- Email: [email protected]
Structural characterization of the solid residue produced by hydrothermal treatment of sunflower stalks and subsequent enzymatic hydrolysis
Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Structural characterization of the solid residue produced by hydrothermal treatment of sunflower stalks and subsequent enzymatic hydrolysis In-Yong Eom, Ju-Hyun Yu * Research Center for Biobased Chemistry, Korea Research Institute of Chemical Technology, Daejeon 305-600, Republic of Korea
A R T I C L E I N F O
Article history: Received 24 June 2014 Received in revised form 25 July 2014 Accepted 25 July 2014 Available online 2 August 2014 Keywords: Sunflower stalks Hydrothermal treatment Solid residue Lignin-enriched fraction Lignin valorization
A B S T R A C T
This study was to structurally characterize solid residues obtained from sunflower stalks hydrothermally treated at 180 and 200 8C for 30 min, followed by enzymatic hydrolysis. Recovered solid residue were 25.3% and 24.1% of fresh biomass, respectively. Each ethanol soluble fraction could be obtained up to 30% of the solid residue. The fraction from the solid residue at 200 8C was composed of smaller-sized lignin macromolecules with higher phenolic hydroxyl group, but lower aliphatic hydroxyl due to enhanced cleavage reactions related to side chain and aryl ether linkages of lignin, compared to that of the solid residue at 180 8C. ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Recently, due to rising concern over the depletion of fossil fuels and accelerating global warming, the demand for lignocellulosic biomass has increased dramatically worldwide. Lignocellulosic biomass has been considered a sustainable and renewable primary energy resource that can be converted to alternative transportation fuels, including bioethanol or biodiesel, bio-chemicals, and bio-polymers [1]. Among the promising lignocellulosic biomass, sunflower is being spotlighted as a huge energy source. Sunflower is commonly cultivated due to its high oil content in the seeds, representing up to 80% of its economic value, and its oilseed yield accounts for 8% of the total global production of oilseeds (404 million tons in 2008/2009), corresponding to the fourth largest production in the world [2]. In addition, sunflower can be cultivated throughout the world due to their relatively short cultivation period with a high yield up to 3 ton/ha. Top sunflower producing countries are in the order of Russia (5.7 billion tons per year), Ukraine (4.2 billion tons per year), and Argentina (3.5 billion tons per year) [2]. Even though it is also useful to produce pectins and essential oil from sunflower heads, sunflower stalks have been studied as an attractive lignocellulosic biomass [3].
* Corresponding author. Tel.: +82 42 860 7438; fax: +82 42 861 4913. E-mail address: [email protected] (J.-H. Yu).
In contrast to 1st generation feedstocks, such as starch or sugarbased crops, lignocellulosic biomass, a promising 2nd generation renewable material, is composed of three major biopolymers – cellulose, hemicelluloses and lignin, the amount of each which varies depending on the lignocellulosic biomass species. Generally, cellulose accounts for around 35–50%, hemicellulose around 15– 35%, and lignin around 15–25% based on the dry weight biomass, and these components construct the plant cell wall structure in which lignin and polysaccharides are linked together. Therefore, to use cellulose and hemicelluloses as carbohydrate sources, pretreatment is an essential processing step in biochemical conversion of the lignocellulosic biomass to biofuels and biochemicals. The purpose of pretreatment is to break down the plant cell wall structure, due to the recalcitrance nature of lignocellulosic biomass, and then make cellulose more accessible to hydrolytic enzymes [4]. Among the various existing pretreatment methods, hydrothermal treatment, also called liquid hot water or autohydrolysis, has been evaluated as being a suitable process for producing fermentable sugars. Hydrothermal treatment is favorable for lower capital and operating investment, due to its use of only water as a reaction medium, lack of any requirement for acid or alkaline catalyst, and lower inhibitor generation such as furfural, 5hydroxymethyl-2-furaldehyde (HMF) and phenolics that may adversely affect microbial fermentation [1]. To date, various studies have evaluated hydrothermal treatment as an effective
http://dx.doi.org/10.1016/j.jiec.2014.07.044 1226-086X/ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
pretreatment for converting cellulose from various herbaceous biomasses such as switchgrass, sunflower stalks and wheat straw to fermentable sugars for subsequent fermentation [3,5–8]. These studies reported that complicated conditioning for subsequent enzymatic hydrolysis was not needed, due to the absence of any mineral acid and alkali catalyst. Moreover, these biomasses can be effectively converted to fermentable sugars with low yields of inhibitory products interfering with downstream operations, such as microbial fermentation. During fermentable sugar production, non-hydrolyzable solid residue that is mainly composed of lignin can be recovered but, unlike fermentable sugars for subsequent microbial fermentation, it is utilized as a low-grade boiler fuel. Lignin is a polyphenolic polymer derived from three phenylpropanoid units as precursors for lignin biosynthesis: p-coumaryl, coniferyl and sinapyl alcohol. The relative portions of each monolignol in lignocellulosic biomass differ depending on the biomass species. The lignin macromolecule is linked by carbon-carbon and carbon–oxygen bonds via radical coupling polymerization of three monolignols. The aryl ether bond (b-O-4) is major interunit linkage, and other linkages, such as biphenyl, b-5, and a-O-4 can also be detected [9]. Hydrothermal treatment has less affect on delignification, thus most of the lignin in the starting material remains in the pretreated biomass. Lignin has been regarded as the most recalcitrant component to enzymatic and microbial attack. Therefore, with the steady acceleration in industrial biofuel and biochemical production from lignocellulosic biomass, large quantities of solid residue are also generated as a byproduct [10]. Improved cost-effectiveness and optimized biomass usage will necessitate effective valorization of solid residue. After enzymatic hydrolysis of the pretreated biomass, the remaining solid residue, with lignin as its main component, is an attractive aromatic resource for producing aromatic chemicals and materials. To date, a number of researchers have attempted to develop novel valorization of lignin, including fractionation methods of lignin, and synthesis of lignin-based polymers [11– 13]. However, significant challenges remain in the application of lignin due to its structural complexity. Moreover, the physicochemical properties of lignin differ depending on the biomass and recovery processes. Therefore, it is important to evaluate the potential use of solid residue after hydrothermal treatment, and subsequent enzymatic hydrolysis. The majority of the lignin research, to date, has focused on organosolv lignin recovered from organosolv-pretreated supernatants, and organosolv lignin has been comprehensively characterized and evaluated for further use [6,10,14–17]. On the other hand, little research has investigated the valorization of solid residue as a by-product obtained from hydrothermally treated lignocellulosic biomass, and subsequent enzymatic hydrolysis for producing fermentable sugars. During hydrothermal treatment without adding any chemicals, lignin is rarely thermally decomposed and hydrolyzed, whereas it retains its native lignin properties [18]. As the lignin in the solid residue could be also as heterogeneous and insoluble as natural lignin, it might not be suitable for chemical purpose. If the solid residue is fractionated into more suitable materials for value added applications, this solid residue could potentially serve as an attractive source for phenolicbased chemical industries [19,20]. The objectives of this study were to structurally characterize the solid residue recovered from the hydrothermally treated sunflower stalks at 180 and 200 8C, followed by enzymatic hydrolysis. In addition, lignin-enriched fractions were collected and characterized by compositional analysis, elemental analysis, thermogravimetric analysis (TGA), gel permeation chromatography (GPC), and 31P nuclear magnetic resonance (NMR) spectroscopy in order to elucidate the structural information on a starting material for bio-based polymers and fine/bulk chemicals.
73
Materials and methods Raw material Sunflower stalks (Helianthus annuus L.) were supplied by Chungbuk Agricultural Research and Extension Services (Ochang, Chungcheongbuk-Do, Korea) and air-dried in a greenhouse. The stalks were ground, using a knife mill equipped with a 20-mesh aperture screen, and the ground stalks containing 2.3% moisture were stored at 20 8C in a refrigerator in sealed plastic bags until use. Cellic1 CTec2 as a cellulase, Cellic1 HTec2 as a xylanase, and Novozyme 188 as a beta-glucosidase were purchased from Novozymes Korea (Seoul, Korea). All organic solvents at the highest purity grades were purchased from Sigma–Aldrich. Hydrothermal treatment followed by enzymatic hydrolysis Hydrothermal treatment was carried out using a 2 L-scale Parr reactor (Parr Instruments, Moline, IL). The pretreatment method has been specifically described in a previous study [5]. The powdered sunflower stalks equivalent to 120 g dry weight were added to reactor vessels, and then filled with de-ionized water, to a total weight of 1500 g. The reactor vessels were sealed, kept at room temperature through the night, and pretreated at 180 and 200 8C for 30 min, respectively. All the pretreated biomasses were transferred to 7 L-scaled batch-type fermenters (BioTron, Seoul, Korea), and a mixture of Cellic1 CTec2, HTec2, and b-glucosidase (volumetric ratio 18:2:1) was loaded at a volume equivalent to 0.1 ml (ca. 8.7 FPU) per 1 g dry biomass. Enzymatic hydrolysis was carried out at 50 1 8C with a rotating speed of 200 rpm for 72 h. The hydrolyzates were separated into solid and liquid fractions by centrifugation at 3727 g for an hour, with a swing-type centrifuge (Model Combi-514, Hanil Scientific Co., Seoul, Korea). The solid residue was washed out with de-ionized water, followed by centrifugation. This washing/centrifugation cycle was repeated until the initial supernatants of the hydrolyzates remained less than 1%. The residue was then dried at 45 8C until the moisture content dropped below 10%. The solid residues obtained from sunflower stalks hydrothermally treated at 180 and 200 8C, followed by enzymatic hydrolysis, are herein referred to as 180 8C-solid and 200 8C-solid, respectively. Sequential extraction of solid residue To prepare homogeneous lignin-enriched fractions, the solid residues were sequentially extracted with water and organic solvents using a Soxhlet apparatus. Around 5 g of the solid residues were extracted with 150 ml de-ionized water, and then oven-dried at 45 8C, until the moisture content dropped below 10%. The water extractives (WE)-free solid residues were sequentially extracted with 150 ml of organic solvents, including ethanol, methanol, acetone, and isopropyl alcohol. The dissolved fractions in water and organic solvents were evaporated and freeze-dried. Compositional and elemental analysis The compositional analysis of the solid residues and extractives with water or organic solvents was performed according to National Renewable Energy Laboratory (NREL) standard procedures [21,22]. The method is briefly described as follows. The carbohydrates and lignin in the sample were measured by the twostep acid hydrolysis. First, concentrated acid hydrolysis was conducted with 72% (w/w) H2SO4 at 30 8C for 60 min. In the second step, the above hydrolyzates were diluted to 4% (w/w) H2SO4 with de-ionized water, and then diluted acid hydrolysis was performed at 121 8C for 60 min in an autoclave. The hydrolyzates
74
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
were filtered with crucibles to obtain the supernatants and solid residues. The monosaccharides in the supernatants were determined using high performance liquid chromatography (HPLC) equipped with an autosampler (Waters 2707), a refractive index detector (Waters 2414), and a binary HPLC pump (Waters 1525). An Aminex HPX-87H column (BIO-Rad, Hercules, CA) operating at 50 8C was used, and the mobile phase was 5 mM H2SO4, at a flow rate of 0.6 ml/min. The acid-soluble lignin (ASL) was also determined by DU 800 UV/Vis spectrophotometer (Beckman Coulter Inc., USA) at 320 nm. The acid-insoluble lignin (AIL) on the crucibles was gravimetrically determined, based on the remaining insoluble residue, by subtracting the ash content from them. Elemental analysis (C, H and N) was performed, using a Thermo Scientific Flash 2000 (Thermo Fisher Scientific Inc., USA). The oxygen content was calculated by subtracting the sum of the weight of C, H, and N from the total. All compositional and elemental analyses were run in triplicate, and reported as average results. Thermogravimetric analysis (TGA) Thermogravimetric (TG) and differential thermogravimetric (DTG) analyses of 180- and 200 8C-solids and organic solvent extractives were performed using the Q-500 IR Instrument (TA Instruments, USA). TG analysis of the samples (2.0 0.5 mg) was carried out at a constant heating rate of 10 8C/min, at temperatures ranging from 30 to 700 8C. To maintain an inert atmosphere during pyrolysis, a purified nitrogen as a carrier gas was flowed at a rate of 100 ml/min. Gel permeation chromatography (GPC) The number-average molecular weight (Mn) and weightaverage molecular weight (Mw) of the organic soluble fractions were determined by GPC, after their acetylation. Acetylation of organic soluble fractions was performed according to Hage et al. [14], and the method is briefly described as follows. The organic solvent-soluble fraction (20–25 mg) was dissolved in an acetic anhydride/pyridine mixture of 2.00 ml (1:1, v/v), stirred for 24 h at room temperature and then added to ethanol (25 ml). The addition and evaporation of ethanol was repeated, until the acetic acid and pyridine were completely evaporated. The acetylated fraction was dissolved in chloroform (2 ml), added drop-wise to diethyl ether (100 ml), and then centrifuged. The precipitate was repeatedly washed with diethyl ether, followed by centrifugation, and then freeze-dried. GPC analysis for lignin was performed, using a NP4000 GPC system (Futecs Co., Ltd) equipped with an autosampler, refractive index detector (Shodex RI-101), and LF-404 column (Shodex Inc.). The acetylated sample was dissolved in tetrahydrofuran (THF) (1 mg/ml), and the solution was filtered through a 0.45 mm filter. GPC was operated with the eluent of THF, at a flow rate of 0.5 ml/min, injection volume of 120 mm, and retention time of 30 min. Standard polystyrene samples were used to determine the calibration curves. Data were analyzed with the Clarity GPC Data System. 31
P NMR spectroscopy
To investigate the phenolic hydroxyl group present in lignin, organic solvent-soluble fractions were phosphitylated according to Pu et al. [23]. To prepare phosphitylated lignin, 2-chloro-4,4,5,5,tetramethyl-1,3,2-dioxaphospholane (TMDP) was used as the phosphitylating reagent. Cyclohexanol was used as an internal standard. 31P NMR experiments were performed on a Bruker AVANCE 500 MHz. Quantitative 31P NMR spectra were acquired, using an inverse-gated decoupling pulse sequence with a 30 pulse
angle and 25 s pulse delay. Chemical shifts were calibrated, based on the sharp and stable signal for phosphitylated water at 132.2 ppm. 31P NMR spectra were quantified, using the SpinWorks 2.5 software. Results and discussion Characterization of the 180- and 200 8C-solids The composition of sunflower stalks, which had been determined in a previous study, contained 35.8 0.5 g of cellulose as glucose, 19.7 0.1 g of hemicellulosic sugars, including the sum of xylose, galactose and mannose (XGM, 19.3 0.1 g), and arabinose (0.4 0.1 g), and 5.1 0.1 g of acetyl group on g per 100 g dry biomass [5]. As other components, AIL and ASL (14.8 0.0 g and 0.1 0.0 g, respectively), water extractives (WE) and ethanol extractives (EE) (20.5 0.2 g and 1.1 0.1 g, respectively), and ash (6.3 0.0 g) were determined [5]. The elemental composition of the sunflower stalks was as follows: C (43.0%), H (5.4%), O (50.8%), N (0.8%) on a dry weight basis before the extraction with water and ethanol, and C (44.8%), H (5.8%), O (49.0%), N (0.4%) after the extraction. The chemical composition of 180- and 200 8C-solids is presented in Table 1. Their elemental compositions before and after sequential extraction with water and ethanol, are also reported. The pretreatment temperatures that applied for sunflower stalks to recover 180- and 200 8C-solids were determined according to a previous study [5], which showed that the pretreatment temperature at 180 8C offers the advantages of maximizing the yield of hemicellulosic sugars as well as producing less carbohydrate decomposition products that act as inhibitors of microbial fermentation. However, the enzymatic digestibility of cellulose in the biomass treated at 180 8C reached only 67% of its theoretical yield. In contrast, sunflower stalks hydrothermally treated at 200 8C can undergo conversion of almost all of the cellulose into glucose through enzymatic hydrolysis, which yielded 76.8% of the theoretical yield. when the pretreated biomass at 200 8C was washed with water, the enzymatic digestibility of cellulose was increased to 88.8% [5]. The solid residue (25.3 g) based on 100 g dry sunflower stalks could be recovered after hydrothermal treatment at 180 8C, and subsequent enzymatic hydrolysis, while it was decreased to 24.1 g after hydrothermal treatment at 200 8C. Both recovered solids were mainly composed of AIL and extractable components with water and ethanol (WE and EE), but contained a small portion of cellulose and hemicelluloses. The amount of glucose liberated from the 180 8C-solid was 13.2 g/100 g dry 180 8C-solid, which was equivalent to 9.3% of total cellulose as glucose in the sunflower stalks; whereas that for the 200 8C-solid was 7.1 g/100 g dry 200 8C-solid, accounting for 4.8% of cellulose as glucose. These results indicated that the solid recovery of the 200 8C-solid was decreased than that of the 180 8C-solid, due to increasing glucose liberation from the biomass. During hydrothermal treatment in which water is used as the only reaction medium, almost all of the lignin in the starting material remained in the pretreated biomass, due to its low solubility in water. However, interesting results were observed. The 180 8C-solid consists of 47.5 g of AIL based on the 100 g dry 180 8C-solid, accounting for 80.7% of lignin in the sunflower stalks, while 38.5 g of AIL was determined from the 200 8C-solid, which accounts for 62.3% of lignin in the biomass. This reveals that even though the lignin in the biomass barely becomes soluble in water during hydrothermal treatment at relatively low temperature, it undergoes significant changes in the point of physicochemical properties. Prior to ethanol extraction, the 180- and 200 8C-solids were extracted with water using Soxhlet apparatus to remove the remaining free-sugars and proteins. Significant portion of
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
75
Table 1 Composition of solid residues hydrothermally treated at 180 and 200 8C for 30 min and subsequent enzymatic hydrolysis. 200 8C-Solid
180 8C-Solid Yielda
25.3
Component
Before extractionb
After extractionc
Before extraction
After extraction
Cellulose as glucose XGMd ASLe AILf WEg EEh Ash
13.2 0.4 1.1 0.0 0.3 0.0 47.5 0.4 7.5 0.6 19.4 0.1 2.1 0.0
18.6 0.5 1.6 0.0 0.4 0.0 66.9 0.5
7.1 0.1 0.9 0.0 0.1 0.0 38.5 0.2 13.6 0.1 29.0 0.3 4.1 0.1
13.3 0.1 1.6 0.0 0.2 0.0 72.3 0.3
Element (%) C H O N H/C (mol/mol)
56.2 5.9 36.9 1.0 1.26
54.6 5.7 38.7 1.0 1.25
58.8 5.7 34.1 1.4 1.16
55.6 5.3 37.5 1.6 1.13
a b c d e f g h
24.1
On the basis of 100 g dry raw biomass. On the basis of 100 g dry solid residue. On the basis of 100 g dry solid residue extracted with water and sequential ethanol. Sum of xylose, galactose, and mannose as hemicellulosic sugars. Acid-soluble lignin. Acid-insoluble lignin. Water extractives. Ethanol extractives.
the 180- and 200 8C-solids could be extracted with ethanol, resulting in the EE being the second largest component, after AIL. The EE content of the 180- and 200 8C-solids was 19.4 and 29.0 g based on the 100 g dry solid, respectively, which could be related with the extent of further degradation of lignin by pretreatment temperature rise. A larger portion of the 200 8C-solid could be extracted with ethanol than that of the 180 8C-solid, and, in turn, the AIL content in the 200 8C-solid was lower than that of the 180 8C-solid, as aforementioned, which indicated that lignin macromolecules in the 200 8C-solid could be further degraded into small fragments to be soluble in organic solvents. Therefore, these EEs, which are believed to be composed mainly of lignin from the 180- and 200 8C-solids, were characterized further and are discussed in the following section. As shown in the elemental composition in Table 1, both the carbon and hydrogen contents increased in the 180- and 200 8Csoilds compared to that of the raw biomass, which could be attributed to not only the increased lignin content in the solid, but also the decreased cellulose content, due to its conversion to glucose [24]. The nitrogen content in the 200 8C-solid was higher than that of the 180 8C-solid. This high nitrogen content in the 180- and 200 8Csolids was thought to be affected by irreversible adsorption of cellulolytic enzyme on the solid surface during enzymatic hydrolysis. Despite adding the same amount of enzyme for enzymatic hydrolysis, the enzyme could be irreversibly bound to the 200 8Csolid more than the 180 8C-solid due to the structural change of lignin during the hydrothermal treatment at higher temperature. The average molecular weight of the 180- and 200 8C-solids could not be determined, because their acetylated samples were not dissolved in THF that was used as an organic solvent for GPC analysis. The insoluble fraction of the solids was due to its highly condensed structure with a high molecular weight and heterogeneous composition such as cellulose and proteins besides lignin.
residues that were thought to have a more condensed structure due to their being more resistant to thermal decomposition than the extractable fractions. TG and DTG curves for the extractivesfree 180- and 200 8C-solids are plotted in Fig. 1. In each TG curve for the 180- and 200 8C-solids, both solid residues started to decompose at around 200 8C, and then thermal decomposition was significantly accelerated at between around 280 and 390 8C. After the active decomposition, the 180- and 200 8C-solids continued to decompose slowly. In one distinct difference in the DTG curves, the decomposition rate (8C/min) of the 180 8C-solid was higher than that of the 200 8C-solid in the active decomposition region ranging from 280 to 390 8C. This result indicated that the main pyrolysis reactions, such as depolymerization, fragmentation, and decarboxylation, are more favorable during thermal decomposition of the 180 8C-solid, whereas the thermal stability of the 200 8C-solid was higher than that of the 180 8C-solid. In addition, a remarkable difference between the two TG curves was observed in the slow decomposition region above 390 8C. There was a significant difference in char amount, referred to as the remaining residue after TG analysis. It was reported that the char forming reaction of lignin began at around 600 8C, with the
Thermal decomposition behaviors of the 180- and 200 8C-solids To investigate the thermal decomposition behaviors of the 180and 200 8C-solids, TGA was carried out at a constant heating rate of 10 8C/min. Prior to TGA of the solid residue, the 180- and 200 8Csolids were sequentially extracted with water and ethanol to evaluate the thermal stability of the non-extractable solid
Fig. 1. TG and DTG curves of the extractives-free 180- and 200 8C-solids in N2 at a heating rate of 10 8C/min.
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
76
Table 2 Compositions of water and ethanol extractives (WW and EE). Composition
180 8C-Solid
200 8C-Solid
WE
EE
WE
EE
Yielda Glucoseb XGMb ASLb AILb
7.5 0.6 5.5 0.0 5.7 0.0 6.1 0.1 54.2 0.7
19.4 0.1 0.4 0.0 ND 0.9 0.0 90.1 0.3
13.6 0.1 8.6 0.1 2.0 0.1 8.4 0.1 57.0 0.7
29.0 0.3 ND ND 1.2 0.1 96.2 0.1
Element (%) C H O N H/C (mol/mol)
53.2 0.2 5.5 0.1 39.4 0.3 1.9 0.1 1.24
64.4 0.2 6.6 0.0 28.4 0.2 0.6 0.0 1.23
54.8 0.3 5.5 0.0 38.2 0.3 1.6 0.0 1.20
66.5 0.1 6.4 0.1 26.3 0.0 0.9 0.0 1.15
a b
On the basis of 100 g dry solid. On the basis of 100 g dry extractives.
emission of methane [25]. The study also found that char yield tends to decrease when lignin contains higher amount of functional groups, such as methoxyl, hydroxyl and carboxyl groups. From these results, the large difference in char yields between the 180- and 200 8C-solids could be closely associated with the structural change, depending on the different pretreatment temperatures. It was, therefore, expected that the nonextractable residue in the 200 8C-solid could be made up of a highly condensed lignin structure through condensation reaction between lignin fragments, due to harsh pretreatment condition, compared to that of the 180 8C-solid. It was also suggested that the hydrothermal treatment of sunflower stalks at higher temperatures led to both condensation and fragmentation reactions that resulted in the 200 8C-solid having a highly condensed lignin structure and, as shown above, being also abundant in extractable components that reached 29% of the total 200 8C-solid, which was more than that of the 180 8C-solid. This result was in accordance with the GPC result discussed in more detail below. Characterization of water and ethanol extractives (WE and EE) from the solid residue The compositions of WE and EE are summarized in Table 2. The 180- and 200 8C-solids were composed of extractable fractions with water and ethanol more than 25% of total solid, respectively. In addition, both the WE and EE yields from the 200 8C-solid were approximately 2-fold higher than those of the 180 8C-solid, which was attributed to the higher pretreatment temperature which promoted fragmentation reaction in the lignin. The WE obtained from both the 180- and 200 8C-solids was mainly composed of lignin accounting for more than 60% of total WE. The lignin extracted with water could be related to the lignin droplets that were formed during pretreatment under acidic condition. During pretreatment at high temperatures ranging from 140 to 190 8C, at which lignin can be molten, lignin and further decomposed fragments could migrate, and subsequently coalesce into droplet-like structures rather than be soluble in weakly acidic water, due to its hydrophobicity [26]. After cooling down to room temperature, these lignin droplets may have been adsorbed on the surface of cellulose’s microfibrils, and then recovered by extraction with water and organic solvents. Therefore, a large portion of the 180- and 200 8C-solids that was regarded as lignin could be extracted with ethanol rather than water, due to its solubility being higher in ethanol than water, and the extent of extraction differed depending on pretreatment temperatures. The large amount of lignin in the WEs could be also attributed to pseudo-lignin which carbohydrate moieties can react to form lignin-like fractions during the determination of lignin by 2-step acid hydrolysis
method [27]. Therefore, the lignin content in the WEs may have been overestimated. Carbohydrates of the WEs were mainly non-structural sugars that could be determined without acid-hydrolysis. These sugars were thought to be predominantly derived from adsorbed sugars that could not be recovered while separating the liquid from the enzymatic hydrolyzates. Some of the WE compositions were not able to be characterized, but were estimated to be proteins originating from the cellulolytic enzyme, due to high nitrogen content in the WEs of more than 1.5%. On the contrary, whereas WE consisted of heterogeneous components, such as sugars, proteins and lignin, EE from both the 180- and 200 8C-solids consisted of a homogeneous component in which lignin was the dominant constituent. Moreover, the EE yield from the 200 8C-solid was 1.5-fold higher than that of the 180 8Csolid, due to predominant fragmentation reaction in lignin at higher temperature. During hydrothermal treatment at mild temperatures, the lignin in the biomass was rarely broken down to lower molecular weight fractions, which limits the potential applications of the recovered solid residue for producing industrial chemicals due to its low solubility and low compatibility [28]. To use a large amount of lignin in the solid residue for potential value-added applications, therefore, it is necessary to prepare a homogeneous lignin fraction from solid residue. The EE fractionated from the solid residue could be an alternative for use as a material. Structural features of ethanol extractives (EE) The average molecular weight of ethanol extractives (EE) The average molecular weight (Mw) of the acetylated EEs completely dissolved in THF was determined using GPC. The analytical results are given in Table 3. As aforementioned, a larger portion of the 200 8C-solid could be extracted with ethanol, compared to 180 8C-EE, and the Mw of the resulting 200 8C-EE was lower than that of 180 8C-EE. These results suggest that the fragmentation reaction in lignin predominated over the recondensation between lignin fragments via the fragmentation reaction during hydrothermal treatment at higher temperature, which resulted in increasing cleavage of the aryl ether bond, particularly b-O-4 linkage. Due to the prevailing cleavage of b-O-4 linkage in the 200 8C-solid, more of 200 8C-EE with lower molecular weight could be extracted than the 180 8C-EE, but the polydispersity of the two EEs was similar. Characterization of hydroxyl group in ethanol extractives (EE) by 31P NMR The EEs from the WE-free 180- and 200 8C-solids were considered as lignin-enriched fractions, due to the high lignin content that was determined by the 2-step acid hydrolysis method. To investigate several hydroxyl groups present in lignin, the phosphitylated EEs were analyzed by 31P NMR spectroscopy. The phosphitylation of various hydroxyl groups attached to lignin macromolecules in the EE is depicted in Fig. 2, along with the detectable forms of phosphitylated EE. The results of the 31P NMR are summarized in Table 4. Several hydroxyl groups could be classified as aliphatic hydroxyl groups (aliphatic-OH) or phenolic hydroxyl groups, such as the H-OH: hydroxyl group in the p-hydroxyphenyl unit, G-OH: hydroxyl group in the guaiacyl unit, Table 3 Average molecular weights of the ethanol extractives (EE) obtained from the 180and 200 8C-solids. Mw (g mol EE from 180 8C-solid EE from 200 8C-solid
4736 3540
1
)
Mn (g mol 2317 1696
1
)
Polydispersity 2.04 2.09
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
77
Fig. 2. Phosphitylation of hydroxyl groups in lignin unit monomer with 2-chloro-4,4,5,5,-tetramethyl-1,3,2-dioxaphospholane (TMDP).
Table 4 Quantitative
31
P NMR analysis results of several hydroxyl groups (mmol g
1
) in ethanol extractives (EE) from the 180- and 200 8C-solids.
Assignment/chemical shift d
Component 180 8C-EE 200 8C-EE
Aliphatic-OH
C-Ph–OH
S-OH
G-OH
H-OH
COOH
145.5–150.0 2.96 1.56
145.5–150.0 0.19 0.37
141.8–143.5 1.14 1.47
138.7–140.1 0.58 0.78
137.2–138.1 0.02 0.12
134–135.7 0.28 0.31
C-Ph, condensed phenolic unit; S, syringyl; G, guaiacyl; H, p-hydroxyphenyl; COOH, carboxylic acids.
S-OH: hydroxyl group in the syringyl unit, and C-Ph: hydroxyl group in the condensed phenolic unit or carboxylic hydroxyl groups (COOH). The higher pretreatment temperature led to a 53% decrease in aliphatic hydroxyl, and a 42% increase in phenolic hydroxyl content in 200 8C-EE, compared with those of 180 8C-EE. The difference between the aliphatic and phenolic hydroxyl contents was attributed to side chain and aryl ether bond cleavage, primarily b-O-4 linkages, respectively [29]. In case of organosolv lignin, it was also found in accordance with this study that aliphatic hydroxyl contents in recovered the organoslov lignin were fairly reduced with increasing organosolv pretreatment severity calculated by combined severity from 1.75 to 2.8 due to enhancement of the dehydration reactions on the side chain. Moreover, the formed organosolv lignin during the organosolv at higher severity was composed of extensively cleaved lignin fractions of which molecular weights were decreased in reverse proportion to the severity [15]. According to previous study, the fragmentation of aryl ether linkages catalyzed by acid is the predominant degradation reaction in lignin, after which re-condensation between resulting lignin fragments simultaneously occurs when pretreatment at low pH was carried out [30]. As shown in Table 4, the phenolic hydroxyl content in the 200 8C-solid was higher than that in the 180 8Csolid; therefore, these results suggested that the higher pretreatment temperature favored the promotion of aryl ether bond cleavage. In addition, with increasing pretreatment temperatures, lignin was susceptible to be broken down via side chain and arylether bond cleavage reaction, which affected both its physicochemical properties and the yield of the extractable component that can be recovered from the solid residue. Consequently, the 31P NMR analysis results revealed 200 8C-EE, compared to 180 8C-EE, to be composed of more lignin macromolecules with higher phenolic hydroxyl groups through fragmentation reaction during hydrothermal treatment. This result was in accordance with the lower molecular weight of 200 8C-EE compared to that of 180 8CEE, as confirmed by the GPC result presented in Table 3 above. Fundamental characteristics of diverse extractable lignins from the 200 8C-solid The study results suggest an increased possibility of 200 8C-EE being used as a polymeric material. To maximize the usability of
the solid residue after fermentable sugar production, the extractable lignin needs to be fractionated from the solid residue. Therefore, the lignin fractions were obtained by extracting the 200 8C-solid with organic solvents such as ethanol, methanol, acetone, and isopropyl alcohol from the WE-free solid, and then characterized in order to elucidate the structural properties. As shown in Table 5, the extractable lignin yields reached 29% when extracted with ethanol and methanol. According to previous study, the elemental compositions of milled wood lignin (MWL) and lignin obtained from miscanthus pretreated by organosolv with ethanol (organosolv lignin) were C (63.5%), H (5.7%), O (29.2%), N (0.2%) and C (65.1%), H (5.7%), O (26.6%), N (0.5%) [14]. In other study, moreover, the elemental compositions of the organosolv lignin from poplar wood xylem was determined as C (65.4%), H (5.9%), O (23.4%), and N (0.3%) [24]. Extractable lignins from the 200 8C-solid also showed comparable elemental compositions to the MWL and organosolv lignin isolated from miscanthus and poplar. These previously reported compositions do not differ significantly from the organosolv lignin and extractable lignins of the present study. Even though the above four lignin fractions were extracted with organic solvents that have decreasing polarity from ethanol to isopropyl alcohol, no distinguishable differences in structural
Table 5 Gravimetric yields and fundamental properties of lignin fractions extracted with organic solvents from water extractives (WE)-free 200 8C-solid. Lignin fractiona
EE
ME
AE
IE
Yieldb
29.0 0.3
28.9 0.1
26.3 0.3
24.9 0.8
Element (%) C H O N H/C (mol/mol)
66.5 0.1 6.4 0.1 26.3 0.0 0.9 0.0 1.15
66.3 0.8 6.3 0.1 26.6 0.9 0.9 0.0 1.14
67.7 0.0 6.5 0.1 25.3 0.1 0.6 0.0 1.15
66.8 0.6 6.8 0.1 25.8 0.7 0.7 0.0 1.22
Molecular weight Mw (g mol 1) Mn (g mol 1) Polydispersity
3540 1696 2.09
3720 1797 2.07
3378 1795 1.88
2838 1652 1.72
a Lignin fractions: EE, extracted with ethanol; ME, extracted with methanol; AE, extracted with acetone; IE, extracted with isopropyl alcohol. b On the basis of 100 g dry 200 8C-solid.
78
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
higher pretreatment temperatures, the solid residue was further degraded, and organic solvent solubles could be more extracted. It was found that the obtained fraction was composed of lignins with higher aromatic hydroxyl content but lower aliphatic; however, remaining solid after the extraction showed higher thermal stability due to being more condensed structure. Further studies is needed on developing method for extracting lignin macromolecules with a high yield to be compatible with other commercial polymers and to be used as technical lignin for further application. References
Fig. 3. TG and DTG curves of the lignin fractions extracted with organic solvents from the WE-free 200 8C-solid in N2 at a heating rate of 10 8C/min. EE, ethanol extractable lignin; ME, methanol extractable lignin; AE, acetone extractable lignin; IE, isopropyl alcohol extractable lignin.
properties were found between the four, which implies high similarity in their structural characteristics. Nonetheless, many possibilities remain for producing the most suitable fractions for lignin applications from the solid residue depending on the extracting methods. To investigate thermal properties of the four lignin fractions, TGA was carried out and the TG and DTG curves are plotted in Fig. 3. The thermal decomposition behaviors of the four lignin fractions varied according to the molecular weight. As shown in the TG curves, the decomposition temperatures of the components increased in proportion to their molecular weight. Particularly, the decomposition temperatures corresponding to 10% weight loss were 256 8C for ME, 248 8C for EE, 235 8C for AE and 218 8C for IE, and molecular weight distributions decreased in the order of ME > EE > AE > IE. This result revealed that the thermal stability of the extractable components tended to increase with increasing molecular weight, which has been observed in a previous study [31]. This result also suggested that the thermal stability could be controlled by varying the solvent used to extract the lignin fractions from the solid residue.
Conclusion This study focused on structural characterization of recovered solid residues through hydrothermal treatment of sunflower stalks with different severity and subsequent enzymatic hydrolysis. At
[1] P. Alvira, E. Toma´s-Pejo´, M. Ballesteros, M.J. Negro, Bioresour. Technol. 101 (2010) 4851. [2] I. Punda, Investment Centre Division, Sunflower Crude and Refined Oils, Agribusiness Handbook, Italy, 2010. [3] M.J. Dı´az, C. Cara, E. Ruiz, M. Pe´rez-Bonilla, E. Castro, Fuel 90 (2011) 3225. [4] E. Ruiz, C. Cara, P. Manzanares, M. Ballesteros, E. Castro, Enzyme Microb. Technol. 42 (2008) 160. [5] C.D. Jung, J.H. Yu, I.Y. Eom, K.S. Hong, Bioresour. Technol. 138 (2013) 1. [6] G. Hu, A.J. Ragauskas, Ind. Eng. Chem. Res. 50 (2011) 4225. [7] P. Kaparaju, C. Felby, Bioresour. Technol. 101 (2010) 3175. [8] J. Larsen, M.Ø. Haven, L. Thirup, Biomass Bioenergy 46 (2012) 36. [9] G. Brunow, K. Lundquist, in: C. Heitner, D.R. Dimmel, J.A. Schmidt (Eds.), Functional Groups and Bonding Patterns in Lignin (Including the Lignin–Carbohydrate Complexes), CRC Press, Florida, 2010, p. 267. [10] R.J.A. Gosselink, W. Teunissen, J.E.G. van Dam, E. de Jong, G. Gellerstedt, E.L. Scott, J.P.M. Sanders, Bioresour. Technol. 106 (2012) 173. [11] M.P. Pandey, C.S. Kim, Chem. Eng. Technol. 34 (2011) 29. [12] E.C. Ramires, J.D. Megiatto, C. Gardrat, A. Castellan, E. Frollini, Biotechnol. Bioeng. 107 (2010) 612. [13] A. Tejado, C. Pen˜a, J. Labidi, J.M. Echeverria, I. Mondragon, Bioresour. Technol. 98 (2007) 1655. [14] R.E. Hage, N. Brosse, L. Chrusciel, C. Sanchez, P. Sannigrahi, A.J. Ragauskas, Polym. Degrad. Stab. 94 (2009) 1632. [15] R.E. Hage, N. Brosse, P. Sannigrahi, A.J. Ragauskas, Polym. Degrad. Stab. 95 (2010) 997. [16] M. Nagy, K. David, G.J.P. Britovsek, A.J. Ragauskas, Holzforschung 63 (2009) 513. [17] W. Xu, S.J. Miller, P.K. Agrawal, C.W. Jones, ChemSusChem 5 (2012) 667. [18] C. Liu, C.E. Wyman, Bioresour. Technol. 96 (2005) 1978. [19] A.U. Buranov, K.A. Ross, G. Mazza, Bioresour. Technol. 101 (2010) 7446. [20] J. Ropponen, L. Ra¨sa¨nen, S. Rovio, T. Ohra-aho, T. Liitia¨, H. Mikkonen, D. van de Pas, T. Tamminen, Holzforschung 65 (2011) 543. [21] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D. Templeton, Determination of Ash in Biomass, Technical Report National Renewable Energy Laboratory NREL/ TP-510-42622, ed. Golden, CO, 2008. [22] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D.D. Templeton, Crocker Determination of Structural Carbohydrates and Lignin in, Technical Report NREL/TP-510-42618, ed. Golden, CO, 2008. [23] Y. Pu, S. Cao, A.J. Ragauskas, Energy Environ. Sci. 4 (2011) 3154. [24] J.Y. Kim, S. Oh, H. Hwang, U.J. Kim, J.W. Choi, Polym. Degrad. Stab. 98 (2013) 1671. [25] E. Jakab, O. Faix, F. Till, J. Anal. Appl. Pyrolysis 40–41 (1997) 171. [26] B.S. Donohoe, S.R. Decker, M.P. Tucker, M.E. Himmel, T.B. Vinzant, Biotechnol. Bioeng. 101 (2008) 913. [27] Y. Pu, F. Hu, F. Huang, B.H. Davison, A.J. Ragauskas, Biotechnol. Biofuels 6 (2013) 15. [28] W.O.S. Doherty, P. Mousavioun, C.M. Fellows, Ind. Crops Prod. 33 (2011) 259. [29] G. Hu, C. Cateto, Y. Pu, R. Samuel, A.J. Ragauskas, Energy Fuels 26 (2012) 740. [30] R. Kumar, G. Mago, V. Balan, C.E. Wyman, Bioresour. Technol. 100 (2009) 3948. [31] R.C. Sun, J. Tomkinson, G.L. Jones, Polym. Degrad. Stab. 68 (2000) 111.
Contents lists available at ScienceDirect
Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec
Structural characterization of the solid residue produced by hydrothermal treatment of sunflower stalks and subsequent enzymatic hydrolysis In-Yong Eom, Ju-Hyun Yu * Research Center for Biobased Chemistry, Korea Research Institute of Chemical Technology, Daejeon 305-600, Republic of Korea
A R T I C L E I N F O
Article history: Received 24 June 2014 Received in revised form 25 July 2014 Accepted 25 July 2014 Available online 2 August 2014 Keywords: Sunflower stalks Hydrothermal treatment Solid residue Lignin-enriched fraction Lignin valorization
A B S T R A C T
This study was to structurally characterize solid residues obtained from sunflower stalks hydrothermally treated at 180 and 200 8C for 30 min, followed by enzymatic hydrolysis. Recovered solid residue were 25.3% and 24.1% of fresh biomass, respectively. Each ethanol soluble fraction could be obtained up to 30% of the solid residue. The fraction from the solid residue at 200 8C was composed of smaller-sized lignin macromolecules with higher phenolic hydroxyl group, but lower aliphatic hydroxyl due to enhanced cleavage reactions related to side chain and aryl ether linkages of lignin, compared to that of the solid residue at 180 8C. ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
Introduction Recently, due to rising concern over the depletion of fossil fuels and accelerating global warming, the demand for lignocellulosic biomass has increased dramatically worldwide. Lignocellulosic biomass has been considered a sustainable and renewable primary energy resource that can be converted to alternative transportation fuels, including bioethanol or biodiesel, bio-chemicals, and bio-polymers [1]. Among the promising lignocellulosic biomass, sunflower is being spotlighted as a huge energy source. Sunflower is commonly cultivated due to its high oil content in the seeds, representing up to 80% of its economic value, and its oilseed yield accounts for 8% of the total global production of oilseeds (404 million tons in 2008/2009), corresponding to the fourth largest production in the world [2]. In addition, sunflower can be cultivated throughout the world due to their relatively short cultivation period with a high yield up to 3 ton/ha. Top sunflower producing countries are in the order of Russia (5.7 billion tons per year), Ukraine (4.2 billion tons per year), and Argentina (3.5 billion tons per year) [2]. Even though it is also useful to produce pectins and essential oil from sunflower heads, sunflower stalks have been studied as an attractive lignocellulosic biomass [3].
* Corresponding author. Tel.: +82 42 860 7438; fax: +82 42 861 4913. E-mail address: [email protected] (J.-H. Yu).
In contrast to 1st generation feedstocks, such as starch or sugarbased crops, lignocellulosic biomass, a promising 2nd generation renewable material, is composed of three major biopolymers – cellulose, hemicelluloses and lignin, the amount of each which varies depending on the lignocellulosic biomass species. Generally, cellulose accounts for around 35–50%, hemicellulose around 15– 35%, and lignin around 15–25% based on the dry weight biomass, and these components construct the plant cell wall structure in which lignin and polysaccharides are linked together. Therefore, to use cellulose and hemicelluloses as carbohydrate sources, pretreatment is an essential processing step in biochemical conversion of the lignocellulosic biomass to biofuels and biochemicals. The purpose of pretreatment is to break down the plant cell wall structure, due to the recalcitrance nature of lignocellulosic biomass, and then make cellulose more accessible to hydrolytic enzymes [4]. Among the various existing pretreatment methods, hydrothermal treatment, also called liquid hot water or autohydrolysis, has been evaluated as being a suitable process for producing fermentable sugars. Hydrothermal treatment is favorable for lower capital and operating investment, due to its use of only water as a reaction medium, lack of any requirement for acid or alkaline catalyst, and lower inhibitor generation such as furfural, 5hydroxymethyl-2-furaldehyde (HMF) and phenolics that may adversely affect microbial fermentation [1]. To date, various studies have evaluated hydrothermal treatment as an effective
http://dx.doi.org/10.1016/j.jiec.2014.07.044 1226-086X/ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
pretreatment for converting cellulose from various herbaceous biomasses such as switchgrass, sunflower stalks and wheat straw to fermentable sugars for subsequent fermentation [3,5–8]. These studies reported that complicated conditioning for subsequent enzymatic hydrolysis was not needed, due to the absence of any mineral acid and alkali catalyst. Moreover, these biomasses can be effectively converted to fermentable sugars with low yields of inhibitory products interfering with downstream operations, such as microbial fermentation. During fermentable sugar production, non-hydrolyzable solid residue that is mainly composed of lignin can be recovered but, unlike fermentable sugars for subsequent microbial fermentation, it is utilized as a low-grade boiler fuel. Lignin is a polyphenolic polymer derived from three phenylpropanoid units as precursors for lignin biosynthesis: p-coumaryl, coniferyl and sinapyl alcohol. The relative portions of each monolignol in lignocellulosic biomass differ depending on the biomass species. The lignin macromolecule is linked by carbon-carbon and carbon–oxygen bonds via radical coupling polymerization of three monolignols. The aryl ether bond (b-O-4) is major interunit linkage, and other linkages, such as biphenyl, b-5, and a-O-4 can also be detected [9]. Hydrothermal treatment has less affect on delignification, thus most of the lignin in the starting material remains in the pretreated biomass. Lignin has been regarded as the most recalcitrant component to enzymatic and microbial attack. Therefore, with the steady acceleration in industrial biofuel and biochemical production from lignocellulosic biomass, large quantities of solid residue are also generated as a byproduct [10]. Improved cost-effectiveness and optimized biomass usage will necessitate effective valorization of solid residue. After enzymatic hydrolysis of the pretreated biomass, the remaining solid residue, with lignin as its main component, is an attractive aromatic resource for producing aromatic chemicals and materials. To date, a number of researchers have attempted to develop novel valorization of lignin, including fractionation methods of lignin, and synthesis of lignin-based polymers [11– 13]. However, significant challenges remain in the application of lignin due to its structural complexity. Moreover, the physicochemical properties of lignin differ depending on the biomass and recovery processes. Therefore, it is important to evaluate the potential use of solid residue after hydrothermal treatment, and subsequent enzymatic hydrolysis. The majority of the lignin research, to date, has focused on organosolv lignin recovered from organosolv-pretreated supernatants, and organosolv lignin has been comprehensively characterized and evaluated for further use [6,10,14–17]. On the other hand, little research has investigated the valorization of solid residue as a by-product obtained from hydrothermally treated lignocellulosic biomass, and subsequent enzymatic hydrolysis for producing fermentable sugars. During hydrothermal treatment without adding any chemicals, lignin is rarely thermally decomposed and hydrolyzed, whereas it retains its native lignin properties [18]. As the lignin in the solid residue could be also as heterogeneous and insoluble as natural lignin, it might not be suitable for chemical purpose. If the solid residue is fractionated into more suitable materials for value added applications, this solid residue could potentially serve as an attractive source for phenolicbased chemical industries [19,20]. The objectives of this study were to structurally characterize the solid residue recovered from the hydrothermally treated sunflower stalks at 180 and 200 8C, followed by enzymatic hydrolysis. In addition, lignin-enriched fractions were collected and characterized by compositional analysis, elemental analysis, thermogravimetric analysis (TGA), gel permeation chromatography (GPC), and 31P nuclear magnetic resonance (NMR) spectroscopy in order to elucidate the structural information on a starting material for bio-based polymers and fine/bulk chemicals.
73
Materials and methods Raw material Sunflower stalks (Helianthus annuus L.) were supplied by Chungbuk Agricultural Research and Extension Services (Ochang, Chungcheongbuk-Do, Korea) and air-dried in a greenhouse. The stalks were ground, using a knife mill equipped with a 20-mesh aperture screen, and the ground stalks containing 2.3% moisture were stored at 20 8C in a refrigerator in sealed plastic bags until use. Cellic1 CTec2 as a cellulase, Cellic1 HTec2 as a xylanase, and Novozyme 188 as a beta-glucosidase were purchased from Novozymes Korea (Seoul, Korea). All organic solvents at the highest purity grades were purchased from Sigma–Aldrich. Hydrothermal treatment followed by enzymatic hydrolysis Hydrothermal treatment was carried out using a 2 L-scale Parr reactor (Parr Instruments, Moline, IL). The pretreatment method has been specifically described in a previous study [5]. The powdered sunflower stalks equivalent to 120 g dry weight were added to reactor vessels, and then filled with de-ionized water, to a total weight of 1500 g. The reactor vessels were sealed, kept at room temperature through the night, and pretreated at 180 and 200 8C for 30 min, respectively. All the pretreated biomasses were transferred to 7 L-scaled batch-type fermenters (BioTron, Seoul, Korea), and a mixture of Cellic1 CTec2, HTec2, and b-glucosidase (volumetric ratio 18:2:1) was loaded at a volume equivalent to 0.1 ml (ca. 8.7 FPU) per 1 g dry biomass. Enzymatic hydrolysis was carried out at 50 1 8C with a rotating speed of 200 rpm for 72 h. The hydrolyzates were separated into solid and liquid fractions by centrifugation at 3727 g for an hour, with a swing-type centrifuge (Model Combi-514, Hanil Scientific Co., Seoul, Korea). The solid residue was washed out with de-ionized water, followed by centrifugation. This washing/centrifugation cycle was repeated until the initial supernatants of the hydrolyzates remained less than 1%. The residue was then dried at 45 8C until the moisture content dropped below 10%. The solid residues obtained from sunflower stalks hydrothermally treated at 180 and 200 8C, followed by enzymatic hydrolysis, are herein referred to as 180 8C-solid and 200 8C-solid, respectively. Sequential extraction of solid residue To prepare homogeneous lignin-enriched fractions, the solid residues were sequentially extracted with water and organic solvents using a Soxhlet apparatus. Around 5 g of the solid residues were extracted with 150 ml de-ionized water, and then oven-dried at 45 8C, until the moisture content dropped below 10%. The water extractives (WE)-free solid residues were sequentially extracted with 150 ml of organic solvents, including ethanol, methanol, acetone, and isopropyl alcohol. The dissolved fractions in water and organic solvents were evaporated and freeze-dried. Compositional and elemental analysis The compositional analysis of the solid residues and extractives with water or organic solvents was performed according to National Renewable Energy Laboratory (NREL) standard procedures [21,22]. The method is briefly described as follows. The carbohydrates and lignin in the sample were measured by the twostep acid hydrolysis. First, concentrated acid hydrolysis was conducted with 72% (w/w) H2SO4 at 30 8C for 60 min. In the second step, the above hydrolyzates were diluted to 4% (w/w) H2SO4 with de-ionized water, and then diluted acid hydrolysis was performed at 121 8C for 60 min in an autoclave. The hydrolyzates
74
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
were filtered with crucibles to obtain the supernatants and solid residues. The monosaccharides in the supernatants were determined using high performance liquid chromatography (HPLC) equipped with an autosampler (Waters 2707), a refractive index detector (Waters 2414), and a binary HPLC pump (Waters 1525). An Aminex HPX-87H column (BIO-Rad, Hercules, CA) operating at 50 8C was used, and the mobile phase was 5 mM H2SO4, at a flow rate of 0.6 ml/min. The acid-soluble lignin (ASL) was also determined by DU 800 UV/Vis spectrophotometer (Beckman Coulter Inc., USA) at 320 nm. The acid-insoluble lignin (AIL) on the crucibles was gravimetrically determined, based on the remaining insoluble residue, by subtracting the ash content from them. Elemental analysis (C, H and N) was performed, using a Thermo Scientific Flash 2000 (Thermo Fisher Scientific Inc., USA). The oxygen content was calculated by subtracting the sum of the weight of C, H, and N from the total. All compositional and elemental analyses were run in triplicate, and reported as average results. Thermogravimetric analysis (TGA) Thermogravimetric (TG) and differential thermogravimetric (DTG) analyses of 180- and 200 8C-solids and organic solvent extractives were performed using the Q-500 IR Instrument (TA Instruments, USA). TG analysis of the samples (2.0 0.5 mg) was carried out at a constant heating rate of 10 8C/min, at temperatures ranging from 30 to 700 8C. To maintain an inert atmosphere during pyrolysis, a purified nitrogen as a carrier gas was flowed at a rate of 100 ml/min. Gel permeation chromatography (GPC) The number-average molecular weight (Mn) and weightaverage molecular weight (Mw) of the organic soluble fractions were determined by GPC, after their acetylation. Acetylation of organic soluble fractions was performed according to Hage et al. [14], and the method is briefly described as follows. The organic solvent-soluble fraction (20–25 mg) was dissolved in an acetic anhydride/pyridine mixture of 2.00 ml (1:1, v/v), stirred for 24 h at room temperature and then added to ethanol (25 ml). The addition and evaporation of ethanol was repeated, until the acetic acid and pyridine were completely evaporated. The acetylated fraction was dissolved in chloroform (2 ml), added drop-wise to diethyl ether (100 ml), and then centrifuged. The precipitate was repeatedly washed with diethyl ether, followed by centrifugation, and then freeze-dried. GPC analysis for lignin was performed, using a NP4000 GPC system (Futecs Co., Ltd) equipped with an autosampler, refractive index detector (Shodex RI-101), and LF-404 column (Shodex Inc.). The acetylated sample was dissolved in tetrahydrofuran (THF) (1 mg/ml), and the solution was filtered through a 0.45 mm filter. GPC was operated with the eluent of THF, at a flow rate of 0.5 ml/min, injection volume of 120 mm, and retention time of 30 min. Standard polystyrene samples were used to determine the calibration curves. Data were analyzed with the Clarity GPC Data System. 31
P NMR spectroscopy
To investigate the phenolic hydroxyl group present in lignin, organic solvent-soluble fractions were phosphitylated according to Pu et al. [23]. To prepare phosphitylated lignin, 2-chloro-4,4,5,5,tetramethyl-1,3,2-dioxaphospholane (TMDP) was used as the phosphitylating reagent. Cyclohexanol was used as an internal standard. 31P NMR experiments were performed on a Bruker AVANCE 500 MHz. Quantitative 31P NMR spectra were acquired, using an inverse-gated decoupling pulse sequence with a 30 pulse
angle and 25 s pulse delay. Chemical shifts were calibrated, based on the sharp and stable signal for phosphitylated water at 132.2 ppm. 31P NMR spectra were quantified, using the SpinWorks 2.5 software. Results and discussion Characterization of the 180- and 200 8C-solids The composition of sunflower stalks, which had been determined in a previous study, contained 35.8 0.5 g of cellulose as glucose, 19.7 0.1 g of hemicellulosic sugars, including the sum of xylose, galactose and mannose (XGM, 19.3 0.1 g), and arabinose (0.4 0.1 g), and 5.1 0.1 g of acetyl group on g per 100 g dry biomass [5]. As other components, AIL and ASL (14.8 0.0 g and 0.1 0.0 g, respectively), water extractives (WE) and ethanol extractives (EE) (20.5 0.2 g and 1.1 0.1 g, respectively), and ash (6.3 0.0 g) were determined [5]. The elemental composition of the sunflower stalks was as follows: C (43.0%), H (5.4%), O (50.8%), N (0.8%) on a dry weight basis before the extraction with water and ethanol, and C (44.8%), H (5.8%), O (49.0%), N (0.4%) after the extraction. The chemical composition of 180- and 200 8C-solids is presented in Table 1. Their elemental compositions before and after sequential extraction with water and ethanol, are also reported. The pretreatment temperatures that applied for sunflower stalks to recover 180- and 200 8C-solids were determined according to a previous study [5], which showed that the pretreatment temperature at 180 8C offers the advantages of maximizing the yield of hemicellulosic sugars as well as producing less carbohydrate decomposition products that act as inhibitors of microbial fermentation. However, the enzymatic digestibility of cellulose in the biomass treated at 180 8C reached only 67% of its theoretical yield. In contrast, sunflower stalks hydrothermally treated at 200 8C can undergo conversion of almost all of the cellulose into glucose through enzymatic hydrolysis, which yielded 76.8% of the theoretical yield. when the pretreated biomass at 200 8C was washed with water, the enzymatic digestibility of cellulose was increased to 88.8% [5]. The solid residue (25.3 g) based on 100 g dry sunflower stalks could be recovered after hydrothermal treatment at 180 8C, and subsequent enzymatic hydrolysis, while it was decreased to 24.1 g after hydrothermal treatment at 200 8C. Both recovered solids were mainly composed of AIL and extractable components with water and ethanol (WE and EE), but contained a small portion of cellulose and hemicelluloses. The amount of glucose liberated from the 180 8C-solid was 13.2 g/100 g dry 180 8C-solid, which was equivalent to 9.3% of total cellulose as glucose in the sunflower stalks; whereas that for the 200 8C-solid was 7.1 g/100 g dry 200 8C-solid, accounting for 4.8% of cellulose as glucose. These results indicated that the solid recovery of the 200 8C-solid was decreased than that of the 180 8C-solid, due to increasing glucose liberation from the biomass. During hydrothermal treatment in which water is used as the only reaction medium, almost all of the lignin in the starting material remained in the pretreated biomass, due to its low solubility in water. However, interesting results were observed. The 180 8C-solid consists of 47.5 g of AIL based on the 100 g dry 180 8C-solid, accounting for 80.7% of lignin in the sunflower stalks, while 38.5 g of AIL was determined from the 200 8C-solid, which accounts for 62.3% of lignin in the biomass. This reveals that even though the lignin in the biomass barely becomes soluble in water during hydrothermal treatment at relatively low temperature, it undergoes significant changes in the point of physicochemical properties. Prior to ethanol extraction, the 180- and 200 8C-solids were extracted with water using Soxhlet apparatus to remove the remaining free-sugars and proteins. Significant portion of
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
75
Table 1 Composition of solid residues hydrothermally treated at 180 and 200 8C for 30 min and subsequent enzymatic hydrolysis. 200 8C-Solid
180 8C-Solid Yielda
25.3
Component
Before extractionb
After extractionc
Before extraction
After extraction
Cellulose as glucose XGMd ASLe AILf WEg EEh Ash
13.2 0.4 1.1 0.0 0.3 0.0 47.5 0.4 7.5 0.6 19.4 0.1 2.1 0.0
18.6 0.5 1.6 0.0 0.4 0.0 66.9 0.5
7.1 0.1 0.9 0.0 0.1 0.0 38.5 0.2 13.6 0.1 29.0 0.3 4.1 0.1
13.3 0.1 1.6 0.0 0.2 0.0 72.3 0.3
Element (%) C H O N H/C (mol/mol)
56.2 5.9 36.9 1.0 1.26
54.6 5.7 38.7 1.0 1.25
58.8 5.7 34.1 1.4 1.16
55.6 5.3 37.5 1.6 1.13
a b c d e f g h
24.1
On the basis of 100 g dry raw biomass. On the basis of 100 g dry solid residue. On the basis of 100 g dry solid residue extracted with water and sequential ethanol. Sum of xylose, galactose, and mannose as hemicellulosic sugars. Acid-soluble lignin. Acid-insoluble lignin. Water extractives. Ethanol extractives.
the 180- and 200 8C-solids could be extracted with ethanol, resulting in the EE being the second largest component, after AIL. The EE content of the 180- and 200 8C-solids was 19.4 and 29.0 g based on the 100 g dry solid, respectively, which could be related with the extent of further degradation of lignin by pretreatment temperature rise. A larger portion of the 200 8C-solid could be extracted with ethanol than that of the 180 8C-solid, and, in turn, the AIL content in the 200 8C-solid was lower than that of the 180 8C-solid, as aforementioned, which indicated that lignin macromolecules in the 200 8C-solid could be further degraded into small fragments to be soluble in organic solvents. Therefore, these EEs, which are believed to be composed mainly of lignin from the 180- and 200 8C-solids, were characterized further and are discussed in the following section. As shown in the elemental composition in Table 1, both the carbon and hydrogen contents increased in the 180- and 200 8Csoilds compared to that of the raw biomass, which could be attributed to not only the increased lignin content in the solid, but also the decreased cellulose content, due to its conversion to glucose [24]. The nitrogen content in the 200 8C-solid was higher than that of the 180 8C-solid. This high nitrogen content in the 180- and 200 8Csolids was thought to be affected by irreversible adsorption of cellulolytic enzyme on the solid surface during enzymatic hydrolysis. Despite adding the same amount of enzyme for enzymatic hydrolysis, the enzyme could be irreversibly bound to the 200 8Csolid more than the 180 8C-solid due to the structural change of lignin during the hydrothermal treatment at higher temperature. The average molecular weight of the 180- and 200 8C-solids could not be determined, because their acetylated samples were not dissolved in THF that was used as an organic solvent for GPC analysis. The insoluble fraction of the solids was due to its highly condensed structure with a high molecular weight and heterogeneous composition such as cellulose and proteins besides lignin.
residues that were thought to have a more condensed structure due to their being more resistant to thermal decomposition than the extractable fractions. TG and DTG curves for the extractivesfree 180- and 200 8C-solids are plotted in Fig. 1. In each TG curve for the 180- and 200 8C-solids, both solid residues started to decompose at around 200 8C, and then thermal decomposition was significantly accelerated at between around 280 and 390 8C. After the active decomposition, the 180- and 200 8C-solids continued to decompose slowly. In one distinct difference in the DTG curves, the decomposition rate (8C/min) of the 180 8C-solid was higher than that of the 200 8C-solid in the active decomposition region ranging from 280 to 390 8C. This result indicated that the main pyrolysis reactions, such as depolymerization, fragmentation, and decarboxylation, are more favorable during thermal decomposition of the 180 8C-solid, whereas the thermal stability of the 200 8C-solid was higher than that of the 180 8C-solid. In addition, a remarkable difference between the two TG curves was observed in the slow decomposition region above 390 8C. There was a significant difference in char amount, referred to as the remaining residue after TG analysis. It was reported that the char forming reaction of lignin began at around 600 8C, with the
Thermal decomposition behaviors of the 180- and 200 8C-solids To investigate the thermal decomposition behaviors of the 180and 200 8C-solids, TGA was carried out at a constant heating rate of 10 8C/min. Prior to TGA of the solid residue, the 180- and 200 8Csolids were sequentially extracted with water and ethanol to evaluate the thermal stability of the non-extractable solid
Fig. 1. TG and DTG curves of the extractives-free 180- and 200 8C-solids in N2 at a heating rate of 10 8C/min.
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
76
Table 2 Compositions of water and ethanol extractives (WW and EE). Composition
180 8C-Solid
200 8C-Solid
WE
EE
WE
EE
Yielda Glucoseb XGMb ASLb AILb
7.5 0.6 5.5 0.0 5.7 0.0 6.1 0.1 54.2 0.7
19.4 0.1 0.4 0.0 ND 0.9 0.0 90.1 0.3
13.6 0.1 8.6 0.1 2.0 0.1 8.4 0.1 57.0 0.7
29.0 0.3 ND ND 1.2 0.1 96.2 0.1
Element (%) C H O N H/C (mol/mol)
53.2 0.2 5.5 0.1 39.4 0.3 1.9 0.1 1.24
64.4 0.2 6.6 0.0 28.4 0.2 0.6 0.0 1.23
54.8 0.3 5.5 0.0 38.2 0.3 1.6 0.0 1.20
66.5 0.1 6.4 0.1 26.3 0.0 0.9 0.0 1.15
a b
On the basis of 100 g dry solid. On the basis of 100 g dry extractives.
emission of methane [25]. The study also found that char yield tends to decrease when lignin contains higher amount of functional groups, such as methoxyl, hydroxyl and carboxyl groups. From these results, the large difference in char yields between the 180- and 200 8C-solids could be closely associated with the structural change, depending on the different pretreatment temperatures. It was, therefore, expected that the nonextractable residue in the 200 8C-solid could be made up of a highly condensed lignin structure through condensation reaction between lignin fragments, due to harsh pretreatment condition, compared to that of the 180 8C-solid. It was also suggested that the hydrothermal treatment of sunflower stalks at higher temperatures led to both condensation and fragmentation reactions that resulted in the 200 8C-solid having a highly condensed lignin structure and, as shown above, being also abundant in extractable components that reached 29% of the total 200 8C-solid, which was more than that of the 180 8C-solid. This result was in accordance with the GPC result discussed in more detail below. Characterization of water and ethanol extractives (WE and EE) from the solid residue The compositions of WE and EE are summarized in Table 2. The 180- and 200 8C-solids were composed of extractable fractions with water and ethanol more than 25% of total solid, respectively. In addition, both the WE and EE yields from the 200 8C-solid were approximately 2-fold higher than those of the 180 8C-solid, which was attributed to the higher pretreatment temperature which promoted fragmentation reaction in the lignin. The WE obtained from both the 180- and 200 8C-solids was mainly composed of lignin accounting for more than 60% of total WE. The lignin extracted with water could be related to the lignin droplets that were formed during pretreatment under acidic condition. During pretreatment at high temperatures ranging from 140 to 190 8C, at which lignin can be molten, lignin and further decomposed fragments could migrate, and subsequently coalesce into droplet-like structures rather than be soluble in weakly acidic water, due to its hydrophobicity [26]. After cooling down to room temperature, these lignin droplets may have been adsorbed on the surface of cellulose’s microfibrils, and then recovered by extraction with water and organic solvents. Therefore, a large portion of the 180- and 200 8C-solids that was regarded as lignin could be extracted with ethanol rather than water, due to its solubility being higher in ethanol than water, and the extent of extraction differed depending on pretreatment temperatures. The large amount of lignin in the WEs could be also attributed to pseudo-lignin which carbohydrate moieties can react to form lignin-like fractions during the determination of lignin by 2-step acid hydrolysis
method [27]. Therefore, the lignin content in the WEs may have been overestimated. Carbohydrates of the WEs were mainly non-structural sugars that could be determined without acid-hydrolysis. These sugars were thought to be predominantly derived from adsorbed sugars that could not be recovered while separating the liquid from the enzymatic hydrolyzates. Some of the WE compositions were not able to be characterized, but were estimated to be proteins originating from the cellulolytic enzyme, due to high nitrogen content in the WEs of more than 1.5%. On the contrary, whereas WE consisted of heterogeneous components, such as sugars, proteins and lignin, EE from both the 180- and 200 8C-solids consisted of a homogeneous component in which lignin was the dominant constituent. Moreover, the EE yield from the 200 8C-solid was 1.5-fold higher than that of the 180 8Csolid, due to predominant fragmentation reaction in lignin at higher temperature. During hydrothermal treatment at mild temperatures, the lignin in the biomass was rarely broken down to lower molecular weight fractions, which limits the potential applications of the recovered solid residue for producing industrial chemicals due to its low solubility and low compatibility [28]. To use a large amount of lignin in the solid residue for potential value-added applications, therefore, it is necessary to prepare a homogeneous lignin fraction from solid residue. The EE fractionated from the solid residue could be an alternative for use as a material. Structural features of ethanol extractives (EE) The average molecular weight of ethanol extractives (EE) The average molecular weight (Mw) of the acetylated EEs completely dissolved in THF was determined using GPC. The analytical results are given in Table 3. As aforementioned, a larger portion of the 200 8C-solid could be extracted with ethanol, compared to 180 8C-EE, and the Mw of the resulting 200 8C-EE was lower than that of 180 8C-EE. These results suggest that the fragmentation reaction in lignin predominated over the recondensation between lignin fragments via the fragmentation reaction during hydrothermal treatment at higher temperature, which resulted in increasing cleavage of the aryl ether bond, particularly b-O-4 linkage. Due to the prevailing cleavage of b-O-4 linkage in the 200 8C-solid, more of 200 8C-EE with lower molecular weight could be extracted than the 180 8C-EE, but the polydispersity of the two EEs was similar. Characterization of hydroxyl group in ethanol extractives (EE) by 31P NMR The EEs from the WE-free 180- and 200 8C-solids were considered as lignin-enriched fractions, due to the high lignin content that was determined by the 2-step acid hydrolysis method. To investigate several hydroxyl groups present in lignin, the phosphitylated EEs were analyzed by 31P NMR spectroscopy. The phosphitylation of various hydroxyl groups attached to lignin macromolecules in the EE is depicted in Fig. 2, along with the detectable forms of phosphitylated EE. The results of the 31P NMR are summarized in Table 4. Several hydroxyl groups could be classified as aliphatic hydroxyl groups (aliphatic-OH) or phenolic hydroxyl groups, such as the H-OH: hydroxyl group in the p-hydroxyphenyl unit, G-OH: hydroxyl group in the guaiacyl unit, Table 3 Average molecular weights of the ethanol extractives (EE) obtained from the 180and 200 8C-solids. Mw (g mol EE from 180 8C-solid EE from 200 8C-solid
4736 3540
1
)
Mn (g mol 2317 1696
1
)
Polydispersity 2.04 2.09
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
77
Fig. 2. Phosphitylation of hydroxyl groups in lignin unit monomer with 2-chloro-4,4,5,5,-tetramethyl-1,3,2-dioxaphospholane (TMDP).
Table 4 Quantitative
31
P NMR analysis results of several hydroxyl groups (mmol g
1
) in ethanol extractives (EE) from the 180- and 200 8C-solids.
Assignment/chemical shift d
Component 180 8C-EE 200 8C-EE
Aliphatic-OH
C-Ph–OH
S-OH
G-OH
H-OH
COOH
145.5–150.0 2.96 1.56
145.5–150.0 0.19 0.37
141.8–143.5 1.14 1.47
138.7–140.1 0.58 0.78
137.2–138.1 0.02 0.12
134–135.7 0.28 0.31
C-Ph, condensed phenolic unit; S, syringyl; G, guaiacyl; H, p-hydroxyphenyl; COOH, carboxylic acids.
S-OH: hydroxyl group in the syringyl unit, and C-Ph: hydroxyl group in the condensed phenolic unit or carboxylic hydroxyl groups (COOH). The higher pretreatment temperature led to a 53% decrease in aliphatic hydroxyl, and a 42% increase in phenolic hydroxyl content in 200 8C-EE, compared with those of 180 8C-EE. The difference between the aliphatic and phenolic hydroxyl contents was attributed to side chain and aryl ether bond cleavage, primarily b-O-4 linkages, respectively [29]. In case of organosolv lignin, it was also found in accordance with this study that aliphatic hydroxyl contents in recovered the organoslov lignin were fairly reduced with increasing organosolv pretreatment severity calculated by combined severity from 1.75 to 2.8 due to enhancement of the dehydration reactions on the side chain. Moreover, the formed organosolv lignin during the organosolv at higher severity was composed of extensively cleaved lignin fractions of which molecular weights were decreased in reverse proportion to the severity [15]. According to previous study, the fragmentation of aryl ether linkages catalyzed by acid is the predominant degradation reaction in lignin, after which re-condensation between resulting lignin fragments simultaneously occurs when pretreatment at low pH was carried out [30]. As shown in Table 4, the phenolic hydroxyl content in the 200 8C-solid was higher than that in the 180 8Csolid; therefore, these results suggested that the higher pretreatment temperature favored the promotion of aryl ether bond cleavage. In addition, with increasing pretreatment temperatures, lignin was susceptible to be broken down via side chain and arylether bond cleavage reaction, which affected both its physicochemical properties and the yield of the extractable component that can be recovered from the solid residue. Consequently, the 31P NMR analysis results revealed 200 8C-EE, compared to 180 8C-EE, to be composed of more lignin macromolecules with higher phenolic hydroxyl groups through fragmentation reaction during hydrothermal treatment. This result was in accordance with the lower molecular weight of 200 8C-EE compared to that of 180 8CEE, as confirmed by the GPC result presented in Table 3 above. Fundamental characteristics of diverse extractable lignins from the 200 8C-solid The study results suggest an increased possibility of 200 8C-EE being used as a polymeric material. To maximize the usability of
the solid residue after fermentable sugar production, the extractable lignin needs to be fractionated from the solid residue. Therefore, the lignin fractions were obtained by extracting the 200 8C-solid with organic solvents such as ethanol, methanol, acetone, and isopropyl alcohol from the WE-free solid, and then characterized in order to elucidate the structural properties. As shown in Table 5, the extractable lignin yields reached 29% when extracted with ethanol and methanol. According to previous study, the elemental compositions of milled wood lignin (MWL) and lignin obtained from miscanthus pretreated by organosolv with ethanol (organosolv lignin) were C (63.5%), H (5.7%), O (29.2%), N (0.2%) and C (65.1%), H (5.7%), O (26.6%), N (0.5%) [14]. In other study, moreover, the elemental compositions of the organosolv lignin from poplar wood xylem was determined as C (65.4%), H (5.9%), O (23.4%), and N (0.3%) [24]. Extractable lignins from the 200 8C-solid also showed comparable elemental compositions to the MWL and organosolv lignin isolated from miscanthus and poplar. These previously reported compositions do not differ significantly from the organosolv lignin and extractable lignins of the present study. Even though the above four lignin fractions were extracted with organic solvents that have decreasing polarity from ethanol to isopropyl alcohol, no distinguishable differences in structural
Table 5 Gravimetric yields and fundamental properties of lignin fractions extracted with organic solvents from water extractives (WE)-free 200 8C-solid. Lignin fractiona
EE
ME
AE
IE
Yieldb
29.0 0.3
28.9 0.1
26.3 0.3
24.9 0.8
Element (%) C H O N H/C (mol/mol)
66.5 0.1 6.4 0.1 26.3 0.0 0.9 0.0 1.15
66.3 0.8 6.3 0.1 26.6 0.9 0.9 0.0 1.14
67.7 0.0 6.5 0.1 25.3 0.1 0.6 0.0 1.15
66.8 0.6 6.8 0.1 25.8 0.7 0.7 0.0 1.22
Molecular weight Mw (g mol 1) Mn (g mol 1) Polydispersity
3540 1696 2.09
3720 1797 2.07
3378 1795 1.88
2838 1652 1.72
a Lignin fractions: EE, extracted with ethanol; ME, extracted with methanol; AE, extracted with acetone; IE, extracted with isopropyl alcohol. b On the basis of 100 g dry 200 8C-solid.
78
I.-Y. Eom, J.-H. Yu / Journal of Industrial and Engineering Chemistry 23 (2015) 72–78
higher pretreatment temperatures, the solid residue was further degraded, and organic solvent solubles could be more extracted. It was found that the obtained fraction was composed of lignins with higher aromatic hydroxyl content but lower aliphatic; however, remaining solid after the extraction showed higher thermal stability due to being more condensed structure. Further studies is needed on developing method for extracting lignin macromolecules with a high yield to be compatible with other commercial polymers and to be used as technical lignin for further application. References
Fig. 3. TG and DTG curves of the lignin fractions extracted with organic solvents from the WE-free 200 8C-solid in N2 at a heating rate of 10 8C/min. EE, ethanol extractable lignin; ME, methanol extractable lignin; AE, acetone extractable lignin; IE, isopropyl alcohol extractable lignin.
properties were found between the four, which implies high similarity in their structural characteristics. Nonetheless, many possibilities remain for producing the most suitable fractions for lignin applications from the solid residue depending on the extracting methods. To investigate thermal properties of the four lignin fractions, TGA was carried out and the TG and DTG curves are plotted in Fig. 3. The thermal decomposition behaviors of the four lignin fractions varied according to the molecular weight. As shown in the TG curves, the decomposition temperatures of the components increased in proportion to their molecular weight. Particularly, the decomposition temperatures corresponding to 10% weight loss were 256 8C for ME, 248 8C for EE, 235 8C for AE and 218 8C for IE, and molecular weight distributions decreased in the order of ME > EE > AE > IE. This result revealed that the thermal stability of the extractable components tended to increase with increasing molecular weight, which has been observed in a previous study [31]. This result also suggested that the thermal stability could be controlled by varying the solvent used to extract the lignin fractions from the solid residue.
Conclusion This study focused on structural characterization of recovered solid residues through hydrothermal treatment of sunflower stalks with different severity and subsequent enzymatic hydrolysis. At
[1] P. Alvira, E. Toma´s-Pejo´, M. Ballesteros, M.J. Negro, Bioresour. Technol. 101 (2010) 4851. [2] I. Punda, Investment Centre Division, Sunflower Crude and Refined Oils, Agribusiness Handbook, Italy, 2010. [3] M.J. Dı´az, C. Cara, E. Ruiz, M. Pe´rez-Bonilla, E. Castro, Fuel 90 (2011) 3225. [4] E. Ruiz, C. Cara, P. Manzanares, M. Ballesteros, E. Castro, Enzyme Microb. Technol. 42 (2008) 160. [5] C.D. Jung, J.H. Yu, I.Y. Eom, K.S. Hong, Bioresour. Technol. 138 (2013) 1. [6] G. Hu, A.J. Ragauskas, Ind. Eng. Chem. Res. 50 (2011) 4225. [7] P. Kaparaju, C. Felby, Bioresour. Technol. 101 (2010) 3175. [8] J. Larsen, M.Ø. Haven, L. Thirup, Biomass Bioenergy 46 (2012) 36. [9] G. Brunow, K. Lundquist, in: C. Heitner, D.R. Dimmel, J.A. Schmidt (Eds.), Functional Groups and Bonding Patterns in Lignin (Including the Lignin–Carbohydrate Complexes), CRC Press, Florida, 2010, p. 267. [10] R.J.A. Gosselink, W. Teunissen, J.E.G. van Dam, E. de Jong, G. Gellerstedt, E.L. Scott, J.P.M. Sanders, Bioresour. Technol. 106 (2012) 173. [11] M.P. Pandey, C.S. Kim, Chem. Eng. Technol. 34 (2011) 29. [12] E.C. Ramires, J.D. Megiatto, C. Gardrat, A. Castellan, E. Frollini, Biotechnol. Bioeng. 107 (2010) 612. [13] A. Tejado, C. Pen˜a, J. Labidi, J.M. Echeverria, I. Mondragon, Bioresour. Technol. 98 (2007) 1655. [14] R.E. Hage, N. Brosse, L. Chrusciel, C. Sanchez, P. Sannigrahi, A.J. Ragauskas, Polym. Degrad. Stab. 94 (2009) 1632. [15] R.E. Hage, N. Brosse, P. Sannigrahi, A.J. Ragauskas, Polym. Degrad. Stab. 95 (2010) 997. [16] M. Nagy, K. David, G.J.P. Britovsek, A.J. Ragauskas, Holzforschung 63 (2009) 513. [17] W. Xu, S.J. Miller, P.K. Agrawal, C.W. Jones, ChemSusChem 5 (2012) 667. [18] C. Liu, C.E. Wyman, Bioresour. Technol. 96 (2005) 1978. [19] A.U. Buranov, K.A. Ross, G. Mazza, Bioresour. Technol. 101 (2010) 7446. [20] J. Ropponen, L. Ra¨sa¨nen, S. Rovio, T. Ohra-aho, T. Liitia¨, H. Mikkonen, D. van de Pas, T. Tamminen, Holzforschung 65 (2011) 543. [21] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D. Templeton, Determination of Ash in Biomass, Technical Report National Renewable Energy Laboratory NREL/ TP-510-42622, ed. Golden, CO, 2008. [22] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D.D. Templeton, Crocker Determination of Structural Carbohydrates and Lignin in, Technical Report NREL/TP-510-42618, ed. Golden, CO, 2008. [23] Y. Pu, S. Cao, A.J. Ragauskas, Energy Environ. Sci. 4 (2011) 3154. [24] J.Y. Kim, S. Oh, H. Hwang, U.J. Kim, J.W. Choi, Polym. Degrad. Stab. 98 (2013) 1671. [25] E. Jakab, O. Faix, F. Till, J. Anal. Appl. Pyrolysis 40–41 (1997) 171. [26] B.S. Donohoe, S.R. Decker, M.P. Tucker, M.E. Himmel, T.B. Vinzant, Biotechnol. Bioeng. 101 (2008) 913. [27] Y. Pu, F. Hu, F. Huang, B.H. Davison, A.J. Ragauskas, Biotechnol. Biofuels 6 (2013) 15. [28] W.O.S. Doherty, P. Mousavioun, C.M. Fellows, Ind. Crops Prod. 33 (2011) 259. [29] G. Hu, C. Cateto, Y. Pu, R. Samuel, A.J. Ragauskas, Energy Fuels 26 (2012) 740. [30] R. Kumar, G. Mago, V. Balan, C.E. Wyman, Bioresour. Technol. 100 (2009) 3948. [31] R.C. Sun, J. Tomkinson, G.L. Jones, Polym. Degrad. Stab. 68 (2000) 111.