processes

Article
Evaluation of Napier Grass for Bioethanol Production
through a Fermentation Process
Mallika Boonmee Kongkeitkajorn 1, * , Chanpim Sae-Kuay 1 and Alissara Reungsang 1,2
1 Department of Biotechnology, Faculty of Technology, Khon Kaen University, Khon Kaen 40002, Thailand;
[email protected] (C.S.-K.); [email protected] (A.R.)
2 Research Group for Development of Microbial Hydrogen Production Process, Khon Kaen University,
Khon Kaen 40002, Thailand
* Correspondence: [email protected]




Received: 12 April 2020; Accepted: 7 May 2020; Published: 11 May 2020


Abstract: Ethanol is one of the widely used liquid biofuels in the world. The move from sugar-based
production into the second-generation, lignocellulosic-based production has been of interest due to
an abundance of these non-edible raw materials. This study interested in the use of Napier grass
(Pennisetum purpureum Schumach), a common fodder in tropical regions and is considered an energy
crop, for ethanol production. In this study, we aim to evaluate the ethanol production potential
from the grass and to suggest a production process based on the results obtained from the study.
Pretreatments of the grass by alkali, dilute acid, and their combination prepared the grass for further
hydrolysis by commercial cellulase (Cellic® CTec2). Separate hydrolysis and fermentation (SHF),
and simultaneous saccharification and fermentation (SSF) techniques were investigated in ethanol
production using Saccharomyces cerevisiae and Scheffersomyces shehatae, a xylose-fermenting yeast.
Pretreating 15% w/v Napier grass with 1.99 M NaOH at 95.7 ◦ C for 116 min was the best condition
to prepare the grass for further enzymatic hydrolysis using the enzyme dosage of 40 Filter Paper
Unit (FPU)/g for 117 h. Fermentation of enzymatic hydrolysate by S. cerevisiae via SHF resulted in the
best ethanol production of 187.4 g/kg of Napier grass at 44.7 g/L ethanol concentration. The results
indicated that Napier grass is a promising lignocellulosic raw material that could serve a fermentation
with high ethanol concentration.

Keywords: Napier grass; bioethanol; biomass fractionation; enzyme hydrolysis; acid pretreatment;
alkali pretreatment

1. Introduction
Napier grass (Pennisetum purpureum Schumach), known also as elephant grass, is a perennial grass
found in tropical regions. Its high yield, easy cultivation, nutrient availability and versatility make
it widely popular for use as a fodder crop. Annual production yields of the grass vary depending
on cultivars, harvest cycles, fertilization, and climates. The reported Napier grass yield in temperate
climates was 20–40 ton/ha [1], while the yield in tropical climates was higher at 50–67 ton/ha [2,3].
Napier grass is classified as a lignocellulosic biomass. Its structural compositions varies depending
on weather, variety, and age. Reports on its composition covered 31–41% cellulose and 15–47%
hemicellulose [4–7]. Hydrolysis of its structure yields glucose and xylose, which are monomeric sugars
that serve as substrates in microbial fermentation to produce various biochemical including biofuels.
Napier grass plantation in Thailand has been for agricultural purposes. As the country
has continuously promoted biofuel production, Napier grass is considered as an energy crop.
Thailand’s 10 Year (2012–2021) Alternative Energy Development Plan has placed the grass as one of
the focused energy plants [8]. Napier grass has been successfully used in biogas production [8–10].

Processes 2020, 8, 567; doi:10.3390/pr8050567 www.mdpi.com/journal/processes

Processes 2020, 8, 567 2 of 19

Studies on the use of Napier grass for other biofuels production also exist, including for bioethanol
production. Pretreatments play an important role in production of liquid fuels. A study showed a
higher ethanol yield when using pretreated versus non-pretreated grasses, mainly due to increases in
sugar yields obtained in hydrolysis [11].
Pretreatment of Napier grass is an important step to prepare it for further hydrolysis. The process
alters physical structure of the biomass so that enzyme has a better access to the cellulose chain of
the biomass and improves hydrolysis [12,13]. Studies have been investigating various methods of
pretreatment, primarily physical and chemical methods, with the aim to obtain high sugar yields.
Some examples of the pretreatments focused on Napier grass included the uses of various alkalis,
e.g., aqueous ammonia, Ca(OH)2 , NaOH, alkali H2 O2 [14–16], and ammonia gas [17,18], dilute acid [14],
acid-peroxide [19], acid-alkali [20], steam explosion [14,21], hydrothermal treatment [11], and nitrogen
explosive decompression [22].
In the current study, we attempted to determine a simple and viable process for ethanol
production from Napier grass. The works in each pretreatment and hydrolysis step involved the use of
response surface methodology in order to determine the suitable conditions for each pretreatment and
hydrolysis routes.
Despite various pretreatment methods available, we chose to investigate the chemical pretreatment
of mechanically ground Napier grass due to its operational simplicity. We also chose NaOH and
H2 SO4 as they are common chemicals that are already in use by industries. The suitable conditions for
enzymatic hydrolysis of the each pretreatment methods were then determined to obtain the process
that resulted in the highest sugar released from the grass. Upon obtaining suitable pretreatment and
hydrolysis conditions, ethanol production was investigated using hydrolysate obtained from enzymatic
hydrolysis of pretreated grass, through separated hydrolysis and fermentation (SHF), and using the
pretreated grass, through simultaneous saccharification and fermentation (SSF). Two yeast strains
were used in fermentations, a common yeast and a xylose-fermenting yeast. We evaluated the ethanol
production using such combination of substrates and yeasts and suggested the process for ethanol
production from Napier grass based on the results of the study.

2. Materials and Methods

2.1. Preparation of Napier Grass

Napier grass (Pennisetum purpureum Schumach) Pakchong 1 aged 90–150 days was harvested from
the Demonstration field of the Faculty of Agriculture, Khon Kaen University, Thailand. Stems were cut
approximately 5 cm above the ground. They were then chopped and oven dried at 70 ◦ C. The dried
stems were then milled and sieved through a 10-mesh screen to obtain dried grass used in the study.

2.2. Alkali and Dilute Acid Pretreatments

Pretreatments using alkali and acid followed a similar procedure. In the case of alkali pretreatment,
dried Napier grass was soaked in 200 mL sodium hydroxide solution at 15% w/v loading in 500 mL
high-pressure laboratory bottles and heated in an autoclave. For acid pretreatments, dried Napier grass
or alkali-pretreated grass was soaked in sulfuric acid and heated in an autoclave. The concentration,
time, and temperature of the pretreatment were varied according to the values shown in Table 1.
After pretreatment, solid fractions were collected and washed until the pH was neutral. They were
then dried at 70 ◦ C prior to analysis and enzymatic hydrolysis.
Design Expert® (version 7.0 demo, Stat-Ease, Minneapolis, MN, USA, 2005) was used in designing
the experimental runs and analysis of results. Box–Benkhen design was applied for all experiments.
Treated grasses were analyzed for their susceptibility to cellulase hydrolysis which was used as
responses in all pretreatments. Mathematical models obtained from program analysis were used to
predict the conditions that resulted in the highest susceptibility for each of the pretreatment methods.
Processes 2020, 8, 567 3 of 19

Table 1. Factors and levels used in pretreatment and enzymatic hydrolysis studies.

Treatment Methods Factors Low Level (−1) High Level (+1)

Alkali pretreatment NaOH, Molarity (M) 0.25 3
Temperature, ◦ C 50 100
Time, min 30 180
Acid pretreatment H2 SO4 , % v/v 1 5
Temperature, ◦ C 80 120
Time, min 30 120
Acid pretreatment H2 SO4 , % v/v 1 5
of alkali-treated grass Temperature, ◦ C 50 100
Time, min 60 180
Enzyme dosage (Filter Paper
Enzymatic hydrolysis 10 50
Unit (FPU)/g substrate)
Incubation time, h 24 120
Substrate loading, % w/v 5 15

2.3. Enzymatic Hydrolysis

Pretreated grasses were hydrolyzed using a cellulase, Cellic® CTec2 (Novozymes, Bagsvaerd,
Denmark). Grass was added into 100 mL of 50 mM citrate buffer solution (pH 5.0) in a 500-mL
laboratory bottle. The enzyme was then added and mixed, and the mixture was incubated in a water
bath at 50 ◦ C with constant mixing. The hydrolysis reaction was stopped by boiling the content for
5 min. The liquid fraction was analyzed for reducing sugar.
Operation factors that were evaluated in enzymatic hydrolysis included enzyme dosage, incubation
time and substrate loading (Table 1). The Box–Benkhen design was used in designing experimental
runs with the help of Design Expert® software (version 7.0 demo, Stat-Ease, Minneapolis, MN, USA,
2005). The analysis of data was also carried out using the program. Suitable conditions for hydrolysis
by the enzyme were determined based on the maximum amount of reducing sugars obtained from
model prediction.

2.4. Ethanol Fermentation via SHF and SSF

Saccharomyces cerevisiae TISTR 5339 (Thailand Institute of Scientific and Technological Research,
Bangkok, Thailand) and Scheffersomyces shehatae ATCC 22984 (The American Type Culture Collection,
Manassas, VA, USA) were used in all fermentations. Yeast stocks were maintained in 30% glycerol.
They were propagated twice in Yeast Malt (YM) agar. Liquid medium used in fermentation was
prepared from enzymatic hydrolysate of alkali-treated grass supplemented with 2 g/L autolyzed yeast
powder (FM801, Angel Yeast, Yichang, China) for S. cerevisiae or 6 g/L for S. shehatae. Fermentations
were carried out in 250 mL Erlenmeyer flasks.
In inoculum preparation, a few single colonies were inoculated in diluted hydrolysate medium
whose glucose was adjusted to 10 g/L. It was incubated at 30 ◦ C with shaking at 200 rpm for 24 h.
Ten percent of the first seed inoculum was transferred to a fresh hydrolysate medium with 20 g/L
glucose and incubated at the same conditions for another 24 h.
In separate hydrolysis and fermentation (SHF), 10 mL of inoculum was transferred to 100 mL of
hydrolysate medium. For simultaneous saccharification and fermentation (SSF), 100 mL of 15% w/v
slurry of alkali-treated grass in water, supplemented with autolyzed yeast powder, were used as
substrate and medium for fermentation. Ten mL of inoculum was added together with 40 Filter Paper
Unit (FPU)/ggrass of cellulase, Cellic® CTec2, to start the cultivation. Fermentation conditions were
30 ◦ C with shaking at 100 rpm. Samples were taken at intervals and analyzed for cell growth, sugars,
and ethanol concentrations.
Processes 2020, 8, 567 4 of 19

2.5. Analysis
Grasses after pretreatment were analyzed by assessing their susceptibility to cellulase hydrolysis
(“susceptibility test” in short). One gram of dried grass was hydrolyzed by 30 FPU/g of cellulase
(Cellic® CTec2) in 50 mM citrate buffer (pH 5.0) for 72 h at 50 ◦ C. Results from the analysis were
reported as susceptibility with the unit of gram of reducing sugar per liter (gRS, suscepibility /L).
Structural compositions of grasses before and after pretreatments were analyzed based on
extractive-free biomass following the NREL procedure [23] and ASTM standard test E1758-01(2007).
Surfaces of grasses were observed under scanning electron microscope (S-3000N, Hitachi, Tokyo, Japan).
Number of cells was reported as cells/mL and determined using cell count on hemocytometer.
Reducing sugar was analyzed using dinitrosalicylic acid colorimetric assay. High-performance liquid
chromatography (LC-20A, Shimadzu, Kyoto, Japan) was used to determine concentrations of glucose,
xylose, and ethanol. It was equipped with an Aminex HPX-87H (Bio-Rad, Hercules, CA, USA) column
and a refractive index detector (RID-6A, Shimadzu, Kyoto, Japan) for analysis. Column temperature
was set at 40 ◦ C. The mobile phase was 5 mM sulfuric acid, flowing at 0.75 mL/min.

3. Results

3.1. Pretreatments of Napier Grass

Prior to hydrolysis of Napier grass by enzyme or using it in fermentation, pretreatment of the grass
is necessary in order to obtain biomass that is more vulnerable to further hydrolysis. Investigations in
using alkali, acid and their combinations in pretreatment of Napier grass were carried out to determine
for suitable conditions when applying each pretreatment regime.

3.1.1. Alkali Pretreatment

In order to obtain pretreatment conditions using alkali, NaOH concentrations (A in molarity),
treatment temperature (B in ◦ C) and time (C in min) were varied within the value ranges indicated in
Table 1. By using the Box–Behnken design, experimental runs and their respective results were obtained
as shown in Table 2. The data fitted well with this following quadratic equation with R-squared value
of 0.9843.

y = 7.24 + 19.8 A + 0.045 B + 0.168 C + 0.093 AB + 0.015 AC − 0.0014 BC − 6.46 A2

(1)
+ 0.00081 B2 − 0.00029 C2

Table 2. Experimental runs and results for determining conditions of alkali pretreatment following the
Box–Behnken design.

Run No. NaOH, A Temperature, B Time, C Susceptibility

(M) (◦ C) (min) (gRS, susceptibility /L)
1 3 75 30 24.3
2 1.625 75 105 31.7
3 0.25 75 180 10.8
4 1.625 100 180 35.8
5 1.625 75 105 32.4
6 1.625 75 105 33.4
7 0.25 100 105 10.0
8 3 100 105 39.1
9 0.25 50 105 9.62
10 1.625 50 180 31.2
11 1.625 75 105 31.7
12 0.25 75 30 9.80
13 1.625 50 30 22.4
14 1.625 100 30 37.7
15 3 75 180 31.4
16 3 50 105 26.0
17 1.625 75 105 26.0
Processes 2020, 8, 567 5 of 19

The analysis of variance (ANOVA) for the model (Table 3, NaOH) indicated that all the factors had
a significant effect on the susceptibility of the grass to cellulase hydrolysis. In addition, interactions
between NaOH concentration and temperature (A, B) and between temperature and time (B, C) also had
significant effects as demonstrated in Figure 1. The plots indicate that an increase in the concentration
of NaOH resulted in an increase in reducing sugar obtained from the susceptibility test until the
concentration was approximately 2 M; then the sugar started to decrease. Furthermore, the time taken
in pretreatment has shown its dependence on the temperature used (Figure 1b). At low pretreatment
temperature, the grass was more susceptible to cellulase hydrolysis with longer treatment time whereas
time had little effect at high treatment temperature.

Table 3. P-values obtained from ANOVA (analysis of variance) for response surface quadratic models
used in the prediction of suitable conditions for pretreatments of Napier grass.

Source p-Value Source p-Value

NaOH H2 SO4 NaOH + H2 SO4
Model <0.0001 Model 0.0075 0.0055
A—NaOH <0.0001 A—H2 SO4 0.0022 0.0244
B—Temperature 0.0005 B—Temperature 0.0011 0.0004
C—Time 0.0318 C—Time 0.0979 0.9906
AB 0.0145 AB 0.2714 0.8541
AC 0.1648 AC 0.7242 0.5962
BC 0.0306 BC 0.0766 0.7019
A2 <0.0001 A2 0.0923 0.0210
B2 0.6162 B2 0.7228 0.0061
C2 0.1380 C2 0.1178 0.2965
Lack of Fit 0.1695 Lack of Fit 0.5632 0.0018
R-Squared 0.9843 R-Squared 0.9053 0.9139
Predicted R-Squared 0.8206 Predicted R-Squared 0.3468 −0.3992
Notes: NaOH = pretreatment with NaOH, H2 SO4 = pretreatment with H2 SO4 and NaOH + H2 SO4 = pretreatment
with NaOH
Processes 2020, 8, xfollowed by H
FOR PEER 2 SO4 .
REVIEW 6 of 21

(a) (b)

1. Surface
Figure 1.
Figure Surface plots
plotsshowing
showingsignificant
significantinteractions
interactions = 0.05)
(α (α between
= 0.05) (a) NaOH
between (a) NaOHconcentration and
concentration
temperature (AB) at 116 min and (b) temperature and time (BC) at 1.99 M NaOH in alkali
and temperature (AB) at 116 min and (b) temperature and time (BC) at 1.99 M NaOH in alkali pretreatment
on reducing sugar
pretreatment obtained
on reducing fromobtained
sugar the susceptibility
from the test.
susceptibility test.

The equation predicted the pretreatment conditions that resulted in the highest susceptibility
The equation predicted the pretreatment conditions that resulted in the highest susceptibility
(39.3 gRS, susceptibility /L) of 15% dried grass to be 1.99 M NaOH, 95.7 ◦ C and 116 min. A verification test
(39.3 gRS, susceptibility/L) of 15% dried grass to be 1.99 M NaOH, 95.7 °C and 116 min. A verification test
using these conditions in pretreating Napier grass resulted in 38.4 g/L of reducing sugar obtained from
using these conditions in pretreating Napier grass resulted in 38.4 g/L of reducing sugar obtained
the susceptibility test, which was a 2.2% deviation from the model prediction.
from the susceptibility test, which was a 2.2% deviation from the model prediction.

3.1.2. Dilute Acid Pretreatment

In the investigation for suitable conditions in pretreatment of Napier grass by dilute acid, three
factors of interest were H2SO4 concentration (A in % v/v), temperature (B in °C) and time (C in min).
Their levels as indicated in Table 1 was applied to the Box–Behnken design and the results of 17
experimental runs are shown in Table 4. The data were fitted with a quadratic equation. The resulting
equation gave the best fit with R-squared value of 0.9053:
Processes 2020, 8, 567 6 of 19

3.1.2. Dilute Acid Pretreatment

In the investigation for suitable conditions in pretreatment of Napier grass by dilute acid,
three factors of interest were H2 SO4 concentration (A in % v/v), temperature (B in ◦ C) and time
(C in min). Their levels as indicated in Table 1 was applied to the Box–Behnken design and the
results of 17 experimental runs are shown in Table 4. The data were fitted with a quadratic equation.
The resulting equation gave the best fit with R-squared value of 0.9053:

y = −3.40 − 1.06 A + 0.124 B + 0.110 C + 0.008 AB − 0.0011 AC − 0.00063 BC + 0.130 A2

(2)
− 0.00025 B2 − 0.00023 C2

Table 4. Experimental runs and results for determining conditions of dilute acid pretreatment following
the Box–Behnken design.

Run No. H2 SO4 , A Temperature, B Time, C Susceptibility

(%) (◦ C) (min) (gRS, susceptibility /L)
1 3 120 120 9.57
2 3 100 75 9.17
3 1 100 120 8.36
4 3 100 75 8.02
5 5 100 120 9.77
6 1 80 75 7.76
7 3 100 75 8.62
8 3 100 75 9.37
9 3 80 120 8.32
10 3 80 30 5.95
11 5 120 75 11.5
12 3 120 30 9.47
13 5 80 75 9.13
14 1 120 75 8.86
15 3 100 75 9.33
16 5 100 30 9.73
17 1 100 30 7.92

Analysis of variance for the model (Table 3, H2 SO4 ) suggested that the main factors that posed
significant influence on the susceptibility of the grass to cellulase hydrolysis at 95% confidence level
were H2 SO4 concentration (A) and temperature (B). Time was also significant but at a higher significance
level (α = 0.1). In pretreatment with dilute acid, interaction between factors did not significantly affect
the susceptibility of the grass. Nonetheless, the interaction between time and temperature (BC) was
significant at a 90% confidence level. The surface plot in Figure 2 shows that temperature affected the
susceptibility strongly when using a short pretreatment time. Its effect was lesser at longer treatment
time. Furthermore, combination of high temperature and long treatment time exerted detrimental
effects on the susceptibility of the grass at 5% H2 SO4 .
The quadratic model predicted the maximum reducing sugar from susceptibility test to be 11.6 g/L.
The respective conditions were pretreating the grass at 15% loading with 5% v/v H2 SO4 at 120 ◦ C
for 56 min. The verification test resulted in 11.4 ± 0.1 g/L of reducing sugar from susceptibility test.
Although the predicted R-squared of this model was low, these results of the conditions were still valid
since the conditions were similar to those of run number 11. The small difference in pretreatment time
between run number 11 and the predicting value was insignificant, as the pretreatment time did not
have a significant effect on the susceptibility test result.
According to the susceptibility results in Table 4, it should be noted that dilute acid pretreatment
was generally inferior to alkali pretreatment. Lower reducing sugar released in susceptibility test
indicated that the pretreated grass was less susceptible to cellulase hydrolysis.
significance level (α = 0.1). In pretreatment with dilute acid, interaction between factors did not
significantly affect the susceptibility of the grass. Nonetheless, the interaction between time and
temperature (BC) was significant at a 90% confidence level. The surface plot in Figure 2 shows that
temperature affected the susceptibility strongly when using a short pretreatment time. Its effect was
Processes
lesser at2020, 8, 567treatment time. Furthermore, combination of high temperature and long treatment
longer 7 of 19

time exerted detrimental effects on the susceptibility of the grass at 5% H2SO4.

Figure 2. Surface plot showing interactions (α = 0.10) between temperature and time (BC) at optimal
Figure 2. Surface plot showing interactions (α = 0.10) between temperature and time (BC) at optimal
H2 SO4 concentration (5% v/v) in dilute acid pretreatment on reducing sugar obtained from the
H2SO4 concentration (5% v/v) in dilute acid pretreatment on reducing sugar obtained from the
susceptibility test.
susceptibility test.
3.1.3. Alkali Followed by Acid Pretreatment
The quadratic model predicted the maximum reducing sugar from susceptibility test to be 11.6
In this
g/L. The pretreatment,
respective Napier
conditions weregrass was firstly
pretreating thepretreated
grass at 15%with NaOH
loading under
with 5% vthe best
/v H conditions.
2SO4 at 120 °C
The NaOH-treated grass was subjected to the treatment using the same factors as in
for 56 min. The verification test resulted in 11.4 ± 0.1 g/L of reducing sugar from susceptibility the dilute acid
test.
pretreatment but the levels were adjusted as shown in Table 1. The results from all
Although the predicted R-squared of this model was low, these results of the conditions were still17 experimental
runs
valid showed
since thethatconditions
the difference
werebetween
similar the lowestofand
to those runhighest
number values
11. ofThereducing sugars from
small difference in
susceptibility test was quite narrow (Table 5). This circumstance indicated the more
pretreatment time between run number 11 and the predicting value was insignificant, as the profound effect of
alkali pretreatment
pretreatment time didon not
the have
biomass such thateffect
a significant further
on treatment by acid test
the susceptibility resulted
result.in small changes in
susceptibility values.
According to the susceptibility results in Table 4, it should be noted that dilute acid pretreatment
was generally inferior to alkali pretreatment. Lower reducing sugar released in susceptibility test
Table 5. Experimental runs and results for determining conditions of alkali followed by dilute acid
indicated that the pretreated grass was less susceptible to cellulase hydrolysis.
pretreatments following the Box–Behnken design.

3.1.3. AlkaliRun
Followed
No. byHAcid Pretreatment
2 SO4 , A Temperature, B Time, C Susceptibility
(%)
In this pretreatment, Napier (◦ C) (min)
grass was firstly pretreated with NaOH(gRS,under /L) conditions.
the best
susceptibility
The NaOH-treated 1 grass was5 subjected to the
75 treatment using
180 the same factors 29.5as in the dilute acid
pretreatment but 2 the levels were
1 adjusted as 75shown in Table 1801. The results from33.9 all 17 experimental
runs showed that 3 the difference
5 between 100the lowest and 120highest values of23.7 reducing sugars from
4 1 50 120 30.7
susceptibility test was quite narrow (Table 5). This circumstance indicated the more profound effect
5 5 75 60 28.1
of alkali pretreatment
6 on the3 biomass such 50 that further treatment
60 by acid resulted
34.3 in small changes
in susceptibility7 values. 1 100 120 25.8
Analysis of 8 variance 5(Table 3, NaOH 50 + H2SO4) suggested
120 that acid29.2 concentration and
temperature were 9 the two factors
1 75
that influenced 60
the susceptibility of the grass31.1
to cellulase hydrolysis
10 3 75 120 32.0
at 97.6% and more than 99.9% confidence levels, respectively. The ANOVA also indicated that there
11 3 50 180 32.7
was no interaction
12 between 3factors. 100 180 25.2
The experimental
13 data 3was fitted to a quadratic
75 model120
with the resulting32.3equation:
14 3 75 120 31.8
15.2 + 2.82 A 3+ 0.522 B − 0.02775C − 0.0026 AB 120
y = 15 − 0.0031 AC − 0.00018
31.5 BC − 0.490
16 3 A − 0.0041
2
75 B + 0.00021
2
120 C2
31.6 (3)
17 3 100 60 27.9

Analysis of variance (Table 3, NaOH + H2 SO4 ) suggested that acid concentration and temperature
were the two factors that influenced the susceptibility of the grass to cellulase hydrolysis at 97.6%
and more than 99.9% confidence levels, respectively. The ANOVA also indicated that there was no
interaction between factors.
Processes 2020, 8, 567 8 of 19

The experimental data was fitted to a quadratic model with the resulting equation:

y = 15.2 + 2.82 A + 0.522 B − 0.027 C − 0.0026 AB − 0.0031 AC − 0.00018 BC − 0.490 A2

(3)
− 0.0041 B2 + 0.00021 C2

This equation fitted well with the experimental data with R-squared value of 0.9139. The model
predicted the best conditions that yielded the highest susceptibility of 33.5 gRS,susceptibility /L to be
the use of 2.53% v/v H2 SO4 at 62 ◦ C for 60 min at 15% grass loading. Although the lack of fit test
was shown as significant (p-value = 0.0018) and the predicted R-squared was negative, the result of
the independent susceptibility test using the predicted conditions (33.0 ± 0.5 gRS,susceptibility /L) was
very close to the predicted value. This situation was possible regardless of the significant lack of fit
and due to the following supporting reason. Since the time was not a significant factor and there
was no interaction between factors, it would not affect the response. Considering runs with the
same temperature, the optimum values of acid concentration (2.53% v/v) laid between run 9 and 10
(Table 5) where the temperature was 75 ◦ C. The responses of the two runs were very similar at 31.6
and 31.9 gRS,susceptibility /L. Following the same principle, optimum temperature (62 ◦ C) laid between
values in run 10 and 11 (Table 5) where acid concentration was the same. The responses of the 2
runs were also similar at 31.9 and 33.3 gRS,susceptibility /L. From these explanations, it was evident that
changes in acid concentration and temperature between those ranges barely affected the responses.
Therefore, the confirmation test using the predicted conditions was valid. However, low predicted
R-squared values indicated that this equation is not suitable for predicting the response of the conditions
that were not at the experimental points.

3.2. Physical Structure and Composition of Napier Grass after Pretreatments

Napier grass obtained after the three pretreatment regimes were analyzed for their physical
structures and structural compositions. The results in Table 6 demonstrated that all pretreatments are
able to reduce the amount of lignin in Napier grass, although at different capability. Reduction in
lignin content was more pronounced when treating with NaOH than using the acid. The acid directly
affects the hemicellulose part of the grass as reflected in the more than 50% reduction in hemicellulose
in the pretreated samples. NaOH also dissolves some hemicellulose causing a drop in hemicellulose
content, although not to the same extent caused by the dilute acid. As the results of pretreatment,
an increase in the cellulose fraction was evident.

Table 6. Structural composition of native and pretreated Napier grasses.

Treated Grass Treated Grass Treated Grass with

Composition Native Grass
with NaOH with H2 SO4 (NaOH + H2 SO4 )
% Lignin 19.4 ± 0.1 a 9.90 ± 0.4 c 15.0 ± 0.6 b 4.27 ± 0.34 d
% Cellulose 35.2 ± 1.2 b 46.5 ± 0.6 a 46.6 ± 0.0 a 48.3 ± 1.7 a
% Hemicellulose 18.1 ± 0.3 a 14.0 ± 0.2 b 7.72 ± 0.0 c 13.9 ± 0.2 b
% Ash 1.73 ± 0.18 a 0.58 ± 0.03 c 0.98 ± 0.08 b 0.25 ± 0.07 d
Susceptibility
n/a 38.4 ± 0.14 a 11.4 ± 0.1 c 33.0 ± 0.5 b
(gRS , susceptibility /L)
Notes: (1) The statistical comparisons were between the same compositions of the different treatments. The same
letter indicated that the values were not significantly different at 95% confidence level. (2) All samples were
extractive-free. n/a = not available.

When matching the susceptibility data to the structural composition of the pretreated grasses, it is
evident that the presence of lignin relates to a much lesser vulnerability of the biomass to hydrolysis by
cellulase. Although the pretreated grasses have similar polysaccharide contents, they do not result in
the same level of hydrolysis as indicated by the resulting reducing sugar from the susceptibility test.
Surfaces of native grass observed under SEM shows smooth and packed-layer surface (Figure 3a).
The layering surface still exists but with loose, wrinkled, and swelling appearance after the pretreatment
Processes 2020, 8, 567 9 of 19

with NaOH, both with and without acid pretreatment (Figure 3b,d). A different effect when using
acid pretreatment is observed such that the grass appeared fibrous with signs of degradation but the
appearance
Processes 2020,is
8, still
x FOR packed (Figure 3c).
PEER REVIEW 10 of 21

(a) (b)

(c) (d)

Figure 3.
Figure Physical appearances
3. Physical appearances of of Napier
Napier grass
grass (a)
(a) in
in native
native form,
form, (b)
(b) when
when pretreated
pretreated with
with NaOH,
NaOH,
(c) when pretreated with H 2 SO4 and (d) pretreated with NaOH followed by H 2 SO4 under
(c) when pretreated with H2SO4 and (d) pretreated with NaOH followed by H2SO4 under a scanning a scanning
electron microscope
electron microscope atat 1500× magnification.
1500× magnification.

According to the results on susceptibility to cellulase hydrolysis, the structural compositions and
3.3. Enzymatic Hydrolysis of Pretreated Grass
the physical structures of the grass, we conclude that the pretreatment of 15% dried grass by 1.99 M
NaOH Although
at 95.7 ◦the previous
C for 116 min results on pretreatment
is suitable as it makes methods
the grasshave
mostsuggested thetouse
susceptible of NaOH
further alone
enzymatic
in pretreating the Napier grass, we decided to investigate the enzymatic hydrolysis
hydrolysis. The treatment with dilute acid following the NaOH pretreatment is not necessary, as of the grass from
the
all three pretreatment
susceptibility methods.with
did not improve Pretreated grasses
the extra were dried prior to the hydrolysis. The hydrolysis
treatment.
conditions followed those shown in Table 1. The reducing sugar concentration in the liquid
3.3. Enzymatic
hydrolysate wasHydrolysis of Pretreated
the response used forGrass
the analysis of the results and the prediction of the suitable
hydrolysis conditions.
Although the previous results on pretreatment methods have suggested the use of NaOH
aloneThe experimental
in pretreating theresults
Napierof grass,
enzymatic hydrolysis
we decided of grass pretreated
to investigate with the
the enzymatic three methods
hydrolysis of the
were demonstrated in Table 7. In general, both pretreatments involving NaOH resulted
grass from all three pretreatment methods. Pretreated grasses were dried prior to the hydrolysis. in similar
reducing sugar concentrations
The hydrolysis and the
conditions followed values
those werein
shown significantly
Table 1. The higher than that
reducing sugar treated with dilute
concentration in
acid. The ANOVA of the results in Table 8 show that all factors (enzyme dosage, hydrolysis
the liquid hydrolysate was the response used for the analysis of the results and the prediction time,ofand
the
substrate loading) affect
suitable hydrolysis the amount of sugar obtained in the liquid hydrolysate, except for the grass
conditions.
pretreated with H 2SO4 where enzyme dosage does not affect the reducing sugar obtained.
The experimental results of enzymatic hydrolysis of grass pretreated with the three methods
were demonstrated in Table 7. In general, both pretreatments involving NaOH resulted in similar
Processes 2020, 8, 567 10 of 19

reducing sugar concentrations and the values were significantly higher than that treated with dilute
acid. The ANOVA of the results in Table 8 show that all factors (enzyme dosage, hydrolysis time,
and substrate loading) affect the amount of sugar obtained in the liquid hydrolysate, except for the
grass pretreated with H2 SO4 where enzyme dosage does not affect the reducing sugar obtained.

Table 7. Experimental runs based on the Box–Behnken design and results for determining conditions
of enzymatic hydrolysis of pretreated Napier grass.

Run Enzyme, A Time, B Substrate Loading, C Reducing Sugars (g/L)

(FPU/g) (h) (%) NaOH H2 SO4 NaOH + H2 SO4
1 30 120 5 65.5 19.3 57.1
2 30 72 10 83.6 33.4 99.9
3 50 120 10 98.9 36.1 112.5
4 10 120 10 71.6 32.5 89.5
5 30 24 5 53.2 11.6 50.9
6 10 72 5 43.3 9.85 46.6
7 30 72 10 83.0 32.3 98.2
8 10 72 15 56.9 52.1 75.9
9 10 24 10 50.3 18.8 82.9
10 30 72 10 84.5 32.8 98.1
11 30 120 15 128.6 54.5 125.7
12 50 72 5 28.7 23.2 73.1
13 30 72 10 84.2 33.7 98.2
14 50 24 10 57.5 15.3 74.9
15 30 72 10 83.2 34.2 99.9
16 50 72 15 110.3 38.1 107.3
17 30 24 15 81.6 31.0 93.6

Table 8. ANOVA for response surface quadratic models used in prediction of suitable conditions for
enzyme hydrolysis of pretreated Napier grass.

Source p-Value
NaOH H2 SO4 NaOH + H2 SO4
Model <0.0001 <0.0001 <0.0001
A—Enzyme dosage <0.0001 0.7787 <0.0001
B—Time <0.0001 <0.0001 <0.0001
C—Substrate loading <0.0001 <0.0001 <0.0001
AB <0.0001 0.0009 <0.0001
AC <0.0001 <0.0001 <0.0001
BC <0.0001 <0.0001 <0.0001
A2 <0.0001 <0.0001 0.0100
B2 <0.0001 <0.0001 <0.0001
C2 <0.0001 0.0789 <0.0001
Lack of Fit 0.0904 0.7141 0.2125
R-Squared 0.9993 0.9989 0.9987
Predicted R-Squared 0.9910 0.9939 0.9865
Notes: NaOH = grass pretreated with NaOH, H2 SO4 = grass pretreated with H2 SO4 and NaOH + H2 SO4 = grass
pretreated with NaOH followed by H2 SO4 .

Quadratic equations for predicting the suitable hydrolysis conditions for the pretreated grass
using different pretreatment regimes are as follows:
Grass pretreated with NaOH:

y = 34.32 + 1.12 A − 0.459 B + 1.47 C + 0.0052 AB + 0.170 AC + 0.036 BC − 0.046 A2

(4)
+ 0.0018 B2 − 0.224 C2
Processes 2020, 8, 567 11 of 19

Grass pretreated with H2 SO4 :

y = −25.60 + 1.01 A + 0.221 B + 3.07 C + 0.0018 AB − 0.068 AC + 0.018 BC − 0.0077 A2

(5)
+ 0.0019 B2 − 0.026 C2

Grass pretreated with H2 SO4 followed by NaOH:

y = −29.77 + 0.879 A − 0.131 B + 14.59 C + 0.0081 AB − 0.096 AC + 0.027 BC − 0.0052 A2

(6)
− 0.0030 B2 − 0.409 C2

where y = reducing sugar (g/L), A = enzyme dosage (FPU/g), B = hydrolysis time (h) and C = substrate
loading (% w/v). All model equations have high R-squared values, which indicate that the models fit
well with the experimental data. In addition, the high values of predicted R-squared suggest that the
model equations can be used to predict the reducing sugar obtained from the hydrolysis.
The surface plots in Figure 4 show the reducing sugar concentrations obtained from varying enzyme
dosages and times when hydrolyzing grasses from different pretreatments. Hydrolysis time greatly
affects the sugar obtained, such that longer time results in higher sugar concentration. By observing
the surface shapes, the pretreatment involving the dilute acid causes a different response pattern of
sugar obtained from hydrolysis when compared with the pretreatment with NaOH alone. Pretreating
the grass with dilute acid seems to gain some benefit from the use of lower enzyme dosage with a
long hydrolysis time, as shown in Figure 4b,c. While removing lignin by NaOH could loosen the grass
structure and allow for more cellulase accessibility to the cellulose, dissolving hemicellulose by dilute
acid could also increase the cellulose accessibility [24]. Since the cellulose accessibility is the major
Processes 2020, 8, x FOR PEER REVIEW 13 of 21
factor on cellulase hydrolysis [12,13], the extra increase in this accessibility allows better penetration of
the enzyme during hydrolysis even at a lower enzyme dosage.

(a) (b)

(c)
Figure4.4.Surface
Figure Surface plots
plots showing
showingsignificant
significantinteractions
interactions (α(α
= 0.05) between
= 0.05) enzyme
between enzymedosage and and
dosage time time
( AB ) for the enzymatic hydrolysis of Napier grass (a) pretreated with NaOH, (b) pretreated with
(AB) for the enzymatic hydrolysis of Napier grass (a) pretreated with NaOH, (b) pretreated with H2 SO4 ,
H2SO4, and (c) pretreated with NaOH followed by H2SO4, each at their optimal conditions.
and (c) pretreated with NaOH followed by H2 SO4 , each at their optimal conditions.

Table 9. Optimal values of operating conditions for enzymatic hydrolysis and reducing sugar
obtained.

NaOH H2SO4 NaOH + H2SO4

Optimal Values Sugar (g/L) Optimal Values Sugar (g/L) Optimal Values Sugar (g/L)
Enzyme dosage
Processes 2020, 8, 567 12 of 19

The hydrolysis conditions obtained from the model equations for pretreated grasses of different
pretreatment regimes are shown in Table 9, together with the predicted and actual reducing sugar
concentrations obtained from the hydrolysis. Pretreatment solely with NaOH resulted in the highest
reducing sugar concentrations when the pretreated grass was hydrolyzed with cellulase. Confirmation
tests using the predicted optimal conditions were carried out for all pretreated grasses and the reducing
sugar concentrations were very close to the predicting values.

Table 9. Optimal values of operating conditions for enzymatic hydrolysis and reducing sugar obtained.

NaOH H2 SO4 NaOH + H2 SO4

Optimal Optimal Optimal
Sugar (g/L) Sugar (g/L) Sugar (g/L)
Values Values Values
Enzyme dosage
40 11 33.6
(FPU/g)
Time 140.5 57.3 125.8
117 105 119.2
(h) (139.7 ± 0.8) (57.2 ± 1.0) (124.7 ± 0.7)
Substrate loading
15 14.8 14.9
(% w/v)
Note: Sugar concentrations are reducing sugar. Values are the predicted concentration from model equations.
Values in parentheses are from confirmation tests.

According to the results, cellulase hydrolysis of Napier grass pretreated with NaOH and with
NaOH followed by H2 SO4 resulted in similar reducing sugar concentrations, with the slightly
higher concentration in the pretreatment with NaOH. The sugar concentrations obtained from the
NaOH-pretreated grasses were high and in accordance with the polysaccharide contents in the
pretreated grasses. However, the amount of sugar obtained from the acid-pretreated grass was much
lower, regardless of the similar polysaccharide contents. These results emphasized the importance of
cellulase susceptibility of the biomass on the hydrolysis rather than the polysaccharide contents of
the biomass.
Both regimes involving the use of NaOH in pretreatment could be used for preparing liquid
hydrolysates for fermentations due to the high sugar concentrations obtained. However, based on
the number of steps involved and the higher sugar concentration obtained, we chose to pretreat the
Napier grass in a single step with NaOH followed by cellulase hydrolysis to prepare the hydrolysate
as a substrate for ethanol fermentation. To pretreat the Napier grass, 15% dried grass is to be soaked in
1.99 M NaOH and subjects to wet heat at 95.7 ◦ C for 116 min. Dried pretreated grass is hydrolyzed
using cellulase (Cellic® CTec2) at 40 FPU/g for 117 h and the grass loading of 15%.

3.4. Ethanol Production Potentials of Napier Grass

In evaluating the ethanol production potential, we employed two yeast strains, which are
S. cerevisiae TISTR 5339 and S. shehatae ATCC 22984, and two fermentation techniques viz. SHF (separate
hydrolysis and fermentation) and SSF (simultaneous saccharification and fermentation). The SHF
technique utilizes liquid hydrolysate obtained from enzymatic hydrolysis of NaOH-pretreated Napier
grass, while the SSF technique uses the NaOH-pretreated grass in the fermentation.

3.4.1. Ethanol Production by S. cerevisiae TISTR 5339

Ethanol production by S. cerevisiae was carried out using two cultivation techniques, SHF and SSF.
The liquid hydrolysate used in SHF contained 134.9 g/L reducing sugar, of which 90.8 g/L was glucose
and 18.8 g/L was xylose. In SHF (Figure 5a), the maximum ethanol of 44.7 g/L was obtained after 24 h
cultivation. Glucose utilization was in accord with increasing ethanol concentration. Yeast growth also
related to glucose utilization with the maximum specific growth rate of 2.79 ± 0.09 1/d. The growth
ceased when glucose consumption ended. In addition, xylose decreased throughout the fermentation
with the total xylose uptake of 6.71 g/L, which was 32.9% of the total amount. Slight increase of xylitol
 Ethanol production by S. cerevisiae was carried out using two cultivation techniques, SHF and
SSF. The liquid hydrolysate used in SHF contained 134.9 g/L reducing sugar, of which 90.8 g/L was
glucose and 18.8 g/L was xylose. In SHF (Figure 5a), the maximum ethanol of 44.7 g/L was obtained
after 24 h cultivation. Glucose utilization was in accord with increasing ethanol concentration. Yeast
growth also related to glucose utilization with the maximum specific growth rate of 2.79 ± 0.09 1/d.
Processes 2020, 8, 567 13 of 19
The growth ceased when glucose consumption ended. In addition, xylose decreased throughout the
fermentation with the total xylose uptake of 6.71 g/L, which was 32.9% of the total amount. Slight
increase of xylitol
was observed withwas observed
the final with theoffinal
concentration 2.47concentration of 2.47 g/L.
g/L. The fermentation The suggested
profiles fermentation
thatprofiles
ethanol
suggested
was mainlythat ethanolfrom
produced was mainly
glucoseproduced fromof
with the yield glucose with the yield of 0.49 g/g.
0.49 g/g.

(a) (b)

Figure 5.
Figure Fermentationprofiles
5. Fermentation ofS.S.cerevisiae
profilesof cerevisiae TISTR
TISTR 5339
5339 when cultivated in
when cultivated in (a)
(a) liquid
liquid hydrolysate
hydrolysate
of NaOH-pretreated Napier grass using SHF and (b) NaOH-pretreated grass using
of NaOH-pretreated Napier grass using SHF and (b) NaOH-pretreated grass using SSF. Symbols:SSF. Symbols:
●-
•—glucose; #—xylose; —ethanol; —xylitol and N—cell
glucose; ○-xylose; ■- ethanol; □- xylitol and ▲-cell number. number.

In ethanol
In ethanol production
productionusingusingSSFSSFby byS.S.cerevisiae
cerevisiae TISTR
TISTR 5339
5339 (Figure 5b), glucose
(Figure 5b), glucose remained
remained at at aa
low level
low level throughout
throughout the the fermentation.
fermentation. Glucose
Glucose waswas gradually
gradually released
released and
and taken
taken up up byby the
the yeast
yeast to
to
produce ethanol and 40.8 g/L of ethanol was produced after 36 h. Yeast growth
produce ethanol and 40.8 g/L of ethanol was produced after 36 h. Yeast growth appeared to be very appeared to be very
slow at
slow at the
the beginning
beginning of of the
the fermentation.
fermentation. The The use
use of
of high
high solid
solid loading
loading inin the
the fermentation
fermentation couldcould
result in
result in this
this initial
initial apparent
apparent slow
slow growth
growth where
where clear
clear separation
separation of of the
the solid
solid and
and liquid
liquid phases
phases was
was
evident. The growing yeast could attach to the surface of the grass and undetected
evident. The growing yeast could attach to the surface of the grass and undetected in the liquid phase. in the liquid phase.
After 12
After 12 h,h, the
the physical
physical appearance
appearance of of the
the fermentation
fermentation became
became slurry
slurry and
and the
the cells
cells could
could detach
detach from
from
the grass
the grass into
into the
the liquid
liquid resulting
resulting inin an
an increase
increase inin cell
cell count.
count. The
The maximum
maximum specific
specific growth
growth rate
rate of
of
3.08 ± 0.03 1/d was slightly higher than that of SHF and the final cell count was
3.08 ± 0.03 1/d was slightly higher than that of SHF and the final cell count was slightly higher than slightly higher than
that in
that in SHF.
SHF. Xylose
Xylose accumulated
accumulated during
during the the first
first 24
24 h
h together
together with
with xylitol
xylitol accumulation.
accumulation. The The final
final
xylitol concentration was 5.96 g/L. The simultaneous increase of xylose and xylitol
xylitol concentration was 5.96 g/L. The simultaneous increase of xylose and xylitol indicated that indicated that xylitol
was produced
xylitol was producedalong with
alongthe release
with of xylose
the release of from
xylosethe grass.
from the grass.
3.4.2. Ethanol Production by S. shehatae ATTC 22984
3.4.2. Ethanol Production by S. shehatae ATTC 22984
Both SHF and SSF were also employed in the study of ethanol production by S. shehatae ATTC
Both SHF and SSF were also employed in the study of ethanol production by S. shehatae ATTC
22984 (Figure 6a,b). The sugar concentrations in the Napier grass hydrolysate was similar to those
22984 (Figure 6a,b). The sugar concentrations in the Napier grass hydrolysate was similar to those
used in the SHF study using S. cerevisiae. In SHF, production of ethanol directly related to glucose
used in the SHF study using S. cerevisiae. In SHF, production of ethanol directly related to glucose
utilization with the highest concentration of 31.3 g/L, which was significantly lower than that obtained
using S. cerevisiae. Xylose utilization observed during the fermentation was as expected as the yeast is a
xylose-fermenting yeast. However, profile analysis suggested that xylose did not contribute to ethanol
production but to xylitol production with the final xylitol concentration of 2.5 g/L. The growth profile
in SHF showed the continuous growth during the first 48 h of fermentation. The uncoupling ethanol
production after 48 h suggested a possibility of product inhibition effect on growth of S. shehatae.
When cultivating S. shehatae using SSF (Figure 6b), the growth profile in relation to ethanol profile
was similar to that occurred in SHF, suggesting also a possibility of growth inhibition by ethanol.
The glucose profile showed a slight increase in concentration during the first 24 h before dropped and
remained at low level throughout the fermentation. The small accumulation of glucose suggested a
lower growth rate of the yeast as compared to the rate of sugar released from the grass by the action of
enzyme. Accumulation of xylose confirmed the ineffective xylose utilization, especially at high ethanol
concentration. The final ethanol concentration obtained from SSF was 34 g/L, which was significantly
 When cultivating S. shehatae using SSF (Figure 6b), the growth profile in relation to ethanol
profile was similar to that occurred in SHF, suggesting also a possibility of growth inhibition by
ethanol. The glucose profile showed a slight increase in concentration during the first 24 h before
dropped and remained at low level throughout the fermentation. The small accumulation of glucose
suggested a lower growth rate of the yeast as compared to the rate of sugar released from the grass
Processes 2020, 8, 567 14 of 19
by the action of enzyme. Accumulation of xylose confirmed the ineffective xylose utilization,
especially at high ethanol concentration. The final ethanol concentration obtained from SSF was 34
g/L, which
higher thanwasthatsignificantly higher
obtained from than that
SHF. Along withobtained from SHF. Along
ethanol production, withalso
xylitol was ethanol production,
produced and the
xylitol was also produced
final concentration and
was 3.33 the final concentration was 3.33 g/L.
g/L.

(a) (b)

Fermentation
Figure6.6.Fermentation
Figure profiles
profiles of S. of S. shehatae
shehatae whenwhen cultivated
cultivated in (a) hydrolysate
in (a) liquid liquid hydrolysate of
of NaOH-
NaOH-pretreated Napier grass using SHF and (b) NaOH-pretreated grass using SSF.
pretreated Napier grass using SHF and (b) NaOH-pretreated grass using SSF. Symbols: ●-glucose; ○-Symbols:
•—glucose;
xylose; #—xylose;
■-ethanol; —ethanol;
□-xylitol and ▲-cell number.and N—cell number.
—xylitol

The summary in Table 10 shows that S. cerevisiae TISTR 5339 is able to deliver a superior ethanol
The summary in Table 10 shows that S. cerevisiae TISTR 5339 is able to deliver a superior ethanol
production than S. shehatae ATCC 22984, in terms of concentration, productivity, and ethanol yield.
production than S. shehatae ATCC 22984, in terms of concentration, productivity, and ethanol yield.
S. shehatae, a xylose-fermenting yeast, did not deliver the expected xylose-fermenting ability when
S. shehatae, a xylose-fermenting yeast, did not deliver the expected xylose-fermenting ability when
fermenting in hydrolysate that contained high glucose concentration. The overall ethanol production
fermenting in hydrolysate that contained high glucose concentration. The overall ethanol production
yield from Napier grass suggested that the grass could be used for ethanol production with the best
yield from Napier grass suggested that the grass could be used for ethanol production with the best
yield of 187.4 g/kg of dried Napier grass obtained with SHF fermentation by S. cerevisiae. The results
yield of 187.4 g/kg of dried Napier grass obtained with SHF fermentation by S. cerevisiae. The results
of this study are in the range, or above the average, of other studies both in terms of concentration
of this study are in the range, or above the average, of other studies both in terms of concentration
obtained during fermentation and in terms of ethanol yield based on the amount of biomass used
obtained during fermentation and in terms of ethanol yield based on the amount of biomass used
(Table 11).
(Table 11).
Table 10. Summary on ethanol production from pretreated Napier grass using SSF and hydrolysate of
4. Discussion
the pretreated grass using SHF by S. cerevisiae and S. shehatae.
Regarding the use of dilute acid or NaOH for pretreatment of Napier or elephant grass, the
S. cerevisiae S. shehatae
results varied among several studies but
Parameters TISTRthey5339
were all towards the sameATCCtrend.
22984 A study showed that
dilute acid reduced both lignin and hemicellulose contents of the grass, while NaOH treatment only
SHF SSF SHF SSF
reduced lignin content but not hemicellulose [14]. Our results on dilute acid pretreatment followed a
µ (l/day) 2.79 ± 0.09 b 3.08 ± 0.03 a 0.94 ± 0.05 a 0.91 ± 0.03 a
similar reductionmaxtrend of both lignin and hemicellulose. However, we found that NaOH solubilized
Ethanol (g/L) 44.7 ± 0.4 a 40.8 ± 0.8 b 31.3 ± 0.6 b 34.0 ± 0.6 a
not only lignin, but also(g/L.h)
Qp, fermentation the hemicellulose fraction
1.87 ± 0.02 a (Table 6).
1.14 ± 0.02 These
b results agreed
0.26 ± 0.01 b with
0.47 some
± 0.01 a other
studies where hemicellulose
Qp, overall reduction0.31
process (g/L.h) was reported
± 0.00 b in alkali
1.14 ± 0.02treatment
a
0.13 but to ba smaller
± 0.00 0.47 ±extent
0.01 a when
Yps (mg/g treated grass) 298.0 ± 0.0 a 272.0 ± 0.0 b 208.7 ± 0.0 b 227.3 ± 0.0 a
Yps (mg/g grass) 187.4 ± 0.0 a 171.1 ± 0.0 b 131.2 ± 0.0 b 142.6 ± 0.0 a
Notes: (1) Qp, fermentation calculation is based only on fermentation time. (2) Qp, overall process calculation includes
hydrolysate preparation time (5 days) and fermentation time. (3) Comparisons are between different fermentation
methods and yeast strains within the same parameter. The same letter indicated that the values were not significantly
different at 95% confidence level. (4) Yps (mg/g treated grass) = ethanol yield based on pretreated grass (5) Yps (mg/g
grass) = ethanol yield based on non-pretreated grass.

4. Discussion
Regarding the use of dilute acid or NaOH for pretreatment of Napier or elephant grass, the results
varied among several studies but they were all towards the same trend. A study showed that dilute
acid reduced both lignin and hemicellulose contents of the grass, while NaOH treatment only reduced
lignin content but not hemicellulose [14]. Our results on dilute acid pretreatment followed a similar
reduction trend of both lignin and hemicellulose. However, we found that NaOH solubilized not only
Processes 2020, 8, 567 15 of 19

lignin, but also the hemicellulose fraction (Table 6). These results agreed with some other studies
where hemicellulose reduction was reported in alkali treatment but to a smaller extent when compared
to acid pretreatment [20,24]. Both lignin and hemicellulose removals by alkali pretreatment were also
evident in other biomass, such as sweet sorghum straw [25] and sugarcane bagasse [26].

Table 11. Comparison of ethanol production from Napier grass in various studies.

Microorganism Ethanol, g/L

Pretreatment References
(Mode of Fermentation) (Yield)
S. cerevisiae NBRC 2044
34.2 g/L **
+ E. coli KO11 Low-moisture anhydrous ammonia [18]
(241 mg/g # )
(SSF+PF *)
S. cerevisiae NBRC 2044
19.4 g/L
+ E. coli KO11 Low-moisture anhydrous ammonia [17]
(204 mg/g)
(SSCF)
Ethanol Red 26.05 g/L
NaOH [14]
(SSF) (-)
S. cerevisiae 24 g/L
H2 SO4 + NaOH [20]
(SSF) (127.9 mg % /g)
S. cerevisiae CAT-1 ~4.5 g/L
Steam explosion [21]
(SHF) (87.2 mg/g)
S. cerevisiae CAT-1 6.1 g/L
Milling + enzyme [27]
(SHF) (-)
Z. mobilis + fungi 0.51 g/L
None [28]
(SSCF) (30 mg/g)
Ethanol Red 30.2 g/L
NaOH [29]
(SSF) (143 mg/g)
Scheffersomyces stipitis
22.7 g/L
(Evolved strain) Low-moisture ammonia hydroxide [1]
(247 mg % /g)
(SHF)
44.7 g/L
S. cerevisiae TISTR 5339
NaOH (187.4 mg/g) This study
(SHF)
(298 mg/g # )
Notes: (1) SSF = simultaneous saccharification and fermentation, SSCF = simultaneous saccharification and
co-fermentation, SHF = separate hydrolysis and fermentation. * PF = pentose fermentation, where ethanol from SSF
was removed prior to PF, ** ethanol concentration from glucose fraction via SSF by S. cerevisiae, # pretreated biomass
and % converted from volume using ethanol density at 20 ◦ C (789.2 kg/m3 ).

Treatment with NaOH alters Napier grass structure by causing swelling of the structures, while the
acid-treated grass maintained a smooth surface, as evident in Figure 3b,c. The swelling is the result of a
saponification reaction at the ester bonds between 4-O-methylglucuronic acid and xylan by NaOH [30].
The swelled structure allows enzymes to penetrate, resulting in a more efficient hydrolysis [31].
This explanation supports the results shown in Tables 2, 4 and 5 where susceptibility to cellulase
hydrolysis of the grass was higher when treated with NaOH as compared to dilute acid. Severity of
pretreatment has shown effects on the hydrolysis step. The use of mild pretreatments such as hot
water and dilute NaOH resulted in low ethanol concentrations obtained even with further enzymatic
hydrolysis of pretreated biomass [28,32]. However, it should be noted that a suitable strength of
pretreatment is also important. A study on the pretreatment of switch grass showed that a high level
of delignification reduces the cellulose accessibility to cellulase due to the change in cellulose structure
that led to the reduction in available surface area for enzyme adsorption [13]. The result from that
study explains our results in Figure 1a where the susceptibility appeared to decrease when NaOH
exceeded 2 M.
Processes 2020, 8, 567 16 of 19

In addition, extra lignin removal in the grass pretreated with NaOH followed by H2 SO4 (Table 6)
did not enhance the susceptibility of the grass to cellulose hydrolysis. The susceptibility indicates an
ability of cellulase to access and hydrolyse a structure. The results agreed with those reported earlier
that the key parameter for hydrolysis of lignocellulosic materials was cellulose accessibility, rather than
lignin content or crystallinity of the biomass [12,13,16].
The effect of alkali and dilute acid on the susceptibility of the grass during pretreatment process
reflects on the results of hydrolysis studies. Since the dilute acid pretreatment resulted in the pretreated
grass with the structure that was less susceptible to hydrolysis by cellulase, as indicated by its low
susceptibility value, the enzyme dosage required for its hydrolysis was much lower than that for the
alkali-treated grasses (Table 9). Similar to the case of excessive delignification, a lesser area accessible
by the enzyme resulted in lower sugars obtained from the hydrolysis of the grass pretreated by dilute
acid. In addition, the lesser accessible area could cause the limit in hydrolysis even at low enzyme
dosage, such that increase in enzyme dosage did not result in a further increase in reducing sugars as
evident in Figure 4b. ANOVA of the acid pretreated grass in Table 8 also confirmed this result as the
enzyme dosage did not exert a significant effect on the hydrolysis.
Regarding the fermentation for ethanol production, decrease in xylose during S. cerevisiae
fermentation was not expected, as the yeast is not reported as a xylose-fermenting yeast. The decrease
may be the result of xylose transport into the cells. Xylose intake through glucose transport system was
formerly reported in S. cerevisiae [33]. In addition, S. cerevisiae also inherits xylose reductase and xylitol
dehydrogenase activities. Xylose reductase is induced by xylose and xylitol produced accumulates
without further conversion to ethanol [33,34].
When comparing the cultivations of S. cerevisiae via SHF and SSF (Table 10), it is evident that SHF
yields higher ethanol concentration and productivity. The hydrolysis of Napier grass at its optimal
temperature yields high sugar concentration, whereas the hydrolysis in SSF occurs at 30 ◦ C at which
the enzyme has lower activity. A 24-h hydrolysis test of pretreated Napier grass at 30 ◦ C resulted
in 119.1 g/L reducing sugar in which contained 82.1 g/L glucose and 14.1 g/L xylose. These values
were slightly lower than the hydrolysate used in SHF. The maximum specific growth rate in SSF
was, in contrary, slightly higher than that in SHF. Substrate inhibition on growth could explain the
result obtained, as glucose at the concentration higher than 80 g/L showed an inhibition effect on S.
cerevisiae [35]. Despite the higher ethanol obtained from SHF, the process involved the hydrolysis
period, which resulted in a significantly lower overall ethanol productivity for the process. Reducing
the hydrolysis time in SHF would improve this overall productivity value.
In SHF of S. shehatae, which is a xylose-fermenting yeast, the yeast utilized only 33% of the xylose.
The issue should not relate to the presence of inhibitors normally found in acid hydrolysate, as we used
enzymatic hydrolysis. This result contradicts our initial assumption that the yeast could effectively
utilize xylose. We based the assumption on our previous study that showed the yeast’s ability to
ferment both glucose and xylose to ethanol [36]. However, the current work involved the use of higher
concentration of sugars (in hydrolysate). It could be possible that the higher ethanol concentration at
the time glucose depleted in this study could have a more negative effect to xylose transport than in
our previous work.
The use of SHF and SSF in ethanol production by S. shehatae resulted differently from the
cultivations of S. cerevisiae. In this case, the SSF by S. shehatae resulted in higher ethanol concentration
regardless of a similar specific growth rate of the yeast between both fermentation techniques (Table 10).
By comparing the ethanol production profiles during the first 24 h of fermentation (Figure 6), it is
evident that the ethanol produced in SSF was higher than that in SHF, resulting also in a higher ethanol
concentration obtained in SSF at the end of the fermentation. The faster ethanol production could be
the result of lower sugar concentration during the fermentation that alleviated the substrate inhibition
effect on product formation. The conversion yield of sugar in the hydrolysate appears more effective in
SSF for S. shehatae as the ethanol was higher than that obtained from SHF regardless of the lower sugar
obtained from hydrolysis at 30 ◦ C.
Processes 2020, 8, 567 17 of 19

5. Conclusions
Napier grass has delivered satisfactory results as a second-generation feedstock for ethanol
production. Pretreatment of the grass by NaOH is an important step for further hydrolysis by cellulase,
both in SHF and SSF for ethanol production. The process for pretreatment involved the use of 1.99 M
NaOH at 95.7 ◦ C for 116 min with the grass loading of 15% w/v. After washing off the alkali, the grass
could be used in ethanol production by S. cerevisiae via SHF or SSF, with SHF yielded a slightly higher
ethanol concentration. The enzyme (Cellic® CTec2) at the dosage of 40 FPU/g could be applied directly
in SSF. The hydrolysate for SHF could be prepared using the same enzyme dosage to hydrolyse 15%
w/v of the pretreated grass at 50 ◦ C for 117 h to obtain fermentable sugars. The overall process could
yield 171–187 g of ethanol/kg of Napier grass at the ethanol concentrations of 41–45 g/L.

Author Contributions: Conceptualization: M.B.K.; data curation: M.B.K.; formal analysis: M.B.K. and C.S.-K.;
funding acquisition: M.B.K. and A.R.; investigation: C.S.-K.; methodology: M.B.K. and C.S.-K.; project
administration: M.B.K. and C.S.-K.; resources: M.B.K. and A.R.; supervision: M.B.K. and A.R.; validation:
M.B.K. and C.S.-K.; visualization: M.B.K.; writing—original draft: M.B.K.; writing—review and editing: M.B.K.
and A.R. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by (1) Thailand Research University Network under Bio-conversion group (2)
Energy Conservation and Promotion Fund, Energy Policy and Planning Office (EPPO), Ministry of Energy Thailand
(3) National Research Council of Thailand (NRCT), the Higher Education Research Promotion, and National
Research University Project of Thailand through Biofuels Research Cluster of Khon Kaen University, project
number NRU543030 and (4) TRF Senior Research Scholar, grant number RTA6280001.
Acknowledgments: The authors appreciate the donation of Napier grass from Department of Agronomy, Faculty of
Agriculture, Khon Kaen University. C.S.-K. is grateful for her partial financial support from Center for Alternative
Energy Research and Development. We would also like to thank Prawphan Yuvadetkun for her valuable assistance
during the course of the project operation.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to
publish the results.

References
1. Dien, B.S.; Anderson, W.F.; Cheng, M.H.; Knoll, J.E.; Lamb, M.; O’Bryan, P.J.; Singh, V.; Sorensen, R.B.;
Strickland, T.C.; Slininger, P.J. Field productivities of Napier grass for production of sugars and ethanol. ACS
Sustain. Chem. Eng. 2020, 8, 2052–2060. [CrossRef]
2. De Favare, H.G.; de Abreu, J.G.; de Barros, L.V.; da Silva, F.G.; Ferreira, L.M.M.; Barelli, M.A.A.; Neto, I.M.d.S.;
Cabral, C.E.A.; Peixoto, W.M.; Campos, F.I.d.S.; et al. Effect of Elephant grass Genotypes to Bioenergy
production. J. Exp. Agric. Int. 2019, 38, 1–11. [CrossRef]
3. Wijitphan, S.; Lowilai, P. Effects of cutting interval on yields and nutritive values of King Napier grass
(Pennisetum purpureum cv. King grass) under Irrigation Supply. KKU Res. J. 2011, 16, 215–224.
4. Cardona, E.; Rios, J.; Pena, J.; Penuela, M.; Luis Rios, L. King Grass: A very promising material for the
production of second generation ethanol in tropical countries. Biomass Bioenergy 2016, 95, 206–213. [CrossRef]
5. Kamarullah, S.H.; Marlina Mohd Mydin, M.M.; Omar, W.S.A.W.; Harith, S.S.; Noor, B.H.M.; Alias, N.Z.A.;
Manap, S.; Mohamad, R. Surface morphology and chemical composition of Napier grass fibers. Malays. J.
Anal. Sci. 2015, 19, 889–895.
6. Li, Y.; Zhang, Y.; Zheng, H.; Du, J.; Zhang, H.; Wu, J.; Huang, H. Preliminary evaluation of five elephant
grass cultivars harvested at different time for sugar production. Chin. J. Chem. Eng. 2015, 23, 1188–1193.
[CrossRef]
7. Mohammed, I.Y.; Abakr, Y.A.; Kazi, F.K.; Yusup, S.; Alshareef, I.; Soh, A.; Chin, S.A. Comprehensive
characterization of Napier grass as a feedstock for thermochemical conversion. Energies 2015, 8, 3403–3417.
[CrossRef]
8. Napier Grass. Available online: http://weben.dede.go.th/webmax/content/napier-grass (accessed on 6
April 2020).
9. Narinthorn, R.; Choorit, W.; Chisti, Y. Alkaline and fungal pretreatments for improving methane potential of
Napier grass. Biomass Bioenergy 2019, 127, 105262. [CrossRef]
Processes 2020, 8, 567 18 of 19

10. Suaisom, P.; Patiroop Pholchana, P.; Aggarangsi, P. Holistic determination of suitable conditions for biogas
production from Pennisetum purpureum x Pennisetum americanum liquor in anaerobic baffled reactor. J. Environ.
Manag. 2019, 247, 730–737. [CrossRef]
11. Bensah, E.C.; Kádár, Z.; Mensah, M.Y. Ethanol production from hydrothermally-treated biomass from West
Africa. BioResources 2015, 10, 6522–6537. [CrossRef]
12. Arantes, V.; Saddler, J.N. Cellulose accessibility limits the effectiveness of minimum cellulase loading on
the efficient hydrolysis of pretreated lignocellulosic substrates. Biotechnol. Biofuels 2011, 4, 3. [CrossRef]
[PubMed]
13. Rollin, J.A.; Zhu, Z.; Sathitsuksanoh, N.; Zhang, Y.-H.P. Increasing cellulose accessibility is more important
than removing lignin: A comparison of cellulose solvent-based lignocellulose fractionation and soaking in
aqueous ammonia. Biotechnol. Bioeng. 2011, 108, 22–30. [CrossRef] [PubMed]
14. Eliana, C.; Jorge, R.; Juan, P.; Luis, R. Effects of the pretreatment method on enzymatic hydrolysis and ethanol
fermentability of the cellulosic fraction from elephant grass. Fuel 2014, 118, 41–47. [CrossRef]
15. Pensri, B.; Aggarangsi, P.; Chaiyaso, T.; Chandet, N. Potential of fermentable sugar production from Napier cv.
Pakchong 1 grass residue as a substrate to produce bioethanol. Energy Procedia 2016, 89, 428–436. [CrossRef]
16. Phitsuwan, P.; Sakka, K.; Ratanakhanokchai, K. Structural changes and enzymatic response of Napier
grass (Pennisetum purpureum) stem induced by alkaline pretreatment. Bioresour Technol. 2016, 218, 247–256.
[CrossRef]
17. Yasuda, M.; Nagai, H.; Takeo, K.; Ishii, Y.; Ohta, K. Bio-ethanol production through simultaneous
saccharification and co-fermentation (SSCF) of a low-moisture anhydrous ammonia (LMAA)-pretreated
napiegrass (Pennisetum purpureum Schumach). Springerplus 2014, 3, 333. [CrossRef]
18. Yasuda, M.; Takeo, K.; Nagai, H.; Uto, T.; Yui, T.; Matsumoto, T.; Ishii, Y.; Ohta, K. Enhancement of ethanol
production from Napiergrass (Pennisetum purpureum Schumach) by a low-moisture anhydrous ammonia
pretreatment. J. Sustain. Bioenergy Syst. 2013, 3, 179–185. [CrossRef]
19. Bohórquez, C.; González, E.A.; Reina, M.R. Effect of pretreatment dilute acid—Peroxide on napier grass
(Pennisetum purpureum schumach) to enhance reducing sugar yield by enzymatic hydrolysis. AVANCES
Investigación en Ingeniería 2014, 11, 48–55. [CrossRef]
20. Camesasca, L.; Ramírez, M.B.; Guigou, M.; Ferrari, M.D.; Lareo, C. Evaluation of dilute acid and alkaline
pretreatments, enzymatic hydrolysis and fermentation of napiergrass for fuel ethanol production. Biomass
Bioenergy 2015, 74, 193–201. [CrossRef]
21. Scholl, A.L.; Menegol, D.; Pitarelo, A.P.; Fontana, R.C.; Filho, A.Z.; Ramos, L.P.; Dillon, A.J.P.; Camassola, M.
Ethanol production from sugars obtained during enzymatic hydrolysis of elephant grass (Pennisetum
purpureum, Schum.) pretreated by steam explosion. Bioresour Technol. 2015, 192, 228–237. [CrossRef]
22. Meneses, L.R.; Otor, O.F.; Bonturi, N.; Orupõld, K.; Kikas, T. Bioenergy yields from sequential bioethanol
and biomethane production: An optimized process flow. Sustainability 2020, 12, 272. [CrossRef]
23. Sluiter, A.; Hames, B.; Hyman, D.; Payne, C.; Ruiz, R.; Scarlata, C.; Sluiter, J.; Templeton, D.; Wolfe, J.
Determination of Total Solids in Biomass and Total Dissolved Solids in Liquid Process Samples; Technical Report
NREL/TP-510-42618; National Renewable Energy Laboratory: Golden, CO, USA, 2011.
24. Sun, S.; Sun, S.; Cao, X.; Sun, R. The role of pretreatment in improving the enzymatic hydrolysis of
lignocellulosic materials. Bioresour. Technol. 2016, 199, 49–58. [CrossRef] [PubMed]
25. Dong, M.; Wang, S.; Xu, F.; Wang, J.; Yang, N.; Li, Q.; Chen, J.; Li, W. Pretreatment of sweet sorghum straw
and its enzymatic digestion: Insight into the structural changes and visualization of hydrolysis process.
Biotechnol. Biofuels 2019, 12, 276. [CrossRef] [PubMed]
26. Zhang, H.; Wei, W.; Zhang, J.; Huang, S.; Xie, J. Enhancing enzymatic saccharification of sugarcane bagasse
by combinatorial pretreatment and Tween 80. Biotechnol. Biofuels 2018, 11, 309. [CrossRef]
27. Menegol, D.; Fontana, R.C.; Dillon, A.J.P.; Camassola, M. Second-generation ethanol production from
elephant grass at high total solids. Bioresour. Technol. 2016, 211, 280–290. [CrossRef]
28. Liu, Y.K.; Chen, W.C.; Huang, Y.C.; Chang, Y.K.; Chu, I.M.; Tsai, S.L.; Wei, Y.H. Production of bioethanol
from Napier grass via simultaneous saccharification and co-fermentation in a modified bioreactor. J. Biosci.
Bioeng. 2017, 124, 184–188. [CrossRef]
29. Tsai, M.H.; Lee, W.C.; Kuan, W.C.; Sirisansaneeyakul, S.; Savarajara, A.A. Evaluation of different pretreatments
of Napier grass for enzymatic saccharification and ethanol production. Energy Sci. Eng. 2018, 6, 683–692.
[CrossRef]
Processes 2020, 8, 567 19 of 19

30. Tarkow, H.; Feist, W.C. A mechanism for improving the digestibility of lignocellulosic materials with dilute
alkali and liquid ammonia. In ACS Advances in Chemistry Series; Hajny, G.J., Reese, E.T., Eds.; American
Chemical Society: Washington, DC, USA, 1969; Volume 95, pp. 197–218.
31. Hendriks, A.T.W.M.; Zeeman, G. Review Pretreatments to enhance the digestibility of lignocellulosic biomass.
Bioresour. Technol. 2009, 100, 10–18. [CrossRef]
32. Soares, I.B.; Marques, O.M.; Benachour, M.; Abreu, C.A.M.d. Ethanol production by enzymatic hydrolysis of
elephant grass. J. Life Sci. 2011, 5, 157–161.
33. Van Zyl, C.; Prior, B.A.; Kilian, S.G.; Kock, J.L.F. D-xylose utilization by Saccharomyces cerevisiae. J. Gen.
Microbiol. 1989, 135, 2791–2798. [CrossRef]
34. Batt, C.A.; Carvallo, S.; Eassond, D.D.; Akedom, M.; Sinskeya, A.J. Direct evidence for a xylose metabolic
pathway in Saccharomyces cerevisiae. Biotechnol. Bioeng. 1986, 28, 549–553. [CrossRef] [PubMed]
35. Zhang, Q.; Wu, D.; Lin, Y.; Wang, X.; Kong, H.; Tanaka, S. Substrate and product inhibition on yeast
performance in ethanol fermentation. Energy Fuels 2015, 29, 1019–1027. [CrossRef]
36. Yuvadetkun, P.; Boonmee, M. Ethanol production capability of Candida shehatae in mixed sugars and rice
straw hydrolysate. Sains Malays. 2016, 45, 581–587.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).