Comparative study of corn stover pretreated by dilute acid and cellulose solvent-based lignocellulose fractionation: Enzymatic hydrolysis, supramolecular structure, and substrate accessibility.

Abstract

Liberation of fermentable sugars from recalcitrant biomass is among the most costly steps for emerging cellulosic ethanol production. Here we compared two pretreatment methods (dilute acid, DA, and cellulose solvent and organic solvent lignocellulose fractionation, COSLIF) for corn stover. At a high cellulase loading [15 filter paper units (FPUs) or 12.3 mg cellulase per gram of glucan], glucan digestibilities of the corn stover pretreated by DA and COSLIF were 84% at hour 72 and 97% at hour 24, respectively. At a low cellulase loading (5 FPUs per gram of glucan), digestibility remained as high as 93% at hour 24 for the COSLIF-pretreated corn stover but reached only approximately 60% for the DA-pretreated biomass. Quantitative determinations of total substrate accessibility to cellulase (TSAC), cellulose accessibility to cellulase (CAC), and non-cellulose accessibility to cellulase (NCAC) based on adsorption of a non-hydrolytic recombinant protein TGC were measured for the first time. The COSLIF-pretreated corn stover had a CAC of 11.57 m(2)/g, nearly twice that of the DA-pretreated biomass (5.89 m(2)/g). These results, along with scanning electron microscopy images showing dramatic structural differences between the DA- and COSLIF-pretreated samples, suggest that COSLIF treatment disrupts microfibrillar structures within biomass while DA treatment mainly removes hemicellulose. Under the tested conditions COSLIF treatment breaks down lignocellulose structure more extensively than DA treatment, producing a more enzymatically reactive material with a higher CAC accompanied by faster hydrolysis rates and higher enzymatic digestibility.

Full text

Page 1

ARTICLE
Comparative Study of Corn Stover Pretreated by
Dilute Acid and Cellulose Solvent-Based
Lignocellulose Fractionation: Enzymatic
Hydrolysis, Supramolecular Structure,
and Substrate Accessibility
Zhiguang Zhu,1Noppadon Sathitsuksanoh,1,2Todd Vinzant,3Daniel J. Schell,3
James D. McMillan,3Y.-H. Percival Zhang1,2,4
1Biological Systems Engineering Department, Virginia Polytechnic Institute and
State University (Virginia Tech), 210-A Seitz Hall, Blacksburg, Virginia 24061;
telephone: 540-231-7414; fax: 540-231-3199; e-mail: ypzhang@vt.edu
2Institute for Critical Technology and Applied Science (ICTAS),
Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061
3National Bioenergy Center, National Renewable Energy Laboratory, Golden, Colorado
4DOE BioEnergy Science Center (BESC), Oak Ridge, Tennessee 37831
Received 23 October 2008; revision received 11 February 2009; accepted 17 February 2009
Published online 23 February 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/bit.22307
ABSTRACT: Liberation of fermentable sugars from recalci-
trant biomass is among the most costly steps for emerging
cellulosic ethanol production. Here we compared two pre-
treatment methods (dilute acid, DA, and cellulose solvent
and organic solvent lignocellulose fractionation, COSLIF)
for corn stover. At a high cellulase loading [15 filter paper
units (FPUs) or 12.3 mg cellulase per gram of glucan],
glucan digestibilities of the corn stover pretreated by DA
and COSLIF were 84% at hour 72 and 97% at hour 24,
respectively. At a low cellulase loading (5 FPUs per gram of
glucan), digestibility remained as high as 93% at hour 24 for
the COSLIF-pretreated corn stover but reached only ?60%
for the DA-pretreated biomass. Quantitative determinations
of total substrate accessibility to cellulase (TSAC), cellulose
accessibility to cellulase (CAC), and non-cellulose accessi-
bility to cellulase (NCAC) based on adsorption of a non-
hydrolytic recombinant protein TGC were measured for the
first time. The COSLIF-pretreated corn stover had a CAC of
11.57 m2/g, nearly twice that of the DA-pretreated biomass
(5.89 m2/g). These results, along with scanning electron
microscopy images showing dramatic structural differences
between the DA- and COSLIF-pretreated samples, suggest
that COSLIF treatment disrupts microfibrillar structures
within biomass while DA treatment mainly removes hemi-
cellulose. Under the tested conditions COSLIF treatment
breaks down lignocellulose structure more extensively than
DA treatment, producing a more enzymatically reactive
material with a higher CAC accompanied by faster hydro-
lysis rates and higher enzymatic digestibility.
Biotechnol. Bioeng. 2009;103: 715–724.
? 2009 Wiley Periodicals, Inc.
KEYWORDS: biofuels; biomass; cellulose accessibility to
cellulase; cellulose solvent- and organic solvent-based
lignocellulose fractionation (COSLIF); dilute acid pretreat-
ment; substrate accessibility
Introduction
Production of second generation biofuels such as cellulosic
ethanol from renewable lignocellulosic biomass will lead the
bioindustrial revolution necessary to the transition from a
fossil fuel-based economy to a sustainable carbohydrate
economy (Lynd et al., 2008; Zhang, 2008). Use of biofuels
will offer several benefits, including reduced greenhouse
gas emissions, decreased competition with tightening
food supplies, enhanced rural economic development,
and increased national energy security (Demain et al.,
2005; Himmel et al., 2007; Lynd et al., 2002; Zhang et al.,
2006b).
Lignoellulosic biomass, such as agricultural and forestry
residues, municipal and industrial solid wastes, and herba-
ceous and woody bioenergy plants, is a natural complex
composite primarily consisting of three biopolymers:
cellulose, hemicelluloses, and lignin (Fengel and Wegener,
Correspondence to: Y.-H.P. Zhang
Contract grant sponsor: USDA-Sponsored Bioprocessing and Biodesign Center
Contract grant sponsor: DOE BioEnergy Science Center
Contract grant sponsor: U.S. Department of Energy’s Office of the Biomass Program
? 2009 Wiley Periodicals, Inc.
Biotechnology and Bioengineering, Vol. 103, No. 4, July 1, 2009 715

Page 2

1984; Himmel et al., 2007; Lynd et al., 2002; Zhang, 2008).
Efficient, cost-competitive production of fermentable sugars
from recalcitrant biomass remains the largest obstacle to
emerging cellulosic ethanol biorefineries (Lynd et al., 2008;
Wyman, 2007; Zhang, 2008). Biomass saccharification via
biological conversion involves two steps—lignocellulose
pretreatment or fractionation followed by enzymatic
cellulose (and perhaps hemicellulose) hydrolysis. Dilute
acid pretreatment (DA), typically using sulfuric acid, is the
most investigated pretreatment method (Bernardez et al.,
1993; Grethlein, 1985; Grethlein and Converse, 1991; Lloyd
and Wyman, 2005; Ooshima et al., 1990; Schell et al., 2003;
Thompson et al., 1992). Conducted at relatively high
temperatures (150–2008C) and pressures (120–200 psia),
DA pretreatment solublizes acid-labile hemicellulose and
thereby disrupts the lignocellulosic composite linked by
covalent bonds, hydrogen bonds and van der Waals forces
(Burns et al., 1989; Lloyd and Wyman, 2005; Ooshima et al.,
1990). As a result, the condensed lignin remains on the
surface of crystalline cellulose following DA, potentially
hindering subsequent enzymatic hydrolysis (Bernardez
et al., 1993; Jeoh et al., 2007; Liu and Wyman, 2003; Lloyd
and Wyman, 2005). In cellulose solvent and organic solvent
lignocellulose fractionation (COSLIF), a cellulose solvent
(e.g., concentrated phosphoric acid or ionic liquid) enables
the crystalline structure of cellulose to be disrupted. This
type of pretreatment can also be carried out at low
temperatures (e.g., ?508C) and the atmospheric pressure
where sugar degradation is minimized (Moxley et al., 2008;
Zhang et al., 2007a). Subsequent washing steps are used to
fractionate biomass; a first washing with an organic solvent
to remove lignin; and a second washing with water to
removefragmentsofpartiallyhydrolyzedhemicellulose.The
COSLIF approach produces highly reactive amorphous
cellulose, which can be enzymatically hydrolyzed quickly
with high glucan digestibility yield (Moxley et al., 2008;
Zhang et al., 2007a).
The root causes of biomass recalcitrance could be
attributed to a number of factors, such as substrate
accessibility to cellulase, cellulose degree of polymerization
(DP), cellulose crystallinity, lignin content and structure,
and hemicellulose content (Chandra et al., 2007; Himmel
et al., 2007; Kim and Holtzapple, 2005; Zhang and Lynd,
2004; Zhang et al., 2006b). A functionally based mathema-
tical model of fungal enzyme-based enzymatic cellulose
hydrolysis has been developed, accounting for cellulose
characteristics (DP and substrate accessibility) and different
modes of action for endoglucanase and cellobiohydrolase
enzyme system components (Zhang and Lynd, 2006). This
model not only correlates disparate phenomena reported in
the literature but also clearly suggests that low cellulose
accessibility is the most important substrate characteristic
limiting enzymatic hydrolysis rates (Zhang and Lynd, 2006).
More recently, a quantitative assay for determining cellulose
accessibility to cellulase (CAC) has been established based
on adsorption of a non-hydrolytic fusion protein (TGC)
containing a cellulose-binding module and a green
fluorescence protein (Hong et al., 2007). This new approach
more accurately assesses substrate characteristic related to
enzymatic cellulose hydrolysis than traditional methods
such as nitrogen adsorption-based Brunauer–Emmett–
Teller (BET), size exclusion, andsmallangleX-ray scattering
(Hong et al., 2007; Zhang and Lynd, 2004). Regenerated
amorphous cellulose (RAC) that is prepared from micro-
crystalline cellulose (Avicel) has ?20-fold higher CAC
(Hong et al., 2007, 2008b) and exhibits much faster
enzymatic hydrolysis rates than microcrystalline cellulose
(Zhang et al., 2006a), which is in agreement with the model
prediction that increasing CAC is more important for
increasing hydrolysis rates than decreasing DP (Zhang and
Lynd, 2006). The CAC value of Avicel (m2per gram of
Avicel) based on the TGC adsorption was only one-tenth of
that based on the BET method (Marshall and Sixsmith,
1974), implying that about 90% gross surface area measure
based on nitrogen adsorption cannot be accessible to large-
size cellulase protein molecules, at least initially. Traditional
size exclusion techniques are labor intensive and cannot
distinguish the real cellulase binding area (110 face) or
ignore the external surface area (Hong et al., 2007; Zhang
and Lynd, 2004).
Enzymatic hydrolysis of pretreated lignocellulose is more
challenging than enzymatic hydrolysis of pure cellulose
because any remaining lignin and residual hemicellulose
could adsorb cellulase components and thereby block or
impede cellulose hydrolysis (Berlin et al., 2005; Bernardez
et al., 1993; Converse et al., 1990; Grethlein and Converse,
1991; Kurabi et al., 2005; Ooshima et al., 1990; Wyman,
2007). The total substrate accessibility has been measured
previously by using cellulase-size molecule exclusion (Burns
et al., 1989; Esteghlalian et al., 2001; Grethlein, 1985;
Thompson et al., 1992), low-temperature cellulase adsorp-
tion (Gerber et al., 1997; Kumar and Wyman, 2008; Lee
et al., 1994; Lu et al., 2002; Mooney et al., 1998) or labeled
cellulase(Jeoh et al., 2007; Palonen etal.,2004). However,in
pretreated lignocellulose materials it remains relatively
challenging to quantitatively differentiate accessibilities for
cellulose and non-cellulose fractions.
In this study, we compared the enzymatic hydrolysis
behaviors (enzymatic cellulose hydrolysis rates and yields)
of corn stover pretreated by DA and COSLIF approaches.
We also used scanning electron microscopy (SEM) to
examinethe supramolecular structures of DA- and COSLIF-
pretreated corn stover samples. Additionally, we developed
and applied new quantitative assays for substrate accessi-
bility by distinguishing cellulose and non-cellulose fractions
of pretreated lignocellulose.
Materials and Methods
Chemicals and Microorganism
All chemicals were reagent grade, purchased from Sigma
(St. Louis, MO) and Fisher Scientific (Pittsburgh, PA),
unless otherwise noted. Fungal cellulase Spezyme CP was
716
Biotechnology and Bioengineering, Vol. 103, No. 4, July 1, 2009

Page 3

gifted from Genencor (Palo Alto, CA). Novozymes 188
b-glucosidase was purchased from Sigma. Microcrystalline
cellulose (Avicel PH105) was purchased from FMC
(Philadelphia, PA). RAC was prepared through Avicel
dissolution in concentrated phosphoric acid followed by
regeneration in water (Zhang et al., 2006a). Corn stover was
obtained from the National Renewable Energy Laboratory
(NREL, Golden, CO). Corn was grown from biomass
AgriProducts (Harlan, IA). The tub-ground materials were
approximately 9 months old (harvested in fall 2005). The
recombinant thioredoxin–green fluorescent protein–cellu-
lose binding module (TGC) fusion protein was produced in
Escherichia coli BL21 (pNT02) (Hong et al., 2007) and
purified by affinity adsorption on RAC followed by modest
desorption using ethyl glycol (EG) (Hong et al., 2008b). The
EG was removed by using membrane dialysis in a 50 mM
sodium citrate buffer (pH 6.0). The TGC protein solution
was re-concentrated using a 10,000 Da molecular weight
cut-off centrifugal ultrafilter columns (Millipore Co.,
Billerica, MA).
Corn Stover Pretreatments
Dilute sulfuric acid pretreatedcornstover was producedin a
pilot-scale continuous vertical reactor at 1908C, 0.048 g
acid/g dry biomass, 1 min residence time, and a 30% (w/w)
total solid loading by using procedures discussed elsewhere
(Schell et al., 2003). The acidic slurry was stored at 48C prior
to use. Before experiments, the slurry was dewatered and the
pretreated solids were washed with deionized water until the
pH of the washed water reached pH ?6.
COLSIF was conducted as described previously (Moxley
etal.,2008;Zhangetal.,2007a).Onegramofdrycornstover
(particle size >60 mesh and <40 mesh screen, i.e., 0.25–
0.42 mm in diameter) in a 50-mL plastic centrifuge tube was
mixed with 8 mL of 84% phosphoric acid using a glass rod,
and the mixed slurry was then incubated at a 508C water
bath for 20 or 45 min. The reaction was then stopped by an
ice-water bath. Forty milliliters of acetone was then added to
precipitate dissolved cellulose and hemicellulose. The slurry
was spun down for 20 min at room temperature using a
swinging bucket centrifuge operating at 3,600 rpm. The
pellets were re-suspended and washed in 40 mL of acetone
twice more. After three acetone washes, the pellets were
washedtwo more times with water. The residual amorphous
solid pellet was neutralized to pH 5–7 using 2 M sodium
carbonate.
The amounts (dry weight percentages) of sugars and
lignins in the DA- and COSLIF-pretreated corn stover were
measured using a modified quantitative saccharification
method, which can determine acid-labile hemicellulose
composition more accurately (Moxley and Zhang, 2007).
The compositions of ashes and extractives in the intact and
pretreated biomass samples were measured according to
the protocol from NREL LAP-002 (Sluiter et al., 2006). The
protein contents in the biomass samples were measured by
the ninhydrin assay (Starcher, 2001).
Enzymatic Hydrolysis of Pretreated Corn Stover
The pretreated corn stover samples were diluted to 10 g
glucan per liter in a 50 mM sodium citrate buffer (pH 4.8)
supplemented with 0.5 g/L sodium azide for enzymatic
hydrolysis. All hydrolysis experiments were carried out in a
rotary shaker at 250 rpm and 508C. Two enzyme loadings
were used: (1) 15 FPUs cellulase and 30 units Novozyme
188 b-glucosidase per gram of glucan (12.3 mg cellulase and
9.4 mg b-glucosidase per gram of glucan) as well as (2)
5 FPUs cellulase and 30 units b-glucosidase per gram of
glucan (4.1 mg cellulase and 9.4 mg b-glucosidase per gram
of glucan). Eight hundred microliters of evenly-mixing
slurry were taken at different time and centrifuged at
13,000 rpm for 5 min. Exactly 500 mL of the supernatant was
transferred to another microtube and held at room
temperature for 30 min, ensuring conversion of nearly all
cellobiose to glucose by free b-glucosidase. The supernatant
was acidified by adding 30 mL of 10% (w/w) sulfuric acid
and then frozen overnight. The freshly thawed liquid
samples were mixed well and then centrifuged at 13,000 rpm
for 5 min to remove any solid sediments. Glucose
concentration in the clear supernatants was measured by
HPLC using a Bio-Rad HPX-87H column operating at 658C
with a mobile phase of 0.005 M sulfuric acid at a flow rate of
0.6 mL/min (Zhang and Lynd, 2003a, 2005a). After 72-h
hydrolysis, the remaining hydrolysis slurries were trans-
ferred to 50-mL centrifuge tubes and centrifuged at 3,600
rpm for 20 min. After decanting the supernatant, the pellets
werere-suspendedin20mLofwaterandthencentrifuged to
remove soluble sugars. After centrifugation, the remaining
sugars and lignin in the lyophilized pellets were measured by
quantitative saccharification. The soluble glucose and xylose
was measured by HPLC as described above.
Glucan digestibility (XG) at the end of hydrolysis
(hour 72) was calculated using the ratio of soluble glucose
(Gsol) in the supernatant to the sum of Gsoland the residual
glucan expressedintermsofglucoseequivalents(Gres)inthe
solid phase (Eq. 1) (Moxley et al., 2008; Zhang et al., 2007b).
XG¼
Gsol
Gsolþ Gres? 100%(1)
Scanning Electron Microscopy
Supramolecular structures of the intact and pretreated corn
stover samples were examined by scanning electron
microscopes, as described elsewhere (Selig et al., 2007;
Zhang et al., 2006a).
Zhu et al.: Dilute Acid Versus Lignocellulose Fractionation
717
Biotechnology and Bioengineering

Page 4

Protein Mass Concentration Assays
The mass concentrations of the non-adsorbed proteins—
bovine serum albumin (BSA), b-glucosidase, and cellu-
lase—were measured by the BioRad Bradford protein kit
(Richmond, CA) with BSA as the protein standard. Mass
concentration of the non-adsorbed TGC protein was
measured based on fluorescence reading using a BioTek
multi-detection microplate reader, as described elsewhere
(Hong et al., 2007, 2008a).
Protein Adsorption
Adsorption of cellulase, b-glucosidase, and BSA on pure
cellulose samples or birch xylan was conducted at room
temperature for 1 h in a 50 mM sodium citrate buffer at
various concentrations (0.5–100 g Avicel/L, 0.1–10 g RAC/L
or 0.5–50 g birch xylan/L). After centrifugation at 13,000g
for 5 min, the protein concentrations in the supernatant
were measured by the Bradford method, as described
previously (Zhang and Lynd, 2003b, 2005b).
The maximum TGC adsorption capacity was calculated
based on the Langmuir isotherm (i.e., a fixed amount of
adsorbent in terms of various concentrations of TGC).
Eight hundred microliters of pretreated corn stover slurry
solutions containing 1 g glucan/L and a final TGC
concentration from 0.05 to 0.3 g/L was well mixed in a
50 mM sodium citrate buffer (pH 6.0) at room temperature
for 1 h. After centrifugation, the free TGC concentrations
were measured by the BioTek multidetection microplate
reader. Amax,TGCwas calculated based on the maximum
TGC adsorption capacity of the pretreated samples.
Amax,BSA/TGCwas calculated based on the maximum TGC
adsorption capacity of the pretreated samples that had been
blocked by adding an excess of BSA (5 g/L, final) for 1 h
before adding TGC. The BAS blocking was conducted at a
pH 4.80. After BSA blocking, the pH was adjusted back to
6.00 by adding sodium carbonate, and TGC adsorption was
assessed by fluorescence measurement as described above.
Quantitative Substrate Accessibility Determination
Cellulase adsorption on the surface of cellulose can be
described by the Langmuir equation:
Ea¼WmaxKpEf
1 þ KpEf
(2)
in which Ea is adsorbed protein (mmol/L), Wmax the
maximum protein adsorption per L (mmol/L), Efthe free
cellulase (mmol/L), and Kp the dissociation constant
(Kp¼Ea/EFS) in terms of L/g cellulose. The Wmaxand Kp
values in Equation (2) can be calculated by a number of
mathematical data fitting methods.
CAC (m2/g cellulose) has been defined previously (Hong
et al., 2007; Zhang and Lynd, 2004, 2006)
CAC ¼ aAmaxNAAG2
(3)
where a is 21.2 cellobiose lattices occupied by a TGC
molecule (Hong et al., 2007), Amaxthe maximum cellulase
adsorption capacity (mole cellulase/g cellulose), Amaxthe
Wmax/(106?S), S the cellulose concentration (g cellulose/
L), NAthe Avogadro’s constant (6.023?1023molecules/
mol),andAG2theareaof thecellobioselattice inthe110face
(0.53?1.04 nm¼5.512?10?19m2).
Total (biomass) substrate accessibility to cellulase
(TSAC), including CAC and non-cellulose accessibility to
cellulase (NCAC), represents cellulase adsorption capacity
for the entire pretreated biomass. For pure cellulosic
samples, TSAC equals CAC since NCAC equals zero.
For pretreated lignocellulosic biomass, TSAC (m2/g
biomass) can be estimated from direct adsorption of the
TGC protein,
TSAC ¼ aAmax;TGCNAAG2
(4)
whereAmax,TGCisthemaximumTGCadsorptioncapacityof
the biomass (mmol TGC/g biomass).
CAC (m2/g biomass) can be measured based on the
maximum TGC adsorption capacity after first blocking
adsorption by using a large amount of BSA (e.g., 5 g/L) that
can non-specifically bind on the surface of lignin (Berlin
et al., 2005; Yang and Wyman, 2006)
CAC ¼ aAmax;BSA=TGCNAAG2
(5)
where Amax,BSA/TGC is the maximum TGC adsorption
capacity of biomass after BSA blocking (mmol TGC/g
biomass).
Therefore, NCAC (m2/g biomass) can be calculated as
NCAC ¼ TSAC ? CAC
(6)
Results
For COSLIF pretreatment, we have previously found that
(1) phosphoric acid at concentrations beyond the critical
value (>?83%) acts as a cellulose solvent, (2) reaction time
should be sufficient to dissolve biomass but be short enough
to prevent complete hydrolysis, and (3) reaction tempera-
ture is set below 608C for no detectable xylose degradation
(Moxley et al., 2008). The optimal reaction condition for
corn stover is 84% phosphoric acid, 508C, and 45 min.
Although the DA pretreatment conditions used in this study
are known to produce enzymatically digestible material,
they have not been optimized. Table I shows glucan,
hemicellulose, and lignincontents of the COSLIF-pretreated
and DA-pretreated corn stover. COSLIF pretreatment
718
Biotechnology and Bioengineering, Vol. 103, No. 4, July 1, 2009

Page 5

removes more lignin than DA pretreatment, producing a
material with lower levels of residual lignin (19.7% vs.
30.3%, respectively). On the other hand, COSLIF pretreat-
ment removes less hemicellulose than DA pretreatment;
residual levels of hemicellulose are 6.2% and 3.4%,
respectively.
Enzymatic Hydrolysis
Figure 1 shows the glucan digestibility profiles for the corn
stover pretreated by DA and COSLIF at two different
enzyme loadings (A, 15 FPUs cellulase per gram of glucan;
and B, 5 FPUs cellulase per gram of glucan). At the high
enzyme loading, glucan digestibility of the COSLIF-
pretreated corn stover (45 min) was greater than 90% at
hour 12 and reached 97% at hour 24. When COSLIF
pretreatment time was decreased to 20 min, hydrolysis rates
were slower and digestibility was lower. But a long COSLIF
reaction time was not recommended because it resulted in
low solid glucan retention. If concentrated phosphoric acid
completely hydrolyzed cellulose to soluble sugars, cost-
effective separation of soluble sugars and soluble acid would
be challenging, similar to what occurs using concentrated
sulfuric acid for cellulose saccharification (Fengel and
Wegener, 1984; Zhang et al., 2007a). In contrast, DA-
pretreated corn stover exhibited considerably slower
enzymatichydrolysisrates,withglucandigestibilityreaching
84% at hour 72. At a low enzyme loading (5 FPUs per gram
of glucan), final glucan digestibility of the COSLIF-
pretreated biomass was 93% within 24 h, while digestibility
of DA-pretreated biomass only reached 60% at hour 72. The
significant difference in observed enzymatic hydrolysis
behaviors between the COSLIF-pretreated and DA-pre-
treated biomass samples motivated additional studies to
develop a better understanding of the causes.
Supramolecular Structures
The supramolecular structure changes in corn stover before
and after the different pretreatments are shown by using
SEM (Fig. 2). The intact plant cell wall structure of corn
stover shows evidence of plant cell wall vascular bundles and
a highly fibrillar structure (A). Dilute acid pretreatment
disrupts the lignocellulosic structure by mainly dissolving
hemicellulose. As a result, major microfibrous cellulose
structures remain (Fig. 2B) and some lignin or lignin-
carbohydrate complexes may be condensed on the surfaceof
the cellulose fibers. Treatment with concentrated H3PO4
significantly alters the fibrillar structure. A well-pretreated
lignocellulose (corn stover) COSLIF sample (84.0% H3PO4,
508C and 45 min) shows no clear fibrous structure (Fig.2C).
These qualitative images are consistent with the observa-
tions that faster hydrolysis rates and higher glucan
digestibilities are obtained for COSLIF-pretreated biomass
than for DA-pretreated biomass.
Protein Adsorption on Pure Cellulosic Substrates
Adsorption of three proteins (cellulase, b-glucosidase, and
BSA) was conducted on pure crystalline cellulose—Avicel
Table I.
The compositions of corn stover samples.
MethodCellulose (%) Hemicelluloses (%)Total lignin (%) Acid-soluble lignin (%)Klason lignin (%) Ashes (%)Protein (%)
Non-pretreated
COSLIF
DA
38.9?1.9
58.2?2.5
53.7?1.5
28.3?3.0
6.2?0.3
3.4?0.2
18.3?0.9
19.7?0.3
30.3?0.7
7.0?0.5
0.7?0.0
1.2?0.1
11.3?0.5
18.9?0.3
29.2?0.7
6.6?0.2
7.1?0.1
3.4?0.2
5.9?0.5
7.5?0.8
4.3?0.4
Figure 1.
(84%H3PO4,508C,and45minor20min)andDA(1.6%sulfuric acid,1908C,and1min)at
enzyme loadings of (A) 15 FPU cellulaseþ30 IU/g glucan b-glucosidase per gram of
glucan and (B) 5 FPU cellulaseþ30 IU/g glucan b-glycosidase per gram of glucan.
Enzymatic hydrolysis profiles of corn stover pretreated by COSLIF
Zhu et al.: Dilute Acid Versus Lignocellulose Fractionation
719
Biotechnology and Bioengineering

Page 6

(Fig. 3). The free cellulase concentration decreased with
increasingcelluloseconcentrations.Incontrast,therewasno
adsorption of b-glucosidase and BSA because they do not
contain cellulose-binding modules. Similarly, cellulase was
adsorbed by RAC but neither were b-glucosidase and BSA
(data not shown). Because significant cellulose hydrolysis
occurs (especially for amorphous cellulose fraction) during
the active cellulase adsorption process, accompanied by a
change in substrate accessibility (Steiner et al., 1988), we
have proposed to determine CAC based on adsorption of a
non-hydrolysis fusion protein, TGC, containing a cellulose-
binding module (CBM) and a green fluorescent protein
(GFP) (Hong et al., 2007).
Figure 4 shows the adsorption equilibrium curves of
the TGC protein on Avicel and RAC. The maximum
protein adsorption capabilities (Amax) after data fitting were
determined to be 7.38?0.13 mmol TGC per gram of RAC
and 0.32?0.01 mmol TGC per gram of Avicel. The CAC
valueofRAC(51.94?0.91m2pergramofRAC)wasgreater
than 20-fold higher than that of Avicel (2.25?0.07 m2per
gram of Avicel). The effects of adsorption temperature and
substrate concentration on maximum adsorption capacity
were also investigated. No significant change was observed
in Amaxover ranges of 1–5 g RAC/L and 2–50 g Avicel/L
at both room temperature and 508C (data not shown).
In addition, the TGC maximum binding capacity on
hemicellulose was determined to be 0.17 mmol TGC per
gram of birch xylan.
Lignocellulosic Substrate Accessibilities
In order to quantitatively determine pretreated lignocellu-
losic substrate accessibility to cellulase in the presence of
Figure 2.
DA (B), and 45-min COSLIF (C).
SEM images of corn stover before pretreatment (A) and pretreated by
Figure 3.
Adsorption of BSA, b-glucosidase, and cellulase on Avicel.
Figure 4.
and 20 g Avicel/L. The dashed curves were fitted by the Langmuir equations.
Adsorption of TGC fluorescence-tagged fusion protein on 1 g RAC/L
720
Biotechnology and Bioengineering, Vol. 103, No. 4, July 1, 2009

Page 7

residual lignin and hemicellulose(Table I), itis important to
distinguish substrate accessibility for cellulose and non-
cellulose (lignin-rich) fractions. Here we have applied a new
scheme for quantitatively determining CAC and NCAC for
pretreated lignocellulosic substrates (Fig. 5), based on the
facts that (i) BSA can irreversibly bind with the accessible
lignin fraction of lignocellulosic biomass (Berlin et al., 2005;
Yang and Wyman, 2006) and (ii) BSA cannot bind with
cellulose (Fig. 3). Similarly to several substrate accessibility
assays, TSAC can be determined based on one-protein
adsorption, where in this case TGC is used rather than a
hydrolytic cellulase. For CAC measurement, a high con-
centration of BSA (5 g/L, final) was mixed with the
pretreated biomass for blocking accessible lignin, where 5 g
BSA/L was much higher than Amax,lignin?non-cellulose
content?1 g biomass/L prior to the TGC adsorption.
Consequentially, the TGC protein was then added to assess
the maximum adsorption capacity of the blocked pretreated
biomass, and the maximum TGC adsorption capacity of
the BSA-blocked biomass was used to represent CAC.
The difference between TSAC and CAC was NCAC that
represented the accessibility of the non-cellulose (lignin-
rich) fraction. Figure 6 shows TGC adsorption equilibrium
curves obtained using cornstover pretreated byeither DA or
COSLIF with or without BSA blocking. The Amax,TGCand
Amax,BSA/TGC are 2.05?0.15 and 1.64?0.13 mmol per
gram of COSLIF-pretreated biomass and 1.09?0.08 and
0.84?0.05 mmol per gram of DA-pretreated biomass,
respectively.TheimpactofremaininghemicelluloseonCAC
was very low because of low hemicellulose contents in
theCOSLIF-and DA-pretreatedsamples relative tocellulose
contents and low TGC binding capacity of hemicellulose
compared to the Amaxof pretreated cellulose.
Table II shows the TSAC, CAC and NCAC values of the
pretreated corn stover. For COSLIF-pretreated sample,
TSAC was 14.44?1.09 m2per gram of biomass, where CAC
and NCAC were 11.57?0.90 and 2.88?0.20 m2per gram
of biomass, respectively. The TSAC, CAC, and NCAC of
the DA-pretreated sample were 7.66?0.55, 5.89?0.34,
and 1.78?0.09 m2per gram of biomass, respectively.
The normalized CAC of the COSLIF-pretreated and DA-
pretreated biomass were 19.94?1.53 and 10.90?0.63 m2
pergram ofcellulose, respectively.The normalized NCAC of
the COSLIF-pretreated and DA-pretreated biomass were
6.89?0.45 and 3.48?0.19 m2per gram of non-cellulose,
respectively, suggesting that non-cellulose fraction of
the COSLIF-pretreated biomass had higher substrate
accessibility based on mass weight than that of the dilute
acid-pretreated biomass.
Discussion
Hydrolysis results clearly showed that soluble sugars were
released faster and to a greater extent in the COSLIF-
pretreatedcorn stover thanin theDA-pretreatedcorn stover
studied here. Such differences (93% digestibility for COSLIF
samples achieved within 24 h vs. 60% for DA samples within
72 h) were more significant at a low enzyme loading of
5 FPUs. Although intensive efforts have been made to
increase specific cellulase activity and decrease cellulase
production costs (Himmel et al., 2007; Lynd et al., 2008;
Zhang et al., 2006b), additional reductions in enzyme
usage costs are important to promote the economy of
biorefineries.
Beyond different enzymatic hydrolysis characteristics,
these substrates exhibited significant differences in their
supramolecular structures and substrate accessibilities.
Qualitative SEM images clearly indicate that the cellulose
solvent (concentrated phosphoric acid) treatment conducted
Figure 5.
pretreated biomass. A: Direct TGC adsorption for determining TSAC, including the
cellulose and non-cellulose (lignin) fractions, and (B) second TGC adsorption for
determining CAC after BSA blocking for the lignin fraction. [Color figure can be seen in
the online version of this article, available at www.interscience.wiley.com.]
Scheme for quantitative determination of TSAC and CAC for the
Figure 6.
stover with and without BSA blocking. The dashed curves were fitted by the Langmuir
equations.
Adsorption of TGC protein on the DA- and COSLIF-pretreated corn
Zhu et al.: Dilute Acid Versus Lignocellulose Fractionation
721
Biotechnology and Bioengineering

Page 8

at a low temperature substantially disrupts the biomass
fibrillarstructure(Fig.2C),whereasdiluteacidpretreatment
at a higher temperature does not (Fig. 2B). The CAC of the
COSLIF-pretreated corn stover was nearly double that of
DA-pretreated biomass, consistent with the hypothesis that
CAC was one of the most important (rate-limiting) factors
influencing enzymatic hydrolysis (Zhang and Lynd, 2006).
However, while changes in CAC appear to be a major
causative factor for the changes in hydrolysis rates
(especially initial rates), additional factors involving feed-
stock structure and composition undoubtedly contribute to
the increased glucan digestibility (enzymatic hydrolysis
yields) observed in COSLIF-pretreated as compared to
DA-pretreated samples.
Quantitative determination of lignocellulosic substrate
accessibility to enzymes (or chemicals) is important to
understand the mechanism of cellulose hydrolysis and
modeling hydrolysis kinetics (Zhang and Lynd, 2004, 2006).
Such information may help improve the methods for
evaluating pretreatment efficiency. As discussed in the
introduction, total substrate accessibility has been measured
previously using several approaches. The TGC-based CAC
method used here has been previously applied to determine
the CAC of pure cellulosic samples (Hong et al., 2007). This
work found that a transition from substrate excess to
substrate limitation occurred over the course of the process
of enzymatically hydrolyzing crystalline cellulose (Hong
et al., 2007).
Although it is well-known that cellulase adsorption
on cellulose often can be empirically described using a
Langmuir isotherm model (Lynd et al., 2002; Zhang and
Lynd, 2004), Jeoh et al. (2007) attempted to estimate total
substrate accessibility based on the linear range of cellulase
adsorption. This estimation can be good only when a very
low concentration cellulase is used for adsorption experi-
ments because any small variations in free protein
measurement may result in large deviations in calculated
Amaxvalues in the first order equation for approximation of
the Langmuir equation. Another deviation could be large
especially for the easily hydrolyzed pretreated biomass
studied here (COSLIF samples) because some hydrolysis
occurs during the active stage of cellulase adsorption, even
whenexperimentsarecarriedoutatadecreasedtemperature
(Beldman et al., 1987; Ooshima et al., 1983; Steiner et al.,
1988).
But not all of the accessible surface in pretreated
lignocellulosic biomass can be hydrolyzed by cellulase, as
reflected here by the measurement of NCAC. Previous
efforts reported in the literature to assay lignin accessibility
to enzymes (lignin being the dominant component in
the non-cellulose fraction) can be divided into two classes:
(1) extracting lignin from lignocellulosic materials using
chemicals (Bernardez et al., 1993; Converse et al., 1990;
Gerber et al., 1997; Lee et al., 1994; Mooney et al., 1998;
Palonen et al., 2004) and (2) leaving primarily only lignin
remaining by hydrolyzing (dissolving away) the hemicellu-
lose and cellulose fractions (Bernardez et al., 1993; Converse
et al., 1990; Ooshima et al., 1990; Palonen et al., 2004).
Both classes of methods may suffer from the large changes
in lignin (or non-cellulose) substrate properties that occur
during extraction or hydrolysis.
Here we attempted to distinguish cellulose accessibility
andnon-cellulose (lignin-rich) fractionaccessibility without
substrate hydrolysis or lignin extraction. We blocked
accessible lignin by using 5 g BSA/L prior to the TGC
adsorption because BSA can non-specifically irreversibly
bind with lignin (Berlin et al., 2005) and cannot bind with
cellulose (Fig. 3). The results shown in Table II suggest that
TSAC overestimated CAC by 24.8% for the COSLIF-
pretreated biomass (14.44 vs. 11.57 m2/g) and by 30.0% for
the DA-pretreated biomass (7.66 vs. 5.89 m2/g).
DA pretreatment efficiency is often correlated with the
extent of hemicellulose and/or lignin removal (Burns et al.,
1989; Converse et al., 1990; Grethlein and Converse, 1991;
Liu and Wyman, 2003, 2005; Ooshima et al., 1990). But this
comparative study of DA- and COSLIF-pretreated corn
stover suggests that hemicellulose removal efficiency is not a
primary determinant of the rate or extent of glucan
enzymaticdigestibility.Inparticular,theCOSLIF-pretreated
corn stover has nearly twofold higher hemicellulose
composition (6.2%) than has DA-pretreated corn stover
(3.4%) but the former has much higher digestibility and
faster hydrolysis rates. A common belief is that lignin
removal promotes faster and more efficient enzymatic
cellulose hydrolysis (Berlin et al., 2005; Chang and
Holtzapple, 2000; Esteghlalian et al., 2001; Lee et al.,
1994; Mooney et al., 1998; Ohgren et al., 2007; Zhu et al.,
2008), but the data presented here suggest that increasing
CAC is more important for achieving fast hydrolysis
rates and high glucan digestibility because NCAC of the
Table II.
values of corn stover samples before pretreatment and after COSLIF and DA pretreatment.
Total substrate accessibility to cellulase (TSAC), cellulose accessibility to cellulase (CAC), and non-cellulose accessibility to cellulase (NCAC)
TSACCAC NCAC
mmol/g
biomass
m2/g
biomass
mmol/g
biomass
m2/g
biomass
m2/g
cellulose
mmol/g
biomass
m2/g
biomass
m2/g
non-cellulose
Non-pretreated
COSLIF
DA
0.16?0.008
2.05?0.15
1.09?0.08
1.13?0.006
14.44?1.09
7.66?0.55
0.06?0.001
1.64?0.13
0.84?0.05
0.42?0.007
11.57?0.90
5.89?0.34
1.07?0.18
19.94?1.53
10.90?0.63
0.10?0.005
0.41?0.02
0.25?0.03
0.70?0.05
2.88?0.20
1.78?0.09
1.15?0.06
6.89?0.45
3.84?0.19
722
Biotechnology and Bioengineering, Vol. 103, No. 4, July 1, 2009

Page 9

COSLIF-pretreated biomass (2.88 m2/g biomass) is higher
than that of the DA-pretreated biomass (1.78 m2/g
biomass).
In conclusion, these results improved the understanding
of how DA- and COSLIF-pretreatments have different
mechanisms of reducing biomass recalcitrance to enzymatic
hydrolysis. DA pretreatment substantially removes hemi-
cellulose and thereby breaks recalcitrance of biomass.
COSLIF pretreatment partially removes lignin and hemi-
cellulose but also substantially disrupts the fibrillar structure
of biomass. The resulting faster hydrolysis rates and higher
glucan enzymatic digestibility of COSLIF-pretreated corn
stover as compared to DA pretreated corn stover are in a
good agreement with (i) more efficient biomass structure
destruction qualitatively shown by SEM images and
(ii) the almost twofold higher CAC levels measured by
quantitative TGC adsorption. COSLIF pretreatment pro-
ducesmorehighlydigestiblematerialthanDApretreatment,
but similar to DA pretreatment it is not yet commercially
proven and its economic viability for use in large scale
biorefining remains to be demonstrated. The COSLIF
technology remains at an earlier stage of development than
DA pretreatment technology and more detailed economic
analysis based on rigorous Aspen-plus models are needed to
understand its potential for practical applications.
Support for this work was provided to YHPZ from the USDA-
sponsored Bioprocessing and Biodesign Center, DOE BioEnergy
Science Center, DuPont Young Professor Award, ICTAS, and ACS
PRF. TV, DJS, and JDMgratefully acknowledge fundingfrom the U.S.
Department of Energy’s Office of the Biomass Program.
References
Beldman G, Voragen AGJ, Rombouts FM, Searle-van Leeuwen MF, Pilnik
W. 1987. Adsorption and kinetic behavior of purified endoglucanases
andexoglucanasesfrom Trichodermaviride. BiotechnolBioeng30:251–
257.
Berlin A, Gilkes N, Kurabi A, Bura R, Tu M, Kilburn D, Saddler J. 2005.
Weak lignin-binding enzymes: A novel approach to improve activity of
cellulases for hydrolysis of lignocellulosics. Appl Biochem Biotechnol
121/124:163–170.
Bernardez TD, Lyford K, Hogsett DA, Lynd LR. 1993. Adsorption of
clostridium thermocellum cellulases onto pretreated mixed hardwood,
avicel, and lignin. Biotechnol Bioeng 42(7):899–907.
Burns DS, Ooshima H, Converse AO. 1989. Surface-area of pretreated
lignocellulosics as a function of the extent of enzymatic-hydrolysis.
Appl Biochem Biotechnol 20–21:79–94.
Chandra R, Bura R, Mabee W, Berlin A, Pan X, Saddler J. 2007. Substrate
pretreatment: The key to effective enzymatic hydrolysis of lignocellu-
losics? Adv Biochem Eng Biotechnol 108:67–93.
Chang VS, Holtzapple MT. 2000. Fundamental factors affecting biomass
enzymatic reactivity. Appl Biochem Biotechnol 84–86:5–37.
Converse AO, Ooshima H, Burns DS. 1990. Kinetics of enzymatic-hydro-
lysis of lignocellulosic materials based on surface-area of cellulose
accessible to enzyme and enzyme adsorption on lignin and cellulose.
Appl Biochem Biotechnol 24–25:67–73.
DemainAL,NewcombM,WuJHD.2005.Cellulase,clostridia,andethanol.
Microbiol Mol Biol Rev 69:124–154.
Esteghlalian A, Bilodeau M, Mansfield S, Saddler J. 2001. Do enzymatic
hydrolyzability and Simons’ stain reflect the changes in the accessibility
of lignocellulosic substrates to cellulase enzymes? Biotechnol Prog 17:
1049–1054.
Fengel D, Wegener G. 1984. Wood: Chemistry, ultrastructure, reactions.
Berlin: Walter de Gruyter & Co.
Gerber P, Joyce T, Heitmann J, Siika-Aho M, Buchert J. 1997. Adsorption
of a Trichoderma reesei endoglucanase and cellobiohydrolase onto
bleached Kraft fibres. Cellulose 4(4):255–268.
Grethlein HE. 1985. The effect of pore size distribution on the rate of
enzymatic hydrolysis of cellulose substrates. Bio/Technol 3:155–160.
Grethlein HE, Converse AO. 1991. Common aspects of acid prehydrolysis
and steam explosion for pretreating wood. Biores Technol 36:77–82.
Himmel ME, Ding S-Y, Johnson DK, Adney WS, Nimlos MR, Brady JW,
FoustTD.2007.Biomassrecalcitrance:Engineeringplantsandenzymes
for biofuels production. Science 315(5813):804–807.
Hong J, Ye X, Zhang YHP. 2007. Quantitative determination of cellulose
accessibility to cellulase based on adsorption of a nonhydrolytic fusion
protein containing CBM and GFP with its applications. Langmuir
23(25):12535–12540.
Hong J, Wang Y, Ye X, Zhang Y.-HP. 2008a. Simple protein purification
through affinity adsorption on regenerated amorphous cellulose fol-
lowed by intein self-cleavage. J Chromatogr A 1194(2):150–154.
Hong J, Ye X, Wang Y, Zhang Y.-HP. 2008b. Bioseparation of recombinant
cellulose binding module-protein by affinity adsorption on an ultra-
high-capacity cellulosic adsorbent. Anal Chem Acta 621:193–199.
JeohT, IshizawaCI, DavisMF,HimmelME,Adney WS,Johnson DK.2007.
Cellulase digestibility of pretreated biomass is limited by cellulose
accessibility. Biotechnol Bioeng 98:112–122.
Kim S, Holtzapple MT. 2005. Effect of structural features on enzyme diges-
tibility of corn stover. Biores Technol 97:583–591.
Kumar R, Wyman CE. 2008. An improved method to directly estimate
cellulase adsorption on biomass solids. Enzyme Microb Technol
42(5):426–433.
Kurabi A, Berlin A, Gilkes NR, Kilburn DG, Bura R, Robinson J, Markov A,
Skomarovsky A, Gusakov AV, Okunev O, Sinitsyn A, Gregg D, Xie D,
Saddler J. 2005. Enzymatic hydrolysis of steam-exploded and ethanol
organosolv-pretreated Douglas-Firby novel and commercial fungal
cellulases. Appl Biochem Biotechnol 121/124:219–230.
Lee D, Yu AHC, Wong KKY, Saddler JN. 1994. Evaluation of the enzymatic
susceptibility of cellulosic substrates using specific hydrolysis rates and
enzyme adsorption. Appl Biochem Biotechnol 45/46:407–415.
Liu C, Wyman CE. 2003. The effect of flow rate of compressed hot water on
xylan, lignin, and total mass removal from corn stover. Ind Eng Chem
Res 42(21):5409–5416.
LiuC,WymanCE.2005.Partialflowofcompressed-hotwaterthroughcorn
stover to enhance hemicellulose sugar recovery and enzymatic digest-
ibility of cellulose. Biores Technol 96:1978–1985.
Lloyd TA, Wyman CE. 2005. Combined sugar yields for dilute sulfuric acid
pretreatment of corn stover followed by enzymatic hydrolysis of the
remaining solids. Biores Technol 96:1967–1977.
Lu Y, Yang B, Gregg D, Saddler JN, Mansfield SD. 2002. Cellulase adsorp-
tion and an evaluation of enzyme recycle during hydrolysis of steam-
exploded softwood residues. Appl Biochem Biotechnol 98/100:641–
654.
Lynd LR, Weimer PJ, van Zyl WH, Pretorius IS. 2002. Microbial cellulose
utilization: Fundamentals and biotechnology. Microbiol Mol Biol Rev
66:506–577.
Lynd LR, Laser MS, Bransby D, Dale BE, Davison B, Hamilton R, Himmel
M, Keller M, McMillan JD, Sheehan J, Wyman CE. 2008. How biotech
can transform biofuels. Nat Biotechnol 26(2):169–172.
MarshallK,SixsmithD.1974.Somephysicalcharacteristicsofmicrocrystal-
line cellulose 1. Powders for pharmaceutical use. Drug Dev Comm
1:57–71.
Mooney CA, Mansfield SD, Touhy MG, Saddler JN. 1998. The effect of
initial pore volume and lignin content on the enzymatic hydrolysis of
softwoods. Biores Technol 64(2):113–119.
Moxley G, Zhang Y-HP. 2007. More accurate determination of acid-labile
carbohydrate composition in lignocellulose by modified quantitative
saccharification. Energy Fuels 21:3684–3688.
Zhu et al.: Dilute Acid Versus Lignocellulose Fractionation
723
Biotechnology and Bioengineering

Page 10

Moxley GM, Zhu Z, Zhang Y-HP. 2008. Efficient sugar release by the
cellulose solvent based lignocellulose fractionation technology and
enzymatic cellulose hydrolysis. J Agric Food Chem 56(17):7885–
7890.
Ohgren K, Bura R, Saddler J, Zacchi G. 2007. Effect of hemicellulose and
ligninremovalonenzymatichydrolysisofsteampretreatedcornstover.
Biores Technol 98(13):2503–2510.
Ooshima H, Sakata M, Harano Y. 1983. Adsorption of cellulase from
Trichoderma viride on cellulose. Biotechnol Bioeng 25:3103–3114.
Ooshima H, Burns DS, Converse AO. 1990. Adsorption of cellulase from
trichoderma reesei on cellulose and lignacious residue in wood pre-
treated by dilute sulfuric-acid with explosive decompression. Biotech-
nol Bioeng 36(5):446–452.
Palonen H, Tjerneld F, Zacchi G, Tenkanen M. 2004. Adsorption of
Trichoderma reesei CBH I and EG II and their catalytic domains on
steam pretreated softwood and isolated lignin. J Biotechnol 107(1):65–
72.
Schell DJ, Farmer J, Newman M, McMillan JD. 2003. Dilute-sulfuric acid
pretreatment of corn stover in pilot-scale reactor: Investigation of
yields, kinetics, and enzymatic digestibilities of solids. Appl Biochem
Biotechnol 105/108:69–85.
Selig MJ, Viamajala S, Decker SR, Tucker MP, Himmel ME, Vinzant TB.
2007. Deposition of lignin droplets produced during dilute acid pre-
treatment of maize stems retards enzymatic hydrolysis of cellulose.
Biotechnol Prog 23(6):1333–1339.
Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D, Crocker D.
2006. Determination of structural carbohydrates and lignin in biomass.
http://devafdc.nrel.gov/pdfs/9572.pdf. Laboratory Analytic Procedure
LAP-002.
Starcher B. 2001. A ninhydrin-based assay to quantitate the total protein
content of tissue samples. Anal Biochem 292(1):125–129.
Steiner W, Sattler W, Esterbauer H. 1988. Adsorption of Trichoderma reesei
cellulase on cellulose: Experimental data and their analysis by different
equations. Biotechnol Bioeng 32:853–865.
Thompson DN, Chen HC, Grethlein HE. 1992. Comparison of pretreat-
ment methods on the basis of available surface-area. Biores Technol
39(2):155–163.
Wyman CE. 2007. What is (and is not) vital to advancing cellulosic ethanol.
Trends Biotechnol 25(4):153–157.
Yang B, Wyman CE. 2006. BSA treatment to enhance enzymatic hydrolysis
of cellulose in lignin containing substrates. Biotechnol Bioeng 94:611–
617.
Zhang Y-HP. 2008. Reviving the carbohydrate economy via multi-product
biorefineries. J Ind Microbiol Biotechnol 35(5):367–375.
Zhang Y-HP, Lynd LR. 2003a. Cellodextrin preparation by mixed-acid
hydrolysis and chromatographic separation. Anal Biochem 322:225–
232.
Zhang Y-HP, Lynd LR. 2003b. Quantification of cell and cellulase mass
concentrations during anaerobic cellulose fermentation: Development
of an ELISA-based method with application to Clostridium thermo-
cellum batch cultures. Anal Chem 75:219–227.
Zhang Y-HP, Lynd LR. 2004. Toward an aggregated understanding of
enzymatic hydrolysis of cellulose: Noncomplexed cellulase systems.
Biotechnol Bioeng 88:797–824.
Zhang Y-HP, Lynd LR. 2005a. Cellulose utilization by Clostridium thermo-
cellum: Bioenergetics and hydrolysis product assimilation. Proc Natl
Acad Sci USA 102:7321–7325.
ZhangY-HP,LyndLR. 2005b.Regulationofcellulasesynthesisin batchand
continuous cultures of Clostridium thermocellum. J Bacteriol 187:99–
106.
Zhang Y-HP, Lynd LR. 2006. A functionally-based model for hydrolysis of
cellulose by fungal cellulase. Biotechnol Bioeng 94:888–898.
Zhang Y-HP, Cui J-B, Lynd LR, Kuang LR. 2006a. A transition from
cellulose swelling to cellulose dissolution by o-phosphoric acid: Evi-
dences from enzymatic hydrolysis and supramolecular structure. Bio-
macromolecules 7(2):644–648.
Zhang Y-HP, Himmel M, Mielenz JR. 2006b. Outlook for cellulase
improvement: Screening and selection strategies. Biotechnol Adv
24(5):452–481.
Zhang Y-HP, Ding S-Y, Mielenz JR, Elander R, Laser M, Himmel M,
McMillan JD, Lynd LR. 2007a. Fractionating recalcitrant lignocellulose
at modest reaction conditions. Biotechnol Bioeng 97(2):214–223.
Zhang Y-HP, Schell DJ, McMillan JD. 2007b. Methodological analysis for
determination of enzymatic digestibility of cellulosic materials. Bio-
technol Bioeng 96(1):188–194.
ZhuL,O’DwyerJP,ChangVS,GrandaCB,HoltzappleMT.2008.Structural
features affecting biomass enzymatic digestibility. Biores Technol
99(9):3817–3828.
724
Biotechnology and Bioengineering, Vol. 103, No. 4, July 1, 2009