Vapor–solid interaction among cellulose, hemicellulose and lignin

Fuel xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Vapor–solid interaction among cellulose, hemicellulose and lignin ⁎

Haiping Yanga, Ming Liua, Yingquan Chena, , Shanzhi Xinb, Xiong Zhanga, Xianhua Wanga, Hanping Chena a b

State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan, Hubei 430074, China Hubei Key Laboratory of Industrial Fume and Dust Pollution Control, Jianghan University, Wuhan 430056, China

A R T I C LE I N FO

A B S T R A C T

Keywords: Biomass pyrolysis Vapor–solid interaction Cellulose Hemicellulose Lignin

Secondary reactions, including those caused by interactions between vapor and solid phases, are unavoidable during biomass pyrolysis. In this study, the vapor–solid interaction between biomass components (cellulose, hemicellulose, and lignin) was investigated in a two-stage fixed-bed reactor. The results indicated that volatiles from hemicellulose promoted the breakdown of cellulose glycoside bonds and pyran rings as well as the removal of branched aliphatic chains and O-containing functional groups of lignin at 280 °C. Moreover, volatiles from cellulose produced abundant anhydrosugar, which was more prone to re-polymerization and to form aromatic rings on the lignin structure at 315 °C. As a result of the vapor–char interactions at 650 °C, the secondary decomposition of cellulose volatiles to gas products (decreasing by ~8 wt%) was inhibited, but carbonized products (increasing by ~3 wt%) tended to form, whereas hemicellulose vapor was more prone to decompose into low-molecular-weight liquid compounds, resulting in a high liquid yield (increasing by ~6 wt%). In addition, vapor–solid interactions accelerated the removal of O-containing functional groups of lignin volatiles, such as carbonyl and carboxyl, but inhibited the decomposition of H-containing functional groups, such as methyl and methylene. The finding is conducive to the understanding of the interactive mechanisms of biomass pyrolysis.

1. Introduction Due to its sustainability, abundance, and near-zero CO2 emission, biomass will play a significant role in future energy supply. As the only renewable carbon source, biomass has attracted attention from researchers regarding its conversion into value-added chemicals and functionalized carbon materials [1]. Pyrolysis is an important process in energy, chemical, or material production from biomass, as it converts biomass into products in three states (biochar, liquid oil, and gas products) [2]. To control the yields and quality of the pyrolysis products, great efforts have been made to clarify the mechanism of biomass pyrolysis [2,3]. The complicated transformation process during biomass pyrolysis is considered to be caused by the complex composition and structure of biomass, which is composed of cellulose, hemicellulose, and lignin [1,2]. Hence, the mechanism of biomass pyrolysis is often simplified by investigating the pyrolysis mechanism of cellulose, hemicellulose, and lignin individually, and summarizing biomass pyrolysis with superposition of the main components as the assumption that no interactions

among cellulose, hemicellulose, and lignin. Under this hypothesis, the approach of superposition rule appeared to behave well in some cases [4–7]. However, Couhert et al. [8] pointed out that simple superposition via the component additivity rule fails to predict the gas yield of biomass in flash pyrolysis at 950 °C because of the interactions between pyrolysis components. Furthermore, several studies have clarified the influences of the interactions between cellulose, hemicellulose, and lignin. With the three components simply mixing, Wang et al. [9] found that the interaction of hemicellulose or lignin upon the pyrolysis of cellulose extended the temperature range of levoglucosan and cellulose-lignin interactions enhanced the formation of 2-furfural and acetic acid. Based on simplex-lattice design method, Liu et al. [10] synthesized 15 biomass samples to investigate interactions between components and reported that hemicellulose-cellulose interactions decreased the yield of levoglucosan and largely promoted the formation of hydroxyacetaldehyde, while the presence of lignin decreased the yield of 2-furaldehyde and C]O containing compounds. With the comparisons of pyrolysis products from physical mixture and native mixture, Wu et al. [11,12] reported that cellulose-hemicellulose interactions

⁎

Corresponding author at: 1037 Luau Road, Wuhan, Hubei 430074, China. E-mail addresses: [email protected] (M. Liu), [email protected] (Y. Chen), [email protected] (X. Wang), [email protected] (H. Chen). https://doi.org/10.1016/j.fuel.2019.116681 Received 23 April 2019; Received in revised form 12 September 2019; Accepted 16 November 2019 0016-2361/ © 2019 Elsevier Ltd. All rights reserved.

Please cite this article as: Haiping Yang, et al., Fuel, https://doi.org/10.1016/j.fuel.2019.116681

Fuel xxx (xxxx) xxxx

H. Yang, et al.

2.2.2.

promoted the formation of hemicellulose-derived products but inhibited cellulose-derived products, while cellulose-lignin interactions enhanced the formation of low weight molecular products but inhibited the formation of anhydrosugar. In terms of the interaction intensities of the different components, Hosoya et al. [13] found that cellulose-lignin interacted more intensely than cellulose-hemicellulose. Without a premixed feedstock, Chen et al. [14] investigated volatiles interaction between variant components and found that hemicellulose volatiles enhanced the decomposition of cellulose volatiles, while the decomposition of lignin volatiles was inhibited by cellulose volatiles. The interactions between these components start to be recognized, which is important for the integral comprehension of biomass pyrolysis. However, more attention is paid to the interactions between volatiles, and the interaction among volatiles and char or solid components is lacking. As a typical multiphase reaction, pyrolysis involves significant vapor–solid interactions. In biomass pyrolysis, Song et al. [15] reported that volatile–char interactions significantly decreased the amount of tar from biomass pyrolysis, especially at relatively high temperatures (e.g. > 800 °C). While during volatile–char interactions, steam–char reactions can produce additional active sites on the char to facilitate tar reformation [16]. Keown et al. [17] reported that volatile–char interactions could also lead to the additional volatilization of alkali and alkaline earth metallic species. However, vapor–solid interactions between cellulose, hemicellulose, and lignin is rarely reported. Furthermore, the influence of vapor–solid interaction on the pyrolysis behavior of biomass components is still unclear. In this work, two kinds of vapor–solid interaction were investigated. Firstly, the influence of the volatiles initially released from hemicellulose or cellulose on the structure of cellulose or lignin were investigated at the low temperature, and the pyrolysis characteristic after vapor–solid interaction was conducted. The second kind of vapor–solid interaction investigated is between volatile and char from components pyrolysis. The effect of char on volatile reformation was investigated at high temperature (650 °C). The study should contribute greatly to the understanding of biomass pyrolysis process.

2.2. Experimental setup and method

2.2.1. Vapor-solid interaction The first part of the experiment was designed to investigate the vapor–solid interactions between the biomass components at lower temperatures. The specific experimental procedure was as follows: Prior to each trial, two zones were heated up to the same designated temperature (280 °C or 315 °C) with N2 (99.999% 500 mL/min) as carrier gas, which was kept constant for 10 min. After that, 500 ± 5 mg hemicellulose or cellulose was placed in the bottom of the sample holder, and with quartz fibers separating the materials, the same mass of cellulose or lignin was placed on top of the quartz fibers. The inhibition of solid–solid interactions was consequently guaranteed. Finally, quartz fibers were placed on top of the sample holder in case the particles spilled during the reaction process. Once the temperature had stabilized at the selected value, the carrier was rapidly pushed to the center of the first heating zone (Fig. 1) with the N2 flow rate adjusted to 250 mL/min. Since the vapor was released from the bottom reactants and contacted the upper stage, vapor–solid interactions occurred. Simultaneously, the cellulose or lignin samples were pyrolyzed individually at 280 °C or 315 °C for comparison. After 30 min, the upper and lower samples were weighed. The upper solid samples were named HC280, HL280, and CL315, where for example, HC280 represents that the solid sample derived from cellulose pyrolysis and interacted with hemicellulose vapor at 280 °C and H is for hemicellulose, C is for cellulose, and L is for lignin while the solid samples from without vapor–solid interactions were named C280, L280, and L315, where C280 represents solid products from cellulose pyrolysis at 280 °C. All abbreviations are defined in Table S1. The second part of the experiment was designed to investigate the vapor–solid interactions between the volatiles and chars at high temperatures. In this part, the char of the biomass components was preprepared in a fixed-bed reactor with 1 g each of cellulose, hemicellulose, or lignin, a temperature of 650 °C, an N2 flow rate of 200 mL/ min, and a pyrolysis time of 15 min. The chars were subsequently crushed to smaller than 42 μm. Then, 500 ± 5 mg pre-prepared char was placed in the sample holder (height: 20 mm), which was then moved to the center of the upper heating zone. Then, 500 ± 5 mg of the cellulose, hemicellulose, or lignin sample was placed in another sample holder, which was placed at the bottom heating zone. The interaction procedure was similar with that of the first part at lower temperature. However, in case additional volatiles was released of the char, it is worth noting that the interaction temperature was set at 650 °C. To ensure the accuracy of the data, each trial was repeated three times with an error of less than 5%, and the measured data presented herein are the average values.

The experimental apparatus mainly consisted of a two-stage fixedbed reactor (inner diameter: 20 mm; thickness: 2 mm; height: 600 mm), a quartz tube sample holder (external diameter: 19 mm) in a carrier (height: 40 mm), a temperature controller system, an ice–water mixture for condensation, and gas purification and drying equipment followed by a gas analysis system to test the gas composition (Fig. 1). One end of the carrier was sealed with a perforated quartz plate. The experiments (Fig. S1) were divided into two parts described in Sections 2.2.1 and

2.2.2. Pyrolysis behavior of samples after interactions The solid sample derived from low temperature interaction were further pyrolyzed at 500 °C to investigate the influence of the vapor–solid interaction on the pyrolysis behavior. Once the temperature stabilized at 500 °C, the sample holder with 500 mg of the collected solid sample was pushed to the center of the first heating zone. The volatiles were then swept into the ice–water condenser so that the condensable volatiles could be collected as liquid products. After

2. Materials and methods 2.1. Materials Cellulose, hemicellulose, and lignin were obtained as indicated in our previous study. The results of proximate and ultimate analyses of the samples are shown in Table 1. Before each trial, the sample was dried for 12 h at 100 ± 5 °C.

Table 1 Proximate and ultimate analysis of biomass components. Proximate analysis (wt%, db.)

Cellulose Hemicellulose Lignin

Ultimate analysis (wt%, daf.)

LHV (MJ/kg) d

V

FC

A

C

H

N

S

O

95.50 76.80 58.90

4.50 21.40 36.90

0.00 1.80 4.20

42.70 41.60 48.30

6.20 5.70 4.90

0.03 0.02 0.10

0.05 0.03 3.10

51.02 52.65 43.60

15.47 15.31 19.31

Notes: d: O was obtained from the difference; db: dry basis; daf: dry and ash free basis; LHV: Low heating value; V: Volatile; FC: Fixed carbon; A: Ash. 2

Fuel xxx (xxxx) xxxx

H. Yang, et al.

Fig. 1. Schematic of experimental apparatus used to investigate volatile-solid interactions.

cleaning, the non-condensable gas was collected in a gas bag for further analysis. The solid product yield was taken from the sample holder weight difference. The gas yield was calculated from the sum of the volumes of all the gases collected during pyrolysis. However, to avoid interactions resulted by solid sample and volatile released in advance, the sample should not be spread too thick and the small amount of biomass sample is used (500 mg), but it might result in big error of the liquid product yield obtained from the condenser weight difference. Hence, the liquid yield was finally determined via the difference between repeatable solid and gaseous yields based on mass constant of the system. 2.3. Analysis methods of pyrolysis product The gas products were qualitatively and quantitatively analyzed using a dual-channel micro-gas chromatograph (Micro-GC 3000A, Agilent Technologies, USA) equipped with thermal conductivity detectors. Gas chromatography–mass spectrometry (HP7890 series GC with a HP5975 MS detector) with a capillary column (Agilent HP-5MD, 19091s-433; length: 30 m; inner diameter: 0.25 mm; film thickness: 0.25 μm) was employed to determine the main organic products of biooil. The analysis method is described in detail elsewhere [14,18]. A CHNS/O elementary analyzer (Vario Micro Cube, Germany) was used to determine the composition of the solid char. A Fourier transform–infrared spectrometer (VERTEX 70, Bruker, Germany) was used to perform 32 scans with 4 cm−1 resolution between 400 cm−1 and 4000 cm−1 to determine the functional groups of the solid char [18].

Fig. 2. Fourier transform–infrared spectra (normalization) for samples under different conditions. Table 2 Ultimate analysis for samples under different conditions.

3. Results and discussion 3.1. Effect of vapor–solid interactions on components structure

C H *O

Fig. 2 shows the variations in functional groups of treated cellulose and lignin with and without interaction, and the elemental composition is listed in Table 2. In the case of hemicellulose–cellulose interactions (Fig. 2(a)), the intensity of the OeH group at 3348 cm−1 was significantly reduced upon interaction with hemicellulose vapor, it might be attributed to that hemicellulose volatiles promoted the dehydration of cellulose. It is consistent with the relatively lower contents of H and O in HC280 compared to those in the original sample. In addition, the

C280

HC280

L280

HL280

L315

CL315

45.50 6.00 48.20

49.70 5.30 44.60

54.10 4.30 41.20

55.00 4.90 39.80

60.40 4.30 34.80

57.10 3.80 38.60

Notes: *: O was obtained from the difference; N and S were omitted as the limited content.

3

Fuel xxx (xxxx) xxxx

H. Yang, et al.

intensity of methylene (1325 cm−1, 1370 cm−1, and 1430 cm−1) decreased. It might be attributed to that the volatile vapor evolved out from hemicellulose pyrolysis promoted the dehydration and breakdown of the intermolecular hydrogen bonding of cellulose. Moreover, the intensity at 1113 cm−1, 1060 cm−1, and 1163 cm−1 decreased, as the degradation of hemicellulose promoted the breakdown of glycosidic bonds and pyranose rings from cellulose. Respected to lignin, variant phenomena were displayed after hemicellulose interaction (Fig. 2(b)), the intensity of the aromatic ring skeleton (1593 cm−1), alkyl ether bond on the branched chain of aliphatic hydrocarbons (1212 cm−1), and CeH bond of guaiacyl (1137 cm−1) were reduced obviously. In contrast, the intensity of the CeH bond of alkane (1392 cm−1, 1457 cm−1, and 2934 cm−1) increased slightly. These observations might be caused by the breakdown of aliphatic hydrocarbon and promotion of lignin decomposition upon interaction with hemicellulose vapor. Moreover, the hemicellulose vapor catalyzed the removal of guaiacyl or O-containing groups of lignin, decreasing the O content of lignin (Table 2). Unlike the effects of hemicellulose vapor, those of cellulose vapor were observed in cellulose–lignin interactions (Fig. 2(c)). After interaction with cellulose pyrolysis vapor, the intensity of the lignin aromatic ring at 1585 cm−1 increased, while the intensity of CeH bond of alkane (1385 cm−1, 1315 cm−1) and aromatic ether (1198 cm−1, 1126 cm−1, and 1052 cm−1) decreased to some extent. There might exist two possible mechanisms. On one hand, volatile vapor evolved out from cellulose pyrolysis at 315 °C might promote the decomposition of lignin to a certain degree, thus resulting in peak intensity decreasing in the above position. On the other hand, polymerization and condensation of levoglucose or anhydrosugar evolved out from cellulose might be converted to aromatic ring [19]. In addition, the cellulose vapor might promote lignin decomposition to a certain degree, and lead to the breakdown of aliphatic ether bonds and aryl or alkyl ether bonds. Moreover, the abundant O-containing functional groups of the oligosaccharide devolatilized by cellulose resulted in a greater O content in CL315. Interestingly, the intensity of eOH at 3426 cm−1 increased, which may be due to the oligosaccharides deposition [1]. In conclusion, from the structure evolution of biomass constituents after vapor–solid interactions at lower temperature (280 °C), hemicellulose vapor could promote the dehydration of cellulose, and accelerate the decrease of polymerization degree, and intra-molecular hydrogen-bond breaking. Similarly, the decomposition of lignin could also be enhanced by hemicellulose and cellulose volatile. This phenomena may be explained by a theory proposed by Wang et al. [20] that there are many active compounds formed by hemicellulose, which would easily decompose the cellulose chain and lignin structure and branches to some degree.

Fig. 3. Weight loss from vapor–solid interactions at low temperatures.

Fig. 4. Product distributions of samples with and without vapor–solid interactions.

and form more solid residues [24].

3.2.2. Products evolution property Fig. 4 shows the product distributions for the pyrolysis of treated cellulose or lignin with hemicellulose and cellulose vapor interactions. With interaction of hemicellulose, the solid and gas yields from cellulose pyrolysis increased obviously (from 15.37 wt% to 19.31 wt% and from 11.08 wt% to 15.12 wt%, respectively), whereas the liquid yield decreased accordingly from 73.55 wt% to 65.57 wt%. It might be attributed to that the anhydrosugar remaining in HC280 was prone to decompose directly or polymerize to form solid products because of the high degree of hydration and depolymerization, which could be referred to Fig. 2(FTIR spectra). For original lignin, the solid yields of lignin at 280 °C and 315 °C were 70.09% and 75.12%, respectively. Before and after interaction with hemicellulose vapor, similar solid yields were obtained (70.09% and 70.22%, respectively). However, interaction with cellulose vapor increased the solid yield from 75.12% to 79.23%, with a change of 7.18%. It might be attributed to that xylan vapor accelerated lignin depolymerization. Alternatively, cellulose vapor increased the content of aromatic nucleus of CL315 (referred to Fig. 2), and resulted in intense aromatic ring condensation and a higher solid yield. Fig. 5 shows the effect of vapor–solid interactions on gas composition. For cellulose interaction with hemicellulose volatiles, the CO content increased from 58.30% to 63.67%, and the CH4 content

3.2. Influence of vapor–solid interaction on biomass components pyrolysis 3.2.1. Weight loss property Fig. 3 shows the weight loss profile of the biomass components and samples after vapor–solid interactions at 280 °C and 315 °C, respectively. At 280 °C, the weight loss of cellulose and lignin was 9.27% and 13.25%, respectively, whereas that of hemicellulose reached 63%. After interaction with hemicellulose vapor, the weight loss of cellulose and lignin increased slightly to 11.99% and 15.62%, respectively. This might suggest that hemicellulose vapor promoted the decomposition of cellulose and lignin. It might be attributed to that the acids and ketones evolved out from hemicellulose vapor may provide free H radicals and promoted lignin chemical conversion [21,22] or catalyzed polysaccharide decomposition and accelerated the ring-opening reaction of pyran rings from cellulose degradation [23]. On the other hand, the opposite result was obtained lignin decomposition after cellulose vapor interaction at 315 °C. The weight loss of lignin without interaction was 23.07%, whereas it was 20.24% with interaction. This reduction might be attributed to that cellulose volatiles was trapped by lignin particles 4

Fuel xxx (xxxx) xxxx

H. Yang, et al.

Fig. 5. Gas compositions of samples with and without vapor–solid interactions.

Fig. 6. Yield distributions from vapor–solid interactions at high temperatures.

increased from 8.71% to 9.97%, whereas the CO2 content decreased obviously from 31.26% to 24.56%. From the low molecular gaseous evolution mechanism, the small releasing of CO2 from cellulose pyrolysis might be due to the lowest content of C]O group [25]. It might be attributed to that hemicellulose vapor accelerated the decomposition and ring-opening of cellulose with generating more C]O or COOH, whereas CO2 was primarily generated in the decomposition process at low temperature [26]. Therefore, the CO2 content decreased with increasing degree of cellulose decomposition. Also the dehydration of cellulose monomers may lead to the formation of ether bonds (CeOeC) between the hydroxyl groups of the pyran rings and carboxyl opening (C]O) after ring-opening, and the breakdown of these chemical bonds may lead to the increment of CO content [25]. However, the contrary result was observed for lignin upon interaction with hemicellulose vapor. CO2 content increased from 24.44% to 31.38%, whereas CH4 content decreased from 39.50% to 31.86%. It might be attributed to that the acids in hemicellulose may cause the removal of the guaiacyl and methoxy group (O-CH3), and resulted in CH4 diminishing, which could be deduced by the decrement of relative peak area (Fig. 2). Similar trend showed for lignin upon interaction

with cellulose vapor in HL280. CO2 content increased largely from 7.13% to 22.61%, whereas the H2 and CH4 content decreased from 52.09% to 44.88% and from 38.70% to 30.17%, respectively. There are two possible mechanisms. On the one hand, cellulose vapor may have caused the removal of methyl (-CH3 ) or methylene (-CH2 ) groups from the lignin structure, thus reducing the H2 and CH4 content. This can also be explained by H-donors of cellulose volatile proposed by Hosoya [21]. On the other hand, cellulose vapor may introduced O-containing functional groups, such as carboxyl (–COOH), into the lignin structure, and increased CO2 formation [27]. Table 3 shows the liquid composition distribution of treated cellulose and lignin pyrolysis. For cellulose pyrolysis after interacted with hemicellulose volatile, the carbohydrate content decreased (e.g., from 76.1% to 63.1% in LG), whereas the content of small molecule oxygenates increased (e.g., the acetic acid content increased from 1.3% to 2.1%, and that of furfural increased from 1.7% to 2.5%). It might be caused by that hemicellulose vapor promoted the dehydration and cleavage of cellulose glycosidic bonds, and resulted in the removal of short glucose chains from cellulose amorphous regions. Moreover, secondary decomposition may have occurred in the anhydrosugar

Table 3 Compositions of liquid products of different samples at 500 °C (peak area, %). Serial number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Substance

Hydroxyl-acetone 4-Hydroxy-4-methyl-2-Pentanone Acetic acid Furfural 2-Cyclopentene-1,4-Dione 1,3-Cyclopentanedione 3-Methyl-1,2-cyclopentanedione guaiacol 3-Methylcatechol 2,6-Dimethylphenol 2-Methoxy-4-methylphenol phenol 2-Ethylphenol 4-Hydroxytoluene 3,4-Dimethylphenol 1,4:3,6-Dianhydro-α-D-glucopyranose 5-Hydroxymethylfurfural Levoglucoseenone Levoglucosan 3,6-Dimethylphenanthrene 9–1-Methylethyl-anthracene 1-Methyl-7-isopropylphenanthrene

Samples C280

HC280

L280

HL280

L315

CL315

0.7 – 1.3 1.7 0.4 1.4 0.8 – – – – 0.7 – – – 8.4 2.6 0.6 76.1 – – –

0.9 – 2.1 2.5 1.4 2.9 2.3 – – – – 1.8 – – – 6.8 2.5 0.4 63.1 – – –

– 19.1 – – – – – 18.7 1.3 1.5 2.2 30.6 – 18.1 3.4 – – – – – – –

– 12.8 – – – – – 12.1 1 2.7 1.3 34.8 – 21.1 4.3 – – – – – – –

– 18.2 – – – – – – – – – 39.1 3.3 29.4 6.9 – – – – – – –

– 12.9 – – – – – 1.4 – 1.8 – 28.2 1.4 22 5.9 – – – – 5.2 3.3 7.4

5

Fuel xxx (xxxx) xxxx

H. Yang, et al.

aromatic rings rich in O-containing aliphatic hydrocarbon branches was formed after interaction with cellulose, resulting in more fused aromatic hydrocarbon compounds generated in the liquid product, such as 3,6-dimethylphenanthrene, 9-(1-methylethyl)-anthracene, and 1-methyl-7-isopropylphenanthrene, which are formed by the condensation reaction between mononuclear aromatic rings. 3.3. Vapor–solid interactions at higher temperature (650 °C) 3.3.1. Property of distributions on pyrolysis product The product yields of the pairs of biomass components are shown in Fig. 6. Generally, char from different biomass model compounds have different effects on vapor–solid interactions. For cellulose vapor, a consistent trend was observed upon interaction with char. The gas product yield decreased, whereas the liquid and solid product yields increased. However, the interaction with cellulose char was stronger, which significantly inhibited the secondary decomposition of cellulose volatiles to yield small molecular gas products, whereas repolymerization among oligosaccharides was enhanced to form more solid products. In contrast, the interactions between cellulose volatiles and xylan and lignin char involved weaker polymerization of carbohydrate products but tended to promote their decomposition to produce small molecular products. For hemicellulose vapor, a uniform trend was observed upon interaction with char. The gas yield decreased, whereas the liquid yield increased. These findings may have resulted from the char inhibiting the formation of small molecules by hemicellulose vapor. For lignin volatiles, the opposite effect was observed upon interaction with char. The gas yield increased, whereas the liquid yield decreased. The lignin volatiles were mainly composed of aromatic compounds, indicating that the vapor–solid interactions promoted the decomposition of lignin volatiles to generate more small molecular gaseous products. 3.3.2. Property of gas composition The gaseous product compositions of the pairs of vapor–solid interactions are shown in Fig. 7. Generally, for cellulose interacting with cellulose char, the change in gaseous composition was not significant, and only the CO content increased slightly. However, upon interaction with char from hemicellulose and lignin, the volume fraction of H2 increased significantly from 9.22% to 12.73% and 14.26%, with changes of 38.07% and 54.67%, respectively, and that of CH4 decreased from 10.90% to 7.88% and 9.76%, respectively. Moreover, the volume fraction of CO decreased dramatically from 61.18% to 46.95% and 54.65%, whereas that of CO2 increased from 15.19% to 30.00% and 18.66%. It is reported that CO mainly originated from the breakdown of CeOeC or C]O from anhydrosugars in cellulose vapor such as levoglucose [25,28,29]. The decreased CO may indicate further polymerization and aromatic cyclization of oligosaccharides occurred in cellulose vapor to form char rather than decomposition, which is consistent with the reacting pathway proposed by Banyasz [30]. Additionally, CO2 was mainly generated during decomposition and ringopening of cellulose at low temperatures [26]. The increased CO2 content can be explained by that the oligosaccharides in the cellulose volatiles were further dehydrated and carbonization, increasing the liquid product on the one hand. Furthermore, CO and H2O formed from carbohydrates decomposition might undergo water shifting and converted into H2 and CO2 under the interactions with char, as shown in follows.

Fig. 7. Gaseous product distributions from the pairs of vapor–solid interactions.

remaining in the solid sample to form O-containing compounds. Moreover, the content of precursors of potential aromatic rings increased, e.g., cyclopentene increased from 0.4% to 1.4%, and cyclopentane increased from 0.8% to 2.3%. This exactly explained that the linear hydrocarbon after the pyran ring opening of the anhydrosugar was also converted into solid product due to aromatic cyclization and condensing, which is consistent with the higher solid yield in HC280. For lignin, the guaiacol content decreased from 18.7% to 12.1% upon interaction with xylan volatiles, and the content of its alkyl-substituted derivative also decreased slightly. In contrast, the phenol and methyl-substituted phenol content significantly increased (e.g., the phenol content increased from 30.6% to 34.8%). This could be illustrated by that hemicellulose vapor catalyzed the removing of lignin methoxyl, which can also be referred to FTIR (Fig. 2). As for aliphatic hydrocarbon such as 4- hydroxyl-4-methyl-2-pentanone, the content was higher without interactions. This indicates that short-chain alkane on lignin structure after interacted by xylan volatile was more prone to break down and combine with phenols as radicals. This can be confirmed by the diminishing of CH4. However, for lignin with cellulose vapor interaction at 315 °C, an opposite change was observed. The phenol and alkyl-phenol content decreased (e.g. the phenol content decreased from 39.1% to 28.2%). It might be attributed to that cellulose vapor catalyzed the removal of guaiacyl units. Simultaneously, more

CO + H2 O→ CO2 + H2 ,

ΔHθ298.15 = −41.2kJ/mol

In the case of hemicellulose pyrolysis at 650 °C, the CO2 volume fraction was highest (38.48%), followed with H2 (25.76%) and CO (27.67%). Upon interaction with cellulose char, the volume fractions of H2 and CO2 decreased slightly, whereas that of CO increased. However, the interactions with hemicellulose and lignin char were more intense than cellulose char. The CO content decreased from 27.67% to 22.91% 6

Fuel xxx (xxxx) xxxx

H. Yang, et al.

Table 4 Liquid composition distributions with and without vapor–solid interactions at 650 °C (peak area, %). Number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

Substance

Hydroxy-acetone Glycolaldehyde 2-Cyclopenten-1-one Acetoin Acetic acid 2-Furaldehyde propionic acid 5-Methyl furfural Furfuryl alcohol 2H-Pyran-2-one 3-Methyl-1,2-cyclopentanedione Phenol 4-Hydroxytoluene 2,3-Dihydrobenzofuran 1,4:3,6-Dianhydro-α-d-glucopyranose 5-Hydroxymethylfurfural Methylhydroquinone 2,5-Anhydro D-mannose oxime Levoglucosan Propyl alcohol Formic acid 2,3-Dimethyl-2-Cyclopenten-1-one 2(5H)-Furanone 2-Hydroxycyclopent-2-en-1-one 3-Methyl-1,2-cyclopentanedione Phenol 4-Hydroxytoluene 6,7-Dihydro-4(5H)-Benzofuranone 4-Hydroxy-4-Methyl-2-pentanone Guaiacol 3,4-Dimethyl phenol Vanillin 2,3-Benzofuran

Operating condition C

CC

CH

CL

H

HC

HH

HL

L

LC

LH

LL

3.4 0.7 0.7 0.4 2.8 1.2 1.3 0.5 0.6 0.8 1.2 1.6 0.3 0.4 3.2 0.9 0.4 6.4 65.8 – – – – – – – – – – – – – –

2.8 2.6 0.8 0.6 3 1.7 1.3 0.5 0.4 1.5 0.9 1.6 0.7 – 3.7 1.8 0.4 1.1 67.5 – – – – – – – – – – – – –

2.5 2.2 0.7 0.5 2.5 1.4 1.1 0.5 0.5 1.3 0.9 1.4 0.8 0.4 3.5 1.9 0.4 1.8 57.3 – – – – – – – – – – – – –

2.5 1.9 0.5 0.3 2.4 1.1 1.2 0.5 0.6 1.6 0.9 1.4 0.7 0.4 3.8 1.9 0.4 1.7 60.4 – – – – – – – – – – – – – –

9.8

7.9

13.7

13.1

1.2 10.3 8.7 3.9 6.5

1.1 8.5 7.3 3.3 6.2

1.4 14.6 11.6 5.6 8.7

1.5 15.5 10.5 4.9 8.2

– – – – –

1.8 – – – – – – – 1.8 – – – 2.2 0.8 0.6 2.4 – 2.5 0.9 11.4 – – – – –

1.4 – – – – – – – 0.7 – – 3.1 8.1 0.7 0.6 1.7 1.9 2.3 0.8 3.9 – – – – –

2.4 – – – – – – – 1 – – 2.1 2.9 1 0.8 2.9 2.7 2.7 0.9 4.1 – – – – –

2.6 – – – – – – – 1.9 – – 2 2.5 1.2 0.7 2.4 3.9 3.7 1.3 4.2 – – – – –

– – – – – – – – – – – – – – – – – – – – – – – – – 32.8 15.9 – 8.4 8.1 3.5 2.5 2.3

– – – – – – – – – – – – – – – – – – – – – – – – – 39.5 18.7 – 5.3 4.4 1.7 2.2 2.9

– – – – – – – – – – – – – – – – – – – – – – – – – 46.8 21.9 – 5.5 4.6 5.1 – 2.4

– – – – – – – – – – – – – – – – – – 38.5 16.9 – 10.9 3.9 5 1.3 2.4

be attributed that the interaction promoted the decomposition of unstable saccharides in cellulose volatiles, thereby increasing the small molecular products such as hydroxyacetaldehyde and 5-hydroxymethylfurfural content. Moreover, in the case of hemicellulose vapor, the liquid product was mainly composed of small molecular oxygenates, such as acid and ketone. Upon interaction with cellulose char, the hydroxyacetone, hydroxybutanone, and acetic acid content in the liquid product decreased. In contrast, the formic acid content increased. It might be attributed to that cellulose char promoted the decarbonylation reaction of the ketone product but inhibited the decomposition of formic acid. Consequently, CO content increased but CO2 content decreased. However, upon interaction with hemicellulose and lignin char, the content of ketone products, such as hydroxyacetone and hydroxybutanone, in the liquid product significantly increased, and led to the CO content in gas phase decreased accordingly. What’s more, there was another interesting phenomenon. For cellulose vapor interacted with char, benzofuranone diminished largely from 15.4% to ~4% while the monomers phenol and furan increased gently. This can be explained by vapor–solid interactions enhanced the cracking of hemicellulose volatiles to small molecules rather than undergoing cyclization and aromatization. The main liquid product of the lignin pyrolysis were phenols and their alkyl substitutes. Comparatively, the small molecular oxygenates (e.g., 4-hydroxy-4-methyl-2-pentanone) were less in abundance and content. When the lignin volatiles interacted with cellulose char, the 4hydroxy-4-methyl-2-pentanone content increased slightly from 8.4% to 10.9%. Upon interaction with hemicellulose and lignin char, the opposite was observed, and the content of 4-hydroxy-4-methyl-2-pentanone decreased obviously to ~5%. It might be attributed to that the existence of hemicellulose and lignin char accelerated the cleavage of C–C bonds in aliphatic hydrocarbons. However, methyl groups resulting

(hemicellulose char) and 21.90% (cellulose char), respectively, whereas the CO2 content increased significantly from 38.48% to 48.19% and 44.80%, respectively. For xylan pyrolysis, CO2 is mainly originated from the decarboxylation of eCOOH groups in glucuronic acid units, cracking and reforming of C]O and eCOOH groups, or decarboxylation of O-acetyl linked to xylan [31]. It might be attributed to that hemicellulose and lignin char promoted the conversion of carboxyl groups in hemicellulose volatiles but inhibited the removal of carbonyl groups. For lignin interacting with char, a uniform trend was observed. The H2 and CH4 content decreased, whereas that of CO and CO2 increased. It might be attributed to that solid char inhibited the reactions with Hcontaining groups in the lignin volatiles, such as dehydrogenation, demethylation, and de-methylene, but promoted the removing of Ocontaining groups, and finally formed CO and CO2 evolved out, thereby increased the yield of gaseous product. 3.3.3. Property of liquid product Table 4 shows the liquid composition distributions (relative peak area) for cellulose, hemicellulose, and lignin pyrolysis with and without vapor–solid interactions. It can be found that the liquid products were dominated by levoglucose and minor differences were found before and after vapor–solid interactions. This might indicate that vapor–solid interactions had a weaker decomposition effect on levoglucose in cellulose volatiles. Interestingly, when cellulose vapor interacted with hemicellulose and lignin char, the peak area of levoglucosan decreased from 65.8% to 57.3% and 60.4%, respectively, unlike increased by cellulose char. It might be attributed to that the ash in hemicellulose and lignin char catalyzed the decomposition of levoglucosan [32]. In addition, the 2,3-dehydrate-D-mannose content in the cellulose volatiles decreased significantly after the vapor–solid interactions. It might 7

Fuel xxx (xxxx) xxxx

H. Yang, et al.

from the breaking of aliphatic hydrocarbons may participate in other reaction pathways instead of generating the gaseous product CH4. This could be deduced from the increase of methyl-containing substances. In addition, the content of characteristic compounds, such as guaiacol and vanillin in the liquid product, decreased upon vapor–solid interactions, and conversely, the phenols content increased significantly. This could be explained by that char promoted the demethoxylation of aromaticring compounds, such as guaiacyl or syringyl, which is consistent with the gas evolution analysis. However, the removal of methoxy groups did not increase the CH4 content in the gas product. This could be explained by that the methyl radical in the lignin volatiles was prone to combine with phenol to form alkyl-phenol after vapor–solid interactions, which was confirmed by the decreased CH4 content and increased 4-methylphenol and 3,4-dimethylphenol content.

[5] [6]

[7] [8]

[9] [10] [11]

[12]

4. Conclusions

[13]

Vapor–solid interactions throughout biomass pyrolysis were investigated in terms of the effects of volatiles on solid-phase components at low temperature and the effects of three-component char on volatiles at high temperature. The hemicellulose volatile evolved at lower temperature (280 °C) could promote the decomposition of cellulose and lignin. In addition, the cellulose volatiles caused depolymerization of lignin to some degree, but abundant anhydrosugar evolved from cellulose volatiles was more prone to thermal polymerization to generate carbonized products. At higher temperature (650 °C), vapor–solid interactions could promote the decomposition of unsteady anhydrosugar. Hemicellulose and lignin char catalyzed the decomposition of hemicellulose volatiles into small molecular liquid products. Instead, lignin volatile was promoted to decompose more gaseous products. Vapor–solid interactions between volatiles and solid phases at low or high temperatures both significantly influenced the composition of the final bio-oil.

[14]

[15]

[16]

[17]

[18]

[19]

[20]

Acknowledgments

[21]

The authors wish to express their great appreciation of the financial support from National Key Research and Development Plan of China (2018YFB1501403) and National Nature Science Foundation of China (51622604 and 51876061), and the technical support from Analytical and Testing Center in Huazhong University of Science & Technology (http://atc.hust.edu.cn).

[22] [23]

[24]

[25]

Appendix A. Supplementary data

[26]

Supplementary data to this article can be found online at https:// doi.org/10.1016/j.fuel.2019.116681.

[27] [28]

References [29] [1] Chen X, Che Q, Li S, Liu Z, Yang H, Chen Y, et al. Recent developments in lignocellulosic biomass catalytic fast pyrolysis: Strategies for the optimization of biooil quality and yield. Fuel Process Technol 2019;196:106180. [2] Wang S, Dai G, Yang H, Luo Z. Lignocellulosic biomass pyrolysis mechanism: A state-of-the-art review. Prog Energy Combust Sci 2017;62:33–86. [3] Collard FX, Blin JJR, Reviews SE. A review on pyrolysis of biomass constituents: Mechanisms and composition of the products obtained from the conversion of cellulose, hemicelluloses and lignin. 2014;38(5):594–608. [4] Yang H, Yan R, Chen H, Zheng C, Lee DH, Liang DT. In-depth investigation of

[30] [31] [32]

8

biomass pyrolysis based on three major components: hemicellulose, cellulose and lignin. Energy Fuels 2006;20(1):388–93. Biagini E, Barontini F, Tognotti L. Devolatilization of biomass fuels and biomass components studied by TG/FTIR technique. Ind Eng Chem Res 2006;45:134486–93. Zhou H, Long Y, Meng A, Li Q, Zhang Y. The pyrolysis simulation of five biomass species by hemi-cellulose, cellulose and lignin based on thermogravimetric curves. Thermochim Acta 2013;566:36–43. Anca-Couce A. Reaction mechanisms and multi-scale modelling of lignocellulosic biomass pyrolysis. Prog Energy Combust Sci 2016;53:41–79. Couhert C, Commandré J-M, Salvador S. Failure of the component additivity rule to predict gas yields of biomass in flash pyrolysis at 950°C. Biomass Bioenergy 2009;33(2):316–26. Wang S, Guo X, Wang K, Luo Z. Influence of the interaction of components on the pyrolysis behavior of biomass. J Anal Appl Pyrol 2011;91(1):183–9. Liu Q, Zhong Z, Wang S, Luo Z. Interactions of biomass components during pyrolysis: A TG-FTIR study. J Anal Appl Pyrol 2011;90(2):213–8. Wu S, Shen D, Hu J, Zhang H, Xiao R. Cellulose-hemicellulose interactions during fast pyrolysis with different temperatures and mixing methods. Biomass Bioenergy 2016;95:55–63. Wu S, Shen D, Hu J, Zhang H, Rui XJB, Bioenergy. Cellulose-lignin interactions during fast pyrolysis with different temperatures and mixing methods. 2016;90:209–17. Hosoya T, Kawamoto H, Saka S. Cellulose–hemicellulose and cellulose–lignin interactions in wood pyrolysis at gasification temperature. J Anal Appl Pyrol 2007;80(1):118–25. Chen Y, Fang Y, Yang H, Xin S, Zhang X, Wang X, et al. Effect of volatiles interaction during pyrolysis of cellulose, hemicellulose, and lignin at different temperatures. Fuel 2019;248:1–7. Song Y, Wang Y, Hu X, Hu S, Xiang J, Zhang L, et al. Effects of volatile–char interactions on in situ destruction of nascent tar during the pyrolysis and gasification of biomass. Part I. Roles of nascent char. Fuel 2014;122:60–6. Song Y, Wang Y, Hu X, Xiang J, Hu S, Mourant D, et al. Effects of volatile–char interactions on in-situ destruction of nascent tar during the pyrolysis and gasification of biomass. Part II. Roles of steam. Fuel 2015;143:555–62. Keown DM, Hayashi J-i, Li C-Z. Effects of volatile–char interactions on the volatilisation of alkali and alkaline earth metallic species during the pyrolysis of biomass. Fuel 2008;87(7):1187–94. Xia S, Xiao H, Liu M, Chen Y, Yang H, Chen H. Pyrolysis behavior and economics analysis of the biomass pyrolytic polygeneration of forest farming waste. Bioresour Technol 2018;270:189–97. Xin S, Yang H, Chen Y, Wang X, Chen H. Assessment of pyrolysis polygeneration of biomass based on major components: Product characterization and elucidation of degradation pathways. Fuel 2013;113:266–73. Wang G, Li W, Li B, Chen H. TG study on pyrolysis of biomass and its three components under syngas. Fuel 2008;87(4):552–8. Hosoya T, Kawamoto H, Saka S. Solid/liquid- and vapor-phase interactions between cellulose- and lignin-derived pyrolysis products. J Anal Appl Pyrol 2009;85(1):237–46. Mukkamala S, Wheeler MC, van Heiningen ARP, DeSisto WJ. Formate-assisted fast pyrolysis of lignin. Energy Fuels 2012;26(2):1380–4. Kawamoto H, Morisaki H, Saka S. Secondary decomposition of levoglucosan in pyrolytic production from cellulosic biomass. J Anal Appl Pyrol 2009;85(1):247–51. Kawamoto H, Murayama M, Saka S. Pyrolysis behavior of levoglucosan as an intermediate in cellulose pyrolysis: polymerization into polysaccharide as a key reaction to carbonized product formation. J Wood Sci 2003;49(5):469–73. Yang H, Yan R, Chen H, Lee DH, Zheng C. Characteristics of hemicellulose, cellulose and lignin pyrolysis. Fuel 2007;86(12):1781–8. Li S, Lyons-Hart J, Banyasz J, Shafer K. Real-time evolved gas analysis by FTIR method: an experimental study of cellulose pyrolysis. Fuel 2001;80(12):1809–17. Liu Q, Wang S, Zheng Y, Luo Z, Cen K. Mechanism study of wood lignin pyrolysis by using TG–FTIR analysis. J Anal Appl Pyrol 2008;82(1):170–7. Shen DK, Gu S. The mechanism for thermal decomposition of cellulose and its main products. Bioresour Technol 2009;100(24):6496–504. Wang S, Guo X, Liang T, Zhou Y, Luo Z. Mechanism research on cellulose pyrolysis by Py-GC/MS and subsequent density functional theory studies. Bioresour Technol 2012;104:722–8. Banyasz JL, Li S, Lyons-Hart J, Shafer KH. Gas evolution and the mechanism of cellulose pyrolysis. Fuel 2001;80(12):1757–63. Werner K, Pommer L, Broström M. Thermal decomposition of hemicelluloses. J Anal Appl Pyrol 2014;110:130–7. Ronsse F, Bai XL, Prins W, Brown RCJEP, Energy S. Secondary Reactions of Levoglucosan and Char in the Fast Pyrolysis of Cellulose. 2012;31(2):256–60.