6618
Ind. Eng. Chem. Res. 2009, 48, 6618–6626
Processing of Acacia dealbata in Aqueous Media: First Step of a Wood Biorefinery Remedios Ya´n˜ez,†,‡ Aloia Romanı´,†,‡ Gil Garrote,*,†,‡ Jose´ Luis Alonso,†,‡ and Juan Carlos Parajo´†,‡ Department of Chemical Engineering, Faculty of Science, Campus de Ourense, UniVersity of Vigo, 32004, Ourense, Spain, and CITI (Centro de InVestigacio´n, Transferencia e InnoVacio´n), UniVersity of Vigo, Tecnopole, San Cibrao das Vin˜as, Ourense, Spain
In order to assess the suitability of autohydrolysis as a first biorefinery stage, Acacia dealbata wood samples were heated in aqueous media to a range of 170-240 °C. Under selected operational conditions (maximum temperature, 215 °C), 70% xylan was converted into xylooligosaccharides, whereas cellulose and lignin remained in solid phase with little alteration. The spent solids from treatments contained 65% cellulose (measured as glucan), 8% hemicelluloses, and 27% lignin. Kinetic models describing the Acacia dealbata wood solubilization as well as the autohydrolysis of the polysaccharide fractions (glucan, xylan, and arabinosyl and acetyl substituents of hemicelluloses) were developed. All the models considered sequential, first-order, pseudohomogeneous kinetics, with Arrhenius type dependence on temperature, and provided a satisfactory interpretation of experimental data. 1. Introduction A variety of commercial products currently manufactured from dwindling fossil resources can be obtained from renewable raw materials compatible with sustainable development. In this context, the chemical processing of biomass provides an interesting alternative for manufacturing organic chemicals at an industrial scale. To this end, lignocellulosic materials (LCM) show interesting features, including their large availability, renewability, and low cost (especially in the case of waste materials). LCM are mainly composed of nonstructural components (such as extractives, ashes, or proteins) and structural components. This latter category includes three types of polymers: cellulose (a homopolysaccharide made up of glucose units), hemicelluloses (heteropolysaccharides made up of sugars and nonsaccharide substituents), lignin (a polymer made up of phenyl propane units), and other compounds of minor importance for the objectives of this study. These polymers have an incredible potential to fulfill the energy and chemical needs of industry while minimizing environmental impact and increasing sustainability.1-4 The chemical utilization of LCM can be accomplished by implementing the “biorefinery” concept:5,6 LCM can be treated to obtain separate streams containing the structural components (cellulose, hemicelluloses, and lignin) or products derived from them, which could be used for specific purposes. A first processing step of a biorefinery could be an autohydrolysis (or hydrothermal) treatment, in which an aqueous suspension of the raw material is heated to cause a variety of effects (including extractive removal and dissolution of hemicelluloses). Upon autohydrolysis, hemicelluloses are converted into soluble products by reactions catalyzed by hydronium ions, which come from water autoionization (particularly, in the earlier stages of the reaction), and also from weak acids generated in the reaction medium (principally, acetic acid generated by cleavage of acetyl groups).7,8 Under selected conditions, the major reaction products from hemicelluloses are sugar oligomers,8,9 whereas * Towhomcorrespondenceshouldbeaddressed.Phone:+34988387075. Fax: +34988387001. E-mail:
[email protected]. † Department of Chemical Engineering. ‡ Centro de Investigacio´n, Transferencia e Innovacio´n.
monosaccharides and sugar-decomposition products (as furfural or hydroxymethylfurfural, HMF) can also be present in the media.10 In the case of hardwoods or agricultural materials, the hemicelluloses are mainly composed of xylans, with a backbone of xylose units substituted with arabinosyl moieties or acetyl groups. Xylan-derived oligomers are called xylooligosaccharides (XO).11 Recent studies have been reported on the manufacture of XO by chemical processing of a variety of feedstocks, including crop residues, woods, or industrial residues.12 XO find applications in a variety of fields, including the food industry. When used as food ingredients, XO cause prebiotic effects, improving the intestinal function by enhancing the growth of healthy Bifidobacteria, suppressing the growth of Clostridium, and exerting bacteriostatic effects.13-16 The spent solids from hydrothermal treatments are mainly composed of cellulose and lignin, which suffer little alteration upon autohydrolysis, and remain in solid phase. Both components can be separated in further processing steps, enabling a variety of possible applications, including cellulose pulp,17,18 manufacture of fermentable sugars by enzymatic hydrolysis of cellulose,19,20 or utilization as fuels.21-23 The utilization of residual LCM as raw materials for industry is especially interesting, due to their low cost and the ecological benefit that involves their removal. Acacia dealbata is an invasive, leguminous, fast-growing woody species24,25 with a high content of polysaccharides, which can be fractionated by chemical methods to yield products suitable for the chemical, food, and pharmaceutical industries. Considered as raw materials, leguminous woody species present advantages derived from their ability to aid in the recovery of degraded grounds (by fixing biological nitrogen) and from their role in preventing erosion, increasing soil fertility, and facilitating the establishment and growth of other plant species.26,27 This work deals with the fractionation of Acacia dealbata wood by autohydrolysis, in the scope of the biorefinery concept. The effects of the treatments on the composition of spent solids and liquors from treatments carried out under a variety of operational conditions were measured, and kinetic models providing a reliable interpretation of data were derived from experimental results.
10.1021/ie900233x CCC: $40.75 2009 American Chemical Society Published on Web 05/29/2009
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
2. Experimental Section 2.1. Raw Material and Chemical Characterization. Acacia dealbata wood samples were collected locally, air-dried, milled, screened to pass a 8-mm sieve, homogenized in a single lot, and stored in a dry, dark place until use. Aliquots from the above homogenized lot were subjected to moisture and ethanol extract determination (standard method TAPPI T-264-om-88) and to quantitative acid hydrolysis with 72% sulfuric acid (standard method TAPPI T-249-em-85). The solid residue after hydrolysis was considered as Klason lignin. Hydrolyzates were analyzed for monosaccharides (glucose coming from cellulose, and xylose and arabinose coming from hemicelluloses) and acetic acid (coming from acetyl groups bound to hemicelluloses) by HPLC using an Agilent 1200 series chromatograph (Agilent, CA) with a refractive index detector (temperature, 40 °C). Other analysis conditions: column, Aminex HPX-87H (Biorad, CA); mobile phase, 0.003 mol/L H2SO4; flow rate, 0.6 mL/min. Elemental N was determined with a Thermo Finnegan Flash EATM 1112 analyzer, using 130 and 100 mL/min of He and O2 and an oven temperature of 50 °C. Protein was calculated from the elemental nitrogen content (conversion factor, 6.25 g protein/g nitrogen). Ashes were determined by calcination (standard method TAPPI T-244-om-93 method). 2.2. Hydrothermal Processing of Acacia dealbata Wood and Analytical Determinations. Acacia dealbata wood samples were reacted with water in a batch stainless steel reactor (Parr reactor model 4563M, 600 mL volume, Parr Instruments Company, Moline, IL) with temperature control using a liquid/ solid ratio of 8 kg water/kg oven-dried wood, heated to the desired temperature following the standard temperature profile of the reactor, and then cooled. In the next sections, the experiments are denoted according to the maximal temperatures achieved in each case. Spent solids from treatments were recovered by filtration, washed with water, air-dried, subjected to gravimetric and moisture determination to measure the solid yield and the amount of dissolved substrate, and subjected to the same analytical processing as the raw material. An aliquot of liquors was filtered through 0.45-µm membranes and used for direct HPLC determination of glucose, xylose, arabinose, furfural, HMF, and acetic acid. A second sample of liquors was oven-dried to a constant weight to determine the content in nonvolatile solids. A third sample of liquors was subjected to quantitative posthydrolysis (with 4% sulfuric acid at 121 °C for 40 min) and assayed by HPLC. The increase in the concentrations of monosaccharides and acetic acid caused by posthydrolysis provided a measure of the oligomer concentration and their degree of substitution with acetyl groups.28 The nomenclature employed in the next sections is as follows: • glucooligosaccharides (GO, expressed as g equiv glucan/ 100 g oven-dried raw material), • xylooligosaccharides (XO, expressed as g equiv xylan/100 g oven-dried raw material), • arabinooligosaccharides (ArO, expressed as g equiv arabinan/100 g oven-dried raw material, • acetyl groups bound to oligomers (AcO, expressed as g/100 g oven-dried raw material). 2.3. Fitting of Data. Empirical equations were used to fit the temperature profiles employed in the numerical solving of the differential equations listed below. The set of differential equations was solved by the fourth-order Runge-Kutta method. The preexponential factors and the activation energies were
6619
Table 1. Chemical Composition of Acacia dealbata Wood (data expressed as weight percent of extract-free, oven-dried wood) component
content ( standard deviation
glucan xylan arabinan Klason lignin acetyl groups proteins ashes extracts content others (by difference)
42.4 ( 0.99 16.4 ( 0.28 0.29 ( 0.10 19.3 ( 0.58 3.90 ( 0.04 1.60 ( 0.00 0.50 ( 0.01 5.85 ( 0.09 10.4
calculated by minimizing the sum of the squares of deviations between experimental and calculated data. A commercial optimization routine dealing with Newton’s method (Solver, Microsoft Excel, Microsoft) was employed for this purpose. 3. Results and Discussion 3.1. Chemical Composition of the Raw Material. The average values of nine replicate analyses of Acacia dealbata wood samples are shown in Table 1. The major fraction was cellulose (accounting for 42.4 weight percent of the oven-dried feedstock), followed by hemicelluloses (including xylan, arabinan, and acetyl groups, which accounted jointly by 20.6 weight percent of the o.d. raw material), and acid insoluble residue (19.3% weight percent). Other components (including extractives, proteins or ashes), which are of minor importance for the purposes of this study, were determined to assess the feedstock composition in depth. The experimental results are close to the ones reported for Eucalyptus globulus29 (a typical fast-growing hardwood) and Arundo donax (an invasive species).30 In the case of hardwoods, hemicelluloses are made up of a xylose chain substituted with arabinosyl moieties and acetyl groups.11 The compositional data in Table 1 indicate that the molar ratio xylose:arabinose:acetyl groups was 10:0.18:7.3, a typical result for hardwood acetylglucuronoxylans. Similar mole ratios were reported by Garrote et al.29 for Eucalyptus globulus xylan (molar ratio xylose:arabinose:acetyl groups ) 10:0.33: 6.6). 3.2. Operational Conditions and Variables. Processing with hot, compressed water (also called hydrothermal treatment or autohydrolysis), is an environmentally friendly technology that employs water and LCM as sole reagents. In the present study, a set of nonisothermal autohydrolysis experiments were carried out with Acacia dealbata wood. The media were heated to achieve temperatures in the range 170-240 °C (see Table 2), in order to cover the whole range of operational conditions with practical interest (from low severity conditions, defined by limited xylan conversion, up to maximum severity conditions defined by a significant extent of sugar-decomposition reactions). This experimental domain includes the operational conditions leading to maximal XO production. The effects caused by processing were measured in terms of: (a) Variables related to biomass fractionation: • Solid yield (SY, g spent solids after autohydrolysis/g raw material, oven dry basis), • solubilized fraction (SF, g solubilized material/g raw material, oven dry basis), SF ) 100 - SY
(1)
• Nonvolatile compounds (NVC, g of nonvolatile compounds in liquid phase/g raw material, oven dry basis). • Other compounds (OC, g of other compounds /g raw material, oven dry basis),
6620
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
OC ) SF - NVC
(2)
(b) Variables employed to measure the composition of spent solid (g/g spent solids, oven dry basis): • Glucan content (Gn), xylan content (Xn), arabinan content (Arn), acetyl group content (AG), and Klason lignin content (KL). (c) Variables employed to measure the composition of the liquid phase (g equiv polymer/100 g raw material, oven dry basis): • Glucose concentration (G), xylose concentration (X), arabinose concentration (Ar), acetic acid concentration (AcH), furfural concentration (F), hydroxymethylfurfural concentration (HMF), oligomers concentration (O), calculated as: O ) XO + ArO + GO + AcO
(3)
3.3. Effect of Hydrothermal Treatments on Acacia dealbata Fractionation. The solid yield and the liquor content of nonvolatile solids and other components (variables SY, NVC, and OC) are listed in Table 2. The values of SY varied within the range 0.93-0.55 g/g. A sharp decrease in SY was observed in the temperature range 170-208 °C, which was ascribed to both extractive removal and hemicelluloses solubilization. Increased severities resulted first in an almost constant SY (average value, 0.69 g/g) and then in a drop to reach values of 0.55-0.58 g/g under the severest conditions assayed. The joint contribution of cellulose and lignin was fairly constant (about 0.61 g/g raw material when expressed in terms of untreated wood) along the first part experimental domain and decreased slightly under conditions defined by high severities. The content of nonvolatile compounds (denoted NVC) increased first with the severity of treatments, to reach a plateau of 0.275 g/g at temperatures in the range 205-227 °C, and dropped up to NVC ) 0.16 g/g under the severest operational conditions assayed. The fraction “other compounds” (denoted OC) increased slightly with temperature, with reaching values below 0.1 g/g in experiments carried out at temperatures 217 °C. 3.5. Composition of Liquors from Hydrothermal Treatments. Oligomers. Figure 1a shows the temperature profiles of oligomer concentration. The glucooligosaccharide concentration (GO) varied in the range 1.2 ( 0.4 g equiv glucan/100 oven-dried raw material, accounting for less than 3% of the glucan present in the raw material, without a defined variation pattern with temperature. Xylooligosaccharides (XO) were the most abundant autohydrolysis products and behaved as reaction intermediates: their concentration first increased with temperature (up to reach 11.6 g equiv xylan/100 g raw material at 217 °C, corresponding to 71% of the initial xylan) and then decreased up to reach a concentration of 2.2 g equiv xylan/100 g raw material in the experiment performed at the highest temperature. Arabinooligosaccharides (ArO) presented a variation pattern similar to XO, but the hydrolysis effects were reached at lower temperatures: the maximum concentration (0.19 g equiv arabinan/100 g o.d. raw material, corresponding to 64% of initial arabinan) was achieved at 190 °C, and higher temperatures resulted in concentration marked drops. The acetyl groups bound to oligomers (AcO) presented a behavior closely related to the one described for XO, with a maximum concentration of 2.53 g acetyl groups/100 g raw material (corresponding to the cleavage of 65% of the initial acetyl groups) achieved under conditions of intermediate severity. The concentration of total oligomers (O, calculated as GO + XO + ArO + AcO) reached a maximum (15.4 g/100 oven-dried raw material) at 214 °C. In experiments performed within the temperature range 210-220 °C (corresponding to the optimum zone for oligosaccharide production), the reaction products showed a molar ratio xylose:arabinose:acetyl groups:glucose fairly constant (average value, 10:0.05:6.7:0.94). In comparison with native xylan, the most remarkable feature of oligomers is their low arabinose
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
6621
Figure 2. Composition of autohydrolysis liquors: fractions of oligomers (OF), sugars (SF), other identified compounds (OICF), unknown (UF), and dry weight content of liquors (DW).
Figure 1. (a) Temperature dependence of glucooligosaccharides (GO), xylooligosaccharides (XO), arabinooligosacharides (ArO), acetyl groups bound to oligomers (AcO), and total oligomers (O). (b) Temperature dependence of glucose, xylose (X), arabinose (Ar), acetic acid (AcH), furfural (F), and hydroxymethylfurfural (HMF).
content, which was related to the high susceptibility of these structural units to the hydrolytic degradation. Monomers. Figure 1b shows the concentration of monosaccharides (glucose, xylose, and arabinose), acetic acid, and products from dehydration of pentoses (F) and hexoses (HMF) in the autohydrolysis liquors. As expected from the temperature profiles of glucan, the glucose concentration remained almost constant with temperature, reaching values in the range 2.0-2.7 g equiv glucan/100 g raw material (average value, 2.3 g/100 g). According to ideas stated above for arabinan, the arabinose concentration showed a fast increase up to reach values in the range 0.2-0.3 g equiv arabinan/100 g raw material in the temperature range 200-220 °C and then decreased to become not detectable under harsh conditions. The xylose concentration first increased slowly with temperature and then faster (owing to the increased concentration of xylooligosaccharides), with a final stage where the degradation reactions were predominant, leading to decreased xylose concentrations (up to reach 2.1 g/100 g at 240 °C). The concentrations of acetic acid, furfural, and HMF increased steadily with temperature to reach 3.3 g acetyl groups
equiv/100 g raw material, 2.7 g equiv pentose/100 g raw material, and 0.64 g equiv hexose/100 g raw material at 240 °C, respectively. Interestingly, the concentrations of these compounds under conditions leading to the maximum oligosaccharide concentrations were comparatively low (0.7, 0.3, and 0.1 in the same units, respectively). Under harsh conditions, the concentration of furfural accounted for 16% of the initial pentoses, whereas the acetic acid concentration accounted for 85% of the acetyl groups present in wood. Additional Compounds. In order to assess the presence of additional compounds in liquors, Figure 2 presents the results achieved for the following variables (calculations referred to the raw material, on an oven-dried basis): • Dry weight of liquors (DW, expressed as grams of nonvolatile compounds in liquid phase/g of liquid phase). This variable provides a measure of total dissolved compounds in liquid phase. • Oligomer fraction (OF, expressed as grams of oligomers in NVC/g of NVC), measuring the total amount oligomers. • Sugars fraction (SF, expressed as grams of sugars in NVC/g of NVC), measuring the joint contribution of monosaccharides (glucose, xylose, and arabinose). • Other identified compounds fraction (OICF, measuring the joint contribution of acetic acid, F and HMF, expressed as g /g of NVC). • Unknown compounds (UF), which is defined to account for products different from saccharides, saccharide-degradation products, and acetic acid, calculated by the difference: UF ) 1 - OF - SF - OICF
(4)
DW showed a fast increase from values of about 0.01 g NVC/g liquor up to values in the range 0.030-0.034 g NVC/g liquor, operating at temperatures in the range 202-227 °C (maximum value, 0.0343 g/g at 215 °C), with a further decrease at T > 227 °C to reach 0.02 g/g. OF presented a variation pattern similar to the one described for variable O, reaching maximal values under conditions of intermediate severity (OF about 0.51 g oligomers/g NVC for T in the range 190-227 °C). The sugar fraction SF presented a different variation pattern, with a decrease from low- to medium-severity conditions (caused by arabinose decomposition) followed by a further increase (owing to the
6622
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
Figure 3. Kinetic models: (a) Acacia dealbata wood solubilization: susceptible solid (SYS), nonvolatile compounds (NVC), and other compounds (OC). (b)Glucan autohydrolysis: susceptible glucan (GnS), glucooligosaccharides (GO), glucose (G), and hydroxymethylfurfural (HMF). (c)Xylan and arabinan autohydrolysis: susceptible xylan (XnS), high molecular weight xylooligosaccharides (XOH), low molecular weight xylooligosaccharides (XOL), xylose (X), furfural (F), susceptible arabinan (ArnS), arabinooligosaccharides (ArO), arabinose (Ar), and decomposition products of pentosans (DPPn). (d)Acetyl group autohydrolysis: susceptible acetyl groups bound to hemicelluloses (AcnS), acetyl groups bound to oligomers (AcO) and acetic acid (AcH).
contribution of xylose). OICF varied in a limited range (0.00-0.07 g/g), operating at temperatures below 222 °C, and increased markedly under the severest conditions assayed (up to reach 0.42 g/g) because of the joint contribution of acetyl group cleavage and dehydration of sugar monomers. Finally, the fraction of unknown compounds decreased with temperature from 0.44 g/g at 170 °C to 0.07 g/g at 240 °C. In the experiment leading to maximum oligomer concentration (217 °C), the mass fractions in the liquid phase (expressed as g/g NVC) were as follows: OF ) 0.534, SF ) 0.131, OICF ) 0.051, and UF ) 0.284 g/g. 3.6. Kinetic Modeling. In order to provide further insight on the process, the kinetics of Acacia dealbata autohydrolysis was assessed using the models shown in Figure 3, which were based on the following assumptions: • The raw material is made up to of two fractions with different susceptibility to autohydrolysis. The fraction “susceptible” to hydrolysis is denoted with the subscript “S”, whereas the “resistant” fraction was not hydrolyzed under the conditions tested. • Similarly, the solid yield (SY) is calculated from the contributions of a fraction “susceptible” to be converted in soluble compounds (SYS), and a fraction (denoted SYNS) which remains in solid phase in treatments performed within the experimental domain. • The “susceptible” fraction of the solid can be partially or totally dissolved in treatments, to give nonvolatile compounds (measured by variable NVC). • The nonvolatile compounds can be decomposed to give “other compounds” (OC). • Glucose-containing polymers (making part or hemicelluloses or low molecular weight cellulose) can be decomposed under
the considered operational conditions. Usually, cellulose degradation upon autohydrolysis is only considered for harsher conditions (maximum temperatures of 230-260 °C or higher),31,32 but there is evidence that some glucose-containing polymers can be solubilized at the temperature range considered in this work.33 • The susceptible glucan (GnS) reacts to give glucooligomers (GO), as reported for experiments in acid-catalyzed media.34 • GO decompose into glucose (G), which can be dehydrated to hydroymethylfurfural (HMF). No experimental evidence of HMF decomposition was found. • The susceptible xylan fraction (XnS) reacts to give xylooligosaccharides (XO). The continuous, random xylan breakdown is a complex process that cannot be modeled with a single reaction.30 This complex behavior has been simplified assuming that XnS yields high molecular weight xylooligosaccharides (XOH), which react to give low molecular weight xylooligosaccharides (XOL) in further reaction stages. • XOL are hydrolyzed to give xylose X, which dehydrates to furfural (F). This latter compound undergoes degradation reaction to give decomposition products of pentosans (DPPn). • The whole amount of arabinan and part of the acetyl groups present in the raw material, which appear as substituents of the xylan backbone, react to give arabinose-containing oligomers (ArO) and acetyl groups linked to oligosaccharides (AcO), respectively. • ArO decompose into arabinose (Ar), which can be converted into furfural. • AcO are cleaved to give acetic acid AcH, which remains in the media without further alteration. • All the above reactions can be interpreted by pseudohomogenous, irreversible, first-order kinetics. • The temperature dependence of the kinetic coefficients follows the Arrhenius law. The above assumptions enable the formulation of the equation set listed below: • Equations describing the solubilization of Acacia dealbata wood: dSYS ) -k1SYS dt
(5)
SY ) SYS + (1 - RRS)SYRM
(6)
RSY )
SYS SY
|
RM
(7)
dNVC ) k1SYS - k2NVC dt
(8)
OC ) 100 - SY - NVC
(9)
• Equations describing the behavior of glucan and its reaction products: dGnS ) -k3GnS dt
(10)
Ind. Eng. Chem. Res., Vol. 48, No. 14, 2009
Gn ) GnS + (1 - RGn)GnRM RGn
GnS ) Gn
|
RM
(12)
dGO ) k3GnS - k4GO dt
(13)
dG ) k4GO - k5G dt
(14)
HMF ) GnRM - Gn - GO - G
(15)
• Equations describing the behavior of xylan, arabinan, and their reaction products: dXnS ) -k6XnS dt
(16)
Xn ) XnS + (1 - RXn)XnRM
(17)
RXn )
XnS Xn
|
RM
(
(11)
(18)
dXOH ) k6XnS - k7XOH dt
(19)
dXOL ) k7XOH - k8XOL dt
(20)
XO ) XOH + XOL
(21)
dX ) k8XOL - k9X dt
(22)
dArn ) -k11Arn dt
(23)
dArO ) k11Arn - k12ArO dt
(24)
dAr ) k12ArO - k13Ar dt
(25)
dF ) k9X + k13Ar - k10F dt
(26)
DPPn ) XnRM + ArnRM - Xn - XO - X - F - Arn ArO - Ar (27) • Equations describing the behavior of acetyl groups autohydrolysis:
ki ) k0i exp -
Eai RT(t)
)
6623
(33)
where k0i (h-1) is the pre-exponential factor, Eai (kJ/mol) is the activation energy, T(t) represents the temperature achieved a time t, and Rj (with j ) SY, Gn, Xn, or Acn) is the susceptible fraction (g susceptible fraction/g total fraction). Calculations are based on 100 g oven-dried raw material, and the respective amounts are expressed in the following units: • SY, NVC, and VC as g solid/100 g oven-dried raw material • Gn, GO, G, HMF, and DPGn as g glucan equiv /100 g oven-dried raw material • Xn, Arn, XOH, XOL, XO, ArO, X, Ar, F, and DPPn as g pentosan equiv/100 g oven-dried raw material • Acn, AcO, and AcH as g acetyl group equiv/100 g ovendried raw material The subscript S refers to the susceptible fraction, and the subscript RM refers to the raw material (GnRM ) 42.4 g glucan/ 100 g oven-dried raw material, XnRM ) 16.4 g pentosan/100 g oven-dried raw material, ArnRM ) 0.29 g pentosan/100 g ovendried raw material, and AcnRM ) 3.90 g solid/100 g oven-dried raw material). The set of equations (eq 5 to eq 33) was solved by fitting the temperature profiles to empirical equations (R2 > 0.999) to enable the application of the fourth-order Runge-Kutta method. The initial conditions (corresponding to the lowest temperature, 170 °C) were fixed in the values determined for these conditions (in the above cited units: SY ) 93.1, NVC ) 7.08, OC ) 0.00; Gn ) 39.5, GO ) 1.03, G ) 1.85, HMF ) 0.00; Xn ) 15.7, Arn ) 0.26; XOH ) 0.26, XOL ) 0.00, ArO ) 0.03, X ) 0.36, Ar ) 0.00, F ) 0.06; Acn ) 3.79, AcO ) 0.05, and AcH ) 0.06). Table 3 summarizes the values determined for pre-exponential factors, susceptible fractions, activation energies, and R2, whereas Figures 4 to 7 display both experimental data and model predictions. Concerning the solid yield, the amount of raw material that could be solubilized by hydrothermal processing accounts 0.329 g/g oven-dried wood (variable RSY). This result agrees fairly with the joint contribution of hemicelluloses, ethanol extract, Table 3. Susceptible Fractions (ri), Pre-exponential Factors (k0i), Activation Energies (Eai), and R2 Determined from Data Analysis reaction
coefficient
ln k0i (k0i in h-1)
Eai (kJ/mol)
R2
Solid Solubilization (RSY ) 0.329 g susceptible solid/g oven-dried solid) SYS f NVC NVC f OC
k1 k2
33.6 35.9
119 141
0.955 0.943
Glucan Autohydrolysis (RGn ) 0.079 g susceptible glucan/g glucan)