Separation and Purification Technology 157 (2016) 93–101
Contents lists available at ScienceDirect
Separation and Purification Technology journal homepage: www.elsevier.com/locate/seppur
Enhanced lignin extraction process from steam exploded corn stalk Guanhua Wang a,b, Hongzhang Chen a,⇑ a b
State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China University of Chinese Academy of Sciences, Beijing 100039, China
a r t i c l e
i n f o
Article history: Received 28 June 2015 Received in revised form 24 November 2015 Accepted 25 November 2015
Keywords: Lignin alkaline extraction Steam exploded corn stalk Enhanced kinetics Neutral solvent extraction Depolymerization characterization
a b s t r a c t In this work, the alkaline extraction processes of lignin from untreated and steam-exploded corn stalk were compared and the enhanced extraction process was investigated by neutral solvent extraction and structural analysis of lignin. It was found that a first order model could readily simulate the lignin extraction from untreated stalk while it was helpless to describe the enhanced process after steam explosion due to the intense dissolution of lignin at the beginning stage of extraction. The lignin extracted by neutral solvent from ball-milled corn stalk and its structural characterization demonstrated that the poor description of extraction process by a first order model was mainly attributed to the considerable depolymerization during steam explosion. Due to the depolymerization of lignin, the alkaline extraction after steam explosion became the direct dissolution of depolymerized lignin and was successfully simulated by a second-order model. Finally, the activation energy of lignin alkali-extraction decreased 37.78% after steam explosion treatment with 1.8 MPa and 5 min, which indicated that milder extraction conditions were available to extract lignin from steam-exploded corn stalk. This study also provides a basic guidance for delignification approaches of lignocelluloses after steam explosion. Ó 2015 Published by Elsevier B.V.
1. Introduction The slow decline of fossil resources and the deterioration of human environment prompt people to rethink the adverse impacts brought by thoughtless use of non-renewable resources and look attentively for sustainable resources and energy. Lignocellulosic biomass is an apt resource to displace fossil resources for energy and useful chemicals production owing to its renewable character and abundance in nature [1]. However, actual lignocellulosic biomass shows a highly tight architecture that contains cellulose fibrils and complex ‘‘matrixing” polymers mainly including hemicelluloses and lignin. Thus, pretreatments of biomass to destroy the compact network structure and enhance the accessibility of saccharification are normally necessary [1]. Among the alternative pretreatments, steam explosion is considered to be an efficient pretreatment due to fewer environmental impacts, investment costs and energy demand compared with other treatments which require the addition of other chemicals [2]. During a steam explosion process, biomass is subjected to high-pressure steam followed by rapid decompression, in which partial hemicelluloses are degraded and the fibrous materials are shattered by the instantaneous expansion force. The degraded hemicelluloses are ⇑ Corresponding author at: 1 North 2nd Street, Zhongguancun, Haidian District, Beijing, China. E-mail address:
[email protected] (H. Chen). http://dx.doi.org/10.1016/j.seppur.2015.11.036 1383-5866/Ó 2015 Published by Elsevier B.V.
eliminated by the subsequent water washing and the lignin-rich loose fibrous materials are proposed for enzymatic hydrolysis and ethanol fermentation. However, the remaining lignin is still a main limitation for the effective enzymatic saccharification of cellulose due to the spatial obstacle [3,4]. In addition, lignin also decreases the activity of cellulase by non productive binding [5]. Therefore, lignin removal is important for steam-exploded lignocelluloses to produce ethanol and other bioenergy. Unlike the other two major components in lignocelluloses which are released as monomeric carbohydrates, lignin is composed of three different phenylpropane monomers and polymerized randomly by ether linkages or C–C bonds. It contributes as much as 30% of the weight and 40% of the energy content of lignocellulosic biomass [6]. Lignin’s native structure allows its versatility of applications as phenol substitution for synthetic resins, dispersant for dyes and pesticides, binder for feeds and fertilizers. Currently, other value-added applications such as lignin-based polymers, carbon fibers, antioxidant and depolymerized chemicals, [7] are also being developed. Since the bioconversion of fractionated cellulose to biofuels is not economically competitive due to the little use of other biomass components, the effective utilization of lignin offers a significant opportunity for the commercial operation of a lignocellulosic biorefinery [8]. Extraction is the first step in isolation and utilization of lignin from steam-exploded lignocelluloses. Alkaline extraction is a relatively facile and efficient process to obtain lignin from
94
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
lignocellulosic biomass compared with other solvent extraction due to the high solubility of the polymer in alkali. Generally, lignin is supposed to be depolymerized by aryl ether breakage in a heated alkaline medium and dissolved in the liquid [9,10]. During this cooking process, depolymerization of lignin into fragments destroys the impact triaxial structure and increases the solubility of lignin by reducing the molecular weight and enhancing the content of hydrophilic groups, such as phenolic hydroxyl groups. Therefore, the depolymerization of lignin is essential for alkaline extraction. When steam explosion is used to assist the extraction of bioactive compounds from plant tissues, it is suggested to enhance the extraction efficiency by increasing the specific surface area due to the formation of large fissures and micropores [11–13]. However, the high temperature and weak acid during steam explosion also induce severe changes of original lignin in the cell wall due to the high sensitivity of protolignin structure [14,15]. The reactions of lignin can change its solubility in an alkaline medium, which also enhances the kinetics of lignin extraction after steam explosion. However, whether the increase of specific surface area or the changes of lignin structure contribute more to the lignin alkaline extraction kinetics after steam explosion is still ambiguous. In this work, the alkaline extraction processes of lignin from untreated and steam-exploded corn stalk were firstly investigated and compared. In order to eliminate the effects of alkali-induced lignin depolymerization, a neutral solvent of dioxane/water (96:4, v/v) was used to extract lignin from ball-milled corn stalk before and after steam explosion treatment. Moreover, gel permeation chromatography (GPC), heteronuclear single quantum coherence (HSQC) and quantitative 13C NMR spectra were applied to characterize the variation of lignin structure and explore the enhancement mechanism of lignin extraction process after steam explosion treatment. Finally, a second-order extraction model was employed to simulate the lignin extraction process according to its extraction behavior from steam-exploded corn stalk and the extraction activation energy (Ea) was calculated. 2. Materials and methods 2.1. Experimental materials Corn stalk harvested in September 2012 was kindly provided by the Chinese Academy of Agricultural Sciences, Beijing. The corn stalk compositions were determined according to the modified National Renewable Energy Laboratory (NREL) analytic methods [16,17]. Specifically, 0.3 g corn stalk was treated with 3 mL 72% H2SO4 at 30 °C for 2 h, then diluted to 4% and autoclaved at 121 °C for 45 min. The hydrolysis solution was filtered and analyzed for sugar content. The sugar content was determined by HPLC (1200 series, Agilent Technologies) equipped with a refractive index detector (RID) and an organic acid analysis column (AminexÒ HPX-87H Ion Exclusion Column). The column was operated at 65 °C and eluted with 5 mM H2SO4 solution at a flow rate of 0.5 mL/min. The solid residue was dried at 105 °C for 12 h and further placed in the muffle furnace at 550 °C for 2 h. The weight of ash was recorded and the content of Klason lignin was calculated by deducting the ash content from the solid residue. Each sample was analyzed in triplicate. The average chemical compositions of dried corn stalk were cellulose 46.54 ± 1.87% (w/w), hemicellulose 28.41 ± 0.87% (w/w), Klason lignin 20.42 ± 1.56% (w/w) and ash 4.62 ± 0.30% (w/w). All laboratory reagents were chemically pure. 2.2. Steam explosion pretreatment Steam explosion pretreatment was performed in a 5 L batch vessel (Weihai Automatic Control Reactor Ltd., China) with
previously established optimum conditions of 1.8 MPa pressure and 5 min treatment time [16]. After pretreatment the corn stalk was cleaned with distilled water to dissolve the water soluble hemicelluloses and squeezed to obtain solid residue. The solid residue, named steam-exploded corn stalk, was dried in a forced-air oven at 55 °C for 48 h and stored in a dry place at room temperature until usage. According to the analysis method in Section 2.1, the average chemical compositions of steam exploded corn stalk were cellulose 55.34 ± 1.03% (w/w), hemicellulose 11.49 ± 0.25% (w/w), Klason lignin 30.31 ± 1.47% (w/w) and ash 2.90 ± 0.34% (w/w). 2.3. Lignin extraction and evaluation 2.3.1. Lignin alkaline extraction The untreated and steam-exploded corn stalks were smashed and screened through a 40 mesh sieve (Zhejiang Yingchao Instrument Co., Ltd., China). About 0.08 g (measured precisely) of dried corn stalk was extracted with 1 ml of 1% NaOH in a sealed pressure-tolerant glass tube placed in a digester (Shanghai Yidian Co., Ltd., China) which was heated to the desired temperature. These experiments were performed at 40, 60, 80, 100, 120 and 140 °C in two repetitions. After extraction, the sealed tube was opened carefully (especially when extraction temperature was over 100 °C) and quickly placed in an ice bath. After that, the extract was diluted to 40 ml with distilled water and centrifuged at 4 °C for 5 min. The supernatant was separated from solid residue, diluted by 10-fold distilled water and detected by ultraviolet (UV) absorption at 280 nm (UVmini-1240, SHIMADZU). 2.3.2. Milled corn stalk lignin (MSL) extraction by neutral solvent Before ball-milling the sample without steam explosion was firstly extracted with ethanol/benzene (1:2, v/v) using Soxhlet extractor to eliminate the extractives which might influence the content of neutral solvent extraction. Ball-milling of untreated and steam-exploded corn stalk was carried out in an agate ball mill (Tianjin Shengyuan Electrical Equipment Co., Ltd., China) for 8 h. In order to avoid increase of temperature, the agate jar with samples was placed in a cold water bath (15 °C) for 10 min after every 30 min ball milling. 5.00 g of ball milled samples were extracted with 125 ml of dioxane/water (96:4, v/v) solvent in sealed flasks placed in a shaker at 120 rpm at 25 °C. When the desired extraction time was attained, 2 ml extract was collected and centrifuged at 4 °C for 5 min immediately. The supernatant was diluted by dioxane/water solvent and detected by UV absorption at 280 nm. The extraction yields of MSL before and after steam explosion were presented by the proportion of extracted lignin to the Klason lignin in the corn stalk. 2.3.3. Establishment of lignin content measurement by UV absorption The dried solid lignin sample from alkali extract was obtained by acid precipitation of alkali extracting solution to pH 2.0 followed by the sufficient water washing to eliminate the ash [18], while the solid MSL sample from neutral solvent extraction was purified according to the method outlined by Björkman [19]. 1.00 g L1 standard lignin solutions were prepared by dissolving 0.10 g alkali-extracted lignin and MSL into 100 ml alkaline water (pH = 11.5) and dioxane/water solvent respectively. The standard lignin solutions were then diluted to obtain different concentrations for the establishment of standard curve with concentration versus UV absorption. The relationship between UV absorption and lignin concentration in extracted liquor was established by linear regression firstly (Fig. 1). It was found that UV absorption precisely predicted the concentration of both alkali-extracted lignin and MSL obtained from untreated and steam-exploded corn stalk. The slightly different absorption of lignin samples before and after
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
Ultraviolet Absorption (280 nm)
a
process was assumed to occur in a system with pseudohomogeneous concentration of NaOH. Thus, the product of k and S in Eq. (1) can be replaced by a new constant k1, and the model can be simply written as:
1.2 Lignin after steam explosion 2 Y=19.711X+0.029 R = 0.999
1.0
dL=dt ¼ k1 ðLe LÞ
0.8
L ¼ Le ½1 expðk1 tÞ
ð3Þ
0.4
2.5. Lignin characterization
Lignin before steam explosion 2 Y=18.555X+0.0324 R = 0.996
0.2
0.01
0.02
0.03
0.04
The molecular weight distributions of the MSL before and after steam explosion pretreatment were detected by gel permeation chromatography (GPC) using Agilent 1200 series HPLC system equipped with a hydrophobic column (Agilent Zorbax PSM 300S) and a Diode Array Detector (280 nm). Lignin samples were dissolved in dioxane/water (9:1, v/v) and 20 ll samples were injected. The column was operated at 25 °C and eluted with dioxane/water solvent (9:1, v/v) at a flow rate of 0.5 ml/min [14]. Monodisperse polystyrene was used as the standard for the molecular weight of lignin. 13C NMR spectra and 2D NMR spectra (HSQC) of MSL before and after steam explosion were acquired using a Bruker Avance 600 MHz. 0.075 g MSL was dissolved sufficiently in 1 ml DMSOd6 and used for quantitative 13C NMR spectra record with an inverse-gated decoupling sequence, 90° pulse width and 1.7 s pulse delay [24]. The HSQC spectra was acquired by applying a 90° pulse, 0.11 s acquisition time, a 1.5 s pulse delay and 1JC–H of 145 Hz [24]. Specific surface area measurements of ball-milled corn stalk were obtained from nitrogen adsorption using an automatic specific surface area analysis instrument (3H-2000A, Beishide Instruments S&T). The parameters were set as follows: gas flow 85 ml/min, program-controlled voltage from 10 to 13 V and electric current from 90 to 110 mA. Each sample was analyzed twice.
0.05
Lignin Concentration (g/L)
Ultraviolet Absorption (280 nm)
ð2Þ
After rearranging Eq. (2), L can be expressed as: 0.6
0.0 0.00
b
95
1.0 Lignin after steam explosion 2 Y=30.299X+0.0023 R = 989 0.8
0.6
0.4 Lignin before steam explosion 2 Y=28.725X+0.0001 R =0.995 0.2 0.010
0.015
0.020
0.025
0.030
Lignin Concentration (g/L) Fig. 1. Calibration curves of ultraviolet absorption at 280 nm to the lignin concentration before (h) and after steam explosion (j). (a) Lignin obtained by alkaline extraction. (b) Lignin obtained by neutral solvent (dioxane/water) extraction. The calibration curve, correlative equation, and R2 value of each correlation were indicated. Each point was averaged from three independent measurements.
steam explosion at 280 nm may be attributed to the structural changes of lignin during steam explosion [20]. 2.4. Kinetics model of lignin alkaline extraction Lignin alkaline extraction is a complex process that includes the NaOH transfer from the solution to the solid surface, heterogeneous chemical reactions and the product transfer from the solid raw material into the solutions [21]. The chemical kinetics of lignin alkaline extraction process has been fully studied in the alkalidominant pulping. It is known in soda pulping that alkaline extraction is a first order reaction and the change rate of lignin removed per unit mass of raw materials is related to the amount of lignin remaining in the solids and the concentration of sodium hydroxide [22,23]. Thus, the following apparent kinetics mechanism for dissolved lignin is proposed:
dL=dt ¼ k ðLe LÞ S
ð1Þ
where dL/dt is the lignin extraction rate; k is the chemical reaction rate constant; Le is the equilibrium concentration of total lignin in the liquid extract (g L1); L is the dissolved lignin concentration at a given extraction time t (g L1); S is the concentration of sodium hydroxide at time t (g L1). In the present work, the liquidto-solid ratio was 12.5:1(v/m, L kg1) and the lignin extraction
3. Results and discussion 3.1. Lignin alkaline extraction process The alkaline extractions of lignin from untreated and steamexploded corn stalk were carried out at different temperatures and the extraction curves were fitted according to Eq. (3) (shown in Fig. 2). It was found that at the same extraction temperature, the lignin concentration after steam explosion was significantly higher than that without explosion. This is mainly due to the higher lignin content in the steam-exploded corn stalk since the content of hemicelluloses is decreased by hydrolysis during steam explosion [16]. In addition, the increased extraction yield of lignin due to the effect of steam explosion also contributes to the rise of lignin concentration in the extracted liquor. The fitting values of Le and k1 and the R2 are calculated and given in Table 1. The relationship between k1 and extraction temperature T can be numerically characterized by the activation energy from linearized Arrhenius equation:
ln k1 ¼ ln A ð1=TÞ ðEa =RÞ
ð4Þ
where A is a pre-exponential factor; Ea is the activation energy for the lignin extraction (kJ mol1); R is the gas constant whose value is 8.314 J/(mol K)1; and T is the absolute temperature (K). The slope of the straight line (Ea/R) is obtained by plotting ln k1 versus 1/T and the activation energy is calculated by using the slope. Fig. 3a shows the relationship between ln k1 and 1/T and it was found that the ln k1 and 1/T of lignin extraction kinetics from untreated corn stalk presented a good linear correction due to the higher
96
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
a
24
-1.5
18 16 14
-1.8
12 10
-2.1
lnk1
Lignin Concentration (g/L)
22 20
8
-2.4
6 4 2 0
Y=-2341X+4.199 2 R =0.9799
-2.7
-2 0
20
40
60 Time (min)
80
100
-3.0
120
0.0024
0.0026
0.0028
0.0030
1/T (1/K)
Table 1 Kinetic constants for alkaline extraction of lignin from untreated and steam-exploded corn stalk at different temperatures. Extraction temperature T (K)
Untreated corn stalk
Steam-exploded corn stalk
b
0.0
-0.1
lnk1
Fig. 2. Lignin concentration in alkali extract of untreated and steam-exploded corn stalk at different extraction temperature and the fitting curves of lignin extraction kinetics according to Eq. (3). h: Extracted at 60 °C without steam explosion; s: extracted at 80 °C without steam explosion; 4: extracted at 100 °C without steam explosion; 5: extracted at 120 °C without steam explosion; q: extracted at 140 °C without steam explosion; r: extracted at 40 °C after steam explosion; j: extracted at 60 °C after steam explosion; d: extracted at 80 °C after steam explosion; ▲: extracted at 100 °C after steam explosion; .: extracted at 120 °C after steam explosion.
-0.2
R2
Eq. (3) Le
k1
333 353 373 393 413
9.126 11.950 13.594 14.694 15.804
0.059 0.085 0.123 0.156 0.256
0.985 0.944 0.974 0.991 0.992
313 333 353 373 393
16.234 18.059 19.449 20.509 21.835
0.989 0.774 0.832 0.917 0.880
0.957 0.975 0.980 0.984 0.986
coefficient R2 value. According to Eq. (4), the activation energy of lignin extraction from corn stalk without steam explosion was 19.47 kJ mol1 which was similar with the result from the bagasse delignification process [25]. However, the ln k1 and 1/T of lignin extraction kinetics from steam-exploded corn stalk showed completely irregular relationship as illustrated in Fig. 3b, which meant that the model Eq. (3) was not suitable for the simulation of lignin extraction process from steam-exploded stalk. Actually, as shown in Fig. 2, the lignin concentration in the extracted liquor of steamexploded corn stalk increased dramatically in the early minutes of the extraction. After 2 h extraction, the concentration only increased by 6–15% over the value obtained at 10 min. Regardless of the temperature, the alkaline extraction of lignin was mainly achieved during the first 10 min. Therefore, compared with untreated corn stalk, steam-exploded corn stalk showed dramatically intensive dissolution of lignin at the beginning stage of extraction, which led to the poor kinetics simulation by the model. 3.2. Changes of lignin neutral solvent extraction after steam explosion Obviously, the physicochemical changes of corn stalk during the steam explosion lead to the variation of lignin extraction process. As described above, steam explosion includes two effects:
-0.3
0.0024
0.0026
0.0028
0.0030
0.0032
1/T (1/K) Fig. 3. Arrhenius plots ln k1 versus 1/T for Eq. (4). (a) Alkaline extraction kinetics of lignin from untreated corn stalk; (b) alkaline extraction kinetics of lignin from steam-exploded corn stalk.
high-temperature cooking and instantaneous expansion force on raw materials. Firstly, the instantaneously expansion force causes the enhancement of specific surface area of the fibrous materials which increases the exchange area between the solid and the alkali liquor. Secondly, the reaction of lignin during the hightemperature cooking may also change its structure and facilitates its dissolution into liquid which is similar to the process in alkaline extraction. In this case, the extraction of lignin is mainly simple dissolution of lignin from the solid to the liquid phase. In order to investigate which is the main reason for the enhancement of extraction kinetics, experiments of lignin neutral solvent extraction were performed. The untreated and steam-exploded corn stalks were both ball-milled into subtle powders to reduce the effect of specific surface area on the solid to liquid mass transfer. The literature from Björkman [19] had demonstrated that milled wood lignin extracted by dioxane/water was equal or close to proto-lignin due to neutral character of the solvent used and the mild conditions during the extraction process. Therefore, lignin from untreated and steam exploded corn stalk was extracted by neutral dioxane/water solvent with a ratio of 25 ml/g at 25 °C in order to eliminate chemical reactions during the extraction, which implied that the extraction process was simply the lignin transfer from the corn stalk into the solvent. Fig. 4 showed the lignin concentration
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
3.3. Changes of lignin structure after steam explosion
MSL extraction after steam explosion MSL extraction before steam explosion
70
MSL extraction yield (%)
60
50
40 10.0 7.5 5.0 2.5 0.0 -30
0
30
60
90 120 Time (min)
150
180
97
210
240
Fig. 4. Lignin concentration in neutral dioxane/water extract of ball milled corn stalk before (h) and after steam explosion (j).
in the extracted solvent at different time. It was found that the lignin concentration reached equilibrium in the early minutes of the extraction. This was possibly attributable to the small diameter of corn stalk particles after ball-milling [26]. As shown in Fig. 4, the yields of MSL extracted from untreated and steam exploded corn stalk were 9.76% and 59.29% respectively, which were in accordance to the previous Refs. [14,19]. However, the specific surface areas of ball-milled corn stalk after steam explosion were 1.17 and 1.64 m2/g, which meant 40.29% increase during the steam explosion treatment. Therefore, the results obviously indicated that the over 6-fold enhancement of lignin extraction yield after steam explosion was not only attributed to the increase of specific surface area but to the chemical reactions during the explosion.
The high-temperature cooking during steam explosion treatment induces hemicelluloses auto-hydrolysis and acetic acid formation to generate the weak acidic conditions. In the acidulous and heated treatment, the predominant reactions of lignin are fragmentation by acidolysis of a-O-40 /b-O-40 linkages and repolymerization by acid-catalyzed condensation between the aromatic C5 or C6 and the carbonium ion of Ca in the side chain (shown in Fig. 5) [14]. Actually, during the alkaline extraction the main reactions of lignin are breakage of b-O-40 linkages and repolymerization of the carbonium ion of Ca and the aromatic ring, which has been fully studied in the soda pulping [9,10]. Thus, a hypothesis is proposed in this work that the steam explosion actually, to some extent, induces certain reactions of lignin which are similar to the alkali reaction of lignin. Therefore, the alkaline extraction process after steam explosion becomes a direct dissolution of lignin from solid steam-exploded corn stalk into the liquid, which leads to the totally different extraction kinetics shown in Fig. 2. In order to test the hypothesis, some lignin analysis techniques including GPC, 13C NMR and HSQC spectra were applied to characterize the structural changes of lignin during the steam explosion treatment. It is notable that the two lignin samples for structural characterization before and after steam explosion treatment were MSL isolated by neutral solvent since they were less damaged during the separation process [19]. Due to the grinding scale of ball milling, it is difficult to destroy the cell wall architecture completely by ball milling and obtain total lignin. Thus the lignin yield is relatively low. However, since few changes of lignin structure occur during the neutral solvent extraction process [19], the extracted lignin (9.76% and 59.29% for untreated and steam exploded corn stalk, respectively) is actually enough to the represent the lignin existing in corn stalk. As the breakage of ether bonds and nucleophilic condensation of carbonium ions firstly change the molecular weight distribution of lignin, GPC is preferentially applied to investigate the molecular
Fig. 5. Depolymerization of a-O-40 and b-O-40 linkages and repolymerization of lignin fragments [14].
98
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
MWL before steam explosion MWL after steam explosion 1
2
Deterctor Response
4 3
10.5
11.0
11.5 12.0 12.5 Elution Time (min)
13.0
13.5
14.0
Fig. 6. Molecular weight distributions of MSL extracted from untreated and steam-exploded corn stalk.
weight variation caused by the steam explosion and the results are shown in Fig. 6. After steam explosion, a clear shift of the main peak (peak 1 and 2) towards a longer retention time in the molecular weight distribution curve was found, which indicated the occurrence of lignin depolymerization. The enhanced peak at shorter retention time (peak 3 and 4) suggested that slight repolymerization might happen during the steam explosion. After calibration by monodisperse polystyrene, the weight-average (M w )
Fig. 7. Aliphatic (a and b) and aromatic regions (c and d) of
13
and number-average (M n ) molecular weight of lignin were 8435 and 3526 g/mol and then decreased to 5673 and 2426 g/mol respectively after steam explosion, which quantitatively verified the depolymerization during the treatment process. HSQC NMR technique is a powerful tool for a detailed understanding of lignin structure. The HSQC spectra of MSL isolated form untreated and steam-exploded corn stalk are compared in Fig. 7 and the main substructures are depicted in Fig. 8 according to Refs. [24,27,28]. From the aliphatic region of the 2D spectra presented in Fig. 7a and b, the dominant side-chain linkages in MSL were the arylglycerol b-aryl ether (b-O-40 , A), the phenyl coumaran (b-50 , B), the pinoresinol (b-b0 , C) and the dibenzodioxocin (D) structures. Although the main linkage of MSL after steam explosion was still bO-40 structure, its intensity showed a substantial decrease due to the bond breakage induced by steam explosion. In addition, the decrease of the other three linkages was observed in the MSL after explosion, which was also caused by the substituent removal of Ca in the side chain. The main cross signals in the aromatic region (Fig. 7c and d) corresponded to the substituted phenyl rings of lignin units. Syringyl (S), guaiacyl (G), p-hydroxyphenyl (H) units and phenolic acid (P) were present in the 2D spectra of samples before and after steam explosion treatment. Fig. 7c and d demonstrated that the pretreated MSL had a slightly decreased signal intensity of G6 due to the condensation reaction during explosion. The condensation could also be confirmed by the observed signal at dC/dH 130.0/7.05 ppm after explosion, which was attributed to the C6H6 of a Ca–C5 condensed H unit. Although the 2D NMR clearly demonstrates the substantial reduction of b-O-40 structure by steam explosion treatment, the quantitative determination is poor from the 2D NMR results. Thus, 13 C spectra were carried out for the quantitative characterization of
C/1H HSOC NMR spectra of MSL before (a and c) and after steam explosion (b and d).
99
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
Fig. 8. Main structures present in MSL.
in Fig. 7a). The MSL before steam explosion had 0.247 b-O-40 linkage per aromatic ring and underwent a significant decrease by 36.44% after steam explosion. Similarly, as shown in Table 2 the b-50 (B) and b-b0 (C) linkages reduced by 75.41% and 75.00% after explosion treatment, respectively. The noteworthy reduction of b-50 (B) and b-b0 (C) structures could be explained by the fragile ether bond of Ca in the acid medium during steam explosion. Since the MSL extracted by neutral solvent is equal or close to the lignin existing in corn stalk due to the less structural damage during extraction process, the data in Table 2 are good enough to represent the actual changes of lignin structure. Consequently, the comparison of 13C NMR spectra indicated that considerable depolymerization of lignin occurred after steam explosion due to the substantial decrease of crosslinking in the polymer.
MWL before steam explosion
MWL after steam explosion
3.4. Kinetics model of lignin alkaline extraction from steam exploded corn stalk 90
85
80
75
70
65
60
55
50
45
ppm Fig. 9. Aliphatic regions of exploded corn stalk.
13
C NMR spectra of MSL from untreated and steam-
lignin changes and the spectra data in the aliphatic region are presented in Fig. 9. The peak assignments based on literature values [29,30] are summarized in Table 2. The integral of 162–102 ppm region was set as the reference including six aromatic carbons and 0.12 vinylic carbons [29]. The quantification of b-O-40 linkage of MSL was calculated based on the Ca resonance peak due to the less overlap with other peaks compared with Cb and Cc (shown
Since lignin undergoes crosslinking breakage during steam explosion, the subsequent alkaline extraction shows totally different kinetics against the directly alkaline extraction in which depolymerization of lignin is the dominant step [22]. Therefore, depolymerized lignin from steam-exploded corn stalk directly dissolves in the alkali solutions and the process can be described by the second-order rate law which is proposed based on the extraction of solvent-soluble compounds from ground plant tissues [31,32]. The second-order kinetic model can be written as:
dL=dt ¼ k2 ðLe LÞ2
where k2 is the second-order extraction rate constant (L(g min)1) and the meanings of the other letters are consistent with that in 3.2.
Table 2 Estimation of interunit linkages in MSL isolated from corn stalk before and after steam explosion via quantitative
a
Assignment
Chemical shift range (ppm)
Aromatic and vinylic carbons Ca in b-5 (B) Cb in b-O-4 (A), Ca and Cb in Dibenzodioxocin (D) Ca in b-O-4 (A) Cc in b-5 (B) and Cc in b-O-4 (A) with Ca = O Cc in b-O-4 (A) without Ca = O Methoxyl group Cb in b-b (C)
162–102 90–86 85–83 71–74 62–64 60–62 57–54 52–49
ð5Þ
13
C NMR.
Number of moieties per aromatic ring MSL before steam explosion
MSL after steam explosion
6.12 0.061 0.200 0.247 0.281 0.121 1.172 0.036
6.12 0.015 0.093 0.157 0.172 0.087 1.199 0.009
Changes of moieties number per aromatic ring of MSL after treatment, in forms of percentage. Negative sign means reduction.
Variation (%)a
0 75.41 53.50 36.44 38.79 28.10 2.30 75.00
100
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101
phenomena during the extraction of lignin from steam exploded corn stalk. Initially, there is intense dissolution in which the maximum extraction takes place. Then a second, much-slower stage occurs, which primarily corresponds to external diffusion and the dissolution of remaining lignin depolymerized in the alkaline extraction process [31]. The equilibrium concentration of lignin, Le, the extraction rate constant, k2, and the coefficient of determination, R2, are given at various temperatures in Table 3. The increase of k2 with increasing temperature also could be described by the Arrhenius equation (Eq. (4)). The good linear relationship of lnk2 versus inverse of absolute temperature (1/T) (shown in Fig. 11) demonstrated that the second-order model could satisfactorily simulate the extraction kinetics of lignin from steam-exploded corn stalk. The activation energy (Ea) of lignin extraction from steam-exploded corn stalk was estimated from the slope of the plot of ln k2 versus 1/T as 12.11 kJ mol1, which decreased by 37.78% compared with that of lignin extraction from corn stalk without steam explosion.
7 6
-1
t/L (minLg )
5 4 3 2 1 0 0
20
40
60
80
100
120
Time (min) Fig. 10. Second-order extraction kinetics of lignin from steam-exploded corn stalk at different temperature (j: 40 °C; d: 60 °C; ▲: 80 °C; .: 100 °C; w: 120 °C).
Table 3 Constants of second-order kinetic model for alkaline extraction of lignin from steamexploded corn stalk. T (K)
Le (g L1)
k2 (L g1 min1)
R2
313 333 353 373 393
18.018 19.305 20.450 21.368 22.727
0.0555 0.0518 0.0489 0.0468 0.0440
0.998 0.999 0.999 0.999 0.999
-2.4
4. Conclusions This work indicates that the considerable depolymerization rather than the increased specific surface area during steam explosion causes the poor representation of lignin alkali-extraction kinetics by a first-order model. Thus, taking into account of the direct dissolution of depolymerized lignin, the alkaline extraction process of lignin from steam-exploded corn stalk is then successfully simulated by a second-order model. The activation energy of lignin alkali-extraction decreases 37.78% after the steam explosion treatment of 1.8 MPa pressure and 5 min treatment time. Consequently, the study demonstrates that milder extraction conditions are available to extract lignin from steam exploded corn stalk and further suggests that some single or mixed solvents with higher solvency for lignin fragments may be feasible to isolate lignin after steam explosion by relatively facile extraction conditions. Acknowledgements
lnk2
-2.8
Financial support for this study was provided by the National Basic Research Program of China (973 Project, No. 2011CB707401), the National High Technology Research and Development Program of China (863 Program, 2012AA021302).
-3.2
-3.6
Y=0.99618-1456.5X 2 R =0.926
-4.0 0.0024
0.0026
References
0.0028
0.0030
0.0032
1/T (1/K) Fig. 11. Arrhenius plots ln k2 versus 1/T of alkaline extraction kinetics of lignin from steam exploded corn stalk.
After integration, the second-order extraction kinetics can be present as a linearized Eq. (6):
t=L ¼ 1=k2 L2e þ t=Le
ð6Þ
It is found from Eq. (6) that in a second-order extraction model, the theoretic t/L shows a linear relationship with extraction time and a series of t/L plots against time are given in Fig. 10. The agreement of the second-order extraction model with the experimental results confirms our assumption that there are primarily two
[1] A.J. Ragauskas, C.K. Williams, B.H. Davison, G. Britovsek, J. Cairney, C.A. Eckert, W.J. Frederick, J.P. Hallett, D.J. Leak, C.L. Liotta, The path forward for biofuels and biomaterials, Science 311 (2006) 484–489. [2] W. Sui, H. Chen, Multi-stage energy analysis of steam explosion process, Chem. Eng. Sci. 116 (2014) 254–262. [3] M.E. Himmel, S.Y. Ding, D.K. Johnson, W.S. Adney, M.R. Nimlos, J.W. Brady, T.D. Foust, Biomass recalcitrance: engineering plants and enzymes for biofuels production, Science 315 (2007) 804–807. [4] C. Mooney, S. Mansfield, M. Touhy, J. Saddler, The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods, Bioresour. Technol. 64 (1998) 113–119. [5] A. Berlin, M. Balakshin, N. Gilkes, J. Kadla, V. Maximenko, S. Kubo, J. Saddler, Inhibition of cellulase, xylanase and b-glucosidase activities by softwood lignin preparations, J. Biotechnol. 125 (2006) 198–209. [6] J. Bozell, J. Holladay, D. Johnson, J. White, Top Value Added Chemicals from Biomass-Volume II: Results of Screening for Potential Candidates from Biorefinery Lignin,
, 2007. [7] P.C.R. Pinto, C.E. Costa, A.E. Rodrigues, Oxidation of lignin from eucalyptus globulus pulping liquors to produce syringaldehyde and vanillin, Ind. Eng. Chem. Res. 52 (2013) 4421–4428. [8] A.J. Ragauskas, G.T. Beckham, M.J. Biddy, R. Chandra, F. Chen, M.F. Davis, B.H. Davison, R.A. Dixon, P. Gilna, M. Keller, P. Langan, A.K. Naskar, J.N. Saddler, T.J. Tschaplinski, G.A. Tuskan, C.E. Wyman, Lignin valorization: improving lignin processing in the biorefinery, Science 344 (2014) 1246843. [9] D.R. Dimmel, L.F. Schuller, Structural/reactivity studies (I): soda reactions of lignin model compounds, J. Wood Chem. Technol. 6 (1986) 535–564.
G. Wang, H. Chen / Separation and Purification Technology 157 (2016) 93–101 [10] F.S. Chakar, A.J. Ragauskas, Review of current and future softwood kraft lignin process chemistry, Ind. Crop. Prod. 20 (2004) 131–141. [11] G. Chen, H. Chen, Extraction and deglycosylation of flavonoids from sumac fruits using steam explosion, Food Chem. 126 (2011) 1934–1938. [12] A. Kurosumi, C. Sasaki, K. Kumada, F. Kobayashi, G. Mtui, Y. Nakamura, Novel extraction method of antioxidant compounds from Sasa palmata (Bean) Nakai using steam explosion, Process Biochem. 42 (2007) 1449–1453. [13] W. Sui, H. Chen, Extraction enhancing mechanism of steam exploded Radix Astragali, Process Biochem. 49 (2014) 2181–2190. [14] J. Li, G. Henriksson, G. Gellerstedt, Lignin depolymerization/repolymerization and its critical role for delignification of aspen wood by steam explosion, Bioresour. Technol. 98 (2007) 3061–3068. [15] S. Shao, G. Wen, Z. Jin, Changes in chemical characteristics of bamboo (Phyllostachys pubescens) components during steam explosion, Wood Sci. Technol. 42 (2008) 439–451. [16] G. Wang, H. Chen, Carbohydrate elimination of alkaline-extracted lignin liquor by steam explosion and its methylolation for substitution of phenolic adhesive, Ind. Crop. Prod. 53 (2014) 93–101. [17] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D. Templeton, D. Crocker, Determination of Structural Carbohydrates and Lignin in Biomass, Laboratory Analytical Procedure, 2008. [18] G. Wang, H. Chen, Fractionation of alkali-extracted lignin from steamexploded stalk by gradient acid precipitation, Sep. Purif. Technol. 105 (2013) 98–105. [19] A. Björkman, Studies on finely divided wood. Part 1. Extraction of lignin with neutral solvents, Svensk Papperstidn. 59 (1956) 477–485. [20] G. Wang, H. Chen, Fractionation and characterization of lignin from steamexploded corn stalk by sequential dissolution in ethanol–water solvent, Sep. Purif. Technol. 120 (2013) 402–409. [21] J.P. Simão, M. Carvalho, C.M. Baptista, Heterogeneous studies in pulping of wood: modelling mass transfer of dissolved lignin, Chem. Eng. J. 170 (2011) 264–269.
101
[22] F. Correia, D. Roy, K. Goel, Chemistry and delignification kinetics of Canadian industrial hemp, J. Wood Chem. Technol. 21 (2001) 97–111. [23] B. Wong, G. Deka, D. Roy, Bulk delignification kinetics at selected vertical heights within jack pine plus trees, Wood Sci. Technol. 29 (1995) 11–18. [24] R. Samuel, Y. Pu, B. Raman, A.J. Ragauskas, Structural characterization and comparison of switchgrass ball-milled lignin before and after dilute acid pretreatment, Appl. Biochem. Biotechnol. 162 (2010) 62–74. [25] K.B. Ansari, V.G. Gaikar, Green hydrotropic extraction technology for delignification of sugarcane bagasse by using alkybenzene sulfonates as hydrotropes, Chem. Eng. Sci. 115 (2014) 157–166. [26] M. Cissé, P. Bohuon, F. Sambe, C. Kane, M. Sakho, M. Dornier, Aqueous extraction of anthocyanins from Hibiscus sabdariffa: experimental kinetics and modeling, J. Food Eng. 109 (2012) 16–21. [27] T. Liitiä, S. Maunu, B. Hortling, M. Toikka, I. Kilpeläinen, Analysis of technical lignins by two-and three-dimensional NMR spectroscopy, J. Agric. Food Chem. 51 (2003) 2136–2143. [28] J. Rencoret, G. Marques, A. Gutiérrez, L. Nieto, J. Jiménez-Barbero, Á.T. Martínez, J.C. del Río, Isolation and structural characterization of the milledwood lignin from Paulownia fortunei wood, Ind. Crop. Prod. 30 (2009) 137–143. [29] E.A. Capanema, M.Y. Balakshin, J.F. Kadla, Quantitative characterization of a hardwood milled wood lignin by nuclear magnetic resonance spectroscopy, J. Agric. Food Chem. 53 (2005) 9639–9649. [30] K.M. Holtman, H.M. Chang, H. Jameel, J.F. Kadla, Quantitative 13C NMR characterization of milled wood lignins isolated by different milling techniques, J. Wood Chem. Technol. 26 (2006) 21–34. [31] Y. Ho, H.A. Harouna-Oumarou, H. Fauduet, C. Porte, Kinetics and model building of leaching of water-soluble compounds of Tilia sapwood, Sep. Purif. Technol. 45 (2005) 169–173. [32] M.G. Bogdanov, I. Svinyarov, Ionic liquid-supported solid–liquid extraction of bioactive alkaloids. II. Kinetics, modeling and mechanism of glaucine extraction from Glaucium flavum Cr. (Papaveraceae), Sep. Purif. Technol. 103 (2013) 279–288.