Fuel 184 (2016) 527–532

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Bioethanol production from cotton stalk: A comparative study of various

pretreatments
Meixia Wang a,1, Dayun Zhou b,1, Yanqin Wang b, Shoujun Wei b, Weihua Yang b, Meng Kuang b, Lei Ma b,
Dan Fang b, Shuangjiao Xu b, Shuang-kui Du a,⇑
a
College of Food Science and Engineering, Northwest A&F University, Yangling 712100, Shaanxi, China
b
Institute of Cotton Research of Chinese Academy of Agricultural Sciences, State Key Laboratory of Cotton Biology, Anyang 455000, Henan, China

a r t i c l e i n f o a b s t r a c t

Article history: Cotton stalk (CS) is a potential biomass for bioethanol production, but the direct conversion without pre-
Received 1 April 2016 treatment always results in an extremely low yield because of the recalcitrant nature of lignocellulose. In
Received in revised form 14 July 2016 this study, the effects of various methods, i.e. dilute sulfuric acid pretreatment (DSAP), ultrasound-
Accepted 16 July 2016
assisted alkali pretreatment (UAAP), and high pressure-assisted alkali pretreatment (HPAP), on chemical
Available online 20 July 2016
composition, physical structure, and subsequent enzymatic hydrolysis and ethanol fermentation for
bioethanol production from CS have been explored. It was found that the intact structures of pretreated
Keywords:
CS were obviously disrupted. The hemicellulose and lignin of biomass were removed and the crystallinity
Cotton stalk
Dilute sulfuric acid pretreatment
of cellulose increased after pretreatments. HPAP led to the highest reducing sugar and ethanol yields
Ultrasound-assisted alkali pretreatment (271.70 mg g
1 and 45.53%, respectively) compared with UAAP and DSAP. HPAP proved to be a potential
High pressure-assisted alkali pretreatment and effective pretreatment method for ethanol production from CS.
Bioethanol Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction hemicellulose, and lignin bonded to one another, so native CS is

Growing concerns over the global energy shortage and rapid
depletion of fossil fuels associated with environmental damages,
such as global warming, acid rain, and urban smog, have led to
the extensive exploration of alternative and renewable energy
sources [1,2]. Bioethanol produced from lignocellulosic biomass
is one of the most promising biofuels as it is abundant, renewable,
and relatively inexpensive [3,4].
Cotton stalk (CS), a by-product of cotton production, is a renew-
able lignocellulosic biomass. It is rich in cellulose (32–46%) and
hemicellulose (20–28%), which make it a potential raw material
for the conversion of cellulose to ethanol [5]. Globally, more than
12 million hectares of cotton is planted across 80 countries [6].
As the world’s largest producer of cotton, China annually produces
40 million tons of CS [5]. However, a large portion of CS is used as
firewood for household energy needs or burned on the ground,
causing serious environmental pollution and biomass waste [7,8].
Some attempts have been made to investigate the potential of
employing the cellulose of CS for fuel ethanol. CS is a complex
and compact network structure consisting primarily of cellulose,
⇑ Corresponding author.
E-mail address: dushuangkui@hotamil.com (S.-k. Du).
1
The authors contributed equally to this work.
recalcitrant to enzymatic accessibility. Therefore, pretreatment is
essential to change the recalcitrant structure [9] and chemical
composition of biomass to facilitate the production of fermentable
sugars and cellulosic ethanol. Pretreatment is necessary to reduce
production and processing costs.
Through many researchers’ efforts, current existing pretreat-
ment techniques of lignocellulosic biomass, including dilute acid,
alkali, ionic liquid, and biological pretreatment or various combi-
nations, have been extensively investigated in laboratories and
under development [10–14]. Although numerous pretreatment
methods exist, each one has its own advantages and disadvantages.
Various pretreatments are better suited for specific feedstocks [15].
Hence, much effort must be made to develop a cheap, efficient, and
environmentally friendly pretreatment technique for CS.
Recently, acid- and alkali-based pretreatment technologies
were extensively used for lignocellulosic biomass, although both
methods demonstrate distinct action mechanisms for cell wall
destruction. Generally, acid-based pretreatment hydrolyzes
hemicellulose components and exposes cellulose for enzymatic
digestion [16]. Dilute acid, such as dilute sulfuric, nitric, and
hydrochloric acids, is commonly used to cost-effectively and
environmentally friendly pretreat lignocellulosic biomass [13,16].
Various lignocellulosic feedstocks were subjected to dilute acid
pretreatment to enhance the production of fermentable sugars

http://dx.doi.org/10.1016/j.fuel.2016.07.061
0016-2361/Ó 2016 Elsevier Ltd. All rights reserved.
528 M. Wang et al. / Fuel 184 (2016) 527–532
via enzymatic hydrolysis, such as cotton gin trash [6] and mustard
stalk [10]. Kapoor et al. [10] investigated dilute acid, steam
exploded, and alkali pretreated mustard stalk and found that dilute
acid pretreatment was the best methodology in terms of the max-
imum sugar yield at low enzyme loading. Dilute acid is favorable
for industrial applications [15]. Alkali-based pretreatment causes
the breakdown of ester bonds cross-linking lignin and xylan,
removal of lignin, cellulose swelling, and partial decrystallization
of cellulose [10,16]. Among alkali-based pretreatments, sodium
hydroxide has been studied by many researchers. Kaur et al. [8]
reported that the enzymatic hydrolysis of 4% alkali-treated CS
(121 °C, 60 min) after 48 h resulted in 65% of the theoretical glu-
cose yield from cellulose. Silverstein et al. [16] compared four
chemical pretreatment methods (sulfuric acid, sodium hydroxide,
hydrogen peroxide, and ozone pretreatments) for improving sac-
charification of CS, and sodium hydroxide pretreatment resulted
in the highest level of cellulose conversion (60.8%, for 2% NaOH,
90 min, 121 °C/15 psi). However, a single pretreatment method
alone is not feasible in consideration of cost and efficiency. There-
fore, complex pretreatment methods need to be extensively inves-
tigated. Excellent efficiency of hydrolysis can be obtained by
integrating acid or alkali with suitable thermomechanical tech-
niques, such as ultrasound and high pressure.
Ultrasound, a sound wave, can produce energy in the form of
cavitation and agitation in liquid [13], which has the potential to
destroy the surface structure of lignocellulosic biomass. Ultra-
sound has been applied to assist in the pretreatment of various lig-
nocellulosic feedstocks with different reaction solutions [13,17].
Although researches on high-pressure pretreatment of biomass
are limited, some researchers have shown that saccharification is
enhanced efficiently with the help of high pressure. Du et al. [5]
reported that HPAP led to the maximum reducing sugar of
0.293 g/g from CS. The above mentioned studies showed that
ultrasound-assisted alkali and high pressure-assisted alkali are
promising pretreatment methods for CS. As different lignocellu-
losic feedstocks have different physicochemical characteristics,
suitable pretreatment techniques based on the properties of each
raw material must be adopted [15].
CS is one of the most abundant agricultural wastes in China [18]
and has the potential to act as a low-cost feedstock for bioethanol
production. To date, studies on the pretreatment of CS are very lim-
ited and lack depth. Dilute sulfuric acid pretreatment (DSAP),
ultrasound-assisted alkali pretreatment (UAAP), and high
pressure-assisted alkali pretreatment (HPAP) were studied respec-
tively for ethanol production from lignocellulosic materials
[5,6,17]. However, further evaluate on the effectiveness of various
pretreatments on specific biomass was required. The purposes of
this work were to compare the effects of these pretreatments on:
(1) the chemical composition and physical structures of pretreated
CS; (2) reducing sugar yields after enzymatic hydrolysis; (3) ethanol
yields after yeast fermentation. The pretreatments were compara-
tively studied to improve ethanol production from CS. All such
efforts on adding value to CS are becoming increasingly necessary
in tackling environmental pollution and global energy shortage.
2. Materials and methods
2.1. Materials and chemicals
CS (Gossypium hirsutum) was obtained from the Institute of Cot-
ton Research of CAAS. The CS were air dried to reduce the moisture
content to 8–9%, shredded to 1–2 cm, and milled to pass through a
40 mesh screen (0.7 mm) using a sawtooth mill. All samples were
stored in air-tight containers at room temperature for composition
analysis and for further use. All the chemical reagents used were of
analytical grade.
2.2. Pretreatment
Three pretreatments (DSAP, UAAP, and HPAP) were employed
according to the following parameters and untreated CS was used
as the control in this study. All of the experiments were carried out
in duplicate.
2.2.1. DSAP
Approximately 2.00 g of dried and ground cotton stalk powders
were mixed with 50 mL of 3.5% H2SO4 to obtain a solid loading of
4% (w/v, grams dry weight per 100 ml). The mixtures were placed
in 250 mL Erlenmeyer flasks, kept at 135 °C for 2.0 h in a laboratory
autoclave, and cooled. The residual solid biomass was collected by
vacuum filtration using a Buchner funnel lined with filter paper
and washed repeatedly with distilled water to pH of 7. The neutral-
ized residues were dried at 60 °C to a constant weight. After cool-
ing, they were weighed to determine weight loss before and after
pretreatment. Finally, the residues were sealed in polybags and
used for composition analysis and enzymatic hydrolysis.
2.2.2. UAAP
Approximately 2.00 g of dried and ground cotton stalk powders
were mixed with 40 mL of 3.5% NaOH to obtain 5% solid loading.
The mixtures were placed in a sealed conical flask and subjected
to ultrasound pretreatment in an ultrasound cleaning bath (KQ-
700DE, Ultrasound Instrument Co., Ltd., Kunshan, China), which
acted as a thermo-stated ultrasound generator for 90 min with
ultrasound at 420 W, and the temperature was controlled at
25 °C. After pretreatment, the pretreated solid biomass was fil-
tered, washed, dried, and collected.
2.2.3. HPAP
Approximately 2.00 g of dried and ground cotton stalk powders
were mixed with 40 mL of 3.0% NaOH to obtain a solid loading of
5% (w/v). The mixtures were placed in a sealed conical flask and
treated at 121 °C by high pressure using a commercial autoclave
(ES-315, Tomy Kogyo Co., Ltd., Japan) within an operating pressure
of 130 kPa for 40 min. After pretreatment, the pretreated solid bio-
mass was filtered, washed, dried, and collected.
2.3. Enzymatic saccharification of CS
Enzymatic saccharification of untreated or pretreated CS was
carried out using commercial cellulase (activity of 60 ± 3.1 FPU/g,
FPU-Filter Paper Unit) from Shanghai Boao Biotech. Corp., China.
The amount of enzyme used was 30 FPU/g dried substrate. The
hydrolysis was performed in a 150 mL flask containing 25 mL
50 mM sodium acetate buffer (pH 5.0, at room temperature).
0.50 g (dry weight) of pretreated residues was added to the acetate
buffer with a resultant substrate concentration of 2% (w/v). The
mixture containing 10 mM sodium azide to prevent microbial con-
tamination was incubated at 48 °C for 24 h with 120 rpm. After
hydrolysis, the samples were filtered and centrifuged at 3000g
for 10 min to remove unhydrolyzed residues. The reducing sugar
(measured as glucose) content of the supernatant was determined
using the 3,5-dinitrosalicylic acid method [19]. Results were
expressed as mg reducing sugar per g dry biomass using the
following equation:
reducing sugar yield ðmg g
1 dry biomassÞ ¼ ðq  VÞ=m
where q is the concentration (mg/mL) of reducing sugars in the
sample hydrolyzed, V is the total volume (mL) hydrolyzed, m is
the initial dry weight (g) of native or pretreated CS.
 M. Wang et al. / Fuel 184 (2016) 527–532
2.4. Ethanol production using CS hydrolysate
Enzymatic hydrolysate was withdrawn and autoclaved at
121 °C for 15 min. Autoclaved supernatants (20 mL from each sam-
ple) were inoculated with Saccharomyces cerevisiae at a solid load-
ing of 5 g/100 mL for ethanol fermentation. The initial pH was 4.8,
and ammonium sulfate was supplemented as a nitrogen source at
a concentration of 1% (w/v). Incubation was carried out in anaero-
bic stoppered flasks for 48 h at 30 °C. After fermentation, a 5 mL
sample was withdrawn and centrifuged at 12,000 rpm for
15 min. The supernatant was then used for ethanol determination
by an ultraviolet spectrometer (UV-1200 Spectrophotometer,
Beckman Coulter, USA.).
The ethanol yields were calculated on the basis of the theoret-
ical maximum ethanol yield that could be obtained from glucose
released during enzymatic hydrolysis using the following
equation:
Ethanol yield% ¼ ½ðc  VÞ=ðm  0:511Þ  100%
where c is the concentration (mg/mL) of ethanol, as calculated by
standard curve method, V is the total volume (mL) of the fermenta-
tion broth, and m is the initial dry weight (mg) of glucose at the
beginning of fermentation. 0.511 (92/180) is the conversion factor
for glucose to ethanol in the biochemical conversion of the sugar.
2.5. Composition analysis
The cellulose and lignin contents of samples were determined
by the HNO3–ethanol method and 72% (w/w) H2SO4 method
according to Liu [20]. The weight loss ratio of pretreated samples
was estimated by weighing the dried materials before and after
pretreatments.
2.6. Scanning electron microscopy (SEM) analysis
The surface morphology and characteristics of the substrate
were studied using SEM. The untreated and pretreated CS samples
were analyzed according to the procedures of Du et al. [5] using
SEM (JSM-6360LV, Japan Electronics Co., Ltd.).
2.7. Fourier transform infrared spectroscopy (FT-IR) analysis
FT-IR analysis was carried out according to the method intro-
duced by Binod et al. [7] with slight modifications. Approximately
1.0 mg of native and pretreated CS samples was dispersed in
100 mg of spectroscopic grade KBr and subsequently pressed into
disks at 10 MPa for 3 min. FT-IR spectra were obtained within
the spectral range of 400–4000 cm
1 using a Vetex70 type Fourier
Spectrometer (Bruker Optics, Ettlingen, Germany) with a detector
at 4 cm
1 resolution and 25 s scan per sample.
2.8. X-ray diffraction (XRD) analysis
The influence of the pretreatment methods on the cellulose
crystalline structure of CS was analyzed by XRD using a Rigaku
Ultima-IV diffractometer (Japan) X-pert Pro diffractometer oper-
ated at 40 kV and 30 mA with Cu/Ka radiation. Samples were
scanned from 2h = 4° to 65° with a step size of 0.02° at a scanning
speed of 1°/min.
2.9. Statistical analysis
The composition data analysis was carried out by one-way anal-
ysis of variance (ANOVA), followed by the Duncan’s tests multiple
comparison using SPSS Statistics 18.0, and statistical significance
was determined at the 0.05 level (P < 0.05).
529
3. Results and discussion
3.1. Composition analysis
Native CS used in the study contained 39.85% ± 0.42% cellulose
and 23.92% ± 0.30% lignin (Table 1). Haykir et al. [21] also showed a
similar result, in which CS collected from Turkey contained
40.9% ± 0.7% cellulose, 22.7% ± 0.6% acid insoluble lignin, and
2.5% ± 0.5% of acid-soluble lignin. Ververis et al. [22] reported that
CS from Greece contained 40% cellulose and 17% lignin. The
detailed composition of CS has been reported in a previous paper
[16]. The main chemical composition of CS varies depending on
the growth location, season, and harvesting and processing meth-
ods [23].
The composition change and weight loss of CS are important
indices for pretreatment effectiveness (Table 1). Differences in
the cellulose content, lignin content and weight loss rate among
various pretreated samples were significant (P < 0.05). Various pre-
treatments have been shown to lead to significant changes in the
composition of CS [7,16,21]. The cellulose contents after pretreat-
ments ranged from 50.50% ± 0.47% (DSAP sample) to
64.20% ± 0.50% (HPAP sample). Evidently, the cellulose contents
of all pretreated CS were significantly higher than that of native
CS (P < 0.05). The findings suggested that amorphous components
(lignin, hemicellulose, or other ash components) were effectively
removed after pretreatments, which increased the percentage level
of cellulose. Compared with the lignin content of native CS, HPAP
CS exhibited a significant reduction in lignin content
(20.98% ± 0.50%). However, the lignin content was observed to
increase in the DSAP sample compared with the untreated sample,
which was consistent with the report of Silverstein et al. [16]. In
their work, the content of acid insoluble lignin of untreated CS
was 27.9%, while the content was 40.68% after sulfuric acid pre-
treatment (2%, 60 min, 121 °C/15 psi). Dilute sulfuric acid was
incapable of degrading lignin. Acid pretreatments removed hemi-
cellulose from the raw biomass, which in turn increased the rela-
tive lignin content [16]. NaOH is known as an efficient reagent
for removing lignin from lignocellulosic biomass, which was con-
firmed by this study. HPAP resulted in high weight loss rate
(35.94% ± 0.60%), followed by DSAP (28.25% ± 0.32%) and UAAP
(24.04% ± 0.15%). The weight losses of the HPAP and UAAP samples
were mainly caused by delignification, hydrolysis of hemicellulose,
and removal of other non-structural components [24]. However,
the weight loss of the DSAP sample was mainly caused by dissolu-
tion of hemicellulose, dissolution of acid-soluble lignin, and
removal of other components. Among the three pretreatment
methods, HPAP had the strongest ability to remove lignin from CS.
3.2. SEM analysis
SEM images showed that untreated CS (Fig. 1a) had a compact,
rough, and nonuniform outer surface. The outer layer of the stalks
Table 1
The main compositions of cotton stalk from different pretreatments.
Pretreatmentsa Compositions (%)b Weight loss rate (%)
Cellulose Lignin
Untreated 39.85d ± 0.42 23.92b ± 0.30 0
DSAP 50.50c ± 0.47 31.52a ± 0.87 28.25b ± 0.32
UAAP 58.02b ± 0.34 22.22c ± 0.66 24.04c ± 0.15
HPAP 64.20a ± 0.50 20.98d ± 0.50 35.94a ± 0.60
Results are mean ± standard deviations of duplicate analysis. Values followed by
different letter in a column are significantly different (P < 0.05).
a
DSAP-Dilute sulfuric acid pretreatment, UAAP-Ultrasound-assisted alkali pre-
treatment, HPAP-High pressure-assisted alkali pretreatment.
b
Composition percentages are on a dry-weight basis.
530 M. Wang et al. / Fuel 184 (2016) 527–532

Fig. 1. SEM images of native and pretreated cotton stalk at 500magnification. (a) Native cotton stalk; (b) DSAP cotton stalk; (c) UAAP cotton stalk; and (d) HPAP cotton stalk.

was mostly composed of lignin, ash, and hemicellulose that

enclosed the interior cellulose fibers [25]. Compared with native (a) 1.0
CS, some discernible changes were found in the outmost surface
of pretreated CS. DSAP CS (Fig. 1b) resulted in apparent abrasion 0.9
Hydrolysis residue
and splitting of fibers, as well as some layering and scaling. The
reason for this may be due to the decomposition of partial hemicel- 0.8
Transmittance

HPAP CS
lulose after DSAP. After UAAP (Fig. 1c), the surface of cellulose fiber UAAP CS
DSAP CS
had some sunken areas and showed more layering and scaling of 0.7
Native CS
the biomass structures. The slight shock and cavitation effect of 1720
1500 897
the ultrasonic waves not only enhanced lignin removal but also 0.6
2900 1635
1360 650
greatly increased hemicellulose degradation [17]. The HPAP sam- 2350 1423
1140
ple (Fig. 1d) showed more clear holes, cracks, and erosion troughs
0.5
on the surface of cellulose fiber than UAAP sample. The increase in 1087
3370 1037
disintegration within the structure may be correlated with the
0.4
enhanced alkali effects on the biomass structures with the help 3500 3000 2500 2000 1500 1000 500
of high temperature and high pressure. Among all samples, the
Wavenumber (cm-1)
morphological structure of HPAP sample was the most destroyed,
making it the most advantageous for enzymatic hydrolysis. This
observation was in accordance with HPAP having the most lignin (b)
removal. The destruction of the stable structure after various pre-
treatments increased the accessibility of cellulose, which may
Intensity (arbitrary unit)

enhance the effective absorption of enzyme in the interior of cellu-

lose [9]. HPAP biomass was more prone to enzyme attack com-
pared with other pretreated samples. Hydrolysis residue

3.3. FT-IR analysis HPAP CS

UAAP CS
Fig. 2a showed the FT-IR spectra of native, pretreated CS, and
DSAP CS
hydrolysis residue of HPAP CS. The positions of absorption peaks
Native CS
were assigned to chemical components according to related litera-
ture data [26–29]. In the FT-IR spectra, some bands at 1850–
500 cm
1 contained lignin-related information [26]. In comparison 5 10 15 20 25 30 35 40 45 50
2θ
with untreated CS, the absorption of corresponding bands (1720,
1500, 1423, 1360, 1140, 1087, and 650 cm
1) [26,27,29] in these Fig. 2. FT-IR spectra (a) and XRD pattern (b) for native cotton stalk; pretreated
regions was reduced in the pretreated substrate, indicating the cotton stalk samples via DSAP, UAAP, and HPAP; and hydrolysis residue of HPAP CS.
 M. Wang et al. / Fuel 184 (2016) 527–532
removal of lignin after pretreatments. A higher reduction of lignin
content in the HPAP sample than in the DSAP and UAAP samples
was testified by a weak absorption of these bands in the HPAP
spectra. These results were in good agreement with the SEM
results (Fig. 1).
A significant decrease in intensity of both the 897 cm
1 band
corresponding to b-D-cellulose linkages (related to removal of
amorphous cellulose, especially in hemicelluloses), and the band
at 1635–1640 cm
1, which was attributed to the absorbed water
bending vibrations [28], was observed after pretreatments and
enzymatic hydrolysis. This finding indicated the breakage of
b-D-cellulose linkages and obvious removal of hemicellulose
(hemicelluloses usually have a strong affinity for water). The
breakage of b-D-cellulose linkages caused by the degradation of
cellulose allowed cellulose to be easily attacked by cellulase,
thereby improving enzymatic hydrolysis. The band at 897 cm
1
of hydrolysis residue was absent, which was related to the decom-
position of cellulose of CS during enzymatic hydrolysis. The band at
around 1037 cm
1 was attributed to CAO stretching from
guaiacyl-type lignin, hemicellulose, or cellulose [29]. The reduction
of a signal at 1037 cm
1 in the spectra of pretreated samples also
implied the decomposition of xylan (hemicellulose) in the pre-
treatment processes. An intensive and sharp band at around
2350 cm
1 indicated C@O bonds in ketone groups [27]. Reduced
absorption at around 2350 cm
1 also may be attributed to the
partial removal of hemicellulose. In the literature, some bands at
2800–3000 cm
1 were related to the CH stretching, and those at
3550–3100 cm
1 were assigned to the hydrogen-bonded OH
stretching of cellulose [29]. The peaks at 2900 and 3370 cm
1 after
all pretreatments decreased significantly, which may be attributed
to the breakdown of intermolecular hydrogen bonding in cellulose
and hemicellulose. The results might cause the changes of
crystallinity in pretreated biomass. The following XRD analyses
of samples also confirmed this argument.
3.4. XRD analysis
The crystalline structure of feedstock is often considered as one
of the factors that affecting enzymatic hydrolysis [30]. The cellu-
lose crystal features of samples were examined by XRD (Fig. 2b).
No new peaks appeared in pretreated cellulose, implying that pre-
treatments did not change the crystalline allomorph of cellulose.
As depicted in Fig. 2b, native biomass exhibited diffraction peaks
near 15–16°, 22.5°, and 35° (2h) originating from cellulose I, this
finding was similar to those obtained in the study by Reddy and
Yang [25]. The peaks were found to retain their positions and
became increasingly sharper for pretreated biomass. Such an
observation, which implied an increase in the crystallinity of the
pretreated biomass, could be attributed to the increase in the rela-
tive amount of crystalline substance due to the removal of amor-
phous, noncellulose components, such as lignin and hemicellulose.
The corresponding crystallinity index (CI) of each sample was
calculated with the method proposed by Segal et al. [31]. The CI
of native CS was 35.2% and it increased to 40.2%, 45.4%, and
46.7% after DSAP, UAAP and HPAP, respectively (Table 2). However,
the CI decreased obviously after enzymatic hydrolysis (38.9%). This
may be due to the degradation of partial crystal cellulose during
enzyme attack. Kim et al. [14] and Kapoor et al. [10] also reported
increased CI for rice straw after dilute sulfuric acid and aqueous
ammonia pretreatment and for mustard stalk after steam explo-
sion pretreatment. A large proportion of amorphous materials,
such as xylan and lignin, were removed after the pretreatments.
Thus, the portion of the exposed crystalline structure of pretreated
samples was increased compared with the native biomass and the
CI of pretreated CS increased. However, the CI of biologically
pretreated CS was observed to decrease by Pandiyan et al. [27].
531
Table 2
Crystallinity index, reducing sugar yield and ethanol yield of different pretreated
cotton stalks.
Samplesa Crystallinity Reducing sugar Ethanol yield/
index/% yield/mg g
1 theoretical yield (%)
Native CS 35.2 55.90 29.63
DSAP CS 40.2 64.36 26.27
UAAP CS 45.4 167.03 32.57
HPAP CS 46.7 271.70 45.53
Hydrolysis 38.9
residue of
HPAP CS
Results are means of duplicate analysis.
a
DSAP-Dilute sulfuric acid pretreatment, UAAP-Ultrasound-assisted alkali pre-
treatment, HPAP-High pressure-assisted alkali pretreatment.
Chemical and biological pretreatments may have different func-
tions in the pretreatment process. There is no general agreement
on correlation between CI and cellulose hydrolysis. Removal of
lignin would invariably improve enzymatic hydrolysis.
3.5. Comparison of pretreatment methods with respect to their effects
on enzymatic saccharification and ethanol fermentation
The reducing sugar yields after enzymatic hydrolysis and etha-
nol yields after yeast fermentation were shown in Table 2. Reduc-
ing sugar released after enzymatic hydrolysis of untreated CS
(55.90 mg g
1 dry solid) was obviously lower than pretreated sam-
ples (64.36–271.70 mg g
1 dry solids). HPAP CS resulted in the
highest reducing sugar yield (271.70 mg g
1), followed by UAAP
(167.03 mg g
1) and then DSAP (64.36 mg g
1). The reducing sugar
yields of HPAP, UAAP, and DSAP were almost 4.8-fold, 3-fold, and
1.2-fold higher than those of native CS, respectively. DSAP signifi-
cantly removed hemicellulose from CS but displayed lower deligni-
fication, which hindered the release of fermentable sugars. This
was in good accordance with the report of Silverstein et al. [16].
After 48 h fermentation, untreated CS resulted in an ethanol
yield 29.63%. Overall ethanol yields of 26.27%, 32.57%, and
45.53% were obtained via DSAP, UAAP, and HPAP, respectively.
These results clearly implied the advantage of HPAP over DSAP
and UAAP for efficient conversion of CS to ethanol. HPAP is an
effective method for CS and requires further study.
3.6. Comparison with other studies
In this study, the HPAP of CS resulted in 271.70 mg g
1 reducing
sugar after 24 h enzymatic hydrolysis, which was higher than
fungal-pretreated CS of 10.91–55.6 mg g
1 and untreated stalk of
67.04 mg g
1 [13]. A higher reducing sugar yield of 0.495 g g
1
was obtained by alkali assisted microwave pretreatment [33].
The high level of sugar yield could be attributed to the lower lignin
content (only 2.12%) retained in solid residues after pretreatment
compared to this study (20.98%). However, the cellulose content
of HPAP CS (64.20%, this study) was obviously higher than pre-
treated cotton plant residue (42.37%) by alkali assisted microwave
on the premise that the two untreated biomass were almost the
same cellulose content [33]. Therefore, HPAP CS has a potential
to release much more reducing sugars if the lignin content is
reduced to the same level (2.12%). Further investigations on delig-
nification of HPAP CS are desirable.
The HPAP CS resulted in 45.53% ethanol yield in this study. This
value was higher than the ethanol yield (31.6%) obtained by ultra-
sonication, liquid hot water, and ligninolytic enzymes pretreat-
ment [32], but was lower than microbial pretreatment (60–85%)
[12]. However, reducing sugar yield obtained by microbial pre-
treatment was only 10.91–55.6 mg g
1 [12]. The absolute ethanol
532 M. Wang et al. / Fuel 184 (2016) 527–532
yield after microbial pretreatment was still lower than this study.
These values were low compared with those reported by Agblevor
et al. [23]. In their report, both xylose and glucose produced etha-
nol, but only glucose was converted to ethanol in the study. Studies
about the utilization of hemicellulose for the production of ethanol
from CS are still deficient. Hence, in addition to improving pre-
treatment techniques and optimizing enzymatic hydrolysis,
enhancement of fermentation is desirable in further work.
4. Conclusions
DSAP, UAAP, and HPAP were comparatively investigated to pro-
duce ethanol from CS. SEM, FT-IR, and XRD analyses confirmed
favorable structural changes in pretreated samples. HPAP CS was
the most destroyed biomass and presented the lowest lignin con-
tent. As a result, HPAP CS led to the highest reducing sugar
(271.70 mg g
1 dry biomass) and ethanol yields (45.53%) com-
pared with DSAP and UAAP CS. HPAP proved to be a potential
and suitable pretreatment method for ethanol production from
CS. Further investigations on delignification of HPAP and optimiza-
tion of yeast fermentation are needed for improvement of this
process.
Acknowledgements
The authors would like to thank the China Agro-Industry Tech-
nology Research System fund of cotton (CARS-18-24) and Special
Fund for Agro-Scientific Research in the Public Interest
(201503135) for the support on this research.
References
[1] Shatalov AA, Pereira H. Xylose production from giant reed (Arundo donax L.):
modeling and optimization of dilute acid hydrolysis. Carbohyd Polym 2012;87
(1):210–7.
[2] Vohra M, Manwar J, Manmode R, Padgilwar S, Patil S. Bioethanol production:
feedstock and current technologies. J Environ Chem Eng 2014;2(1):573–84.
[3] Kim I, Seo YH, Kim G-Y, Han J-I. Co-production of bioethanol and biodiesel
from corn stover pretreated with nitric acid. Fuel 2015;143:285–9.
[4] Nikolić S, Mojović L, Rakin M, Pejin D. Bioethanol production from corn meal
by simultaneous enzymatic saccharification and fermentation with
immobilized cells of Saccharomyces cerevisiae var. ellipsoideus. Fuel 2009;88
(9):1602–7.
[5] Du SK, Zhu XN, Wang H, Zhou DY, Yang WH, Xu HX. High pressure assist-alkali
pretreatment of cotton stalk and physiochemical characterization of biomass.
Bioresour Technol 2013;148:494–500.
[6] McIntosh S, Vancov T, Palmer J, Morris S. Ethanol production from cotton gin
trash using optimised dilute acid pretreatment and whole slurry fermentation
processes. Bioresour Technol 2014;173:42–51.
[7] Binod P, Kuttiraja M, Archana M, Janu KU, Sindhu R, Sukumaran RK, et al. High
temperature pretreatment and hydrolysis of cotton stalk for producing sugars
for bioethanol production. Fuel 2012;92(1):340–5.
[8] Kaur U, Oberoi HS, Bhargav VK, Sharma-Shivappa R, Dhaliwal SS. Ethanol
production from alkali- and ozone-treated cotton stalks using thermotolerant
Pichia kudriavzevii HOP-1. Ind Crop Prod 2012;37(1):219–26.
[9] Chang VS, Holtzapple MT. Fundamental factors affecting biomass enzymatic
reactivity. Appl Biochem Biotechnol 2000;84–6:5–37.
[10] Kapoor M, Raj T, Vijayaraj M, Chopra A, Guptaa RP, Tuli DK, et al. Structural
features of dilute acid, steam exploded, and alkali pretreated mustard stalk
and their impact on enzymatic hydrolysis. Carbohyd Polym 2015;124:265–73.
[11] Haykir NI, Bakir U. Ionic liquid pretreatment allows utilization of high
substrate loadings in enzymatic hydrolysis of biomass to produce ethanol
from cotton stalks. Ind Crop Prod 2013;51:408–14.
[12] Shi J, Sharma-Shivappa RR, Chinn M, Howell N. Effect of microbial
pretreatment on enzymatic hydrolysis and fermentation of cotton stalks for
ethanol production. Biomass Bioenergy 2009;33(1):88–96.
[13] Rehman MSU, Kim I, Kim KH, Han J-I. Optimization of sono-assisted dilute
sulfuric acid process for simultaneous pretreatment and saccharification of
rice straw. Int J Environ Sci Technol 2014;11(2):543–50.
[14] Kim SB, Lee SJ, Lee JH, Jung YR, Thapa LP, Kim JS, et al. Pretreatment of rice
straw with combined process using dilute sulfuric acid and aqueous ammonia.
Biotechnol Biofuels 2013;6(1):109.
[15] Alvira P, Tomás-Pejó E, Ballesteros M, Negro MJ. Pretreatment technologies for
an efficient bioethanol production process based on enzymatic hydrolysis: a
review. Bioresour Technol 2010;101(13):4851–61.
[16] Silverstein RA, Chen Y, Sharma-Shivappa RR, Boyette MD, Osborne J. A
comparison of chemical pretreatment methods for improving saccharification
of cotton stalks. Bioresour Technol 2007;98(16):3000–11.
[17] Zhang Y, Fu E, Liang J. Effect of ultrasonic waves on the saccharification
processes of lignocellulose. Chem Eng Technol 2008;31(10):1510–5.
[18] Zheng JL, Yi WM, Wang NN. Bio-oil production from cotton stalk. Energy
Convers Manage 2008;49(6):1724–30.
[19] Miller GL. Use of dinitrosalicylic acid reagent for determination of reducing
sugar. Anal Chem 1959;31(3):426–8.
[20] Liu SC. Analysis and measurement in papermaking industry. Beijing: Chemical
Industry Press; 2003.
[21] Haykir NI, Bahcegul E, Bicak N, Bakir U. Pretreatment of cotton stalk with ionic
liquids including 2-hydroxy ethyl ammonium formate to enhance biomass
digestibility. Ind Crop Prod 2013;41:430–6.
[22] Ververis C, Georghiou K, Christodoulakis N, Santas P, Santas R. Fiber
dimensions, lignin and cellulose content of various plant materials and their
suitability for paper production. Ind Crop Prod 2004;19(3):245–54.
[23] Agblevor FA, Batz S, Trumbo J. Composition and ethanol production potential
of cotton gin residues. Appl Biochem Biotechnol 2003;105:219–30.
[24] Peng H, Li H, Luo H, Xu J. A novel combined pretreatment of ball milling and
microwave irradiation for enhancing enzymatic hydrolysis of microcrystalline
cellulose. Bioresour Technol 2013;130:81–7.
[25] Reddy N, Yang Y. Properties and potential applications of natural cellulose
fibers from the bark of cotton stalks. Bioresour Technol 2009;100(14):3563–9.
[26] Raiskila S, Pulkkinen M, Laakso T, Fagerstedt K, Löija M, Mahlberg R, et al. FTIR
spectroscopic prediction of Klason and acid soluble lignin variation in Norway
spruce cutting clones. Silva Fenn 2007;41(2):351–71.
[27] Pandiyan K, Tiwari R, Rana S, Arora A, Singh S, Saxena AK, et al. Comparative
efficiency of different pretreatment methods on enzymatic digestibility of
Partheniums sp.. World J Microbiol Biotechnol 2014;30(1):55–64.
[28] Bodirlau R, Teaca CA, Spiridon I. Influence of ionic liquid on hydrolyzed
cellulose material: FT-IR spectroscopy and TG-DTG-DSC analysis. Int J Polym
Anal Charact 2010;15(7):460–9.
[29] Chung C, Lee M, Choe EK. Characterization of cotton fabric scouring by FTIR
ATR spectroscopy. Carbohydr Polym 2004;58:417–20.
[30] Wang K, Yang HY, Xu F, Sun RC. Structural comparison and enhanced
enzymatic hydrolysis of the cellulosic preparation from Populus tomentosa
Carr., by different cellulose-soluble solvent systems. Bioresour Technol
2011;102(6):4524–9.
[31] Segal L, Creely JJ, Martin AEJ, Conrad CM. An empirical method for estimating
the degree of crystallinity of native cellulose using the X-ray diffractometer.
Text Res J 1959;29(10):786–94.
[32] Plácido J, Imam T, Capareda S. Evaluation of ligninolytic enzymes,
ultrasonication and liquid hot water as pretreatments for bioethanol
production from cotton gin trash. Bioresour Technol 2013;139:203–8.
[33] Vani S, Binod P, Kuttiraja M, Sindhu R, Sandhya SV, Preeti VE, et al. Energy
requirement for alkali assisted microwave and high pressure reactor
pretreatments of cotton plant residue and its hydrolysis for fermentable
sugar production for biofuel application. Bioresour Technol 2012;112:300–7.