Fuel 185 (2016) 73–84

Contents lists available at ScienceDirect

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

Full Length Article

Acid post-hydrolysis of xylooligosaccharides from hydrothermal

pretreatment for pentose ethanol production
P.Y.S. Nakasu a,b, L.J. Ienczak b, A.C. Costa a, S.C. Rabelo b,⇑
a
Faculdade de Engenharia Química, Universidade Estadual de Campinas (UNICAMP), Caixa Postal 6066, 13083-970, Campinas, São Paulo, Brazil
b
Laboratório Nacional de Ciência e Tecnologia do Bioetanol (CTBE), Rua Giuseppe Máximo Scolfaro, 10.000, Polo II de Alta Tecnologia, Caixa Postal 6192, 13083-970, Campinas,
São Paulo, Brazil

h i g h l i g h t s g r a p h i c a l a b s t r a c t

The acid post-hydrolysis of a Concentrated

hydrothermal hemicellulosic acid
Sugarcane bagasse
hydrolysate was studied. + water
Inoculum

The performances of three acids: Hydrothermal Detoxified
Hemicel. 4
oxalic, maleic and sulfuric acid was Pretreament substeps Post-hydrolysate Ethanol
Post-hydrolysate
compared. Disllaon
Separaon

The C5-rich post-hydrolysates were Detoxificaon
Post-hydrolysis
C5 Pentose (C5)
fermented with a wild-type yeast. Yeast
Hemicel. Hydrolysate
fermentaon ethanol

Sulfuric acid showed the best kinetic Separaon
Pre-concentraon

of post-hydrolysis of
xylooligosaccharides.

Acid post hydrolysis increases overall
ethanol yield from sugarcane bagasse.

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

Article history: Hydrothermal pretreatment solubilizes about 65% of the hemicelluloses in sugarcane bagasse. However,
Received 29 May 2016 nearly 80% of the xylose recovered from the hemicellulosic hydrolysate is present as xylooligosaccharides
Received in revised form 15 July 2016 (XO’s), which cannot be directly fermented into pentose (C5) ethanol. In this work a kinetic study of the
Accepted 16 July 2016
post hydrolysis process considering the use of three acids – sulfuric, oxalic and maleic – was performed.
Xylose and furfural post-hydrolysis profiles showed the reaction time in which xylose peaked with min-
imum furfural production. Among the three studied acids, sulfuric acid showed the fastest kinetics of
Keywords:
post-hydrolysis with XO’s being fully hydrolyzed in less than 1 h reaction time. C5 fermentation exper-
Sugarcane bagasse
Pentose ethanol
iments showed that the detoxified post-hydrolysates fermented with ethanol yields ranging from 0.10 to
Hydrothermal pretreatment 0.31 g ethanol g1 reducing sugars. Most samples were fermented from 48 to 72 h of experiment with
Post-hydrolysis productivities ranging from 0.02 to 0.43 g L1 h1.
Xylooligosaccharides Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction operations in the production of second generation (2G) ethanol.

Although essential, biomass pretreatment is also an economic
hurdle that greatly impacts the overall costs of other unit
Abbreviations: DPH, detoxified post-hydrolysate; HH, hemicellulosic hydroly-
sate; HPH, hemicellulosic post-hydrolysate; XO, xylooligosaccharide.
⇑ Corresponding author.
E-mail address: sarita.rabelo@bioetanol.org.br (S.C. Rabelo).
Hydrothermal pretreatment is a water-based and environment-
friendly pretreatment that does not require addition of other
reagents, reducing the consumption of chemicals for pH adjust-
ment and the risk of equipment corrosion [23].
Hydrothermal pretreatment of sugarcane bagasse—Brazilian’s
main agricultural residue for ethanol production—solubilizes
hemicelluloses down mainly to xylooligosaccharides (XO’s), in
addition to a small monomeric fraction, leading to less sugars

http://dx.doi.org/10.1016/j.fuel.2016.07.069
0016-2361/Ó 2016 Elsevier Ltd. All rights reserved.
74 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84
degradation and consequently less fermentation inhibitors genera-
tion [9]. However, XO’s cannot be directly fermented by microor-
ganisms. A post-hydrolysis of the hemicellulosic hydrolysate
(HH) is required for the second generation (2G) ethanol production
via the biochemical platform biorefinery concept [11]. Post-
hydrolysis can be carried out via either chemicals or enzymes.
XO’s structure resemble its parent macromolecule, hemicelluloses,
a highly branched and complex heteropolymer. A broad hemicellu-
lolytic enzymes cocktail would be required to depolymerize XO’s,
increasing the process cost mainly due to longer residence times,
which directly impact on the size of reactors, and making it eco-
nomically unfeasible [7].
The acid post-hydrolysis of lignocellulosic hydrolysates from
different biomasses has been widely reported in the literature. It
was first employed by Saeman [25] to quantify the total hexose
(C6) sugars in solid wood samples, and later adapted by Mok and
Antal [22] for pentose sugars (C5) in HH’s naming it Quantitative
Saccharification. However, there is only one report in the literature
on the post-hydrolysis of soluble XO’s from sugarcane bagasse for
2G ethanol production so far [36].
Additionally, most of the post-hydrolysis studies for lignocellu-
losic materials were limited to a very small scale, either high
throughput or small capacity tubes, in which problems related to
mass transfer are barely seen [16,37,36]. Such studies proved the
efficacy of post-hydrolysis but gave no further idea of possible
challenges of process scale-up.
While it may seem contradictory using acid in the post-
hydrolysis of HH’s once the hydrothermal pretreatment was cho-
sen precisely for its greener features, there are several advantages
associated to the hydrothermal pretreatment followed by a post-
hydrolysis in comparison to the diluted acid pretreatment:
i. The foremost fact is the rising attention hydrothermal pre-
treatment itself has been gaining. When compared with
the diluted acid pretreatment, it is considered more
environment-friendly and cheaper [5]. Besides avoiding
hemicelluloses degradation, which may pave the way for a
C5-utilization platform.
ii. In post-hydrolysis, smaller reaction volumes are used—only
the HH—contrary to the whole biomass used in pretreat-
ment. This avoids using additional amounts of acid needed
to neutralize ashes in the whole biomass [18]. Therefore,
the amount of acid used per liter of ethanol is lower.
iii. There’s a heterogeneous medium in pretreatment due to an
abrasive solid substrate—sugarcane bagasse. In post-
hydrolysis there is a pseudo-homogeneous liquid medium
[16] that allows higher mass transfer rates.
iv. Pretreatment is performed at higher process temperatures
(T P 190 °C), which together with the solid substrate may
accelerate equipment wear out. On the other hand, post-
hydrolysis occurs at lower temperatures (120 °C 6
T 6 150 °C), which reduces process severity.
In this work, the post-hydrolysis performances of sulfuric,
maleic and oxalic acids on sugarcane bagasse HH in a bench scale
reactor (2 L) were compared. Maleic and oxalic acids were selected
due to their selectivity to hydrolyze XO’s, i.e. producing more
xylose with less degradation into furfural when compared to sulfu-
ric acid, as previously mentioned by Lu and Mosier [19].
Kinetic profiles of XO’s post-hydrolysis, C5 monomers and
furfural were determined. The post-hydrolysates were then fer-
mented by Scheffersomyces stipitis (wild type yeast able to consume
xylose) and compared in terms of ethanol yield and volumetric
productivities. A quantitative insight into the pentose use in 2G
ethanol production from sugarcane bagasse is therefore provided
in this work.
2. Materials and methods
2.1. Feedstock
Sugarcane bagasse was provided by Usina da Pedra (Serrana –
SP). The material was collected in the 2012/13 crop (May/2012).
It was mechanically harvested and resulted from the last milling
before juice extraction. Samples subjected to hydrothermal
pretreatment were not comminuted. The chemical composition
of raw sugarcane bagasse, as percentage of dry mass, was:
cellulose, 41.38 ± 0.14%; hemicelluloses, 27.84 ± 0.50%; lignin,
22.50 ± 0.43%; extractives 4.01 ± 0.06% and ashes, 5.62 ± 0.45%.
2.2. Hydrothermal pretreatment
For hydrothermal pretreatment, around 15 kg of raw bagasse
(50% w/w moisture content) were fed into a 350 L alloy steel reac-
tor (Pope Scientific Inc., Saukville, USA), without previous milling
or washing, with 9% (w/w) solids loading. No attempts were made
in order to optimize the pretreatment operational parameters. The
reaction was performed at 190 °C by thermal fluid percolation
through the reactor’s jacket, for 10 min at 150 rpm [28]. After the
reaction, the reactor was water-cooled, depressurized and opened.
Pretreated material fractions were separated by a nutsche filter
(Pope Scientific Inc., Saukville, USA) with 140 L capacity. The solid
fraction, mainly composed of cellulignin, was stored in a refriger-
ated container to be later subjected to an enzymatic hydrolysis step
(data not shown). The liquid fraction, HH, was concentrated three
times in a wiped film evaporator (Pope Scientific Inc., Saukville,
USA) with capacity up to 50 kg h1 of water evaporation and stored
in a refrigerated container for the post-hydrolysis assays.
2.3. Acid post-hydrolysis
2.3.1. Experimental design
For three acids—sulfuric, oxalic and maleic—seven post-
hydrolysis assays were performed following a 22 full factorial design
with three central point repetitions. Temperature (120 °C, 135 °C
and 150 °C) and acid loading (0.5%, 1.25% and 2.0% w/w) were con-
sidered as factors; xylose release and furfural production were con-
sidered as response variables. Reaction time ranged from 50 to
100 min, depending on the severity of the condition, and it was
assessed in short time intervals to determine pentose (arabinose
and xylose), acetic acid and furfural kinetic profiles. Sulfuric, oxalic,
maleic acid and all other reagents and chemicals, unless otherwise
noted, were purchased from Sigma-Aldrich (St. Louis, MO, USA).
2.3.2. Post-hydrolysis assays
Approximately 850 mL of the HH were fed into a 2 L alloy steel
reactor (Parr Instrument Company, Moline, United States) and
heated by electrical resistance until the reaction temperature
was reached. Stirring was kept constant at 200 rpm. Sulfuric,
maleic or oxalic acid (50 mL) were added by an external hydraulic
pump over five minutes. After the addition, hemicellulosic post-
hydrolysate (HPH) samples were periodically collected from the
reactor through its dip tube. The reaction was stopped by cooling
the reactor with cold water. The samples were analyzed by HPLC
for sugars (xylose, arabinose, glucose and cellobiose), sugar degra-
dation products (furfural, 5-hydroxymethylfurfural [HMF] and
formic acid) and acetic acid content.
 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84
Concentration profiles of xylose and furfural were evaluated to
determine the reaction time in which the xylose levels were the
highest and furfural levels the lowest for each assay of the factorial
design. Among the three acids, oxalic acid showed the lowest
xylose conversions and, along with its poor solubility in water, it
was decided not to include its post-hydrolysates in the following
fermentations.
2.4. Detoxification of the HPH’s
Preliminary tests showed that the post-hydrolysates obtained
were not fermentable, mainly due to the high acetic acid concen-
trations released during the post-hydrolysis (approximately
5 g L1) and the high phenolics content arising from the soluble
lignin generated during the hydrothermal pretreatment (approxi-
mately 8 g L1). For such reason, the HPH’s were detoxified by a
four-step detoxification. The first step consisted in vacuum
evaporation in a rotary evaporator (IKA, HB 10 control, Staufen,
Germany). HPH’s were evaporated 2.3 times at 100 °C, 450 mbar.
In this step it was possible to remove large amounts of organic
acids and volatile compounds such as acetic and formic acids,
and furfural. The next steps consisted in calcium hydroxide
overliming followed by pH reduction with phosphoric acid and
then treatment with activated charcoal (1% w/v) at 60 °C for
30 min in an orbital shaker at 200 rpm (New Brunswick Sci. Com-
pany Inc., New Jersey, USA). These last three steps were aimed to
remove part of the phenolic compounds. Samples were centrifuged
after overliming, acidification and treatment with activated char-
coal at 4 °C and 9000g according to Marton et al. [21], resulting
in the production of the detoxified post-hydrolysates (DPH).
2.5. Pentose fermentation
The post-hydrolyses assays were reproduced and were inter-
rupted in their respective times of xylose peak in order to generate
the substrates for fermentation. All five operational conditions of
the factorial design of sulfuric and maleic acids were reproduced.
Scheffersomyces stipitis NRRL Y-7124 (formerly known as Pichia
stipitis) was used to metabolize the xylose (C5), arabinose (C5) and
glucose (C6) present in the DPH’s. The yeast was kept in YPDX agar
containing in g L1: yeasts extract, 3.0; glucose, 10.0; xylose, 10.0;
agar, 15.0; and peptone, 10.0. The yeast maintained in agar was
transferred to the first seed culture that had the same medium
composition as described before, but without agar. A second seed
culture was inoculated with 10% (v/v) of the first seed culture
and the composition was based on Silva et al. [31] in g L1: glu-
cose, 12.0; xylose, 20.0; magnesium sulfate heptahydrate, 1.0;
urea, 2.3; yeast extract, 3.0. Then, 10% (v/v) of the second seed cul-
ture was used as inoculum for the propagation step (according to
[27]. Both first and second seed cultures were performed in erlen-
meyer flasks incubated in a shaker (Innova 44, New Brunswick Sci.
Company Inc., New Jersey, USA). Propagation was performed in a
20 L bioreactor (BIOFLO 520, New Brunswick Sci. Company Inc.,
New Jersey, USA) and the cream yeast obtained was stored in cold
storage (7 °C), being reactivated in a 1.5 L bioreactor (BIOFLO 115,
New Brunswick Sci. Company Inc., New Jersey, USA) at 30 °C,
200 rpm, pH 2.5 and 2.0 vvm (gas volume flow per unit of liquid
volume per minute) for 30 min before the fermentation step. After
reactivation, yeasts were centrifuged and the cream yeast was used
as inoculum for the fermentation flask. The fermentation was per-
formed in 250 mL erlenmeyer flasks containing 100 mL of medium
as described before (second seed culture), except for the carbon
source that consisted in 84 mL of DPH. Agitation, temperature
and pH conditions were 200 rpm, 30 °C and 5.0. The yeast’s initial
concentration was 5.0 g L1. Two fermentation sets were
75
performed, one for the non-detoxified post-hydrolysates, other
for the DPH, both sets were performed in duplicates. Reducing sug-
ars content (glucose and xylose) was determined via HPLC
(Section 2.6.2).
2.6. Compositional analyses
2.6.1. Solid samples
Samples of raw or pretreated bagasse were air-dried to less than
10% (w/w) of moisture content and then milled to obtain particle
size of 0.12 mm in a knife mill (Pulverisette 19, Fritsch GmbH, Idar
– Oberstein, Germany). After quantification of extractives (per-
formed only for the raw bagasse) and ashes, the materials were
milled again in a shear and impact mill (Pulverisette 14, Fritsch
GmbH, Idar – Oberstein, Germany), to obtain a particle size <
0.5 mm. Structural carbohydrates (glucan, xylan, arabinan) and
soluble and insoluble lignin were quantified according to the
National Renewable Energy Laboratory (NREL) standard methods
for lignocellulosic feedstocks [33]. Moisture content of biomass
was determined using an automatic infrared moisture analyzer
MA35 (Sartorius Gmbh, Goettingen, Germany).
2.6.2. Liquid samples
Monosaccharides and organic acids (formic and acetic acids) in
the hydrolysates were analyzed using a HPLC system 1260 Infinity
(Agilent, Santa Clara, USA) equipped with refractive index (RI)
detector. The analytical Aminex HPX-87H column
(300 mm  7.8 mm, 5 lm) was used in combination with a guard
column Micro-Guard Cation PC H Refill Cartridges (Bio-Rad Labora-
tories, Hercules, USA). The mobile phase was sulfuric acid
(5 mmol L1) at flow rate of 0.6 mL min1. For the analysis of fur-
fural, 5-hydroxymethylfurfural (HMF) and some of the major phe-
nolic monomers (4-hydroxybenzaldehyde, 4-hydroxybenzoic acid,
syringic acid, vanillinic acid and vanillin), a reversed-phase HPLC
equipped with an Acclaim 120 C18 column (150 mm  4.6 mm,
3 lm) and a single wavelength UV detector were used. The mobile
phase was water-acetonitrile 1:8 (v/v) with 1% acetic acid (v/v) at a
flow rate of 0.8 mL min1.
Xylooligosaccharides (XO’s) were determined according to Mok
and Antal [22] quantitative saccharification procedure with a
quantitative post-hydrolysis with sulfuric acid 4% (w/w) at
120 °C for one hour. The XO’s content was calculated as the differ-
ence between the monosaccharide content measured via HPLC
before and after the acid hydrolysis.
Total phenolic compounds content was determined by UV-
spectrophotometry at the wavelength of 280 nm. Samples ana-
lyzed in such way were diluted, when necessary, and had their
pH increased to 12.0 by a concentrated sodium hydroxide solution
according to Gouveia et al. [12].
3. Calculations
The combined severity factor (CSF) is a parameter useful for
comparisons between pretreatments performed under acidic con-
ditions. Although it usually describes solids decomposition, it can
be used for approximate estimates in xylose recovery [4,35]. The
CSF is described in Eq. (1).
  
T  T ref
CSF ¼ log t  exp  pH ð1Þ
14:75
where ‘‘t” is the residence time of treatment in min, T is the treat-
ment temperature in °C, Tref is the reference temperature (100 °C)
and the pH is the value measured after the pretreatment.
Xylose conversion during the post-hydrolysis was calculated by
Eq. (2).
76 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84
½xylosereleased 
npost-hydrolysis ¼  100%
½xylosemaximum 
where [xyloserelease] is the monomeric xylose concentration
released during the post-hydrolysis and [xylosemaximum] is the max-
imum equivalent xylose concentration obtained from the quantita-
tive saccharification as described by Mok and Antal [22]. Equivalent
xylose concentration stands for the total xylose content in terms of
monosaccharides, oligosaccharides and its degradation product,
furfural, similarly calculated by [16].
Ethanol volumetric productivity and ethanol yield for the fer-
mentation were calculated by Eqs. (3) and (4), respectively.
½Ef  ½E0
QP ¼
tf
where [E] refers to ethanol concentration at a given time and tf is
the fermentation time.
½Ef  ½E0 100%
Yield ¼  ð4Þ
½RS0  ½RSf 0:511
where [E] and [RS] refers to ethanol and reducing sugars concentra-
tions (glucose and xylose) at a given time. The value 0.511 g g1
corresponds to the stoichiometric conversion of xylose and glucose
into ethanol for S. stipitis.
4. Results and discussion
4.1. Composition of pretreated bagasse and HH
The chemical composition of the pretreated bagasse, expressed
as percentage of dry mass, was: cellulose, 60.30 ± 0.16; hemicellu-
loses, 5.30 ± 0.47; lignin, 32.80 ± 0.94 and ashes, 8.90 ± 0.76. The
pretreatment mass yield was 64.80 ± 1.07%. The chemical composi-
tion of the hemicellulosic hydrolysate (HH), expressed in g L1,
was: xylose, 49.00 ± 0.84; glucose, 7.80 ± 0.11; arabinose,
3.6 ± 0.06; furfural, 0.03 ± 0.003; HMF, 0.2 ± 0.02; formic acid,
0.7 ± 0.01; acetic acid, 1.2 ± 0.02. Most of the solubilized carbohy-
drates are present as oligomers: 77.2% of the xylose, 89.3% of the
glucose and 76.3% of arabinose. The total hemicellulose recovery
was 65.0 ± 1.2%.
Approximately 80% of the hemicelluloses solubilized and recov-
ered were presented in oligomeric form in the HH. Duarte et al. [7]
performed a hydrothermal pretreatment of wheat straw and stated
that approximately 80% of soluble sugars in oligomeric form in the
HH is a typical value because the conditions leading to the highest
sugar recovery do not lead to their complete hydrolysis. In fact,
more severe pretreatment conditions would favor sugar degrada-
tion pathways and the total sugar recovery would therefore
decrease.
4.2. 22 Full factorial design of the acid post-hydrolysis
4.2.1. Sulfuric acid post-hydrolysis
Xylose, furfural, arabinose and acetic acid profiles for the seven
assays of the 22 factorial design with sulfuric acid are depicted in
Fig. 1(a–d). With respect to the production of monomeric xylose
(Fig. 1a), there is a tendency towards a plateau formation close
to the maximum xylose concentration that can be obtained (mono-
mers plus oligomers in the pretreatment liquor), 49.0 g L1 (as
presented in the Section 4.1). The plateau—rapidly distinguished
between 15 and 20 min of reaction (except for the mildest condi-
tion, 120 °C and 0.5% w/w, and the most severe condition, 150 °C
and 2% w/w)—indicates the depletion of XO’s that can be hydro-
lyzed and shows that such hydrolysis follows a fast kinetics. Fig. 1c
indicates that the monomeric arabinose recovery ranged between
2.0 and 3.0 g L1 and its hydrolysis pattern was not similar to
ð2Þ
those of xylose or acetic acid.
Fig. 1d shows that XO’s hydrolysis process occurs concomitantly
to the release of acetic acid, as also observed by Aguilar et al. [1] on
the dilute sulfuric acid pretreatment of sugarcane bagasse.
Hemicelluloses in sugarcane bagasse consist mainly of a backbone
of D-xylose residues partially substituted with L-arabinose,
D-glucuronic acid and other minor side groups. The xylose residues
may be O-acetylated to some extent [11].
Maloney et al. [20], in a work on the dilute acid hydrolysis of
paper birch, found that deacetylation occurs at a faster rate than
xylose release at 100–130 °C, whereas the two rates are virtually
equal at higher temperature ranges, 150–170 °C. It is noteworthy
ð3Þ to mention that the aforementioned work dealt with hemicellu-
loses hydrolysis in heterogeneous media. In this work, it was also
found the same temperature trend with XO’s hydrolysis in liquid
medium. At lower temperatures, 120–135 °C, the deacetylation
rate was slower than the xylose release rate, while at higher tem-
perature, 150 °C, both rates were nearly equal.
The most severe condition of the experimental design, 150 °C
and 2.0% (w/w), did not present the xylose plateau due to rapid
degradation of xylose. Furfural profiles (Fig. 1b) clearly indicates
this hypothesis, since concentrations up to 10 g L1 were reached.
These results prove Kim et al. [16] assumption that low tempera-
tures and high acid loadings lead to higher xylose conversion.
The HMF present in the HH’s were generated under hydrothermal
pretreatment by glucose dehydration pathways. This inhibitor was
present in very low concentrations (below 0.3 g L1) and remained
as such throughout the post-hydrolyses because they were not
severe enough to cause more glucose degradation.
During the post-hydrolysis assays performed with sulfuric acid
under severe temperature conditions (150 °C), there was precipita-
tion of large amounts of a black brilliant solid on the impeller and
inner walls of the reactor. At first, it was thought being consisted
only of condensed lignin, as it was also reported by Saska and Ozer
[29], but there was also a high degree of sugar degradation indicat-
ing humins formation as well. Such precipitation may occur by a
synergistic combination between temperature and acid loading;
once by only adding the acid at ambient temperature or by only
increasing the temperature with no acid addition there was no pre-
cipitation. Further characterization of the solids through elemental
analysis, IR, C13 solid-state NMR and pyrolysis-GC–MS would be
required in order to find out whether occurred both lignin and/or
humins precipitation. Post-hydrolysis assays carried out with
maleic and oxalic acids had no significant lignin precipitation com-
pared with sulfuric acid; this was probably due to the much minor
degree of sugar degradation and higher pH of the liquid medium.
4.2.2. Maleic acid post-hydrolysis
Xylose profiles as a function of post-hydrolysis time for maleic
acid are shown in Fig. 2(a–d). The same plateau formation for
xylose release can be observed (Fig. 2a), but due to the slower
kinetics of hydrolysis, the plateaus are achieved with longer reac-
tion times (except in the most severe condition, 150 °C and 2.0%
w/w). However, xylose was less degraded when compared to the
post-hydrolysis with sulfuric acid; showing a higher selectivity
for maleic acid.
Fig. 2b shows that the furfural concentrations were much lower
when compared to Fig. 1b (sulfuric acid). Likewise, the range of
monomeric arabinose formed in Fig. 2c is narrower than in Fig. 1c.
Acetic acid profiles (Fig. 2d) indicate that it was completely
released at 150 °C and 2.0% w/w and 135 °C and 1.25% w/w, though
the latter has not reached the xylose plateau, confirming again that
the hydrolysis of acetyl groups is faster than the post-hydrolysis of
XO’s.
 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84 77

60
10
50
8

Furfural (g·L1)
Xylose (g·L-1)
40
6
30
20 4

10 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)
(a) (b)

5 10

4 8

Acetic acid (g·L-1)

Arabinose (g·L-1)

3 6

2 4

1 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)
(c) (d)
Fig. 1. Profiles of (a) xylose, (b) furfural (c) arabinose and (d) acetic acid as a function of hydrolysis time for the five conditions of the 22 factorial design with sulfuric acid:
120 °C and 0.5% w/w ( ), 120 °C and 2.0% w/w ( ), 135 °C and 1.25% w/w ( ), 150 °C and 0.5% w/w ( ) and 150 °C and 2.0% w/w ( ). The standard
deviation of the experiments was calculated with the center point triplicate of the factorial design (135 °C and 1.25% w/w).

60
10
50
8
Furfural (g·L-1)
Xylose (g·L-1)

40
6
30
4
20

10 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)
(a) (b)
5
10
4
Arabinose (g.L-1)

Acetic acid (g.L-1)

8
3
6
2 4
1 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)

(c) (d)

Fig. 2. Profiles of (a) xylose, (b) furfural (c) arabinose and (d) acetic acid as a function of hydrolysis time for the five conditions of the 22 factorial design with maleic acid:
120 °C and 0.5% w/w ( ), 120 °C and 2.0% w/w ( ), 135 °C and 1.25% w/w ( ), 150 °C and 0.5% w/w ( ) and 150 °C and 2.0% w/w ( ). The standard
deviation of the experiments was calculated with the center point triplicate of the factorial design (135 °C and 1.25% w/w).

Lu and Mosier [19] showed that sulfuric acid in fact degrades enzymes in nature; later reinforced by Kim et al. [16] upon the use
3–10 times more hemicelluloses from corn stover than maleic acid. of dicarboxylic acids (maleic and oxalic) in acid post-hydrolysis, this
Also, according to the authors, the activation energy (Ea) for the hypothesis would require further mechanistic studies.
xylose degradation into furfural with maleic acid is 2.8 times greater With respect to the thermal degradation of maleic acid reported
than the Ea for the same reaction with sulfuric acid. The hypothesis is by Kim et al. [16], the HPLC analysis have shown that there was no
that maleic acid stabilizes xylose during the hydrolysis of XO’s significant degradation over the post-hydrolysis carried out in this
thereby mimicking the role of the active site of hemicellulolytic work. Indeed, Kim et al. [16] performed post-hydrolysis with
78 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84

longer reaction times (30 h), which has favored the thermal degra-
dation of maleic acid to byproducts (malic acid and fumaric acid).
4.2.3. Oxalic acid post-hydrolysis
Xylose and furfural profiles for the 22 factorial design with oxa-
lic acid are shown in Fig. 3(a and b). A tendency for a plateau for-
mation of xylose release can also be observed (150 °C/2.0%,
150 °C/0.5% and 135 °C/1.25% w/w); however, none of the plateaus
reached the maximum xylose available (49 g L1). Furfural, arabi-
nose and acetic acid profiles (Fig. 3(b–d)) are similar to those pro-
duced by the post-hydrolysis with maleic acid, so the same
considerations made for maleic acid are valid for oxalic acid as
well. Indeed, both are dicarboxylic acids and differ only by the car-
bon chain size and the presence of a double bond in maleic acid.
4.2.4. Xylose conversions
The xylose peaks for sulfuric, maleic and oxalic acids post-
hydrolysis were summarized in Table 1. It is worth mentioning
that such peaks do not correspond to the actual xylose maxima,
they correspond to the time when xylose reached a high conver-
sion and further increment in conversion were considered minor
along time. For example, at 135 °C and 1.25%, the xylose peak
(95.5% conversion) was considered in 20 min of reaction in Table 1,
but in fact xylose maximum occurred in 30 min (99% conversion,
data not shown); however, the increment in conversion does not
offset the longer reaction time necessary for this. Reaction time
is a precious operational parameter in industrial processes, short
reaction times are therefore desired.
For sulfuric acid, the maximum conversion values were almost
quantitative for all conditions of the design. Milder temperature
conditions, however, showed slightly higher conversions than the
most severe ones due to less sugar degradation, i.e. furfural forma-
tion (Fig. 1(b and c)), although they required longer reaction times.
In terms of xylose peak, the best condition for sulfuric acid’s design
was at 120 °C, 1.25% and 20 min of reaction time for the highest
conversion and minimal furfural generation (Table 1).
Maximum xylose conversions for maleic and oxalic acids were
not quantitative within the duration of the experiments except
for the most severe condition of maleic acid’s design (150 °C and
2.0% w/w), wherein the conversion was quantitative. In terms of
xylose peak, the best condition for maleic acid’s design was at
150 °C, 2.0% and 20 min; as for oxalic acid’s design, it was at
150 °C, 2.0% and 32 min (Table 1). It may be noticed that more sev-
ere conditions were necessary for maleic acid (CSP of 1.53 ± 0.02)
and oxalic acid (CSP of 1.65 ± 0.05) to obtain fast xylose peaks than
sulfuric acid (CSP of 1.30 ± 0.15).
The boom of studies on hydrothermal pretreatment of biomass
in the 1990s [9,22] was progressively followed by studies on the
downstream processing of the generated oligosaccharides in the
2000s [6,7,10]. The current decade represents the consolidation
of the studies from the past decades. This section aims to provide
a compilation of post-hydrolyses studies of biomass and contextu-
alization of this work in the literature. Table 2 summarizes the
most important studies in the area.
In four of the eight mentioned papers in Table 2 the post-
hydrolysis was carried out at a fixed temperature of 121 °C in an
autoclave (studies 3, 6, 7 and 8). This temperature is the standard
for carrying out the quantitative saccharification with sulfuric acid
4.0% (w/w or v/v) for one hour [22]. Fortuitously, this temperature
is relatively low and requires higher acid concentrations.
All studies varied the employed acid loadings. The ranges varied
from 0.1% (w/w) up to 4% (w/w) (default quantitative post-
hydrolysis condition). In this study (0.5–2.0% range), two organic
acids that are solids at room temperature and with solubility
restraints were used, precluding the study with higher loadings.
Besides, in terms of process, loadings as high as 4% (w/w) would
mean higher costs with reagents and corrosion resistant reactors.
4.3. Fermentation of detoxified and non-detoxified C5 sugars
Fermentation assays were performed in order to provide a feed-
back of the post-hydrolysis in terms of ethanol production. This is

60
10
50
8
Furufral (g·L-1)
Xylose (g·L-1)

40
6
30

20 4

10 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)
(a) (b)

5 10
Acetic acid (g.L-1)

4 8
Arabinose (g.L-1)

3 6

2 4

1 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)
(c) (d)
Fig. 3. Profiles of (a) xylose, (b) furfural (c) arabinose and (d) acetic acid as a function of hydrolysis time for the five conditions of the 22 factorial design with oxalic acid:
120 °C and 0.5% w/w ( ), 120 °C and 2.0% w/w ( ), 135 °C and 1.25% w/w ( ), 150 °C and 0.5% w/w ( ) and 150 °C and 2.0% w/w ( ). The standard
deviation of the experiments was calculated with the center point triplicate of the factorial design (135 °C and 1.25% w/w).
 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84 79

Table 1
Xylose conversion, maximum xylose recovery, furfural and combined severity factor (CSF) for the post-hydrolysis with sulfuric, maleic and oxalic acids. The best conditions for
each acid in terms of xylose peak were marked in bold. The standard deviation was calculated with the center point triplicate of the factorial design (135 °C and 1.25% w/w).

Acids Condition (°C; % w/w; min) Xylose peak (g L1) Furfural (g L1) Final pH CSF Conversion (%)
120; 0.5; 70 46.2 ± 2.0 0.7 ± 0.15 1.21 ± 0.01 1.28 ± 0.15 94.2 ± 0.02
120; 2.0; 20 47.7 ± 2.0 0.8 ± 0.15 0.59 ± 0.01 1.30 ± 0.15 97.3 ± 0.02
Sulfuric acid 135; 1.25; 20 46.8 ± 2.0 1.6 ± 0.15 0.84 ± 0.01 1.36 ± 0.15 95.5 ± 0.02
150; 0.5; 15 44.7 ± 2.0 1.1 ± 0.15 1.24 ± 0.01 1.41 ± 0.15 91.2 ± 0.02
150; 2.0; 15 45.0 ± 2.0 4.0 ± 0.15 0.64 ± 0.01 2.01 ± 0.15 91.8 ± 0.02
120; 0.5; 104 20.4 ± 1.4 0.3 ± 0.03 2.22 ± 0.05 0.39 ± 0.02 41.2 ± 0.02
120; 2.0; 95 44.0 ± 1.4 0.6 ± 0.03 1.41 ± 0.05 1.16 ± 0.02 89.8 ± 0.02
Maleic Acid 135; 1.25; 95 47.4 ± 1.4 1.5 ± 0.03 1.68 ± 0.05 1.38 ± 0.02 97.6 ± 0.02
150; 0.5; 50 42.2 ± 1.4 1.8 ± 0.03 2.10 ± 0.05 1.06 ± 0.02 86.1 ± 0.02
150; 2.0; 20 48.9 ± 1.4 1.4 ± 0.03 1.42 ± 0.05 1.53 ± 0.02 99.8 ± 0.02
120; 0.5; 93 27.7 ± 0.9 0.3 ± 0.05 1.30 ± 0.02 0.35 ± 0.05 53.5 ± 1.80
120; 2.0; 93 43.5 ± 0.9 0.8 ± 0.05 2.03 ± 0.02 1.15 ± 0.05 88.8 ± 1.80
Oxalic Acid 135; 1.25; 65 45.8 ± 0.9 1.2 ± 0.05 2.05 ± 0.02 1.06 ± 0.05 93.5 ± 1.80
150; 0.5; 40 29.9 ± 0.9 1.1 ± 0.05 1.59 ± 0.02 1.16 ± 0.05 61.0 ± 1.80
150; 2.0; 32 43.6 ± 0.9 2.0 ± 0.05 1.32 ± 0.02 1.65 ± 0.05 89.0 ± 1.80

Table 2
Summary of works upon the acid post-hydrolysis of several biomasses compared to the current work (in bold). PT stands for pretreatment. HPT stands for hydrothermal
pretreatment.

Work Biomass Fermentation? Kinetic Post-hydrolysis Acid Scale Reaction Total Ref.
study? temperature loading time (h) xylose
(°C) (% w/w) conc.a
(g L1)
1 HTP and post-hydrolysis Sugarcane bagasse C5 Yes 120–150 0.5–2.0 2.0 L Parr 0–1.5 49 This
with sulfuric, maleic and reactor work
oxalic acids
2 HTP and post-hydrolysis Eucalyptus wood No Yes 100–135 0.5–2.0 Not 0–10 20 [10]
with sulfuric acid mentioned
3 Acid and enzymatic post- Brewery spent grain C5/C6 No 121 1.0–4 Schott flasks 0.25 and 25 [6]
hydrolysis (in autoclave) 1
4 HPT and post-hydrolysis Corn stover No No 101–130 0.25–1.0 High 0–8.3 16 [37]
with sulfuric acid throughput
5 HTP and post-hydrolysis Poplar wood No Yes 120–180 0.2–1.0 4 mL vials 0–30 Not [16]
with sulfuric, maleic and mentioned
oxalic acids
6 HPT and post-hydrolysis Wheat straw No Yes 121 0.1–4.0 Schott flasks 1 9 [7]
with sulfuric acid (in autoclave)
7 PT with steam explosion Douglas fir wood C6 No 121 0.1–3.0 27 mL Schott 1 25 [30]
and post-hydrolysis with flasks (in
sulfuric acid autoclave)
8 HPT and post-hydrolysis Eucalyptus residues, No Yes 121 1–4.0 50 mL Schott 0.1–1 20.5 [32]
with sulfuric acid wheat straw and olive flasks (in
tree pruning autoclave)
a
The acid post-hydrolyses had practically quantitative conversion of XO’s to xylose therefore reaching near the maximum total xylose concentration. The enzymatic post-
hydrolysis conversions in Duarte et al. [6] study varied from 60 to 75%.

necessary because depending on the temperature and acid loading
in the post-hydrolysis, some types of sugars and lignin byproducts
such as glycolaldehyde, furfural, HMF, carboxylic acids and pheno-
lic compounds [13,34] might form and impact negatively on fer-
mentation even if there were high xylose and low furfural
concentrations in the hydrolysates.
The results obtained for the fermentation set without prior
detoxification showed no consumption of xylose and consequently
ethanol production (data not shown). This may be associated with
inhibitory compounds present in the non-detoxified post-
hydrolysates. Ethanol production by S. stipitis is strongly inhibited
by acetic acid and delayed by furfural and HMF [3]. It may by
observed in Tables 3 and 4 that the employed detoxification
method decreased the overall concentrations of the inhibitors eval-
uated in this work. However, it may be also noted that the condi-
tion of 150 °C/0.5% from sulfuric acid’s design still presented high
total phenolics and acetic acid concentrations (Table 4).
Despite being costly and time-demanding, detoxification of
samples was mandatory in order to decrease the amount of
compounds that are harmful to the yeast. There are no works
in the literature that carry out fermentations without prior
detoxification of post-hydrolysates [6,36]. An outlook of the
detoxification method may be seen in a tentative mass balance
shown in Table 5. It may be observed that overliming employed
very low amounts of CaO and they were correlated with the acid
loading in the post-hydrolysates; the highest acid loadings—
120 °C; 2.0% and 150 °C; 2.0%—presented higher solids mass
obtained after centrifugation. Adsorption also employed low
chemicals amount. On the other hand, acidification had the
greatest chemicals amount utilization and maleic acid samples
required more H3PO4 than sulfuric acid probably due to buffer
formation after overliming, once maleic acid is a weak acid. Mass
loss ranged from 7.6 to 13.6% for the overall process mainly
because the detoxification process wasn’t continuous, each step
was performed separately. The detoxification method was able
to remove in average 99.8 ± 0.50% of furfural, 78 ± 9.80% of
HMF, 75.8 ± 9.80% of acetic acid and 60.4 ± 7.9% of total phenolic
compounds (data now shown).
80 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84

Table 3
Inhibitors content in the raw (HPH) and detoxified (DPH) hemicellulosic post-hydrolysates in sulfuric acid samples.

Sulfuric acid
Design conditions °C; % (w/w); min 120; 0.5; 70 120; 2.0; 20 135; 1.25; 20 150; 0.5; 15 150; 2.0; 15
Fermentation yield %a 56.5 ± 8.25 52.1 ± 2.02 57.0 ± 1.22 0 40.6 ± 5.49

HPH DPH HPH DPH HPH DPH HPH DPH HPH DPH
4-Hydroxybenzaldehyde 0.174 0.002 0.137 0.001 0.138 0.001 0.209 0.003 0.119 0.002
4-Hydroxybenzoic acid 0.018 0.015 0.020 0.005 0.018 0.004 0.020 0.02 0.012 0.008
Syringic acid 0.034 0.013 0.037 0.01 0.036 0.009 0.045 0.013 0.035 0.011
Vanillinic acid 0.026 0.016 0.026 0.007 0.025 0.008 0.034 0.019 0.021 0.012
Vanillin gL1 0.109 0.006 0.104 0c 0.119 0c 0.156 0.01 0.101 0.002
HMF 0.238 0.085 0.211 0.044 0.196 0.050 0.189 0.03 0.270 0.054
Furfural 0.525 0c 0.709 0c 1.037 0c 0.865 0.012 2.880 0c
Total phenolicsb 6.327 4.080 7.006 3.515 5.022 3.231 12.974 5.404 8.795 2.623
Acetic acid 5.419 1.750 6.142 1.711 4.842 1.397 5.539 3.100 5.631 1.308
a
Yield for the fermentation of the detoxified post-hydrolysates (DPH’s). The raw post-hydrolysates (HPH’s) did not ferment.
b
Total phenolics determined by UV spectrophotometry.
c
Below the detection limit of the method.

Table 4
Inhibitors content in the raw (HPH) and detoxified (DPH) hemicellulosic post-hydrolysates in maleic acid samples.

Maleic acid
Design conditions °C; % (w/w); min 120; 0.5; 104 120; 2.0; 95 135; 1.25; 95 150; 0.5; 55 150; 2.0; 20
Fermentation yield %a 49.1 ± 7.23 57.7 ± 3.34 46.4 ± 4.48 65.8 ± 3.51 55.8 ± 1.75

HPH DPH HPH DPH HPH DPH HPH DPH HPH DPH
4-Hydroxybenzaldehyde 0.163 0.002 0.173 0.002 0.161 0c 0.152 0c 0.154 0c
4-Hydroxybenzoic acid 0.017 0.009 0.015 0.005 0.019 0.006 0.013 0c 0.014 0c
Syringic acid 0.034 0.019 0.039 0.015 0.045 0.014 0.041 0.004 0.042 0.002
Vanillinic acid 0.029 0.019 0.026 0.013 0.028 0.015 0.022 0.002 0.025 0.001
Vanillin gL1 0.119 0.009 0.113 0.006 0.128 0.005 0.105 0c 0.117 0c
HMF 0.234 0.104 0.214 0.121 0.287 0.135 0.227 0.120 0.303 0.136
Furfural 0.505 0.003 0.687 0.004 1.611 0.013 2.095 0.001 1.552 0.007
Total phenolicsb 9.622 4.272 8.588 8.310 7.995 6.621 9.313 4.526 8.180 7.031
Acetic acid 3.221 2.819 5.577 2.633 5.526 3.752 5.145 2.966 5.634 3.099
a
Yield for the fermentation of the detoxified post-hydrolysates (DPH’s). The raw post-hydrolysates (HPH’s) did not ferment.
b
Total phenolics determined by UV spectrophotometry.
c
Below the detection limit of the method.

For all samples, there was a considerable decrease in total phe-
nolics (at both low molecular fragments and single molecules
level). It is also worth mentioning that not all single phenolic mole-
cules were analyzed due to the multitude of possible phenolics
being generated by lignin degradation pathways. Total phenolics
content determination by UV is commonly used but may be biased
by other UV-absorbing compounds such as humins and furan
derivatives although furfural and HMF absorbance contributions
had been taken into account in total phenolics calculation as
described by [12].
In Fig. 4, xylose, ethanol and dry cell mass profiles were drawn
as a function of fermentation time (0–72 h). Fermentation results
of the detoxified post-hydrolysates (DPH’s) are summarized in
Table 6. It is important to highlight that, for all the samples in this
work, there was no xylose consumption within 24 h of fermenta-
tion, except for the xylose control, as can be seen in Fig. 4(a and
b), probably due to the preferable reduction of HMF into its respec-
tive alcohols over glucose consumption by the yeast [34]. In this
way, it may be verified that the yeast consumed HMF (Table 6).
Furthermore, as arabinose content in the DPH’s was quite low
when compared to xylose (2 g L1 average, data not shown) and
its consumption was not complete for all samples along the fer-
mentation, it was decided not to show arabinose fermentation pro-
files. Ferrari et al. [8] also reported partial arabinose assimilation
by S. stipitis of wheat straw HH’s but without fermentation.
Although it has a slower bioconversion rate, HMF is a less severe
inhibitor for many microorganisms than furfural [26]. In this sense,
the fermentation of maleic acid’s DPH’s was longer than sulfuric
acid’s due to their higher HMF concentration (Tables 3 and 4). At
150 °C and 0.5%, maleic acid’s fermentation sample had almost
seven times more HMF than the sulfuric acid’s one (Table 6). How-
ever, HMF concentrations were in general low in all fermentation
samples, varying from 0.05 to 1.89 g L1 for HMF (Table 6).
For the sulfuric acid’s DPH’s, most samples were fermented
within 48 h and their maximum ethanol titer varied from 1.6
(150 °C, 0.5% and 15 min, Table 6) up to 10.6 g L1 (120 °C, 0.5%
and 70 min, Table 6). The fermentations showed volumetric pro-
ductivities lower than the fermentation control (0.37 g L1 h1).
The fermentation yield that most closely resembles the
fermentation control (62.7%) was for the DPH obtained at
135 °C/1.25%, 57%.
The ethanol production from 48 to 72 h was negligible for all
the DPH’s with sulfuric acid (Fig. 4). The sample generated at
150 °C and 0.5% (w/w) did not ferment, probably due to the
presence of a severe inhibitor known as glycolaldehyde. Hot-
pressurized treatment of hydrolysates may produce glycolalde-
hyde by retro-aldol condensation of glucose and xylose [13]. This
compound is one of the main culprits for inhibiting fermentation
after pressurized hot water degradation of lignocellulosic materials
[13,14,34]. As shown by Jayakodi et al. [13], glycolaldehyde con-
centrations as low as 2 mM in the fermentation medium may
impart strong fermentation inhibition. More investigation would
be necessary, however, in order to study the formation and quan-
tification of glycolaldehyde in the post-hydrolysates.
 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84 81

Table 5
Material balance for the detoxification method employing evaporation, overliming with CaO, acidification with H3PO4 and adsorption with activated charcoal. The input and
output liquid streams correspond to the raw (HPH) and detoxified (DPH) hemicellulosic post-hydrolysates in sulfuric and maleic acid samples.

Evaporationa Overlimingb Acidificationc Adsorptiond Overall

Stream Input Input Output Input Output Input Output Input Output Output Mass Mass
Description HPH H2O H2O CaO Solids H3PO4 Solids A. charc. Solids DPH balancee lossf
(g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (g) (%)
120; 0.5; 70 200.53 159.33 158.34 2.79 0.31 5.01 0.54 1.78 2.55 179.51 207.71 13.6
Sulfuric acid 120; 2.0; 20 200.42 157.43 155.33 2.53 7.41 3.83 0.00 1.34 1.82 181.47 200.98 9.7
(°C, % w/w, min) 135; 1.25; 20 200.32 159.97 157.8 1.50 3.47 3.90 0.00 1.76 2.44 182.65 203.73 10.3
150; 0.5; 15 203.42 157.55 154.26 2.23 1.11 12.44 0.97 1.79 2.31 193.07 218.77 11.7
150; 2.0; 15 200.63 158.26 156.6 3.57 8.38 3.76 0.01 1.46 1.88 185.57 200.80 7.6
120; 0.5; 104 200.12 158.91 156.82 1.32 0.67 8.09 0.01 1.89 2.54 184.68 210.29 12.2
Maleic acid 120; 2.0; 95 200.06 156.32 154.04 3.09 0.39 17.14 0.01 1.98 2.83 197.29 221.33 10.9
(°C, % w/w, min) 135; 1.25; 95 200.21 159.73 157.34 2.21 3.64 13.28 0.01 1.92 2.93 190.49 213.42 10.7
150; 0.5; 50 171.71 136.53 136.18 0.86 0.39 6.12 0.01 1.57 2.07 160.26 178.13 10.0
150; 2.0; 20 200.22 159.62 157.64 3.00 0.35 21.53 0.00 2.03 2.97 201.07 225.44 10.8
a
In the evaporation step, distilled water was the input and the evaporated water was the output.
b
In the overliming step, CaO was the input and the centrifuged solids were the output.
c
In the acidification step, H3PO4 was the input and the centrifuged solids were the output.
d
In the adsorption step, activated charcoal was the input and the centrifuged solids were the output.
e
The mass balance is the overall sum of input and output masses for the detoxification.
f
Mass losses are mainly related to losses in transference from flasks.

There are few works in the literature dealing with the produc-
tion of hemicellulosic hydrolysates with maleic acid and further
fermentation. Indeed, it was only found one work, by Jung et al.
[15], which studied the pretreatment of biomass of oil palm empty
fruit bunches with maleic acid, subsequent enzymatic hydrolysis
and fermentation of the whole slurry with S. cerevisiae. They
reported that maleic acid concentrations of up to 2% (w/w) did
not affect the yield or productivity of the yeast used.
For maleic acid’s DPH’s, most samples were fermented within
72 h with maximum ethanol concentrations varying between 7.1
and 10.6 g L1 (Table 6). This is probably due to the higher furfural
concentrations than in sulfuric acid samples. The volumetric pro-
ductivities achieved were much lower than those of sulfuric acid’
samples and the control, except for the condition of 150 °C/0.5%
(0.22 g L1 h1). Also for this condition, the fermentation yield
was higher than the fermentation control (65.8%). The fermenta-
tion yields were in average high, varying from 49.1 to 65.8%.
In general, fermentation results for both the acid post-
hydrolysates showed similar results than those obtained in the lit-
erature for other types of pretreatments. Ferrari et al. [8] obtained
a yield of 68.6% with a maximum ethanol concentration of
12.6 g L1 within 75 h of fermentation of eucalyptus wood
hemicellulosic hydrolysate with the same yeast, S. stipitis NRRL
Y-7124, and 30 g L1 of initial xylose. In the present study, it
was managed to obtain up to 10.6 g L1 of ethanol within 48 h
of fermentation (sulfuric acid, 120 °C and 0.5%) with a productivity
of 0.20 g L1 h1. The lower productivity obtained by Ferrari et al.
[8], of 0.17 g L1 h1, was probably due to their initial dry cell
mass concentration being lower than in this work, 0.6 gL1
compared to 5.0 gL1. Higher cell concentrations were used in this
work envisioning a fermentation process with cell recycling; such
methodology is employed in order to increase process productiv-
ity, reduce the need for yeast cell propagation, and utilize less
sugar for biomass growth [2].
In Table 6 it may also be noted that the concentration of pheno-
lic compounds had a slight reduction during fermentation. These
compounds can be metabolized by yeasts via oxidative mecha-
nisms [17]. Generally, such enzymes oxidize the phenolic interme-
diates to quinones, which in turn can act as potent inhibitory
agents of quinone reductases [24]. Further analyses of the interme-
diates from phenolics metabolism are needed, i.e., analyses of qui-
nones/hydroquinones and possible correlations between their
amounts and the degree of yeast inhibition.
Samples obtained with sulfuric acid usually showed higher
ethanol yields and productivities than the ones with maleic acid
under low and intermediary temperatures (120 and 135 °C). Dia-
metrically opposite, maleic acid produced more easily fermentable
post-hydrolysates under severe temperature conditions (150 °C).
The key to this fact might rely on the selectivity of the acids.
All samples with sulfuric acid peaked the monomeric xylose
concentration, 49 g L1, or came very close to this (except the con-
dition of 120 °C and 0.5%). Those xylose peaks were reached in dif-
ferent times. The post-hydrolysis carried out at 150 °C was harsher
and probably not only hydrolyzed XO’s, but also produced more
toxic inhibitors than the other conditions. In the post-hydrolysis
with maleic acid, a selective acid, the harsher condition (150 °C)
favored a more efficient hydrolysis of the XO’s compared to the
other conditions. A selective acid, therefore, is favored under more
severe temperature conditions.
This valuable information about the influence of selectivity of
acids in alcoholic pentose fermentation was not found in the liter-
ature. Kim et al. [16] were the ones who came closer to find that
out when performing a post-hydrolysis with sulfuric, maleic and
oxalic acids. However, they did not ferment the post-
hydrolysates after the post-hydrolysis on microscale.
Yu et al. [36], in a study similar to this work, carried out the
post-hydrolysis of the hydrothermal HH of sugarcane bagasse via
either sulfuric acid or enzymes (xylanase cocktails). They did not
vary the conditions of the acid post-hydrolysis such as time and
temperature; instead, they performed a quantitative post-
hydrolysis (4% w/w sulfuric acid at 120 °C). Samples obtained by
acid post-hydrolysis did not ferment even after detoxification
with ion exchange resins due to high inhibitors content. In
contrast, samples from enzymatic post-hydrolysis fermented with
a yield of 52.9% and 0.09 gL1 h1 of ethanol volumetric
productivity.
The results obtained from the C5 fermentation validated the
acid post-hydrolysis and showed that it is possible to efficiently
fractionate the hemicelluloses via hydrothermal pretreatment in
sugarcane bagasse and later use them for 2G ethanol production.
The detoxification of the post-hydrolysates was mandatory in
order to halter the toxic effect of some compounds generated dur-
ing the pretreatment and acid post-hydrolysis. Further analyses of
some specific inhibitors are necessary for the establishment of
cause/effect relationships upon the inhibition of the fermentation
and growth of the employed yeast.
82 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84

Fig. 4. Fermentation profiles of xylose ( ), dry cell mass ( ), and ethanol ( ) for sulfuric acid (a) and maleic acid (b) post-hydrolysates and xylose control. The profiles
were drawn by linking the experimental points, there were no attempts to perform data fitting. The dashed lines indicate the time chosen for the parameters calculation
(ethanol yield and volumetric productivity). The standard deviation was calculated by duplicates of the samples.
Table 6
Data summary for the fermentation of the detoxified post-hydrolysates (DPH) with sulfuric and maleic acid. Deviations were calculated with duplicate of the samples. The subscripts ‘‘0” and ‘‘final” stand for 0 h and 72 h of fermentation
time.

Sulfuric acid Maleic acid

Conditions (°C. % w/w and min) Pure xylose 120; 0.5; 70 120; 2.0; 20 135; 1.25; 20 150; 0.5 15 150; 2.0; 15 120; 0.5; 104 120; 2.0; 95 135; 1.25; 95 150; 0.5; 55 150; 2.0; 20

P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84

Xylose0 (g L1) 36.6 ± 2.38 33.5 ± 0.59 33.3 ± 0.01 28.8 ± 0.06 37.4 ± 0.50 21.4 ± 0.67 19.1 ± 0.06 35.1 ± 2.00 34.2 ± 0.02 31.6 ± 0.50 30.4 ± 0.31
Xylosefinal (g L1) 0.0 0.3 ± 0.02 0.5 ± 0.15 0.5 ± 0.04 29.0 ± 1.57 0.5 ± 0.02 0.6 ± 0.13 2.57 ± 0.02 6.54 ± 0.27 0.58 ± 0.07 0.38 ± 0.02
DCMfinal (g L1) 11.3 ± 2.45 13.8 ± 0.89 13.3 ± 0.21 13.2 ± 0.56 6.6 ± 0.00 13.0 ± 0.31 13.3 ± 0.00 10.9 ± 0.71 11.6 ± 2.52 13.6 ± 0.37 12.2 ± 0.06
Xylitolfinal (g L1) 2.2 ± 0.11 0.9 ± 0.05 0.9 ± 0.02 1.3 ± 0.18 4.2 ± 0.14 1.7 ± 0.38 0.3 ± 0.13 0.6 ± 0.13 1.22 ± 0.17 0.5 ± 0.02 0.6 ± 0.17
Acetic acid0 (g L1) 0 1.8 ± 0.02 1.7 ± 0.01 1.4 ± 0.01 2.2 ± 0.02 1.2 ± 0.01 1.9 ± 0.00 3.0 ± 0.09 3.6 ± 0.08 2.5 ± 0.02 2.9 ± 0.07
Acetic acidfinal (g L1) 0 0c 0c 0c 0.55 ± 0.11 0c 0.3 ± 0.00 1.1 ± 0.12 1.9 ± 0.29 0c 0.3 ± 0.01
Furfural0 (g L1) 0 0c 0c 0c 0c 0c 0c 0c 0c 0c 0c
Furfuralfinal (g L1) 0 0c 0c 0c 0c 0c 0c 0c 0c 0c 0c
HMF0 (g L1) 0 0.08 ± 0.001 0.03 ± 0.009 0.04 ± 0.004 0.02 ± 0.005 0.01 ± 0.001 0.11 ± 0.009 0.08 ± 0.004 0.10 ± 0.05 0.14 ± 0.001 0.10 ± 0.003
HMFfinal (g L1) 0 0c 0c 0c 0c 0c 0c 0c 0c 0c 0c
Phenolics0 (g L1) 0 3.1 ± 0.07 2.4 ± 0.06 2.1 ± 0.03 2.8 ± 0.18 1.5 ± 0.03 5.1 ± 0.07 3.8 ± 0.06 3.5 ± 0.08 4.1 ± 0.17 3.5 ± 0.04
Phenolicsfinal (g L1) 0 3.0 ± 0.03 2.2 ± 0.15 1.5 ± 0.13 3.3 ± 0.09 1.4 ± 0.04 4.7 ± 0.11 3.8 ± 0.21 3.1 ± 0.13 3.3 ± 0.32 3.2 ± 0.29
Fermentation time (h) 24 48 48 48 72 48 72 72 72 48 72
Ethanol (g L1)a 9.9 ± 1.21 10.6 ± 0.13 9.1 ± 0.12 8.3 ± 0.25 1.6 ± 0.15 7.2 ± 0.06 7.1 ± 0.75 10.6 ± 0.84 9.5 ± 0.35 8.7 ± 0.74 8.5 ± 0.55
Qp (g L1 h1)b 0.37 ± 0.01 0.17 ± 0.04 0.21 ± 0.00 0.19 ± 0.01 0.01 ± 0.00 0.09 ± 0.00 0.10 ± 0.01 0.15 ± 0.01 0.13 ± 0.00 0.22 ± 0.00 0.12 ± 0.01
Yieldethanol (%) 62.7 ± 1.52 56.5 ± 8.25 52.1 ± 2.02 57.0 ± 1.22 0 40.6 ± 5.49 49.1 ± 7.23 57.7 ± 3.34 46.4 ± 4.48 65.8 ± 3.51 55.8 ± 1.75
a
Maximum ethanol concentration.
b
Volumetric productivity in ethanol.
c
Below the detection limit of the method.

83
84 P.Y.S. Nakasu et al. / Fuel 185 (2016) 73–84
5. Conclusions
The kinetic study of the post-hydrolysis of sugarcane bagasse
hemicellulosic hydrolysate with diluted acids showed that it is
possible to obtain high levels of monomeric xylose with minimal
production of fermentation inhibitors. Sulfuric acid outperformed
both dicarboxylic acids (maleic and oxalic acids) by showing the
fastest kinetics of post-hydrolysis. Detoxification of the hydroly-
sates produced by sulfuric and maleic acids post-hydrolysis
enabled the C5 fermentation in erlenmeyer flasks, but considerable
chemicals utilization and time spent on such process demands fur-
ther study to avoid it. Ethanol yields ranged from 40.6 to 65.8% and
most samples fermented between 48 and 72 h with volumetric
productivities ranging from 0.02 to 0.43 g L1 h1.
Acknowledgements
The authors would like to thank the financial support from
FAPESP – São Paulo Research Foundation [grant number
2013/05369-2], and also would like to thank CNPEM (National
Centre for Research in Energy and Materials) that provided us
access to the CTBE laboratories (Brazilian Bioethanol Science and
Technology Laboratory).
References
[1] Aguilar R, Ramírez JA, Garrote G, Vázquez M. Kinetic study of the acid
hydrolysis of sugar cane bagasse. J Food Eng 2002;55:309–18.
[2] Basso LC, Basso TO, Rocha SN. Ethanol production in Brazil: the industrial
process and its impact on yeast fermentation. In: Bernardes MAS, editor.
Biofuel production-recent developments and prospects. InTech; 2011. p.
85–100.
[3] Bellido C, Bolado S, Coca M, Lucas S, González-Benito G, García-Cubero MT.
Effect of inhibitors formed during wheat straw pretreatment on ethanol
fermentation by Pichia stipitis. Bioresour Technol 2011;102:10868–74.
[4] Chum HL, Johnson DK, Black SK, Overend RP. Pretreatment-catalyst effects and
the combined severity parameter. Appl Biochem Biotechnol 1990;24–25:1–14.
[5] da Silva ARG, Torres Ortega CE, Rong B-G. Techno-economic analysis of
different pretreatment processes for lignocellulosic-based bioethanol
production. Bioresour Technol 2016;218:561–70.
[6] Duarte LC, Carvalheiro F, Lopes S, Marques S, Parajó JC, Gírio FM. Comparison of
two posthydrolysis processes of Brewery’s spent grain autohydrolysis liquor to
produce a pentose-containing culture medium. Appl Biotechnol Biochem
2004;113–116:1041–58.
[7] Duarte LC, Silva-Fernandes T, Carvalheiro F, Gírio F. Dilute acid hydrolysis of
wheat straw oligosaccharides. Appl Biotechnol Biochem 2009;153:116–26.
[8] Ferrari MD, Neirotti E, Albornoz C, Saucedo E. Ethanol production from
eucalyptus wood hemicellulose hydrolysate by Pichia stipitis. Biotechnol
Bioeng 1992;40:753–9.
[9] Garrote G, Dominguez H, Parajó J, Domínguez H, Parajó JC. Hydrothermal
processing of lignocellulosic materials. J Wood Wood Prod 1999;57:191–202.
[10] Garrote G, Domínguez H, Parajó JC, Domínguez H, Parajó JC. Generation of
xylose solutions from Eucalyptus globulus wood by autohydrolysis–
posthydrolysis processes: posthydrolysis kinetics. Bioresour Technol
2001;79:155–64.
[11] Gírio FM, Fonseca C, Carvalheiro F, Duarte LC, Marques S, Bogel-Lukasik R.
Hemicelluloses for fuel ethanol: a review. Bioresour Technol
2010;101:4775–800.
[12] Gouveia ER, Nascimento RT, Souto-Maior AM, Rocha GJM. Validação de
metodologia para a caracterização química do bagaço da cana-de-açúcar.
Quím 2009;32:1500–3.
[13] Jayakodi LN, Hayashi N, Kitagaki H. Identification of glycolaldehyde as the key
inhibitor of bioethanol fermentation by yeast and genome-analysis of its
toxicity. Biotechnol Lett 2011;33:285–92.
[14] Jayakody LN, Ferdouse J, Hayashi N, Kitagaki H. Identification and
detoxification of glycolaldehyde, an unattended bioethanol fermentation
inhibitor. Crit Rev Biotechnol 2016:1–13.
[15] Jung YH, Kim IJ, Kim HK, Kim KH. Whole slurry fermentation of maleic acid-
pretreated oil palm empty fruit bunches for ethanol production not
necessitating a detoxification process. Bioproc Biosyst Eng 2014;37:659–65.
[16] Kim Y, Kreke T, Ladisch MR. Reaction mechanisms and kinetics of xylo-
oligosaccharide hydrolysis by dicarboxylic acids. AIChE J 2013;59:188–99.
[17] Krug M, Ziegler H, Straube G. Degradation of phenolic compounds by the yeast
Candida tropicalis HP 15. I. Physiology of growth and substrate utilization. J
Basic Microbiol 1985;25:103–10.
[18] Lloyd T, Wyman C. Predicted effects of mineral neutralization and bisulfate
formation on hydrogen ion concentration for dilute sulfuric acid pretreatment.
Appl Biochem Biotechnol 2004;115:1013–22.
[19] Lu Y, Mosier NS. Biomimetic catalysis for hemicellulose hydrolysis in corn
stover. Biotechnol Prog 2007;23:116–23.
[20] Maloney MT, Chapman TW, Baker AJ. Dilute acid hydrolysis of paper birch:
kinetics studies of xylan and acetyl group hydrolysis. Biotechnol Bioeng
1985;27:355–61.
[21] Marton JM, Felipe MGA, Silva JBA, Junior AP. Evaluation of the activated
charcoals and adsorption conditions used in the treatment of sugarcane
bagasse hydrolysate for xylitol production. Braz J Chem Eng 2006;23:9–21.
[22] Mok WSL, Antal MJA. Uncatalyzed solvolysis of whole biomass hemicellulose
by hot compressed liquid water. Ind Eng Chem Res 1992;31:1157–61.
[23] Petersen MØ, Larsen J, Thomsen MH. Optimization of hydrothermal
pretreatment of wheat straw for production of bioethanol at low water
consumption without addition of chemicals. Biomass Bioenergy
2009;33:834–40.
[24] Rodriguez CE, Shinyashiki M, Froines J, Yu RC, Fukuto JM, Cho AK. An
examination of quinone toxicity using the yeast Saccharomyces cerevisiae
model system. Toxicology 2004;201:185–96.
[25] Saeman JF. Hydrolysis of cellulose and decomposition of sugars in dilute acid
at high temperature. Ind Eng Chem 1945;37:43–52.
[26] Sanchez B, Bautista J. Effects of furfural and 5-hydroxymethylfurfural on the
fermentation of Saccharomyces cerevisiae and biomass production from
Candida guilliermondii. Enzyme Microb Technol 1988;10:315–8.
[27] Santos SC, Dionísio SR, Roque LR, Costa AC, Ienczak JL. Fermentation of xylose
and glucose mixture in intensified reactors by Scheffersomyces stipitis to
produce ethanol. Int J Sci Eng Res 2015;9:517–22.
[28] Santucci BS, Maziero P, Rabelo SC, Curvelo AAS, Pimenta MTB. Autohydrolysis
of hemicelluloses from sugarcane bagasse during hydrothermal pretreatment:
a kinetic assessment. BioEnerg Res 2015;8:1778–87.
[29] Saska M, Ozer E. Aqueous extraction of sugarcane bagasse hemicellulose and
production of xylose syrup. Biotechnol Bioeng 1995;45:517–23.
[30] Shevchenko SM, Chang K, Robinson J, Saddler JN. Optimization of
monosaccharide recovery by post-hydrolysis of the water-soluble
hemicellulose component after steam explosion of softwood chips. Bioresour
Technol 2000;72:207–11.
[31] Silva JPA, Mussatto SI, Roberto IC, Teixeira JA. Fermentation medium and
oxygen transfer conditions that maximize the xylose conversion to ethanol by
Pichia stipitis. Renewable Energy 2012;37:259–65.
[32] Silva-Fernandes T, Chorão L, Carvalheiro F, Marques S, Loureiro-dias MC,
Fonseca C, et al. Biorefining strategy for maximal monosaccharide recovery
from three different feedstocks: eucalyptus residues, wheat straw and olive
tree pruning. Bioresour Technol 2015;183:203–12.
[33] Sluiter JB, Chum H, Gomes AC, Tavares RAP, Azevedo V, Pimenta MTB, et al.
Evaluation of brazilian sugarcane bagasse characterization: an interlaboratory
comparison study. J AOAC Int 2016;99(3):579–85.
[34] Taherzadeh MJ, Karimi K. Chapter 12 – Fermentation inhibitors in ethanol
process and different strategies to reduce their effects. In: Pandey A, Larroche
C, Ricke S, Dussap CG, Gnansounou E, editors. Biofuels, Alternative Feedstocks
and Conversion Processes. Elsevier Science & Technology; 2011. p. 287–311.
[35] Wyman CE, Decker SR, Himmel ME, Brady JW, Skopec CE. Hydrolysis of
Cellulose and Hemicellulose. In: Dumitriu S, editor. Polysaccharides Struct.
Divers. Funct. Versatility. CRC Press; 2005.
[36] Yu Q, Xu C, Zhuang X, Yuan Z, He M, Zhou G. Xylo-oligosaccharides and ethanol
production from liquid hot water hydrolysate of sugarcane bagasse.
BioResources 2015;10:30–40.
[37] Zhang T, Kumar R, Tsai Y-D, Elander RT, Wyman CE. Xylose yields and
relationship to combined severity for dilute acid post-hydrolysis of
xylooligomers from hydrothermal pretreatment of corn stover. Green Chem
2015;17:394–403.