were also significantly lower than that of raw SD. With the extraction temperature or residence time increasing, the carbon contents of the solubles and deposits increased and the oxygen contents decreased, respectively. This may further indicate that several conversion reactions occurred with the extraction temperature or residence time rising. 3.2. Kinetic Model Establishment. Owing to the complexity of biomass structures and the varieties of products derived from biomass DSE, it is unlikely for a simple chemical reaction to be employed to describe this process. Therefore, in

Figure 2. Schematic diagram of the separation procedure of biomass DSE products.

isothermal DSE of SD at two different temperatures, i.e., 300 and 350 °C, was carried out with residence time ranging from 0 to 90 min. Tables 2 and 3 show the DSE product yields at 300 Table 2. Product Yields (wt %, daf) of SD DSE at 300 °C time (min)

residue

deposit

soluble

liquid

gas

0 15 30 45 60 90

58.75 51.94 31.68 19.74 15.74 15.28

10.32 14.77 21.43 24.14 24.42 22.81

8.16 10.95 13.23 19.28 21.39 22.20

17.82 15.86 26.44 26.75 26.73 27.99

4.94 6.49 7.23 10.08 11.72 11.71

Table 3. Product Yields (wt %, daf) of SD DSE at 350 °C time (min)

residue

deposit

soluble

liquid

gas

0 15 30 45 60 90

3.51 3.58 4.99 5.09 5.06 5.24

28.71 21.46 16.80 12.73 11.06 8.76

25.62 29.23 37.39 40.91 40.59 40.32

28.28 31.83 25.62 26.96 30.91 30.56

13.88 13.90 15.19 14.31 12.39 15.11

Table 4. Ultimate and Proximate Analyses of Biomass DSE Solid Products ultimate analysis (wt %, daf) a

atomic ratio

T/RT (°C/min)

sample

C

H

N

O

VM

A

FC

H/C

O/C

300/0

soluble deposit residue soluble deposit residue soluble deposit residue soluble deposit residue soluble deposit residue

74.1 69.9 58.6 76.0 74.9 75.3 75.6 71.5 74.0 76.6 73.5 79.1 79.2 75.1 79.2

6.8 6.7 6.2 6.5 5.8 5.6 6.5 5.6 5.1 6.2 5.5 5.1 6.5 5.3 4.9

0.7 0.2 0.1 0.5 0.2 0.2 0.6 0.3 0.2 0.5 0.4 0.3 0.4 0.4 0.2

18.4 23.2 35.2 17.1 19.1 18.9 17.4 22.5 20.7 16.8 20.5 15.5 14.0 19.2 15.7

86.0 70.0 80.8 79.1 68.6 58.2 73.8 61.8 35.6 73.2 56.2 53.8 76.7 47.8 50.2

0.2 0.2 2.9 0.3 0.8 4.6 0.0 0.0 6.2 0.0 0.2 8.4 0.0 0.4 6.4

13.9 29.8 16.4 20.7 30.7 37.2 26.2 38.3 58.2 26.8 43.7 37.9 23.4 51.8 43.4

1.1 1.15 1.27 1.02 0.92 0.9 1.02 0.95 0.83 0.97 0.9 0.77 0.98 0.85 0.74

0.19 0.25 0.45 0.17 0.19 0.19 0.17 0.24 0.21 0.16 0.21 0.15 0.13 0.19 0.15

300/30

350/0

350/30

350/60

a

proximate analysis (wt %, db) b

Temperature/residence time. bBy difference. C

DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

where R, S, D, L, and G represent the yields of residue, deposit, soluble, liquid, and gas, respectively, and k1−k9 refer to the reaction rate constants of the corresponding reactions. The differential eqs 1−5 were solved to obtain the values of the rate constants using the experimental data shown in Tables 2 and 3. The solution method included integration of the differential equations (by MATLAB fourth-order Runge−Kutta method)29 and the least squares fit. The reaction rate constants were determined by minimizing the value of the objective function (OF, as displayed in eqs 6) using the MATLAB lsqnonlin function, which was defined as the minimum square difference between the calculated data and experimental data at a given temperature. Initially, the rate constants (k1−k9) were first given and set in the range of 0.0−0.5 min−1, to start the calculation and avoid overmuch computation time. The diagram of the solving process of the kinetic model was displayed in Figure 4. The computation ended, and the desired

this study, the biomass DSE process was described by a lumped reaction kinetic model, which has been successfully applied to study biomass or coal liquefaction, pyrolysis processes.30−32 Five extraction products were lumped as residue, deposit, soluble, liquid, and gas, respectively. The residence time of 0 min was defined as the beginning of the isothermal stage. DSE of biomass is a complex multiphase reaction process, which may consist of series, parallel, and retrogressive reactions. Because there is no reaction network available for the biomass DSE process up to date, the kinetic model was established on the basis of the following procedure: (1) First, a preliminary reaction model was proposed according to the yield change trends discussed above and our previous study for the reaction mechanism of the biomass DSE process.22 (2) Then, the mathematical formulas to describe the reaction model were developed and solved according to the experimental data, as shown in Tables 2 and 3. Thus, the reaction rate constants were obtained. (3) If the rate constants were non-convergent or unreasonable, the model was changed accordingly. If the rate constant of a certain reaction was close to 0, the corresponding reaction was removed from the model. (4) Repeat these aforementioned steps 1−3, until the finally appropriate model was established. The final kinetic model was established after many trials, as shown in Figure 3, which was found to be most valid and

Figure 3. Possible reaction pathways of DSE of biomasses.

appropriate for the biomass DSE process. This novel model was composed of series, parallel, and retrogressive reactions. There were nine reaction rate constants in this model, which were k1, k2, k3, k4, k5, k6, k7, k8, and k9, as shown in Figure 3. To develop the mathematic equations to describe the model, two assumptions were made for this reaction scheme: (1) All of the reactions were assumed as pseudo-first-order reactions in relation to the corresponding reactants. (2) The influences of mass transfer, particle size, etc. on biomass DSE were neglected, and all of the rate constants fitted the Arrhenius law. On the basis of the model presented in Figure 3 and the assumptions, the mathematical differential equations were obtained as shown in eqs 1−5 ∂R /∂t = k 2D − (k1 + k5 + k 7 + k 8)R

(1)

∂D/∂t = k1R − (k 2 + k 3 + k6)D

(2)

∂S /∂t = k 3D + k5R − k4S

(3)

∂L /∂t = k4S + k 7R + k6D − k 9L

(4)

∂G /∂t = k 8R + k 9L

(5)

Figure 4. Solving process of the kinetic model for DSE of biomass.

values of the rate constants (k1−k9) were obtained until |OFj + 1 − OFj| < 1 × 10−6. All calculation results retained four significant digits after the decimal point OF = min ∑ (Ymi − Yei)2 = min((R m − R e)2 + (Dm − De)2 i

+ (Sm − Se)2 + (Lm − Le)2 + (Gm − Ge)2 )

(6)

where Y represents the product yield, m and e represent calculated and experimental values of the product yield, respectively, and i represents R, D, S, L, or G. Figures 5 and 6 show the comparison of experimental yields and model yields of residue, deposit, soluble, liquid, and gas as a function of the residence time at 300 and 350 °C, respectively. D

DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Figure 7. Detailed reaction pathways of the biomass DSE process after fitting the proposed model at 300 and 350 °C (red color indicates the dominated reaction).

Figure 5. Experimental and model yields of residue, deposit, soluble, liquid, and gas as a function of the residence time at 300 °C.

soluble was apparently higher than that of deposit to soluble. This indicates that the low-molecular-weight extract soluble was preferred to be formed from raw biomass (residue) rather than from deposit at the low temperature of 300 °C. At a higher DSE temperature of 350 °C, the rate constants of deposit to soluble (k3 = 0.0155) were much larger than those of other reactions. Moreover, the yield of deposit was much larger than that of residue, as shown in Figure 6. Therefore, the reaction of deposit to soluble was dominant at 350 °C. The interconversion of the extraction products (mainly deposit to soluble) was the main reaction at 350 °C. Through the comparisons of reaction pathways and the rate constants at different temperatures, it shows that the residue was consumed to form deposit, soluble, liquid, and gas at the same time at 300 °C. It also implies that the residue in the model of 300 °C mainly consists of unreacted biomass. On the contrary, the residue in the model of 350 °C should be mainly the stable residual lignin and polymerized product formed through the cross-linking reaction. The rate constant of deposit to soluble at 350 °C was an order of magnitude higher than that at 300 °C. The liquid and gas were formed only from soluble and liquid at 350 °C, respectively. However, the liquid was formed from residue, deposit, and soluble, and the gas was formed from both residue and liquid at 300 °C. The rate constants of the reactions for the liquid and gas formation at 300 °C were also relatively higher. This indicates that the liquid and gas were formed at relatively mild conditions. The liquid and gas were mainly composed of H 2 O and CO 2 , respectively;27 therefore, the liquid and gas were formed mainly through the deoxygenation reactions of the biomass at relatively mild conditions. Figure 8 presented the comparison of the calculated product yields and the experimental product yields in the form of a parity diagram,34 which can be used to evaluate the accuracy of the model-fitting results.35 As observed, almost all data points clustered randomly on or close to the diagonal line of the parity. The fit quality of kinetic modeling is evaluated by eqs 7 and 836

Figure 6. Experimental and model yields of residue, deposit, soluble, liquid, and gas as a function of the residence time at 350 °C.

The figures show good correlations between the experimental and model results, indicating that the proposed model is wellcapable of describing the evolution trend of the evolution of the DSE product yields. Figures 5 and 6 also show that the change trends of the product yields were completely different, which implies that different conversion pathways exist at 300 and 350 °C. These different reaction pathways can be reflected quantitatively by the rate constants obtained by fitting the proposed reaction model. 3.3. Kinetic Model Analysis. Figure 7 shows the optimized reaction rate constants of the DSE of SD at 300 and 350 °C. The rate constants for the same reactions at 300 and 350 °C were different from each other, implying once again that the reaction pathways at 300 and 350 °C were different, as discussed above. According to the values of rate constants, the dominant reactions were the conversion of residue to deposit (k1R), residue to soluble (k9R), and residue to liquid (k6R) at 300 °C. This was mainly attributed to the decomposition of most hemicellulose and part of cellulose in raw biomass. Because it was reported in previous studies, the decomposition temperature of hemicellulose and cellulose was around 220− 315 and 280−400 °C, respectively.33 The rate constant of deposit to residue (k2) was much smaller than that of residue to deposit (k1), indicating that the inverse reaction was not significant at 300 °C. Hence, the conversions of raw biomass to extraction products were the rate-dominated processes for the biomass DSE at 300 °C. Besides, the reaction rate of residue to

⎛ SSD ⎞ ⎜ N ⎟ × 100% fit (%) = ⎜1 − (Yei)max ⎟⎟ ⎜ ⎝ ⎠

(7)

N

SSD =

∑ (Ymi − Yei)2 i=1

E

(8) DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Combustion (FSKLCCA1602), and Graduates’ Innovation Fund, Huazhong University of Science and Technology. The authors also appreciate the test assistances from the Analytical and Testing Center of Huazhong University of Science and Technology.

■ Figure 8. Parity diagram of the proposed kinetic model on the DSE of SD at all conditions (300 and 350 °C).

where Ymi and Yei represented calculated and experimental values of the product yields, respectively, and N was the total number of experimental points at all conditions (300 and 350 °C). As shown in Figure 8, the fit quality was 96.59%, indicating that the proposed model provided a satisfying description for the experimental results.

4. CONCLUSION The reaction pathways and kinetics of the biomass DSE process were investigated in detail. A general and lumped reaction model was proposed to simulate the biomass DSE process. The kinetic parameters were estimated by the least squares fit based on the experimental data and the model. The calculation results show that the proposed model was rather valid and well-capable of describing the biomass DSE process. The detailed reaction pathways of the biomass DSE process were clarified. Different conversion pathways existed at 300 and 350 °C. At 300 °C, the dominant reactions were the conversion of residue (unreacted biomass) to the extraction products (deposit, soluble, and liquid). The deoxygenation reactions mainly occurred at relatively mild conditions. While at 350 °C, the rate-controlled process is the conversion of deposit to soluble. The interconversion of the extraction products became dominant at relatively severe conditions. A comprehensive understanding of the reaction pathways of the biomass DSE conversion process has been provided in this study. The results of this study are meaningful and helpful in the process optimization or reactor design for a future commercial scale of this biomass DSE method.

■

REFERENCES

(1) Anca-Couce, A. Reaction mechanisms and multi-scale modelling of lignocellulosic biomass pyrolysis. Prog. Energy Combust. Sci. 2016, 53, 41−79. (2) Niu, Y.; Tan, H.; Ma, L.; Pourkashanian, M.; Liu, Z.; Liu, Y.; Wang, X.; Liu, H.; Xu, T. Slagging characteristics on the superheaters of a 12 MW biomass-fired boiler. Energy Fuels 2010, 24, 5222−5227. (3) Proskurina, S.; Sikkema, R.; Heinimö, J.; Vakkilainen, E. Five years leftHow are the EU member states contributing to the 20% target for EU’s renewable energy consumption; the role of woody biomass. Biomass Bioenergy 2016, 95, 64−77. (4) McKendry, P. Energy production from biomass (part 1): Overview of biomass. Bioresour. Technol. 2002, 83, 37−46. (5) Lu, P.; Wang, Y.; Huang, Z.; Lu, F.; Liu, Y. Study on NO reduction and its heterogeneous mechanism through biomass reburning in an entrained flow reactor. Energy Fuels 2011, 25, 2956−2962. (6) Wang, B.; Cao, J.; Zhao, X.; Bian, Y.; Song, C.; Zhao, Y.; Fan, X.; Wei, X.; Takarada, T. Preparation of nickel-loaded on lignite char for catalytic gasification of biomass. Fuel Process. Technol. 2015, 136, 17− 24. (7) Jeong, H. J.; Hwang, I. S.; Park, S. S.; Hwang, J. Investigation on co-gasification of coal and biomass in Shell gasifier by using a validated gasification model. Fuel 2017, 196, 371−377. (8) Abdullah, H.; Mourant, D.; Li, C.; Wu, H. Bioslurry as a fuel. 3. Fuel and rheological properties of bioslurry prepared from the bio-oil and biochar of mallee biomass fast pyrolysis. Energy Fuels 2010, 24, 5669−5676. (9) Demirbas, A. Potential applications of renewable energy sources, biomass combustion problems in boiler power systems and combustion related environmental issues. Prog. Energy Combust. Sci. 2005, 31, 171−192. (10) McKendry, P. Energy production from biomass (part 2): Conversion technologies. Bioresour. Technol. 2002, 83, 47−54. (11) Zhang, L.; Xu, C. C.; Champagne, P. Overview of recent advances in thermo-chemical conversion of biomass. Energy Convers. Manage. 2010, 51, 969−982. (12) Kambo, H. S.; Dutta, A. Strength, storage, and combustion characteristics of densified lignocellulosic biomass produced via torrefaction and hydrothermal carbonization. Appl. Energy 2014, 135, 182−191. (13) Wannapeera, J.; Worasuwannarak, N. Examinations of chemical properties and pyrolysis behaviors of torrefied woody biomass prepared at the same torrefaction mass yields. J. Anal. Appl. Pyrolysis 2015, 115, 279−287. (14) Colin, B.; Dirion, J. L.; Arlabosse, P.; Salvador, S. Quantification of the torrefaction effects on the grindability and the hygroscopicity of wood chips. Fuel 2017, 197, 232−239. (15) Peng, J. H.; Bi, H. T.; Sokhansanj, S.; Lim, J. C. A study of particle size effect on biomass torrefaction and densification. Energy Fuels 2012, 26, 3826−3839. (16) Chen, W.; Cheng, W.; Lu, K.; Huang, Y. An evaluation on improvement of pulverized biomass property for solid fuel through torrefaction. Appl. Energy 2011, 88, 3636−3644. (17) Lynam, J. G.; Reza, M. T.; Vasquez, V. R.; Coronella, C. J. Effect of salt addition on hydrothermal carbonization of lignocellulosic biomass. Fuel 2012, 99, 271−273. (18) Shen, Y.; Yu, S.; Ge, S.; Chen, X.; Ge, X.; Chen, M. Hydrothermal carbonization of medical wastes and lignocellulosic biomass for solid fuel production from lab-scale to pilot-scale. Energy 2017, 118, 312−323.

AUTHOR INFORMATION

Corresponding Authors

*Telephone/Fax: +86-27-87545526. E-mail: [email protected]. cn. *Telephone/Fax: +86-27-87545526. E-mail: [email protected]. cn. ORCID

Hong Yao: 0000-0002-2836-7803 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The work was financially sponsored by the National Natural Science Foundation of China (U1510119, 51661145010, and 21306059), the Foundation of State Key Laboratory of Coal F

DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

■

(19) Fang, J.; Gao, B.; Chen, J.; Zimmerman, A. R. Hydrochars derived from plant biomass under various conditions: Characterization and potential applications and impacts. Chem. Eng. J. 2015, 267, 253− 259. (20) Zhu, X.; Liu, Y.; Qian, F.; Zhang, S.; Chen, J. Investigation on the physical and chemical properties of hydrochar and its derived pyrolysis char for their potential application: Influence of hydrothermal carbonization conditions. Energy Fuels 2015, 29, 5222−5230. (21) Chew, J. J.; Doshi, V. Recent advances in biomass pretreatment−Torrefaction fundamentals and technology. Renewable Sustainable Energy Rev. 2011, 15, 4212−4222. (22) Zhu, X.; Xue, Y.; Li, X.; Zhang, Z.; Sun, W.; Ashida, R.; Miura, K.; Yao, H. Mechanism study of degradative solvent extraction of biomass. Fuel 2016, 165, 10−18. (23) Zhu, X.; Zhang, Z.; Zhou, Q.; Cai, T.; Qiao, E.; Li, X.; Yao, H. Upgrading and multistage separation of rice straw by degradative solvent extraction. J. Fuel. Chem. Technol. 2015, 43, 422−428. (24) Wannapeera, J.; Li, X.; Worasuwannarak, N.; Ashida, R.; Miura, K. Production of High-Grade Carbonaceous Materials and Fuel Having Similar Chemical and Physical Properties from Various Types of Biomass by Degradative Solvent Extraction. Energy Fuels 2012, 26, 4521−4531. (25) Zhu, X.; Li, X.; Xiao, L.; Zhang, X.; Tong, S.; Wu, C.; Ashida, R.; Liu, W.; Miura, K.; Yao, H. Novel carbon-rich additives preparation by degradative solvent extraction of biomass wastes for coke-making. Bioresour. Technol. 2016, 207, 85−91. (26) Li, X.; Zhu, X.-q.; Okuda, K.; Zhang, Z.; Ashida, R.; Yao, H.; Miura, K. Preparation of Carbon Fiber from Low-Molecular-Weight Compounds obtained from Low-Rank Coal and Biomass by Degradative Solvent Extraction. New Carbon Mater. 2017, 32, 41−47. (27) Li, X.; Priyanto, D. E.; Ashida, R.; Miura, K. Two-Stage Conversion of Low-Rank Coal or Biomass into Liquid Fuel under Mild Conditions. Energy Fuels 2015, 29, 3127−3133. (28) Zhu, X.; Tong, S.; Li, X.; Gao, Y.; Xu, Y.; Dacres, O. D.; Ashida, R.; Miura, K.; Liu, W.; Yao, H. Conversion of Biomass into HighQuality Bio-oils by Degradative Solvent Extraction Combined with Subsequent Pyrolysis. Energy Fuels 2017, 31, 3987−3994. (29) Shui, H. F.; Chen, Z. X.; Wang, Z. C.; Zhang, D. X. Kinetics of Shenhua coal liquefaction catalyzed by SO42=/ZrO2 solid acid. Fuel 2010, 89, 67−72. (30) Wang, Z.; Shui, H.; Lei, Z.; Ren, S.; Kang, S.; Zhou, H.; Gu, X.; Gao, J. Study of the preasphaltenes of coal liquefaction and its hydroconversion kinetics catalyzed by SO42−/ZrO2. Fuel Process. Technol. 2011, 92, 1830−1835. (31) Li, X.; Hu, H.; Zhu, S.; Hu, S.; Wu, B.; Meng, M. Kinetics of coal liquefaction during heating-up and isothermal stages. Fuel 2008, 87, 508−513. (32) Vo, T. K.; Lee, O. K.; Lee, E. Y.; Kim, C. H.; Seo, J.; Kim, J.; Kim, S. Kinetics study of the hydrothermal liquefaction of the microalga Aurantiochytrium sp. KRS101. Chem. Eng. J. 2016, 306, 763−771. (33) Yang, H.; Yan, R.; Chen, H.; Lee, D. H.; Zheng, C. Characteristics of hemicellulose, cellulose and lignin pyrolysis. Fuel 2007, 86, 1781−1788. (34) Valdez, P. J.; Tocco, V. J.; Savage, P. E. A general kinetic model for the hydrothermal liquefaction of microalgae. Bioresour. Technol. 2014, 163, 123−127. (35) Hita, I.; Aguayo, A. T.; Olazar, M.; Azkoiti, M. J.; Bilbao, J.; Arandes, J. M.; Castaño, P. Kinetic Modeling of the Hydrotreating and Hydrocracking Stages for Upgrading Scrap Tires Pyrolysis Oil (STPO) toward High-Quality Fuels. Energy Fuels 2015, 29, 7542− 7553. (36) Tran, K.; Bui, H.; Luengnaruemitchai, A.; Wang, L.; Skreiberg, Ø. Isothermal and non-isothermal kinetic study on CO2 gasification of torrefied forest residues. Biomass Bioenergy 2016, 91, 175−185.

Article

NOTE ADDED AFTER ASAP PUBLICATION This paper was published on the Web on April 21, 2017, with errors in Table 1. The corrected version was reposted on April 25, 2017.

G

DOI: 10.1021/acs.energyfuels.6b03442 Energy Fuels XXXX, XXX, XXX−XXX