- Email: [email protected]
Effect of structural features on enzyme digestibility of corn stover
Bioresource Technology 97 (2006) 583–591
Effect of structural features on enzyme digestibility of corn stover Sehoon Kim, Mark T. Holtzapple
*
Department of Chemical Engineering, Texas A&M University, College Station, TX 77843-3122, USA Received 28 March 2005; accepted 29 March 2005 Available online 14 June 2005
Abstract Corn stover was pretreated with excess calcium hydroxide (0.5 g Ca(OH)2/g raw biomass) in non-oxidative and oxidative conditions at 25, 35, 45, and 55 C. The enzymatic digestibility of lime-treated corn stover was affected by the change of structural features (acetylation, lignification, and crystallization) resulting from the treatment. Extensive delignification required oxidative treatment and additional consumption of lime (up to 0.17 g Ca(OH)2/g biomass). Deacetylation reached a plateau within 1 week and there were no significant differences between non-oxidative and oxidative conditions at 55 C; both conditions removed approximately 90% of the acetyl groups in 1 week at all temperatures studied. Delignification highly depended on temperature and the presence of oxygen. Lignin and hemicellulose were selectively removed (or solubilized), but cellulose was not affected by lime pretreatment in mild temperatures (25–55 C), even though corn stover was contacted with alkali for a long time, 16 weeks. The degree of crystallinity slightly increased from 43% to 60% with delignification because amorphous components (lignin, hemicellulose) were removed. However, the increased crystallinity did not negatively affect the 3-d sugar yield of enzymatic hydrolysis. Oxidative lime pretreatment lowered the acetyl and lignin contents to obtain high digestibility, regardless of crystallinity. The non-linear models for 3-d hydrolysis yields of glucan (Yg), xylan (Yx), and holocellulose (Ygx) were empirically established as a function of the residual lignin (L) for the corn stover pretreated with lime and air. 2005 Published by Elsevier Ltd. Keywords: Corn stover; Lime pretreatment; Structural features; Acetyl; Lignin; Crystallinity
1. Introduction The enzymatic hydrolysis of lime-treated biomass is affected by structural features resulting from the treatment. Key structural features that affect digestibility are the extent of acetylation, lignification, and crystallization. Lime pretreatment removes amorphous substances (e.g., lignin and hemicellulose), which increases the crystallinity index. Chang and Holtzapple (2000) reported correlations between enzymatic digestibility and three structural factors: lignin content, crystallinity, and acetyl content. They concluded that (1) extensive delignifica*
Corresponding author. Tel.: +1 409 845 9708; fax: +1 409 845 6446. E-mail address: [email protected] (M.T. Holtzapple).
0960-8524/$ - see front matter 2005 Published by Elsevier Ltd. doi:10.1016/j.biortech.2005.03.040
tion is sufficient to obtain high digestibility regardless of acetyl content and crystallinity; (2) delignification and deacetylation remove parallel barriers to enzymatic hydrolysis; and (3) crystallinity significantly affects initial hydrolysis rates but has less effect on ultimate sugar yields. These results indicate that an effective lignocellulose treatment process should remove all the acetyl groups and reduce the lignin content to about 10% in the treated biomass. Further lignin reduction incurs an extra cost; therefore, it is not justified for enzymatic hydrolysis. Lee and Fan (1982) reported that the rate of enzymatic hydrolysis depends on enzyme adsorption and the effectiveness of the adsorbed enzymes, instead of the diffusive mass transfer of enzyme. Lignin removal increases enzyme effectiveness by eliminating
584
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
stover (Kaar and Holtzapple, 2000), switchgrass (Chang and Holtzapple, 2000), and bagasse (Sinitsyn et al., 1991). In this study, the enzymatic digestibility of untreated and lime-treated corn stovers was correlated with three structural features: acetylation, lignification, and crystallinity.
non-productive adsorption site and by increasing access to holocellulose (cellulose + hemicellulose). In hardwoods, the O-acetyl groups are combined with the xylose units, whereas in the softwoods, they are combined with the mannose and glucose units of glucomannans (Whistler and Jeans, 1943). Acetylation sites are maximally 2 positions per anhydroxylose unit. For natural xylan, the degree of acetylation is approximately 1. Deacetylation in alkaline solution increases moisture content (ÔswellingÕ) (Mitchell et al., 1990). Kong et al. (1992) reported that alkalis remove acetyl groups from hemicellulose (mainly xylan) thereby reducing the steric hindrance of hydrolytic enzymes and greatly enhancing carbohydrate digestibility. The removal of acetyl groups from xylan is not mainly affected by swelling, because there are no cation effects among several different types of alkalis. They concluded that the sugar yield in enzymatic hydrolysis is directly associated with acetyl group content, and not with the swelling feature. The acetyl groups of biomass can be cleaved by hydrothermal treatment (autohydrolysis; P170 C in water), because the hydronium ions from water autoionization removes acetyl groups to give acetic acid in the reaction medium (Garrote et al., 2002). The degree of crystallinity of lignocellulosic biomass has been considered an important factor in resisting enzymatic hydrolysis (Chang and Holtzapple, 2000; Puri, 1984; Rivers and Emert, 1989). It has been reported that the particle size of biomass (excluding big chunks) has no effect on enzymatic conversions of corn R1
R2
R3
R4
R5
R6
R7
2. Methods 2.1. Lime pretreatment Corn stover was supplied from NREL (National Renewable Energy Laboratory, Golden, CO) in two different batches (Source: BioMass AgriProducts, Harlan, IA). The stover was already washed, dried, and milled to pass 1/4-in. round screen before being delivered to our laboratory. Corn stover was pretreated with lime (calcium hydroxide) in the presence of water. Packed-bed PVC columns (D · L = 1 in. · 17 in.) were used for the limepretreatment reaction at 25 (ambient temperature), 35, 45, and 55 C. Raw biomass (15.0 g dry weight of corn stover), calcium hydroxide (7.5 g dry weight), and distilled water (150 mL) were transferred into a reactor after thoroughly being mixed using a spatula. After the pretreatment time elapsed (0–16 weeks), the reactors were moved out of the system randomly and cooled down to ambient temperature. Samples were then collected for various analyses. Lime was used in excess to R8
R9
R10
Reactors
Water Jacket
Manifold (N2 or Air)
Manifold (Water)
Frame
Temperature Controller Tray Pump
Drain Water
Temperature sensor
ATM o
C
Tank
Float Heating Element
P
CO2 N2 scrubber or Air
Fig. 1. Schematic diagram of the jacketed reactor system for lime pretreatment in the non-oxidative (N2 supply) and oxidative (air supply) conditions.
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
600 CrI =
500
Intensity (I)
maintain the alkaline pH (P12.0) and to determine the actual amounts of lime consumed during the pretreatment. Nitrogen and air were purged into each reactor to make non-oxidative and oxidative conditions, respectively, after being preheated and saturated in the immersed cylinder in the heating tank. Air was scrubbed of carbon dioxide by passing it through a lime-water slurry in a bottle. A schematic diagram of the whole reactor system is shown in Fig. 1.
585
I002 – Iam
× 100
I002
400 300 200 I002
Iam
2.2. Determination of lime consumed during the pretreatment
100 0
2.3. Estimation of delignification and deacetylation Klason and acid-soluble lignin content in corn stover were determined by NREL standard procedures nos. 03 and 04, respectively. Delignification was defined by the weight fraction of original lignin removed as follows: Delignification ¼ 1 W L
ð1Þ
Here, WL is the fraction of residual lignin expressed as follows: L YT ð2Þ WL ¼ L0 where L and L0 are the Klason lignin contents of treated biomass and raw biomass, respectively, and YT is the pretreatment yield of the total solids determined after the lime pretreatment. Acid-soluble lignin content was not included in this study, because its variation was relatively large and interfered with the accurate estimation of delignification at lower lignin contents. Acetyl content in untreated and treated corn stover was determined as described in the previous report (Kim and Holtzapple, 2005). Deacetylation can be estimated the same way as delignification defined in Eqs. (1) and (2).
10
12
14
16
18
20
22
24
26
2θ (degree) Fig. 2. X-ray diffraction pattern of raw corn stover.
The definition of crystallinity index is I 002 I am CrI ¼ 100 I 002
ð3Þ
where I002 = intensity of the diffraction from the 0 0 2 plane at 2h = 22.6 and Iam = intensity of the background scatter measured at 2h = 18.7 (Segal et al., 1959), as shown in Fig. 2. 2.5. Enzymatic hydrolysis Substrates for the enzyme reaction were the untreated washed-only (control) and the lime-treated, pH-neutralized, and washed corn stovers. After lime pretreatment, the treated samples were neutralized with 5-N HCl to remove the unreacted lime and washed several times (at least 6 times) with distilled water to remove watersoluble components. The substrate concentration was 10 g cellulose/L. Cellulase enzyme (Spezyme CP, Lot 1.0
0.8
Delignification g lignin removed g initial lignin
The amounts of lime unreacted after the pretreatment at each condition were determined by pH neutralization with a standard solution of acid, 5-N HCl (ACS certified). The amounts of lime consumed for pretreatment were estimated by the subtraction of the amounts of lime unreacted from the initial amounts of lime for each pretreatment.
0.6
0.4
0.2
2.4. Crystallinity The crystallinities of untreated and treated corn stovers were measured by the XRD Laboratory, Department of Geophysics, Texas A&M University (College Station, TX) using a Rigaku Powder X-ray Diffractometer (Rigaku Denki Co., Japan). The specimen was scanned at 2/min from 2h = 10 to 26 with a step size of 0.05.
0.0 0.00
0.05
0.10
0.15
0.20
Lime consumption g Ca(OH)2 reacted g raw biomass Fig. 3. The fractional changes of lignin solubilized as a function of the fraction of lime consumed in the non-oxidative condition (m) and oxidative condition (d).
586
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
301-00348-257) was kindly provided by NREL. Excess amounts of b-glucosidase (Novozyme 188, 250 CBU/g of activity) was added to completely convert cellobiose to glucose, i.e., 40 CBU/g cellulose. Cellulase was added at 15 FPU/g cellulose. Cellulase activity was periodically determined as filter paper units per mL, as described by NREL standard procedure no. 06. Citrate buffer (0.05 M, pH 4.8, in reaction mixture) and sodium azide solution (0.03 (w/v)% in reaction mixture) were used to keep constant pH and prevent microbial contamination, respectively. The enzyme reaction was conducted in a shaking incubator at 50 C and 150 rpm for 3 days. The concentration of sugars (glucose and xylose) was determined by HPLC using Aminex HPX-87P column (BioRad, USA) and RI detector (RefractoMonitor III, Model 1109, LDC/MiltonRoy, USA).
3. Results and discussion
more delignification. In the lower range of lime consumption (60.06 g lime reacted/g raw biomass), approximately 50% of original lignin in raw biomass was removed easily. This phenomenon was independent of the presence of oxygen. However, removing the remaining fraction of the lignin required oxidative treatment with additional lime consumption. 3.2. Effect of lime pretreatment on deacetylation and delignification The acetyl group was removed very quickly regardless of temperature and the oxidation condition, whereas lignin was removed gradually throughout the pretreatment and depended on the pretreatment condition. For example, deacetylation reached a plateau within 1 week and there were no significant differences between the extremes (non-oxidative condition at 25 C and oxidative condition at 55 C). However, the delignification trends between these two conditions were very different (Figs. 4–7).
3.1. Effect of lime consumption on delignification 3.3. Effect of crystallinity
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0
Deacetylation (g acetyl removed/g initial acetyl)
a
2
3
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0 3
0.6
0.6
0.4
0.4
0.2
0.2
0.0
a
1.0
2
0.8
0
1.0
1
0.8
4
Pretreatment time (weeks)
0 b
1
Deacetylation (g acetyl removed/g initial acetyl)
0
1.0
Fig. 4. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 25 C.
4
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2 0.0 0
b
0.0
1.0
0.0
4
Pretreatment time (weeks)
1 2 3 Pretreatment time (weeks)
1
2
3
Delignification (g lignin removed/g initial lignin)
0.8
1.0
Delignification (g lignin removed/g initial lignin)
0.8
Corn stover crystallinity increased after lime pretreatment. It was related to the solubilization of lignin and
Deacetylation (g acetyl removed/g initial acetyl)
1.0
Delignification (g lignin removed/g initial lignin)
1.0
Delignification (g lignin removed/g initial lignin)
Deacetylation (g acetyl removed/g initial acetyl)
In alkaline pretreatment, much of the lignin in lignocellulosic biomass can be solubilized by hydroxide ion, OH. As shown in Fig. 3, more lime is needed to get
4
Pretreatment time (weeks)
Fig. 5. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 35 C.
0.6
0.6
0.4
0.4 0.2
0.2
0.0
0.0
Deacetylation (g acetyl removed/g initial acetyl)
a
b
1
2
3
4
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0 1
2
3
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2 0.0 0
a
1.0
0
1.0
0.0
Pretreatment time (weeks)
4
Deacetylation (g acetyl removed/g initial acetyl)
0
1.0
Fig. 6. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 45 C.
hemicellulose—both are amorphous components. Regardless of the oxidative treatment, as delignification proceeded due to lime pretreatment, the xylan (hemicellulose) contents slightly decreased causing the glucan to xylan ratio (G/X) to increase (Fig. 8). This means that lignin and hemicellulose are selectively removed (or solubilized), but cellulose is not significantly affected by lime pretreatment at mild temperatures (25–55 C), even though corn stover was contacted with alkali for a long time, 16 weeks. The degree of crystallinity slightly increased with delignification due to the increase of glucan content in the pretreated solid fraction of corn stover (Fig. 9). 3.4. Effect of deacetylation and delignification on 3-d sugar yield of enzymatic hydrolysis
1
2
3
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0 b
4
1.0
0.0
Pretreatment time (weeks)
Delignification (g lignin removed/g initial lignin)
0.8
587
1 2 3 Pretreatment time (weeks)
Delignification (g lignin removed/g initial lignin)
0.8
Deacetylation (g acetyl removed/g initial acetyl)
1.0
Delignification (g lignin removed/g initial lignin)
1.0
Delignification (g lignin removed/g initial lignin)
Deacetylation (g acetyl removed/g initial acetyl)
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
0.0 4
Fig. 7. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 55 C.
to more than 0.40 g holocellulose hydrolyzed/g holocellulose in treated biomass. Later in the treatment, complete deacetylation was achieved, and the 3-d enzymatic digestibility increased linearly with delignification. Linear relationships between delignification and 3-d enzymatic digestibility at 15 FPU/g cellulose were better for glucan (R2 = 0.7551) than for xylan (R2 = 0.4321), as shown in Fig. 10. The low value of linear regression coefficient for xylan could be due to insufficient xylanase activity in the cellulase enzyme preparation, as discussed in Kim and Holtzapple (2005). These linear relationships were successfully identified between 3-d enzymatic digestibilities of glucan and/or xylan and delignification. 3.5. Proposed model for corn stover
Pretreated corn stover showed high levels of deacetylation and delignification, which increased its enzymatic digestibility. In the early stages of lime pretreatment, when corn stover was highly deacetylated but with little delignification, its 3-d hydrolysis yield of holocellulose at 15 FPU/ g cellulose of enzyme loading increased from 0.25 g holocellulose hydrolyzed/g holocellulose in raw biomass
Chang and Holtzapple (2000) reported that lignin and acetyl groups in hemicellulose are significant barriers for cellulase enzymes to access the lignocellulosic fiber matrix. Also, they reported that crystallinity affects the efficiency of enzyme contacted with cellulose and hemicellulose. Lime pretreatment significantly removes the acetyl and lignin barriers allowing enzyme to access
588
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591 60
Glucan and xylan contents (g/100 g treated biomass)
50 40
Glucan
30 20 Xylan
10 0 0
20
40 60 Delignification
80
100
g lignin removed 100 g lignin in raw biomass
a
sigmoidal as a function of lignin content remaining in lime-treated woody biomass. By their empirical model, 3-d hydrolysis yields increase sigmoidally rather than linearly at the low range of lignin residuals. In our case, lower lignin residuals (higher delignification) were only obtained from the oxidative lime-pretreatment. In this study, with lime-treated corn stover, the crystallinity and acetyl contents did not vary. The acetyl group content was almost 0% and CrI did not change significantly; therefore, it is expected that the 3-d hydrolysis profiles would be sigmoidal with residual lignin (L). The following empirical equations generate the requisite sigmoidal relationships: a0 ð4Þ Yg ¼ a1 L 1 þ exp a2 and
4
3
b0 b1 L 1 þ exp b2
2
70
1
60
0 0
20
40 60 Delignification
80
100
Crystallinity (%)
Ratio of glucan/xylan in lime-pretreated biomass
Yg ¼
50 Treated
40
g lignin removed
b
ð5Þ
100 g lignin in raw biomass
Untreated 30
Fig. 8. Correlation of delignification with holocellulose (glucan and xylan) content (a) and with the ratio of glucan to xylan (b) of limepretreated solid in non-oxidative (m) and oxidative (d) conditions.
0
20
40 60 Delignification
80
100
g lignin removed a
70
Crystallinity (%)
the substrates, cellulose and hemicellulose. Even though the crystallinity is high, the amount of adsorbed enzyme is sufficient to achieve high digestibility in a 3-d enzymatic hydrolysis. Oxidative lime pretreatment lowers the acetyl and lignin contents to obtain high digestibility, regardless of crystallinity. This result agrees with Chang and HoltzappleÕs (2000) observations of lime pretreatment on poplar wood. Using 147 data sets of pretreated poplar wood, Chang and Holtzapple (2000) suggested an empirical formula for hydrolysis yields (Yg,Yx, and Ygx) for glucose, xylose, and total sugar (glucose + xylose) that is a function of the contents of lignin (L), acetyl (A), glucan (G), xylan (X), and crystallinity (CrI). Their data are summarized in Fig. 11, which shows, for fixed values of acetyl content and crystallinity, the profiles of 3-d hydrolysis yield from holocellulose are
100 g lignin in raw biomass
60
50 Treated 40 Untreated 30 30
b
40 50 Glucan content (g glucan/100 g treated biomass)
60
Fig. 9. Correlation of crystallinity with delignification (a) and glucan content (b) of lime-pretreated corn stover in non-oxidative (h) and oxidative (j) conditions.
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
60 R2 = 0.7551 40
20
0
100 g treated biomass
80
100 g holocellulose hydrolyzed
Ygx, 3-d hydrolysis yield of holocellulose
100 g glucan in treated biomass
g glucan hydrolyzed
3-d hydrolysis yield of glucan, Yg ×100
100
80 60
40 20
20
40 60 Delignification
80
100 g xylan in treated biomass
10 15 20 L, Lignin content
25
30
g lignin remaining 100 g treated biomass Fig. 11. 3-d hydrolysis yield of holocellulose as a function of lignin content in lime-treated woody biomass with air (Chang and Holtzapple, 2000).
100
g gxylan hydrolyzed
5
100
100 g lignin in raw biomass
a
3-d hydrolysis yield of xylan, Yx×100
0 3
0
g lignin removed
Table 1 Parameters of correlations for 3-d hydrolysis yields of glucan (Yg, g glucan hydrolyzed/100 g glucan in treated biomass) and xylan (Yx, g xylan hydrolyzed/100 g xylan in treated biomass)
80
60 R2 = 0.4321 40
Parameters
Yg (Eq. (4))
Yx (Eq. (5))
a0 or b0 a1 or b1 a2 or b2
120 13.3 5.7
90 16.3 5.3
20
0
0
20
40 60 Delignification
80
100
g lignin removed 100 g lignin in raw biomass
b
Acetyl (%)
0 0
589
Fig. 10. Effect of delignification on the hydrolysis yields of glucan (a) and xylan (b) in 3-d enzymatic digestibility at 15 FPU/g cellulose.
and Y g ðG=0.90Þ þ Y x ðX =0.88Þ ðG=0.90Þ þ ðX =0.88Þ Yg Yx ¼ þ 0.90X 0.88G 1þ 1þ 0.88G 0.90X
lysis yield of total sugar (g holocellulose hydrolyzed/ 100 g treated biomass); a0–a2 and b0–b2 are constants. The constants (ai and bi, i = 0, 1, and 2) of the simplified models listed in Table 1 were empirically determined from the oxidative lime-pretreatment data by using nonlinear regression for parameter estimation by minimizing the root mean square errors in Excel. The plots of Eqs. (4)–(6) are shown as solid lines in Fig. 12. Thus, for oxidative lime-pretreatment of corn stover, the hydrolysis yields of glucan (Yg), xylan (Yx), and holocellulose (Ygx) were well described by the simplified nonlinear models with the single parameter (L).
Y gx ¼
4. Conclusion ð6Þ
where G is the glucan content in lime-treated corn stover (g glucan/100 g treated biomass); X is the xylan content in lime-treated corn stover (g xylan/100 g treated biomass); L is the lignin content in lime-treated corn stover (g lignin/100 g treated biomass); Yg is the 3-d hydrolysis yield of glucan (g glucan hydrolyzed/100 g treated biomass); Yx is the 3-d hydrolysis yield of xylan (g xylan hydrolyzed/100 g treated biomass); Ygx is the 3-d hydro-
Lime is a very effective chemical for deacetylation and delignification of corn stover. Acetyl groups were removed within 1 week regardless of temperature and the oxidation condition for lime pretreatment, whereas lignin was removed gradually through the whole period of pretreatment and depended on the pretreatment conditions. The hydrolysis yield of glucan and xylan to glucose and xylose was affected by deacetylation and linearly depended on delignification.
590
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
100
80 Yx
60 40 20
100 g xylan in treated biomass
100 g xylan hydrolyzed
100 g glucan in treated biomass
g glucan hydrolyzed
Yg
120
80
60 40
20
0
0 0
a
5
10
L
15
20
0
25
5
10
L
15
g lignin remaining
g lignin remaining
100 g treated biomass
100 g treated biomass
b
20
25
100 g holocellulose in treated biomass
g holocellulose hydrolyzed
Ygx
120 100 80 60 40 20 0 0
5
10
L
15
20
25
g lignin remaining c
100 g treated biomass
Fig. 12. Correlations between the weight fraction of the residual lignin (WL) and 3-d hydrolysis yields: (a) Yg; (b) Yx; (c) Ygx, for corn stover pretreated with lime and air. The enzyme loading rate is 15 FPU/g cellulose. The solid lines show plots of non-linear regressions using Eqs. (4)–(6), respectively.
The degree of crystallinity increased with delignification due to the increase of glucan content in the pretreated solid fraction of corn stover. Oxidative lime pretreatment lowers the acetyl and lignin contents to obtain high digestibility, regardless of crystallinity. Non-linear models for 3-d hydrolysis yields of glucan (Yg), xylan (Yx), and holocellulose (Ygx) were established as a function of the residual lignin content (L) for the corn stover pretreated with lime and air. Acknowledgement This work was supported by the US Department of Agriculture, contract number 00-52104-9663.
References Chang, V.S., Holtzapple, M.T., 2000. Fundamental factors affecting biomass enzymatic reactivity. Applied Biochemistry and Biotechnology 84–86, 5–37. Garrote, G., Domı´nguez, H., Parajo´, J.C., 2002. Autohydrolysis of corncob: study of non-isothermal operation for xylooligosaccharide production. Journal of Food Engineering 52, 211– 218. Kaar, W.E., Holtzapple, M.T., 2000. Using lime pretreatment to facilitate the enzyme hydrolysis of corn stover. Biomass and Bioenergy 18 (3), 189–199. Kim, S., Holtzapple, M.T., 2005. Lime pretreatment and enzymatic hydrolysis of corn stover. Bioresource Technology, in press, doi:10.1016/j.biortech.2005.01.014. Kong, F., Engler, C.R., Soltes, E.J., 1992. Effects of cell-wall acetate, xylan backbone, and lignin on enzymatic hydrolysis of aspen wood. Applied Biochemistry and Biotechnology 34/35, 23–35.
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591 Lee, Y.H., Fan, L.T., 1982. Kinetic studies of enzymatic hydrolysis of insoluble cellulose: analysis of the initial rates. Biotechnology and Bioengineering 24, 2383–2406. Mitchell, D.J., Grohmann, K., Himmel, M.E., 1990. Effect of the degree of acetylation of the enzymatic digestion of acetylated xylans. Journal of Wood Chemistry and Technology 10 (1), 111– 121. NREL standard procedure no. 03. Determination of acid-insoluble lignin in biomass. NREL standard procedure no. 04. Determination of acid-soluble lignin in biomass. NREL standard procedure no. 06. Measurements of cellulase activities.
591
Puri, V.P., 1984. Effect of crystallinity and degree of polymerization of cellulose on enzymatic saccharification. Biotechnology and Bioengineering 26, 1219–1222. Rivers, D.B., Emert, G.H., 1989. Factors affecting the enzymatic hydrolysis of municipal-solid-waste components. Biotechnology and Bioengineering 31, 278–281. Sinitsyn, A.P., Gusakov, A.V., Vlasenko, E.Y., 1991. Effect of structural and physico-chemical features of cellulosic substrates on the efficiency of enzymatic hydrolysis. Applied Biochemistry and Biotechnology 30, 43–59. Whistler, R.L., Jeans, A., 1943. Quantitative estimation of acetyl in carbohydrate acetates. Industrial and Engineering Chemistry 15 (5), 317–318.
Effect of structural features on enzyme digestibility of corn stover Sehoon Kim, Mark T. Holtzapple
*
Department of Chemical Engineering, Texas A&M University, College Station, TX 77843-3122, USA Received 28 March 2005; accepted 29 March 2005 Available online 14 June 2005
Abstract Corn stover was pretreated with excess calcium hydroxide (0.5 g Ca(OH)2/g raw biomass) in non-oxidative and oxidative conditions at 25, 35, 45, and 55 C. The enzymatic digestibility of lime-treated corn stover was affected by the change of structural features (acetylation, lignification, and crystallization) resulting from the treatment. Extensive delignification required oxidative treatment and additional consumption of lime (up to 0.17 g Ca(OH)2/g biomass). Deacetylation reached a plateau within 1 week and there were no significant differences between non-oxidative and oxidative conditions at 55 C; both conditions removed approximately 90% of the acetyl groups in 1 week at all temperatures studied. Delignification highly depended on temperature and the presence of oxygen. Lignin and hemicellulose were selectively removed (or solubilized), but cellulose was not affected by lime pretreatment in mild temperatures (25–55 C), even though corn stover was contacted with alkali for a long time, 16 weeks. The degree of crystallinity slightly increased from 43% to 60% with delignification because amorphous components (lignin, hemicellulose) were removed. However, the increased crystallinity did not negatively affect the 3-d sugar yield of enzymatic hydrolysis. Oxidative lime pretreatment lowered the acetyl and lignin contents to obtain high digestibility, regardless of crystallinity. The non-linear models for 3-d hydrolysis yields of glucan (Yg), xylan (Yx), and holocellulose (Ygx) were empirically established as a function of the residual lignin (L) for the corn stover pretreated with lime and air. 2005 Published by Elsevier Ltd. Keywords: Corn stover; Lime pretreatment; Structural features; Acetyl; Lignin; Crystallinity
1. Introduction The enzymatic hydrolysis of lime-treated biomass is affected by structural features resulting from the treatment. Key structural features that affect digestibility are the extent of acetylation, lignification, and crystallization. Lime pretreatment removes amorphous substances (e.g., lignin and hemicellulose), which increases the crystallinity index. Chang and Holtzapple (2000) reported correlations between enzymatic digestibility and three structural factors: lignin content, crystallinity, and acetyl content. They concluded that (1) extensive delignifica*
Corresponding author. Tel.: +1 409 845 9708; fax: +1 409 845 6446. E-mail address: [email protected] (M.T. Holtzapple).
0960-8524/$ - see front matter 2005 Published by Elsevier Ltd. doi:10.1016/j.biortech.2005.03.040
tion is sufficient to obtain high digestibility regardless of acetyl content and crystallinity; (2) delignification and deacetylation remove parallel barriers to enzymatic hydrolysis; and (3) crystallinity significantly affects initial hydrolysis rates but has less effect on ultimate sugar yields. These results indicate that an effective lignocellulose treatment process should remove all the acetyl groups and reduce the lignin content to about 10% in the treated biomass. Further lignin reduction incurs an extra cost; therefore, it is not justified for enzymatic hydrolysis. Lee and Fan (1982) reported that the rate of enzymatic hydrolysis depends on enzyme adsorption and the effectiveness of the adsorbed enzymes, instead of the diffusive mass transfer of enzyme. Lignin removal increases enzyme effectiveness by eliminating
584
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
stover (Kaar and Holtzapple, 2000), switchgrass (Chang and Holtzapple, 2000), and bagasse (Sinitsyn et al., 1991). In this study, the enzymatic digestibility of untreated and lime-treated corn stovers was correlated with three structural features: acetylation, lignification, and crystallinity.
non-productive adsorption site and by increasing access to holocellulose (cellulose + hemicellulose). In hardwoods, the O-acetyl groups are combined with the xylose units, whereas in the softwoods, they are combined with the mannose and glucose units of glucomannans (Whistler and Jeans, 1943). Acetylation sites are maximally 2 positions per anhydroxylose unit. For natural xylan, the degree of acetylation is approximately 1. Deacetylation in alkaline solution increases moisture content (ÔswellingÕ) (Mitchell et al., 1990). Kong et al. (1992) reported that alkalis remove acetyl groups from hemicellulose (mainly xylan) thereby reducing the steric hindrance of hydrolytic enzymes and greatly enhancing carbohydrate digestibility. The removal of acetyl groups from xylan is not mainly affected by swelling, because there are no cation effects among several different types of alkalis. They concluded that the sugar yield in enzymatic hydrolysis is directly associated with acetyl group content, and not with the swelling feature. The acetyl groups of biomass can be cleaved by hydrothermal treatment (autohydrolysis; P170 C in water), because the hydronium ions from water autoionization removes acetyl groups to give acetic acid in the reaction medium (Garrote et al., 2002). The degree of crystallinity of lignocellulosic biomass has been considered an important factor in resisting enzymatic hydrolysis (Chang and Holtzapple, 2000; Puri, 1984; Rivers and Emert, 1989). It has been reported that the particle size of biomass (excluding big chunks) has no effect on enzymatic conversions of corn R1
R2
R3
R4
R5
R6
R7
2. Methods 2.1. Lime pretreatment Corn stover was supplied from NREL (National Renewable Energy Laboratory, Golden, CO) in two different batches (Source: BioMass AgriProducts, Harlan, IA). The stover was already washed, dried, and milled to pass 1/4-in. round screen before being delivered to our laboratory. Corn stover was pretreated with lime (calcium hydroxide) in the presence of water. Packed-bed PVC columns (D · L = 1 in. · 17 in.) were used for the limepretreatment reaction at 25 (ambient temperature), 35, 45, and 55 C. Raw biomass (15.0 g dry weight of corn stover), calcium hydroxide (7.5 g dry weight), and distilled water (150 mL) were transferred into a reactor after thoroughly being mixed using a spatula. After the pretreatment time elapsed (0–16 weeks), the reactors were moved out of the system randomly and cooled down to ambient temperature. Samples were then collected for various analyses. Lime was used in excess to R8
R9
R10
Reactors
Water Jacket
Manifold (N2 or Air)
Manifold (Water)
Frame
Temperature Controller Tray Pump
Drain Water
Temperature sensor
ATM o
C
Tank
Float Heating Element
P
CO2 N2 scrubber or Air
Fig. 1. Schematic diagram of the jacketed reactor system for lime pretreatment in the non-oxidative (N2 supply) and oxidative (air supply) conditions.
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
600 CrI =
500
Intensity (I)
maintain the alkaline pH (P12.0) and to determine the actual amounts of lime consumed during the pretreatment. Nitrogen and air were purged into each reactor to make non-oxidative and oxidative conditions, respectively, after being preheated and saturated in the immersed cylinder in the heating tank. Air was scrubbed of carbon dioxide by passing it through a lime-water slurry in a bottle. A schematic diagram of the whole reactor system is shown in Fig. 1.
585
I002 – Iam
× 100
I002
400 300 200 I002
Iam
2.2. Determination of lime consumed during the pretreatment
100 0
2.3. Estimation of delignification and deacetylation Klason and acid-soluble lignin content in corn stover were determined by NREL standard procedures nos. 03 and 04, respectively. Delignification was defined by the weight fraction of original lignin removed as follows: Delignification ¼ 1 W L
ð1Þ
Here, WL is the fraction of residual lignin expressed as follows: L YT ð2Þ WL ¼ L0 where L and L0 are the Klason lignin contents of treated biomass and raw biomass, respectively, and YT is the pretreatment yield of the total solids determined after the lime pretreatment. Acid-soluble lignin content was not included in this study, because its variation was relatively large and interfered with the accurate estimation of delignification at lower lignin contents. Acetyl content in untreated and treated corn stover was determined as described in the previous report (Kim and Holtzapple, 2005). Deacetylation can be estimated the same way as delignification defined in Eqs. (1) and (2).
10
12
14
16
18
20
22
24
26
2θ (degree) Fig. 2. X-ray diffraction pattern of raw corn stover.
The definition of crystallinity index is I 002 I am CrI ¼ 100 I 002
ð3Þ
where I002 = intensity of the diffraction from the 0 0 2 plane at 2h = 22.6 and Iam = intensity of the background scatter measured at 2h = 18.7 (Segal et al., 1959), as shown in Fig. 2. 2.5. Enzymatic hydrolysis Substrates for the enzyme reaction were the untreated washed-only (control) and the lime-treated, pH-neutralized, and washed corn stovers. After lime pretreatment, the treated samples were neutralized with 5-N HCl to remove the unreacted lime and washed several times (at least 6 times) with distilled water to remove watersoluble components. The substrate concentration was 10 g cellulose/L. Cellulase enzyme (Spezyme CP, Lot 1.0
0.8
Delignification g lignin removed g initial lignin
The amounts of lime unreacted after the pretreatment at each condition were determined by pH neutralization with a standard solution of acid, 5-N HCl (ACS certified). The amounts of lime consumed for pretreatment were estimated by the subtraction of the amounts of lime unreacted from the initial amounts of lime for each pretreatment.
0.6
0.4
0.2
2.4. Crystallinity The crystallinities of untreated and treated corn stovers were measured by the XRD Laboratory, Department of Geophysics, Texas A&M University (College Station, TX) using a Rigaku Powder X-ray Diffractometer (Rigaku Denki Co., Japan). The specimen was scanned at 2/min from 2h = 10 to 26 with a step size of 0.05.
0.0 0.00
0.05
0.10
0.15
0.20
Lime consumption g Ca(OH)2 reacted g raw biomass Fig. 3. The fractional changes of lignin solubilized as a function of the fraction of lime consumed in the non-oxidative condition (m) and oxidative condition (d).
586
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
301-00348-257) was kindly provided by NREL. Excess amounts of b-glucosidase (Novozyme 188, 250 CBU/g of activity) was added to completely convert cellobiose to glucose, i.e., 40 CBU/g cellulose. Cellulase was added at 15 FPU/g cellulose. Cellulase activity was periodically determined as filter paper units per mL, as described by NREL standard procedure no. 06. Citrate buffer (0.05 M, pH 4.8, in reaction mixture) and sodium azide solution (0.03 (w/v)% in reaction mixture) were used to keep constant pH and prevent microbial contamination, respectively. The enzyme reaction was conducted in a shaking incubator at 50 C and 150 rpm for 3 days. The concentration of sugars (glucose and xylose) was determined by HPLC using Aminex HPX-87P column (BioRad, USA) and RI detector (RefractoMonitor III, Model 1109, LDC/MiltonRoy, USA).
3. Results and discussion
more delignification. In the lower range of lime consumption (60.06 g lime reacted/g raw biomass), approximately 50% of original lignin in raw biomass was removed easily. This phenomenon was independent of the presence of oxygen. However, removing the remaining fraction of the lignin required oxidative treatment with additional lime consumption. 3.2. Effect of lime pretreatment on deacetylation and delignification The acetyl group was removed very quickly regardless of temperature and the oxidation condition, whereas lignin was removed gradually throughout the pretreatment and depended on the pretreatment condition. For example, deacetylation reached a plateau within 1 week and there were no significant differences between the extremes (non-oxidative condition at 25 C and oxidative condition at 55 C). However, the delignification trends between these two conditions were very different (Figs. 4–7).
3.1. Effect of lime consumption on delignification 3.3. Effect of crystallinity
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0
Deacetylation (g acetyl removed/g initial acetyl)
a
2
3
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0 3
0.6
0.6
0.4
0.4
0.2
0.2
0.0
a
1.0
2
0.8
0
1.0
1
0.8
4
Pretreatment time (weeks)
0 b
1
Deacetylation (g acetyl removed/g initial acetyl)
0
1.0
Fig. 4. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 25 C.
4
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2 0.0 0
b
0.0
1.0
0.0
4
Pretreatment time (weeks)
1 2 3 Pretreatment time (weeks)
1
2
3
Delignification (g lignin removed/g initial lignin)
0.8
1.0
Delignification (g lignin removed/g initial lignin)
0.8
Corn stover crystallinity increased after lime pretreatment. It was related to the solubilization of lignin and
Deacetylation (g acetyl removed/g initial acetyl)
1.0
Delignification (g lignin removed/g initial lignin)
1.0
Delignification (g lignin removed/g initial lignin)
Deacetylation (g acetyl removed/g initial acetyl)
In alkaline pretreatment, much of the lignin in lignocellulosic biomass can be solubilized by hydroxide ion, OH. As shown in Fig. 3, more lime is needed to get
4
Pretreatment time (weeks)
Fig. 5. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 35 C.
0.6
0.6
0.4
0.4 0.2
0.2
0.0
0.0
Deacetylation (g acetyl removed/g initial acetyl)
a
b
1
2
3
4
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0 1
2
3
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2 0.0 0
a
1.0
0
1.0
0.0
Pretreatment time (weeks)
4
Deacetylation (g acetyl removed/g initial acetyl)
0
1.0
Fig. 6. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 45 C.
hemicellulose—both are amorphous components. Regardless of the oxidative treatment, as delignification proceeded due to lime pretreatment, the xylan (hemicellulose) contents slightly decreased causing the glucan to xylan ratio (G/X) to increase (Fig. 8). This means that lignin and hemicellulose are selectively removed (or solubilized), but cellulose is not significantly affected by lime pretreatment at mild temperatures (25–55 C), even though corn stover was contacted with alkali for a long time, 16 weeks. The degree of crystallinity slightly increased with delignification due to the increase of glucan content in the pretreated solid fraction of corn stover (Fig. 9). 3.4. Effect of deacetylation and delignification on 3-d sugar yield of enzymatic hydrolysis
1
2
3
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0 b
4
1.0
0.0
Pretreatment time (weeks)
Delignification (g lignin removed/g initial lignin)
0.8
587
1 2 3 Pretreatment time (weeks)
Delignification (g lignin removed/g initial lignin)
0.8
Deacetylation (g acetyl removed/g initial acetyl)
1.0
Delignification (g lignin removed/g initial lignin)
1.0
Delignification (g lignin removed/g initial lignin)
Deacetylation (g acetyl removed/g initial acetyl)
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
0.0 4
Fig. 7. The profiles of deacetylation (j) and delignification (d) during the lime pretreatment in non-oxidative (a) and in oxidative (b) conditions at 55 C.
to more than 0.40 g holocellulose hydrolyzed/g holocellulose in treated biomass. Later in the treatment, complete deacetylation was achieved, and the 3-d enzymatic digestibility increased linearly with delignification. Linear relationships between delignification and 3-d enzymatic digestibility at 15 FPU/g cellulose were better for glucan (R2 = 0.7551) than for xylan (R2 = 0.4321), as shown in Fig. 10. The low value of linear regression coefficient for xylan could be due to insufficient xylanase activity in the cellulase enzyme preparation, as discussed in Kim and Holtzapple (2005). These linear relationships were successfully identified between 3-d enzymatic digestibilities of glucan and/or xylan and delignification. 3.5. Proposed model for corn stover
Pretreated corn stover showed high levels of deacetylation and delignification, which increased its enzymatic digestibility. In the early stages of lime pretreatment, when corn stover was highly deacetylated but with little delignification, its 3-d hydrolysis yield of holocellulose at 15 FPU/ g cellulose of enzyme loading increased from 0.25 g holocellulose hydrolyzed/g holocellulose in raw biomass
Chang and Holtzapple (2000) reported that lignin and acetyl groups in hemicellulose are significant barriers for cellulase enzymes to access the lignocellulosic fiber matrix. Also, they reported that crystallinity affects the efficiency of enzyme contacted with cellulose and hemicellulose. Lime pretreatment significantly removes the acetyl and lignin barriers allowing enzyme to access
588
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591 60
Glucan and xylan contents (g/100 g treated biomass)
50 40
Glucan
30 20 Xylan
10 0 0
20
40 60 Delignification
80
100
g lignin removed 100 g lignin in raw biomass
a
sigmoidal as a function of lignin content remaining in lime-treated woody biomass. By their empirical model, 3-d hydrolysis yields increase sigmoidally rather than linearly at the low range of lignin residuals. In our case, lower lignin residuals (higher delignification) were only obtained from the oxidative lime-pretreatment. In this study, with lime-treated corn stover, the crystallinity and acetyl contents did not vary. The acetyl group content was almost 0% and CrI did not change significantly; therefore, it is expected that the 3-d hydrolysis profiles would be sigmoidal with residual lignin (L). The following empirical equations generate the requisite sigmoidal relationships: a0 ð4Þ Yg ¼ a1 L 1 þ exp a2 and
4
3
b0 b1 L 1 þ exp b2
2
70
1
60
0 0
20
40 60 Delignification
80
100
Crystallinity (%)
Ratio of glucan/xylan in lime-pretreated biomass
Yg ¼
50 Treated
40
g lignin removed
b
ð5Þ
100 g lignin in raw biomass
Untreated 30
Fig. 8. Correlation of delignification with holocellulose (glucan and xylan) content (a) and with the ratio of glucan to xylan (b) of limepretreated solid in non-oxidative (m) and oxidative (d) conditions.
0
20
40 60 Delignification
80
100
g lignin removed a
70
Crystallinity (%)
the substrates, cellulose and hemicellulose. Even though the crystallinity is high, the amount of adsorbed enzyme is sufficient to achieve high digestibility in a 3-d enzymatic hydrolysis. Oxidative lime pretreatment lowers the acetyl and lignin contents to obtain high digestibility, regardless of crystallinity. This result agrees with Chang and HoltzappleÕs (2000) observations of lime pretreatment on poplar wood. Using 147 data sets of pretreated poplar wood, Chang and Holtzapple (2000) suggested an empirical formula for hydrolysis yields (Yg,Yx, and Ygx) for glucose, xylose, and total sugar (glucose + xylose) that is a function of the contents of lignin (L), acetyl (A), glucan (G), xylan (X), and crystallinity (CrI). Their data are summarized in Fig. 11, which shows, for fixed values of acetyl content and crystallinity, the profiles of 3-d hydrolysis yield from holocellulose are
100 g lignin in raw biomass
60
50 Treated 40 Untreated 30 30
b
40 50 Glucan content (g glucan/100 g treated biomass)
60
Fig. 9. Correlation of crystallinity with delignification (a) and glucan content (b) of lime-pretreated corn stover in non-oxidative (h) and oxidative (j) conditions.
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
60 R2 = 0.7551 40
20
0
100 g treated biomass
80
100 g holocellulose hydrolyzed
Ygx, 3-d hydrolysis yield of holocellulose
100 g glucan in treated biomass
g glucan hydrolyzed
3-d hydrolysis yield of glucan, Yg ×100
100
80 60
40 20
20
40 60 Delignification
80
100 g xylan in treated biomass
10 15 20 L, Lignin content
25
30
g lignin remaining 100 g treated biomass Fig. 11. 3-d hydrolysis yield of holocellulose as a function of lignin content in lime-treated woody biomass with air (Chang and Holtzapple, 2000).
100
g gxylan hydrolyzed
5
100
100 g lignin in raw biomass
a
3-d hydrolysis yield of xylan, Yx×100
0 3
0
g lignin removed
Table 1 Parameters of correlations for 3-d hydrolysis yields of glucan (Yg, g glucan hydrolyzed/100 g glucan in treated biomass) and xylan (Yx, g xylan hydrolyzed/100 g xylan in treated biomass)
80
60 R2 = 0.4321 40
Parameters
Yg (Eq. (4))
Yx (Eq. (5))
a0 or b0 a1 or b1 a2 or b2
120 13.3 5.7
90 16.3 5.3
20
0
0
20
40 60 Delignification
80
100
g lignin removed 100 g lignin in raw biomass
b
Acetyl (%)
0 0
589
Fig. 10. Effect of delignification on the hydrolysis yields of glucan (a) and xylan (b) in 3-d enzymatic digestibility at 15 FPU/g cellulose.
and Y g ðG=0.90Þ þ Y x ðX =0.88Þ ðG=0.90Þ þ ðX =0.88Þ Yg Yx ¼ þ 0.90X 0.88G 1þ 1þ 0.88G 0.90X
lysis yield of total sugar (g holocellulose hydrolyzed/ 100 g treated biomass); a0–a2 and b0–b2 are constants. The constants (ai and bi, i = 0, 1, and 2) of the simplified models listed in Table 1 were empirically determined from the oxidative lime-pretreatment data by using nonlinear regression for parameter estimation by minimizing the root mean square errors in Excel. The plots of Eqs. (4)–(6) are shown as solid lines in Fig. 12. Thus, for oxidative lime-pretreatment of corn stover, the hydrolysis yields of glucan (Yg), xylan (Yx), and holocellulose (Ygx) were well described by the simplified nonlinear models with the single parameter (L).
Y gx ¼
4. Conclusion ð6Þ
where G is the glucan content in lime-treated corn stover (g glucan/100 g treated biomass); X is the xylan content in lime-treated corn stover (g xylan/100 g treated biomass); L is the lignin content in lime-treated corn stover (g lignin/100 g treated biomass); Yg is the 3-d hydrolysis yield of glucan (g glucan hydrolyzed/100 g treated biomass); Yx is the 3-d hydrolysis yield of xylan (g xylan hydrolyzed/100 g treated biomass); Ygx is the 3-d hydro-
Lime is a very effective chemical for deacetylation and delignification of corn stover. Acetyl groups were removed within 1 week regardless of temperature and the oxidation condition for lime pretreatment, whereas lignin was removed gradually through the whole period of pretreatment and depended on the pretreatment conditions. The hydrolysis yield of glucan and xylan to glucose and xylose was affected by deacetylation and linearly depended on delignification.
590
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591
100
80 Yx
60 40 20
100 g xylan in treated biomass
100 g xylan hydrolyzed
100 g glucan in treated biomass
g glucan hydrolyzed
Yg
120
80
60 40
20
0
0 0
a
5
10
L
15
20
0
25
5
10
L
15
g lignin remaining
g lignin remaining
100 g treated biomass
100 g treated biomass
b
20
25
100 g holocellulose in treated biomass
g holocellulose hydrolyzed
Ygx
120 100 80 60 40 20 0 0
5
10
L
15
20
25
g lignin remaining c
100 g treated biomass
Fig. 12. Correlations between the weight fraction of the residual lignin (WL) and 3-d hydrolysis yields: (a) Yg; (b) Yx; (c) Ygx, for corn stover pretreated with lime and air. The enzyme loading rate is 15 FPU/g cellulose. The solid lines show plots of non-linear regressions using Eqs. (4)–(6), respectively.
The degree of crystallinity increased with delignification due to the increase of glucan content in the pretreated solid fraction of corn stover. Oxidative lime pretreatment lowers the acetyl and lignin contents to obtain high digestibility, regardless of crystallinity. Non-linear models for 3-d hydrolysis yields of glucan (Yg), xylan (Yx), and holocellulose (Ygx) were established as a function of the residual lignin content (L) for the corn stover pretreated with lime and air. Acknowledgement This work was supported by the US Department of Agriculture, contract number 00-52104-9663.
References Chang, V.S., Holtzapple, M.T., 2000. Fundamental factors affecting biomass enzymatic reactivity. Applied Biochemistry and Biotechnology 84–86, 5–37. Garrote, G., Domı´nguez, H., Parajo´, J.C., 2002. Autohydrolysis of corncob: study of non-isothermal operation for xylooligosaccharide production. Journal of Food Engineering 52, 211– 218. Kaar, W.E., Holtzapple, M.T., 2000. Using lime pretreatment to facilitate the enzyme hydrolysis of corn stover. Biomass and Bioenergy 18 (3), 189–199. Kim, S., Holtzapple, M.T., 2005. Lime pretreatment and enzymatic hydrolysis of corn stover. Bioresource Technology, in press, doi:10.1016/j.biortech.2005.01.014. Kong, F., Engler, C.R., Soltes, E.J., 1992. Effects of cell-wall acetate, xylan backbone, and lignin on enzymatic hydrolysis of aspen wood. Applied Biochemistry and Biotechnology 34/35, 23–35.
S. Kim, M.T. Holtzapple / Bioresource Technology 97 (2006) 583–591 Lee, Y.H., Fan, L.T., 1982. Kinetic studies of enzymatic hydrolysis of insoluble cellulose: analysis of the initial rates. Biotechnology and Bioengineering 24, 2383–2406. Mitchell, D.J., Grohmann, K., Himmel, M.E., 1990. Effect of the degree of acetylation of the enzymatic digestion of acetylated xylans. Journal of Wood Chemistry and Technology 10 (1), 111– 121. NREL standard procedure no. 03. Determination of acid-insoluble lignin in biomass. NREL standard procedure no. 04. Determination of acid-soluble lignin in biomass. NREL standard procedure no. 06. Measurements of cellulase activities.
591
Puri, V.P., 1984. Effect of crystallinity and degree of polymerization of cellulose on enzymatic saccharification. Biotechnology and Bioengineering 26, 1219–1222. Rivers, D.B., Emert, G.H., 1989. Factors affecting the enzymatic hydrolysis of municipal-solid-waste components. Biotechnology and Bioengineering 31, 278–281. Sinitsyn, A.P., Gusakov, A.V., Vlasenko, E.Y., 1991. Effect of structural and physico-chemical features of cellulosic substrates on the efficiency of enzymatic hydrolysis. Applied Biochemistry and Biotechnology 30, 43–59. Whistler, R.L., Jeans, A., 1943. Quantitative estimation of acetyl in carbohydrate acetates. Industrial and Engineering Chemistry 15 (5), 317–318.