Study on Pyrolysis Behaviors of Various Plant Fibers

Abstract

Pyrolysis is an effective way to convert plant fibers into high-value-added chemicals and bioenergy. The pyrolysis behavior of plant fibers varies with their compositions. A high-performance anion-exchange chromatography integrated pulse amperometric method was established to detect the composition of arabinose, galactose, glucose, xylose, and mannose in plant fiber hydrolysate. The contents of cellulose, hemicellulose, and lignin in six plant fibers were calculated. Furthermore, the pyrolysis kinetic parameters of the plant fibers and their pyrolysis product distribution depending on chemical compositions were analyzed. The pyrolysis of flax fiber with high cellulose content (92.19%) tended to generate ketones, accounting for about 37.3% of the total product distribution, while coniferous and broadleaf fiber with high hemicellulose contents (13.23 and 15.07%, respectively) was more likely to generate aldehydes and hydrocarbons. Furthermore, the result of pyrolysis of a grass fiber demonstrated the interactions between its chemical components, which had been captured during pyrolysis from the perspective of pyrolysis product distribution that inhibits the pyrolysis to generate CO2, and promoted the generation of furan, phenols, and toluene, to different degrees. The research results are expected to provide basic data and theoretical support for obtaining high-value-added chemicals and biomass energy through the pyrolysis of plant fibers.

1. Introduction

Plant fiber is a kind of natural composite, which is mainly composed of cellulose, hemicellulose, and lignin [1]. The three polysaccharides are quite different in the composition of monosaccharides. Cellulose is a linear polymer composed of dehydrated glucose linked by β-1,4-glycosidic bonds. Hemicellulose is much more complex than cellulose. It is a heteropolysaccharide with some branches composed of pentose (xylose and arabinose), hexose (glucose, mannose, and galactose) and hexoic acid (4-O-methyl-D-glucuronic acid, D-glucuronic acid, and D-galacturonic acid). These functional groups can be assembled into various hemicellulose polysaccharides with different structures from linear to highly branched, such as β-1,4-D-xylan, arabinose xylan, mannan, dextran, galactose, and galactomannan. The detailed sugar compositions and chemical structures of these hemicellulosic polysaccharides vary according to plant species [1, 2]. Lignin is an aromatic polymer with highly branched chains, which is composed of phenylpropane derivative monomers (such as coumarin, coniferol, and sinapinol). The structure of lignin is complex and its molecular weight is large [1]. Due to the complex connection between cellulose, hemicellulose, and lignin in plant fibers, it is a challenge to complete the separation of the three components. At present, the “nitric acid-ethanol method,” the “12% hydrochloric acid hydrolysis method,” and the “72% sulfuric acid method” is used to measure the cellulose content, hemicellulose or pentosan content, and lignin content, respectively. However, these methods are cumbersome and the structural composition of hemicellulose could not be distinguished effectively. The National Renewable Energy Laboratory (NREL) of the United States has developed a method to systematically analyze the contents of cellulose, hemicellulose, and lignin in fibers by two-stage acid hydrolysis [3]. This method was improved by us to use high concentration sulfuric acid to convert plant fibers into oligosaccharides at low temperatures firstly, and then using dilute acid to further convert oligosaccharides into monosaccharides at high temperatures. The contents of cellulose and hemicellulose are determined based on analyzing monosaccharides by high-performance liquid chromatography. The contents of acid-insoluble lignin and acid-soluble lignin are determined by weighing the filter residue and detecting acid hydrolysate with UV. It is simple and easy to operate and is widely used by international research institutions [3, 4], and it is also economically viable for scaling up.

Moreover, pyrolysis of plant fibers is a multistep reaction process due to their complex multi-components. At present, calculation of pyrolysis kinetic parameters and analysis of pyrolysis product distribution from the perspective of cellulose, hemicellulose, lignin, and other component groups are effective ways to deeply understand plant fibers’ pyrolysis behavior. The pyrolysis and combustion kinetics of various fibers are essential for the multipurpose utilization of biomass materials, because pyrolysis is an effective way to convert plant fibers into high-value-added chemicals and bioenergy [5, 6, 7, 8, 9]. Compared with model method, model-free kinetics, as Friedman method [10, 11, 12, 13, 14], Vyazovkin method [10, 13, 15], Ozawa method [16, 17], Kissinger-Akahira-Sunose (KAS) method [13, 15, 18], Flynn-Wall-Ozawa (FWO) method [13, 15, 18], and distributed activation energy model (DAEM) [11, 19, 20, 21, 22], could reduce the errors introduced by the model fit. According to Friedman method [11, 13, 16], the pyrolysis and combustion processes of different plant fibers by using multiple heating rate programs were used to obtain reliable kinetics, in order to reveal their pyrolysis and combustion properties. In addition, the pyrolytic characteristics and tendentious pyrolysis reaction path of each polysaccharide (cellulose, hemicellulose, or lignin) could be demonstrated from the perspective of the pyrolytic product distribution of plant fibers with different chemical compositions. Interactions among these different chemical compositions during pyrolysis were also first noticed and speculated by analyzing the pyrolysis product of the actual whole and the sum of pyrolysis product by the accumulation of individual chemical composition. These pyrolysis properties investigated above would be useful references for the production of high-value-added products and biomass energy from plant fibers.

2. Analysis of plant fiber composition

High-performance anion-exchange chromatography with an integrated pulsed amperometric method (HPAEX-IPAM) is a newly developed method for sugar analysis in recent years. Zheng [23, 24] established HPAEX-IPAM for the determination of monosaccharides (arabinose, galactose, glucose, xylose, and mannose), glycuronate acid (galacturonic acid and glucuronic acid), and cellobiose in plant fiber hydrolysate. In order to hydrolyze cellulose, National Renewable Energy Laboratory (NREL) method optimized by the heating method is used to hydrolyze cellulose into monosaccharides. Then, monosaccharides are separated by an efficient anion exchange column in a strong alkaline medium, and then, the current generated by the oxidation reaction of hydroxyl groups in the sugar molecular structure on the gold electrode surface is detected [25, 26]. This method is simple in pretreatment, and the pulsed amperometric detector has high sensitivity, selectivity, accuracy, and reproducibility. Calculate the content of fiber components, according to the concentration of hydrolysis products of each fiber component. Zheng [23, 24] used this method to determine the acid hydrolysis monosaccharides of six common plant fibers (coniferous fiber, broadleaf fiber, bamboo fiber, flax fiber, grass fiber, and cotton fiber), and calculated the composition of cellulose and hemicellulose of each plant fiber, according to the corresponding hydrolysis monosaccharides mass fraction of cellulose and hemicellulose (as shown in Table 1). The lignin content shown in Table 1 is the sum of acid-insoluble lignin and acid-soluble lignin, which are determined by weighing the filter residue and detecting acid hydrolysate with UV.

Samples	Cellulose	Hemicellulose	Lignin
Coniferous fiber	74.56 ± 0.23	13.23 ± 0.43	9.32 ± 1.61
Broadleaf fiber	75.80 ± 0.31	15.07 ± 0.55	5.55 ± 1.21
Bamboo fiber	74.05 ± 0.52	18.28 ± 0.46	5.86 ± 0.71
Flax fiber	92.19 ± 0.34	4.32 ± 0.16	2.33 ± 0.51
Grass fiber	71.90 ± 0.17	17.59 ± 0.55	4.28 ± 0.50
Cotton fiber	95.84 ± 0.97	0.23 ± 0.01	1.17 ± 0.01

Table 1.

Information of plant fiber compositions (%) [27].

3. Pyrolysis behaviors of plant fibers

3.1 Pyrolysis characteristic of plant fibers

In general, the process of biomass pyrolysis includes four stages [5]. The first stage (<150^°C) is dehydration, where the biomass releases external water by absorbing heat. This registers as a slight weight loss on the TG curve. The second stage (150–250°C) is evaporation and distillation of the volatile and semi-volatile components. At the same time, lignin begins to lose weight and releases some small molecules, and the crystalline regions of cellulose transform to amorphous partly. The major pyrolysis stage (250–500°C) shows the maximum weight loss peak on the DTG curve of the biomass. The last stage (>500°C) corresponds to the further slow decomposition of residues that generated from incomplete pyrolysis. Figure 1 shows the TGA curve of coniferous fiber at a heating rate of 10°C/min and it presents the typical pyrolysis characteristics of biomass under inert atmosphere. The initial pyrolysis temperature T_s and terminal temperature T_h are obtained by the tangent method [28, 29], that are 339°C and 362.4°C, respectively.

Figure 1.

TG/DTG curves of coniferous fiber sheet under N₂ atmosphere (heating rate β = 10°C/min) [27].

The pyrolysis behaviors of various plant fibers are quite distinct because of their different morphologies, chemical components, and pyrolysis conditions. Table 2 demonstrates the pyrolysis characteristic parameters of various plant fibers. With an increase in heating rates, the values of these parameters T_s,T_h, and the peak temperature T_max show an increasing trend [27]. Zhao [30] investigated the pyrolysis rate elevated linearly with the increase in heating rates. The heating rate has an appreciable impact on the temperature difference for heat transfer and the temperature gradient between the measuring point and the sample, besides, that causes an additional endothermic amount to balance the thermal hysteresis. The distribution of cellulose, hemicellulose, and lignin demonstrates differences across species of plant fibers [31, 32, 33]. The pyrolysis temperature of the hemicellulose is the lowest, while that of lignin behaves the highest. As shown in Table 2, the maximum mass loss rate −(dm/dt)_max of the fibers rises with the increase of the heating rate, which −(dm/dt)_max of cotton fiber is the highest, and −(dm/dt)_max of grass fiber is the lowest. When the heating rate is 15°C/min, the values of −(dm/dt)_max are 35.2%/min and 24.1%/min, respectively. In addition, the pyrolysis index P [34, 35], which reflects the pyrolysis degree of fibers, enhances with the increase of heating rates.

Samples	β	Ts	Th	Tmax	−(dm/dt)max	△T1/2	P × 10^−6
	/K·min^−1	/K	/K	/K	/(%·min^−1)	/°C	/(%·min^−1·K^−3)
Coniferous fiber	5	602.8	637.1	624.6	12.5	25.3	1.31
	10	612.0	649.3	635.4	22.9	27.9	2.11
	15	617.4	657.3	641.6	32.1	30.5	2.65
Broadleaf fiber	5	597.8	634.0	621.4	11.9	26.2	1.22
	10	608.9	648.1	634.1	22.2	29.1	1.98
	15	613.5	655.5	640.1	31.0	31.7	2.49
Bamboo fiber	5	595.2	633.3	619.2	11.0	27.7	1.08
	10	604.6	646.2	631.2	20.3	30.7	1.74
	15	609.3	652.4	636.5	29.1	32.6	2.30
Flax fiber	5	607.6	645.2	630.8	11.5	29.3	1.03
	10	618.3	658.3	643.2	21.8	31.7	1.73
	15	625.0	666.0	649.1	31.6	27.7	2.81
Grass fiber	5	591.7	637.8	619.8	9.0	33.0	0.74
	10	601.2	651.1	631.9	16.9	36.1	1.23
	15	607.4	659.8	639.7	24.1	39.0	1.59
Cotton fiber	5	610.0	643.4	629.8	13.2	27.1	1.27
	10	620.6	656.4	640.8	24.8	29.9	2.09
	15	627.1	664.0	648.1	35.2	31.4	2.76

Table 2.

Pyrolysis characteristic parameters of fibers [27].

β: heating rate; T_s: initial decomposition temperature; T_h: terminal decomposition temperature; T_max: peak temperature; −(dm/dt)_max: maximum mass loss rate; △T_1/2: peak width at half-height; and P: pyrolysis index.

3.2 Pyrolysis and combustion kinetics of plant fibers

As a typical biomass, the pyrolysis and combustion kinetics of various plant fibers are essential for the multipurpose utilization of biomass materials [6–9]. The TG/DTG data obtained at different heating rates (5, 10, and 15^°C/min) were applied to get the relationship between $\ln \left[\beta {\left(\frac{d\alpha }{dT}\right)}_{\alpha }\right]$ and $\frac{1}{T_{\alpha }}$ for each plant fiber, at selected α values based on the Friedman method, as shown in Figure 2. The activation energy (E_a) corresponding to the α values could be obtained from the slope of the fitting curve, as shown in Table 3 [27]. The value of α ranged from 0.05 to 0.95, and the step size was 0.05.

Figure 2.

Friedman results of pyrolysis of different fiber sheets under N₂ atmosphere (a) coniferous fiber, (b) broadleaf fiber, (c) bamboo fiber, (d) flax fiber, (e) grass fiber, and (f) cotton fiber [27].

α	Coniferous		Broadleaf		Bamboo		Flax		Grass		Cotton
	Eα	R^2	Eα	R^2	Eα	R^2	Eα	R^2	Eα	R^2	Eα	R^2
0.05	224.16	0.9962	185.76	0.9769	214.39	0.9837	182.96	0.9952	183.16	0.9989	206.94	0.9994
0.1	228.85	0.9992	196.07	0.9927	227.60	0.9968	193.57	0.9997	209.72	0.9971	186.28	1.0000
0.15	211.04	1.0000	188.48	0.9910	216.08	0.9949	184.16	1.0000	203.67	0.9985	189.93	0.9998
0.2	210.38	0.9999	188.72	0.9954	218.02	0.9962	188.78	1.0000	184.92	0.9998	188.52	0.9996
0.25	203.58	0.9998	185.37	0.9963	213.00	0.9967	187.00	0.9999	186.89	0.9992	186.82	0.9995
0.3	197.55	0.9995	181.42	0.9972	205.12	0.9976	186.65	1.0000	182.30	0.9993	185.16	0.9998
0.35	193.79	0.9994	178.06	0.9980	201.10	0.9963	184.48	0.9999	176.78	0.9992	189.52	1.0000
0.4	190.80	0.9998	173.30	0.9980	198.15	0.9974	184.77	1.0000	173.92	0.9986	191.19	0.9990
0.45	188.00	0.9994	171.94	0.9983	193.86	0.9962	186.93	0.9999	170.33	0.9987	187.55	0.9994
0.5	186.02	0.9995	169.93	0.9986	193.62	0.9974	187.58	0.9996	168.08	0.9989	181.27	1.0000
0.55	184.37	0.9996	168.42	0.9977	192.50	0.9968	185.69	1.0000	168.19	0.9988	177.59	0.9999
0.6	179.99	0.9987	167.40	0.9982	191.93	0.9979	182.11	1.0000	167.02	0.9986	176.19	1.0000
0.65	177.07	0.9978	165.31	0.9989	191.79	0.9975	179.71	0.9996	167.47	0.9988	176.97	0.9999
0.7	174.73	0.9953	161.26	0.9995	192.96	0.9972	178.66	0.9991	170.44	0.9986	177.59	1.0000
0.75	175.14	0.9898	160.78	1.0000	196.99	0.9981	177.87	0.9987	173.45	0.9983	178.37	1.0000
0.8	177.50	0.9845	162.33	1.0000	205.95	0.9992	180.06	0.9975	179.63	0.9982	184.03	0.9999
0.85	194.01	0.9744	178.59	0.9994	245.74	1.0000	188.62	0.9935	194.08	0.9976	199.28	0.9998
0.9	422.16	0.9265	334.18	1.0000	1227.04	0.8234	261.73	0.9700	287.75	0.9931	267.37	0.9945
0.95	−378.69	0.8906	−396.67	0.5410	−400.47	0.9187	−227.41	0.6301	259.91	0.7083	−232.31	0.8507
Eα (α:0.05–0.85)	193.79		173.30		201.10		184.77		176.78		186.28

Table 3.

The apparent activation energy E_α(kJ/mol) of different fibers pyrolysis under N₂ atmosphere obtained by Friedman method and correlation coefficient R² [27].

Due to the complicated pyrolysis processes of biomass, the apparent kinetics could be described by the appropriate models based on the global weight loss, while the pyrolysis mechanism of a specific component is still an intractable problem. Plant fiber is composed of complex multi-components and its pyrolysis is a multistep reaction process. As indicated in Table 3, the resulting apparent activation energies of plant fibers are various with the conversion rate. The average apparent activation energies of coniferous, broadleaf, bamboo, flax, grass, and cotton fibers are 193.79, 173.30, 201.10, 184.77, 176.78, and 186.28 kJ/mol, respectively, among the conversion range of 0.05–0.85. The average apparent activation energy of broadleaf fiber is the lowest and that of bamboo fiber is the highest.

Although biomass pyrolysis is generally carried out under an inert atmosphere, some research has been performed under an oxygen atmosphere [27]. The characteristic parameters, both of pyrolysis process of fibers and combustion of residual fixed carbon, as Ts, Th, and Tmax, demonstrate a high-temperature-shifting with the increase of the heating rate. The Tmax of all fibers behaves are lower than those under nitrogen condition, and the apparent activation energies show a similar variation trend, especially among the conversion between 0.05 and 0.65. Cotton fiber shows the highest Tmax, compared with grass fiber, which is the lowest. Improving the heating rate benefits the combustion performance of plant fibers and the combustion characteristic index (S) [27] increases accordingly. The result shows that the broadleaf fiber has the largest combustion characteristic index, while grass fiber has the smallest. The apparent activation energy of plant fiber pyrolysis in an oxygen atmosphere is lower than that in nitrogen, indicating that oxygen atmosphere can promote the pyrolysis reactions of plant fibers.

4. Pyrolytic product distribution analysis of plant fibers affected by chemical compositions and interactions among them

Pyrolysis behaviors of coniferous, broadleaf, bamboo, flax, grass, and cotton fibers have been proven to be relevant with the biomass types and chemical compositions [5], showing different pyrolysis kinetic parameters because of varieties of cellulose and hemicellulose compositions in diverse plant fibers [27]. The distribution of pyrolysis products would also be largely restricted by the different chemical compositions of these fibers. The pyrolysis products and their contents of flax fiber, coniferous fiber, and broadleaf fiber at 350°C are shown in Table 4 [36]. The main pyrolysis products of several plant fibers included alcohol, aldehyde, ketone (furanone), acid, ester, hydrocarbon, anhydrosugar, and CO₂. Flax fiber contained a high proportion of cellulose (92.35%), hence, a high proportion of glucose molecular units and its ketone yield in the pyrolysis product is the highest (37.3%). The glucose ring could be easily broken at C₁-O and C₂-C₃ confirmed by Piskorz et al. [37] based on bond energy analysis, forming fragments of two carbon atoms (C₁/C₂) and four carbon atoms (C₃–C₆), of which two-carbon fragments could form hydroxyacetaldehyde, while four-carbon fragments would generate hydroxyacetone and other products after thermal decomposition process. In addition, 1,3-dihydroxy-2-propanone mainly originates from the breaking of the cellulose chain during pyrolysis [38]. Compared with flax fiber, the cellulose contents of coniferous fiber and broadleaf fiber are significantly reduced, the hemicellulose contents are significantly increased, and their ketone yields after pyrolysis show significantly reduced, while the yields of aldehydes and hydrocarbons are enhanced. It could be related to the fact that the pyrolysis of five-carbon sugars (xylose and arabinose) tended to generate furfural, and the pyrolysis of six-carbon sugars (glucose, mannose, and galactose) tended to generate 5-hydroxymethylfurfural [3, 39, 40, 41]. 5-hydroxymethylfurfural is mainly obtained through an acetal reaction between C-2 and C-5, furfural can be obtained through the hydroxymethyl elimination reaction of 5-hydroxymethylfurfural [38], or through the cyclization/dehydration reaction of xylose [39].

Pyrolysis products percent (%)	Aldehydes	Ketones	Hydrocarbons	Esters	Alcohols	Anhydrosugars	Acids	CO2
Flax fiber	17.9	37.3	11.1	7.0	11.7	0.5	2.4	9.7
Coniferous fiber	24.3	13.6	15.6	14.1	12.9	0.5	6.2	7.5
Broadleaf fiber	23.8	20.6	15.1	7.9	9.5	0.4	4.0	10.2

Table 4.

Pyrolysis product distribution of flax fiber, coniferous fiber, and broadleaf fiber at 350°C [36].

Hemicellulose and lignin of lobular seal could be obtained through step-by-step extraction by Xuefei Cao [42], the main components of water-soluble hemicellulose are β-D-glucan and a small amount of pectin. The main component of alcohol-soluble hemicellulose is poly arabinogalactose. The main pyrolysis range of cellulose and hemicellulose spans 200–400°C, and their pyrolysis products are mainly carbonyl compounds. The main pyrolysis range of lignin is 300–700°C, and its pyrolysis products are mainly aromatic compounds. Both glycosidic bonds of cellulose and hemicellulose and the C-C bonds on the sugar ring have low dissociation energy, which is easy to break during pyrolysis to produce low carbon and oxygen-containing small molecule products. The bond dissociation energies of the methyl group on methoxy group of lignin monomer, C_α-C_β connection bond on side chains, as well as β-O-4′ ether bond and C_α-C_β in lignin dimer connection are small, so the lignin pyrolysis is easy to produce phenolic products with short side chains.

Therefore, the distribution of fiber pyrolytic products is closely related to their differences in cellulose/hemicellulose composition, and maybe the interactions between them as well. The production and utilization of reconstituted tobacco leave based on component reconstruction involved the separation and reorganization/reassembling of water-impregnated extracts, ethanol-impregnated extracts, etc. of tobacco biomass. We take a grass fiber as an experimental example, its water-soluble components, ethanol-soluble components, and residual solid-phase components could be obtained through a step-by-step extraction process. The actual pyrolysis product distribution of the grass fiber and the weighted pyrolysis product distribution based on its component distribution are investigated, and the comparison results confirm that there are interactions between the chemical components of the plant fiber during pyrolysis [43], in which the generation of CO₂ is inhibited by the interactions, and the generation of furan, phenol, and toluene was promoted by the interactions between components to varying degrees.

5. Conclusions

In this chapter, a method for the determination of arabinose, galactose, glucose, xylose, mannose, and other monosaccharides in plant fiber hydrolysate is established based on high-performance anion exchange chromatography integrated pulse amperometric technique. The acid hydrolysis monosaccharides in coniferous fiber, broadleaf fiber, bamboo fiber, flax fiber, grass fiber, and cotton fiber could be determined by this method, and cellulose, hemicellulose, and lignin contents of each plant fiber were further calculated. Plant fibers are complex polymers, and their pyrolysis process is very complex. Under inert atmosphere, the average apparent activation energies of coniferous fiber, broadleaf fiber, bamboo fiber, flax fiber, grass fiber, and cotton fiber with different chemical compositions in the conversion rate of 0.05–0.85 are 193.79, 173.30, 201.10, 184.77, 176.78 and 186.28 kJ/mol, respectively. The apparent activation energies of plant fibers pyrolysis in the oxygen atmosphere are lower than those in the nitrogen atmosphere. The oxygen atmosphere can promote the pyrolysis of plant fibers. The pyrolysis of flax fiber with high cellulose content tends to generate ketones, while coniferous fiber and broadleaf fiber with high hemicellulose contents are more likely to generate aldehydes and hydrocarbons. Furthermore, taking a grass fiber as the experimental object, interactions between its chemical components have been captured during pyrolysis from the perspective of pyrolysis product distribution, which inhibits the pyrolysis to generate CO₂, and promote the generation of furan, phenols, toluene, etc. to different degrees.

Conflict of interest

The authors declare no conflict of interest.

Acronyms and abbreviations

NREL

National Renewable Energy Laboratory

E_α

Average apparent activation energy

KAS

Kissinger-Akahira-Sunose method

FWO

Flynn-Wall-Ozawa method

DAEM

Distributed activation energy model

HPAEX-IPAM

High-performance anion-exchange chromatography with integrated pulsed amperometric method