Catalysis Today 319 (2019) 14–24

Contents lists available at ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

The production of furfural directly from hemicellulose in lignocellulosic T

biomass: A review
⁎
Yiping Luoa,b, Zheng Lia, Xiaoling Lia, Xiaofeng Liub, Jiajun Fanc, , James H. Clarkc,
⁎
Changwei Hua,
a
Key Laboratory of Green Chemistry and Technology, Ministry of Education, College of Chemistry, Sichuan University, Chengdu, Sichuan 610064, P.R. China
b
Key Laboratory of Environmental and Applied Microbiology, CAS; Environmental Microbiology Key Laboratory of Sichuan Province, Chengdu Institute of Biology,
Chinese Academy of Sciences, Chengdu, Sichuan 610041, P.R. China
c
Green Chemistry Centre of Excellence, Department of Chemistry, University of York, York, YO10 5DD, UK

A R T I C LE I N FO A B S T R A C T

Keywords: As one main component of lignocellulosic biomass, hemicellulose is a promising alternative for the replacement
Biomass of limited fossil resources to produce furfural, thus preserving a high atom efficiency. However, the complex
Hemicellulose structure of hemicellulose and the interaction between the other two components in lignocellulosic biomass
Furfural (cellulose and lignin) make the effective utilization of naturally formed structure of hemicellulose challenging.
Catalysts
This review presents an overview of the production of furfural directly from hemicellulose in lignocellulosic
Solvents
biomass with special emphasis on achieving the effective utilization of hemicellulose, which includes the se-
lective dissolution of hemicellulose from lignocellulosic biomass and the selective formation of furfural from
hemicellulose derivatives. Whereas the cellulose and lignin structures are retained, which can be utilized se-
parately. Solvents and catalysts are considered as two main factors in this valorization process of hemicellulose.

1. Introduction chemicals. Unlike cellulose, hemicellulose consists of short, highly

branched polymer of five- and six-carbon polysaccharide units, such as
The challenges faced by the over-dependence of petroleum-based xylan, mannan, β-glucans and xyloglucans [2,13,14]. The highly
resources are increasing due to diminishing fossil fuel reserves and branched and amorphous nature of hemicellulose enables it to be easily
environmental concerns [1–5]. Lignocellulosic biomass is viewed as a converted. Noteworthy is that hemicellulose has a much lower degree
sustainable feedstock because of its renewability, abundance and wide of polymerization (100–200 U) compared with that of cellulose and
distribution in nature, which is mainly composed of the three compo- lignin. Hemicellulose is more unstable than cellulose and therefore,
nents, hemicellulose, cellulose, and lignin [2,6–9]. This composition degrades more easily when subjected to heat treatment [15]. Hemi-
holds significant challenges for the effective utilization of biomass due cellulose shows many excellent properties, including biodegradability,
to the rigid structure of polymeric composite and complicated inter- biocompatibility, bioactivity, and so on, which enable it be applied in a
actions connecting the three main components [10,11]. Successfully variety of areas such as food, medicine, energy, chemical industry and
replacing petroleum based fuels and chemicals by lignocellulosic bio- polymeric materials [16–21]. At present, there are many researches
mass-based products will require high-yield, low-cost and energetically about the conversion of cellulose and lignin, however, the reports about
efficient targeted upgrading processes [1]. Therefore, the development hemicellulose conversion are limited.
of green and sustainable technologies is required for the conversion of The effective utilization of hemicellulose in lignocellulosic biomass
lignocellulosic biomass to value-added chemicals and biofuels. includes the selective dissolution of hemicellulose from raw biomass
Hemicellulose is the second most abundant polysaccharide after and the selective formation of target products from hemicellulose de-
cellulose in plant cell walls, accounting for 15–30% of lignocellulosic rivatives. In the plant cell walls, hemicellulose is thought to adhere to
biomass by weight [8]. In general, the amounts of hemicellulose in cellulose through hydrogen bonds and Van der Waal’s interactions,
wood and woody biomass are greater than those in herbaceous and which enable hemicellulose to form highly resistant networks with
agricultural biomass [12]. Among the three main components in bio- cellulose [22,23]. Ferulic acid forms covalent feruloyl ester-ether
mass, hemicellulose is a promising material to produce value-added bridges between hemicellulose and lignin, resulting in the formation of

⁎
Corresponding authors.
E-mail addresses: [email protected] (J. Fan), [email protected] (C. Hu).

https://doi.org/10.1016/j.cattod.2018.06.042
Received 29 December 2017; Received in revised form 11 May 2018; Accepted 28 June 2018
Available online 05 July 2018
0920-5861/ © 2018 Elsevier B.V. All rights reserved.
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Fig. 1. The pathways for the conversion of furfural as platform molecule into chemicals and fuels.

hemicellulose-lignin linkages. The extraction of hemicellulose is re- conversion of biomass into valuable products [30–32]. Among the three
stricted by the physical and /or covalent interactions with other cell- main components in biomass, hemicellulose is a promising resource to
wall components. So, the development of effective methods toward the produce furfural. Therefore, the development of efficient pre-treatment
selective conversion of hemicellulose to target products with high yield methods is of great significance for the effective utilization of hemi-
and selectivity is crucial in facilitating the effective utilization of cellulose in lignocellulosic biomass to produce furfural with high yield
hemicellulose, avoiding significant decomposition of cellulose and and selectivity.
lignin [24–26]. Typically, as shown in Fig. 4(A), pre-treatment methods can be di-
Furfural, identified by the US Department of Energy (DOE) as one of vided into three categories: physical, chemical and biological methods
the top 12 value-added products, is a valuable product with a word [33,34]. Physical pre-treatment methods use mechanical force and/or
market of around 300.000 tons per year [27,28]. Furfural is a typical heat to open the biomass structure and reduce its particle size by
product which could be obtained from hemicellulose in raw biomass grinding, milling or fibre steam-explosion [35]. Biological pre-treat-
and is also a key platform chemical produced in lignocellulosic bior- ment methods use microbes or fungal enzymes for the digestion of plant
efineries that could further be transformed to fuels and useful chemi- biomass [36,37]. Chemical pre-treatment methods use ionic liquids or
cals, which is widely used in oil refining, plastics, pharmaceutical and dilute acids, bases, or other reagents to dissociate the biomass fractions
agrochemical industries (Fig. 1) [29]. With the aim to obtain furfural and alter its composition [38,39]. In past decades, biochemical and
with high yield and selectivity from the selective dissolution and con- thermal-chemical conversion technologies and processes have been
version of hemicellulose, solvent and catalyst were considered as two investigated to produce chemicals and biofuels from biomass [37,40].
main crucial factors. In this paper, different techniques investigated in Thermal liquefaction was identified as the most promising method to
the selective dissolution of hemicellulose from raw biomass using sol- obtain low molecular weight liquid products, gas fuel and solid residue.
vents or catalysts have been reviewed. Additionally, a potential bio- Among thermal liquefaction, pyrolysis and solvolysis were the two
refinery aim is to produce furfural with high yield and selectivity via main methods used for the conversion of biomass to chemicals and fuels
the further depolymerization and conversion of dissolved hemicellulose [40]. The pyrolysis of biomass relied on high temperature for the de-
derivatives obtained directly from raw biomass materials (Fig. 2). gradation of lignocellulosic biomass. The high operating temperature of
pyrolysis (400–1000 °C) could lead to cross-linking reactions between
hydrocarbons and aromatics, resulting in tar, which was difficult to be
2. Solvent-thermal conversion of hemicellulose in biomass
decomposed further [40]. Moreover, the liquid products obtained (bio-
oil) was a mixture of compounds, which made separations difficult
Lignocellulosic biomass has proved to be difficult to degrade, with
prior to upgrading to targeted chemicals and fuels, and additional costs
its complex composition and interactions among the three main com-
were required in the process [41]. Solvent-thermal reactions, generally
ponents, requiring additional energy inputs for their dissolution
in the presence of aqueous solvents, organic solvents or ionic liquids,
[11,30]. The conversion of biomass involves multi-scale complexity
have been proposed as one of the most efficient approaches to improve
from molecular to macro raw biomass level (Fig. 3), which limits the

Fig. 2. The roles of solvents and catalysts on the selective dissolution and conversion of hemicellulose in actual biomass to produce furfural.

15
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Fig. 3. The multi-scale complexity from molecular levels to macro raw biomass level in biomass conversion.

selectivity of chemicals from biomass [42]. Compared with pyrolysis higher productivity and reduced costs [49]. The addition of catalysts
methods, solvolysis exhibited many advantages, such as (1) The pre- promoted the cleavage of inter- and intra-bonds in the reaction process,
sence of solvents diluted the concentration of the products, thus pre- which enhanced the selective dissolution of hemicellulose in biomass
venting the cross-linked reactions and reverse reactions. (2) The heat and facilitate its conversion to furfural. The addition of catalysts also
and mass transfer become easier in the fluid system. (3) Relatively low promoted the formation of furfural via the following effects:(1) En-
temperature (less energy consumption) was needed [40]. Li et al. re- hancing the chemical bond cracking of xylose and/or xylose oligomers
ported that solvents have significant effects in biomass conversion, to form furfural; (2) Enhancing the rearrangement of different species
through reaction rate, reaction pathways, product distributions, and to form furfural; (3) Enhancing the selectivity to furfural. Therefore,
yields [43]. So reaction rates and product selectivity can be controlled catalysts play important roles in hemicellulose conversion to furfural
to a certain degree by simply changing the solvent medium. Further- (Fig. 4(B)). Catalysts can be divided into homogeneous and hetero-
more, because the three main components in biomass showed different geneous. Homogeneous catalysts were dissolved in the reaction
solubility in different solvents, fractional conversion of biomass could medium, facilitating the interaction with the solid biomass. However,
be achieved in solvent-thermal processes. Therefore, the utilization of homogeneous catalysts have difficulties for the recovery of catalysts
solvents in biomass conversion might be benefit for the selective dis- from reaction media. Heterogeneous catalysts, with easy separation and
solution and conversion of hemicellulose in biomass to furfural. recovery, would be beneficial in view of the principles of green
chemistry [50]. However, heterogeneous catalysts, such as solid acids
faced the problem of their easy deactivation in aqueous solution or
3. The role of catalyst in hemicellulose conversion to furfural biphasic systems containing salt [17,51]. To achieve the effective uti-
lization of hemicellulose in biomass to produce furfural with high yield
Breaking down the complex lignocellulose to dissolve and convert and selectivity, the design of effective catalyst capable of being used in
hemicellulose selectively requires pre-treatment under harsh conditions the solvent system was required.
such as high temperature, high pressure and long reaction time
[44–48]. These processes of achieving the selective dissolution and
conversion of hemicellulose were complex, and chemically-energy in-
tensive. Therefore, it required additional upgrading and separation
steps, leaving a distinct negative environmental footprint. Holm et al.
reported that catalysis could improve process efficiently, resulting in

Fig. 4. (A) Categories of pre-treatment methods in biomass conversion; (B) Categories of catalysts and roles of catalysts in biomass conversion.

16
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Table 1 Table 2
The conversion of xylose as model compound to furfural. The conversion of pure hemicellulose to furfural.
Starting Solvent Catalyst Conditions Yield/ Reference Starting Solvent Catalyst Conditions Yield /% Reference
materials mol/a materials

Xylose GVL H2SO4 175 °C 75 [57] Xylan H2O HCl 180 °C 36.5b [63]
Xylose GVL-10% H2SO4 170 °C, > 80 [58] Xylan MIBK-H2O HCl 177 °C 45.79a [64]
H2O Xylan GVL-10% FeCl3·6H2O 170 °C 68.6a [65]
Xylose GVL-10% H-Mordenite 175 °C 80 [59] H2O
H2O Xylan 2-MTHF- SnCl4 150 °C 78.1a [66]
Xylose H2O HCl, NaCl 200 °C 81.3 [60] H2O
a
Xylose GVL-10% H-Beta 160 °C 71 [61] Xylan H2O-THF AlCl3·6H2O- 140 °C 64 [67]
H2O NaCl
a
Xylose SBP-H2O- NaCl-Sn-MMT 180 °C 76.8 [62] Xylan GVL-H2O Al2(SO4)3 130 °C 87.8 [68]
DMSO Xylan 2-MTHF- NaCl-SO42−/Sn- 160 °C 77.35a [69]
H2O MMT
a
Mole yield of furfural from the conversion of xylose.
a
Yield of furfural based on the moles of C5 fraction in pure hemicellulose.
b
4. Dissolution and conversion of hemicellulose in raw biomass to Yield of furfural was based on the weight of the starting materials.
furfural
H2O solvent using H-M catalysts [59]. When xylose was reacted at
4.1. Conversion of xylose and pure hemicellulose to furfural 180 °C for 30 min using Sn-MMT catalyst, 76.8% furfural yield with
82.5% furfural selectivity was obtained in SBP/NaCl-DMSO-H2O bi-
As an abundant and non-edible component of biomass, hemi- phasic system with saturated NaCl [62].
cellulose possessed the potential to serve as a sustainable feedstock for Excepting xylose as model compounds to produce furfural, many
bio-based energy and value-added chemical products [52–55]. Furfural researchers use pure hemicellulose such as xylan as starting materials
can be derived from hemicellulose in raw biomass, generally from C5 for the production of furfural (Table 2). Yemis et al. studied the acid-
sugars, mainly xylose, the monomer unit contained in hemicellulose of catalyzed conversion of xylan to furfural by microwave-assisted reac-
lignocellulosic materials [27,56]. Much attention has been paid to use tion, and obtained 36.5 wt% furfural using 0.1 M HCl at 180 °C for
xylose as model compound to produce furfural (Table 1). Solvent- 30 min [63]. The effect of different Brønsted acids on the conversion
thermal conversion of xylose was one of efficient methods to improve efficiency of xylan was evaluated, and it was found that HCl was the
the yield of furfural. Concern about what was considered to be a “green most effective catalyst. Several organic solvents such as GVL, THF and
solvent” has arisen in the scientific community, and more attention has 2-MTHF, and homogeneous catalysts such as FeCl3·6H2O, SnCl4,
been given to this concept as efforts to replace petroleum-derived sol- AlCl3·6H2O and Al2(SO4)3 were added to promote the conversion of
vents with those from biomass were undertaken [41]. Various solvents xylan to furfural [64–69]. For example, Wang et al. found that a bio-
such as H2O, GVL, 2-MTHF and DMSO were used for the conversion of based 2-MTHF/H2O biphasic system, was more conducive for the
xylose [56–62]. Among these solvents, GVL was considered as a bio- conversion of xylan-type hemicellulose to produce furfural with highest
mass-derived green solvent and it was identified as an effective solvent yield of 78.1% using SnCl4 catalyst at 150 °C for 120 min [66]. Yang
in hemicellulose conversion compared to H2O. Mellmer et al. studied et al. obtained 64% yield of furfural using xylan as the starting material
the effects of polar aprotic organic solvents on the acid-catalyzed con- in H2O-THF medium containing AlCl3·6H2O and NaCl under microwave
version of xylose into furfural by reaction kinetics [57]. Significant heating at 140 °C [67]. Yang et al. tested a variety of metal salt for the
increases in reaction rates as well as increased product selectivity were conversion of xylan under microwave conditions in GVL/H2O solvent
observed using GVL as solvent. The work also showed that the use of system, and 87.8% furfural was obtained at 130 °C using Al2(SO4)3
GVL as the solvent changed the activation energies for the probe re- catalyst [68]. Their results showed that the Brønsted acidity and Lewis
action, that is, the dehydration of xylose to furfural. Gallo et al. re- acidic species were key factors for the high conversion of xylan in GVL/
ported that xylose dehydration reactions occurred faster in GVL than in H2O system. In the same reaction system, other organic solvents such as
water, which minimized the degradation products, and furfural was MIBK, THF and 2-MTHF were also used to compare the solvent effects
thus obtained in higher yields (70%) [58]. Furthermore, some re- on the conversion of xylan. It was found that the other three organic
searchers used GVL with a small amount of water for the conversion of solvents provided lower yields of furfural compared with GVL. This
xylose to furfural with high yield and confirmed that GVL/H2O co- suggested that organic solvent with a specific catalyst was benefit for
solvent systems could further improve hemicellulose conversion to the production of furfural from xylan [68]. Lin et al. prepared SO42−/
furfural [58,59]. Besides GVL/H2O co-solvent, other organic solvents Sn-MMT as a solid acid catalyst for the heterogeneous catalytic trans-
such as SBP and DMSO were also added into water to form a biphasic formation of xylan [69]. The highest furfural yields were achieved up to
system for hemicellulose conversion [62], and a high yield of furfural 77.35% from xylan at 160 °C in 2-MTHF/NaCl-H2O biphasic system.
was obtained. The results also suggested that SO42−/Sn-MMT contained both
In the solvent-thermal conversion of pure hemicellulose process, Brønsted acid and Lewis acid sites, which could improve the reaction
some homogeneous catalysts were added to improve furfural yield. conditions.
[58−6 ]
° For example, with the addition of 0.05 mol L-1 sulfuric acid, The conversion of xylose and pure hemicellulose to furfural gave
furfural could be produced from xylose in higher yields (80%) with valuable ideas for the understanding of the furfural formation me-
short reaction times (15 min) using a mixture of 10 wt% water in GVL chanism, which will be beneficial to the better utilization of hemi-
as solvent [58]. Marcotullio et al. obtained 81.3 mol% furfural with cellulose. A high yield of furfural was also obtained from xylose and
90% selectivity from the conversion of xylose in 50 mM HCl aqueous pure hemicellulose by the use of co-solvent or biphasic system with the
solution with the addition of NaCl [60]. It indicated that the Cl– as addition of catalysts, which opened the possibility of using more com-
catalyst promote the formation of the 1,2-enediol from the acyclic form plex carbohydrate, such as raw biomass, as a starting material.
of xylose, and thus the subsequent acid catalyzed dehydration to fur- However, the use of xylose and pure hemicellulose as a substrate in the
fural. In order to further improve furfural yield, heterogeneous catalysts production of furfural was not an economically viable option, as they
were used. Gurbuz et al. presented a processing strategy for the con- needed to be obtained in pure form from biomass. Besides, there re-
version of xylose into furfural with high yield (80%) in GVL-10 wt% mained a large gap between the structure or state of xylose or pure

17
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Table 3
“One-pot” method for the dissolution and conversion of hemicellulose in raw biomass to furfural.
Starting materials Solvent Catalyst Conditions Yield/% Reference

Sugarcane bagasse H2O – 200-230 °C low [77]

Pubescens 25% GVL/H2O – 160 °C 33.6a [78]
Corn stover 90% GVL-H2O 0.025 M H2SO4 170 °C 96a [79]
Bagasse THF-H2O (2:1) 0.4 M HCl 200 °C 66.3b [80]
Birch wood H2O H2SO4 147 °C 11.09c [81]
Corn stover GVL-SBP 0.05 wt% H2SO4, AlCl3 170 °C 70a [82]
Corncob H2O 0.9 mmol H2SO4 160 °C 71b [83]
Corncob H2O 0.10 wt% H2SO4 190 °C 99.5b [84]
Bagasse H2O, HUSY 170 °C 18b [17]
Bagasse H2O/p-xylene, HUSY 170 °C 56b [17]
Bagasse/ rice husk /wheat straw H2O/toluene SAPO-44, 170 °C 93b [85]
Corn stover GVL SC-CaCt-700 200 °C 93b [86]
Corncob GVL SPTPA 175 °C 73.9b [87]

a
Yield of furfural was based on the weight of C5 fraction in biomass.
b
Yield of furfural was based on the moles of C5 fraction in biomass.
c
Yield of furfural was based on the weight of the starting materials.

hemicellulose and the actual lignocellulosic biomass. Compared with attracted worldwide interests.
pure hemicellulose, the intermolecular bonds linking hemicellulose, Because the three main components in biomass have complicated
cellulose and lignin existed in actual biomass. Moreover, the in- interactions, the dissolution of hemicellulose was always accompanied
tramolecular bonds of hemicellulose in actual biomass were more with the dissolution of cellulose and lignin. In order to study the dis-
complex. These inter- and intra-molecular bonds in actual biomass solution and conversion of hemicellulose in actual biomass for the
would inhibit the conversion of hemicellulose. Therefore, the produc- production of furfural, many researchers used “one-pot” methods
tion of furfural with high yield and selectivity from hemicellulose in (Table 3), including the simultaneous conversion of hemicellulose and
actual biomass was more challenging than from xylose and pure lignin, simultaneous conversion of all the three main components firstly
hemicellulose (Table 3). and then recovered dissolved lignin, and the simultaneous conversion
of all three main components in raw biomass.
Most lignocellulosic fractionations involved delignification, which
4.2. “One-pot” method for the dissolution and conversion of hemicellulose was often accompanied with hemicellulose conversion, leaving a rather
in actual biomass to furfural pure cellulose as solid residue that could be further used to produce
value-added chemicals by fermentation [72]. This is because the dis-
Industrially, furfural was produced from pentosane sugars present in solution of highly structured cellulose usually needed harsher condi-
lignocellulosic materials using homogeneous mineral acid catalysts tions [73]. Moreover, the conversion of pure cellulose to value-added
(H2SO4 catalyst) in aqueous solutions. The C5 polysaccharides were chemicals (e.g. sugars, levulinic acid) have attracted many researchers’
first hydrolyzed to monosaccharides (mainly xylose), which were sub- attention [74–76]. Therefore, some researches focus on the simulta-
sequently dehydrated to furfural (Fig. 5A). Furfural was then recovered neous removal of hemicellulose and lignin in biomass to obtain furfural
from the liquid phase by steam stripping to avoid further degradation and high purity cellulose. For example, Sasaki et al. [77] reported that
and purified by double distillation. However, a yield of only 40%–50% hemicellulose and lignin were removed as a water-soluble fraction at
furfural was obtained [59,70]. Furthermore, conventional furfural 200–230 °C by hydrothermal treatment of sugarcane bagasse to obtain
production required high energy and generated acidic waste streams low yields of furfural, while the cellulose fraction was hydrolyzed at
[71]. Therefore, research on more economic and environmentally be- higher temperatures (230–280 °C) or recovered as solid residue from
nign furfural production strategies from hemicellulose in biomass has

Fig. 5. Reaction scheme for the acid-catalyzed conversion of hemicellulose into furfural (A), a two-step method in H2O/organic solvent to achieve the selective
dissolution of hemicellulose in biomass(B), and a biphasic system to produce furfural from lignocellulosic biomass(C).

18
Y. Luo et al. Catalysis Today 319 (2019) 14–24

this treatment. A GVL-H2O co-solvent system was developed for the catalyst, while only 51.5% furfural yield was achieved in aqueous
conversion of pubescens. At 160 °C for 4 h, H2O promoted the cleavage media under the same conditions [86]. Zhang et al. found that a new
of chemical bonds linking hemicellulose, lignin and cellulose, and GVL porous polytriphenylamine–SO3H (SPTPA) had better catalytic perfor-
further helped the co-dissolution of hemicellulose- (93.6 wt%) and mance on the conversion of corncob in GVL solvent. Under the op-
lignin- derivatives (80.2 wt%), leaving a high purity cellulose (83.3 wt timum reaction conditions, 73.9% furfural was obtained at 175 °C [87].
%). The maximum furfural yield from hemicellulose conversion These results suggested that solvent and catalyst were important for the
reached 33.6 wt% (based on the weight of hemicellulose in pubescens) production of furfural.
[78]. Although high yield of furfural was obtained by a “one-pot” method
One strategy for the simultaneous conversion of the three main for the dissolution and conversion of hemicellulose in actual biomass,
components firstly and then recovery of the dissolved lignin in co-sol- the products obtained were a mixture that composed of many kinds of
vent with the addition of mineral acid, was used to obtain furfural from carboxylic acids, furans, phenols and some oligomers, causing the dif-
hemicellulose. Alonso et al. reported a process for obtaining hemi- ficulty in product separation and the further use of the products [88].
cellulose and cellulose mixture in 90% GVL-H2O solvent using 0.025 M Hence, the selectivity to furfural was low. Therefore, the development
H2SO4, while the dissolved lignin was precipitated upon addition of of new technologies to produce furfural with high yield and selectivity
water and removed by filtration. The cellulosic fraction of corn stover from raw biomass in environmentally sustainable manners deserves
could be converted into levulinic acid (LA, 66%, yield based on C6 further study.
fraction), while under the same conditions the hemicellulose fraction
could be converted into furfural (96%, yield based on C5 fraction) [79]. 4.3. Selective dissolution and conversion of hemicellulose in actual biomass
A biphasic system consisting of THF and water (2:1) was studied to to furfural
achieve the integrated conversion of cellulose and hemicellulose in
bagasse to 66.3 mol% furfural and 44.3 mol% levulinic acid using 0.4 M 4.3.1. One-step method to produce furfural from hemicellulose in raw
HCl at 200 °C, while lignin was precipitated as a solid residue [80]. As biomass
compared to GVL solvent, the utilization of a lower boiling point sol- Most lignocellulosic fractionations involved the conversion of
vent, THF, also achieved the simultaneous hydrolysis of C6 and C5 hemicellulose, which needed to be upgraded, however generally ac-
carbohydrates in lignocellulosic biomass. companied by the conversion of lignin and cellulose, because the
The conversion of all the three components in raw biomass was also structure of hemicellulose–lignin linkages coating on cellulose was
adapted to study the dissolution and conversion of hemicellulose in raw destroyed and then made the degradation of lignin and cellulose easy
biomass to furfural. For instance, a new dilute sulphuric acid hydrolysis [89]. Efficient conversion of hemicellulose from raw biomass prior to
method was used for the conversion of hemicellulose in birch wood, fermentation was crucial, because the cellulase enzyme was sig-
and 11.09 wt% furfural was obtained based on the weight of birch nificantly inhibited by xylose, xylan and especially xylo-oligosacchar-
wood at 147 °C for 90 min [81]. Luterbacher et al. reported a labora- ides [90–92]. Demartini et al. found that the subsets of hemicellulose
tory-scale production of soluble carbohydrates from corn stover, hard- were the key recalcitrance-causing factors in switchgrass, and the
wood, and softwood at high yields (70 to 90%) in a solvent mixture of conversion of hemicellulose could achieve nearly 100% glucose yields
GVL, H2O, and 0.05 wt% H2SO4. The separated aqueous phase from at high enzymatic loading in subsequent enzymatic hydrolysis [93].
GVL/water solvent systems with addition of NaCl, was further treated Dussan et al. reported that the amount of soluble lignin in the liquor
with AlCl3 and SBP, and 70% furfural based on the weight of C5 frac- solutions could significantly affect the selectivity of the pentose dehy-
tion in biomass was obtained at 170 °C [82]. Furfural with the yield of dration reaction to form furfural [94]. Therefore, the selective dis-
71 mol% could be produced by acid hydrolysis of raw corncob [83]. Li solution and conversion of hemicellulose to furfural with high yields
et al. studied the conversion of corn cob to furfural using GVL as the and selectivity avoiding significant influence on other components
solvent with a dilute sulfuric acid catalyst. A maximum furfural yield of (cellulose and lignin) was very important for the effective utilization of
99.50% was obtained at 190 °C for 20 min with 0.10% sulfuric acid biomass to its fullest.
[84]. Current methods for production of furfural from hemicellulose The One-step method to produce furfural from the selective dis-
used mineral acid catalysts, which were corrosive, difficult to recover, solution and conversion of hemicellulose in raw biomass was studied
and posed environmental and health risks [1,59]. (Table 4). Ionic liquids have received increasing attention as en-
To overcome the drawbacks posed by mineral acid methods in vironmentally preferable solvents for the selective conversion of
converting hemicelllulose, Sahu et al. presented a solid-acid catalyzed hemicellulose in biomass. Carvalho and co-workers achieved the se-
one-pot method for the selective conversion of solid hemicellulose in lective catalytic conversion of hemicellulose fraction in wheat straw to
crop waste without its separation from other lignocellulosic compo- produce 36.2 wt% furfural based on the weight of C5 sugars in [bmim]
nents [17]. When bagasse was reacted at 170 °C for 6 h in the presence [HSO4] ionic liquids at 160 °C [94]. Fractionation of corn stover into
of only water and HUSY catalyst, 30% xylose + arabinose and 18 mol% cellulose, hemicellulose and lignin was successfully performed in ul-
furfural were obtained. While in a biphasic reaction system (water +p- trasound-assisted ionic liquids at low reaction temperature of 70 °C for
xylene, 170 °C, 6 h) with HUSY catalyst, 56 mol% yields of furfural with 3 h followed by alkaline extraction [95]. A high yield of reducing sugars
17% xylose + arabinose was obtained. Their work also showed that the was obtained (97.77%), while the yield of furfural was quite low in this
addition of organic solvents increased the furfural yield from 18 to process. However, it has proved challenging using ionic liquids in the
56%. If furfural was formed near the catalytically active acidic sites, it fractionation of biomass. Apart from the expense and chemical waste of
may easy degrade. In contrast, furfural was extracted into the organic losing some of the ionic liquid, it could prevent the enzymes from
phase in the biphasic system, it could be protected from further de- functioning in the subsequent fermentation step [11].
gradation. However, the catalysts recycling study revealed a 20% drop Water was considered as an efficient reaction medium, since it is
in activity for each subsequent run. In an efficient one-pot methodology environmental friendly. Besides, water has good hydrogen bonds ac-
using SAPO-44 as catalysts, a high yield of furfural of about 93 mol% cepting ability, which could disrupt the inter- molecular linkages (e.g.
was obtained by the conversion of hemicellulose from raw biomass hydrogen bonds, ether bonds) connecting hemicellulose, cellulose and
(bagasse, rice husk and wheat straw) at 170 °C in H2O/toluene (1:2) lignin, and then promote the dissolution of hemicellulose. Some re-
biphasic solvent system [85]. Recently, Li et al. used a novel hetero- searchers studied the dissolution of hemicellulose in biomass using
geneous strong acid catalyst (SC-CaCt-700) for the conversion of corn traditional autoclave heating method in water. For example, Nitsos
stover, and 93% furfural was obtained at 200 °C in GVL. In comparison, et al. discovered that 80% hemicellulose from beech wood and various
there was almost no furfural obtained in GVL without the addition of other hard- and softwoods was removed in hydrothermal pre-

19
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Table 4
One-step method to produce furfural from the selective conversion of hemicellulose in raw biomass.
Starting materials Solvent Catalyst Conditions Yield/% Hd Ce Lf Reference

Wheat straw [bmim][HSO4] – 160 °C 36.2 a

24.8/0.0g
38.5/52.4g
17.7/36.7g
[95]
Corn stover ionic liquid – 70 °C low 19.3/0.4-2.6 43.2/31.2-40.0 22.5/21.1-20.7 [96]
Hard- and softwood H2O – 170-220 °C low 25.4/5.1 42.1/37.9 26.1/20.9 [97]
Pubescens H2O – 160 °C 1.2b 19.7/8.0 44.7/41.3 21.1/16.7 [99]
Eucalyptus globulus H2O – 196 °C low Most dissolved remained remained [100]
Pubescens H2O – 180 °C < 1b 17.9/3.0 46.5/41.1 25.4/16.7 [88]
Pubescens H2O, microwave – 200 °C ∼1b 20.6/1.0 42.8/remained 24.7/remained [101]
Pubescens H2O-cyclohexane – 160 °C ∼1b 17.9/1.8 51.9/46.6 24.0/17.9 [102]
olive tree pruning H2O 0.025 M H2SO4 180-230 °C low 20.8/8.23-0 16.5/11.6-10.6 21.3/remained [104]
Corn stover H2O AlCl3 140 °C 26a 18.8/2.6 46.1/33.4 17.3/10.9 [105]
Pubescens H2O AlCl3 120 °C 0.91b 17.9/4.9 46.5/41.7 25.4/22.0 [88]
Wheat straw H2O SO42−/Fe2O3 141.97 °C low 27.5/17.5 45.6/remained 19.4/remained [106]
Cardoon CPME/H2O 1 wt% H2SO4 170 °C ∼100c 21.0/unknow 37.0/unknow unknow [109]
Maple wood THF/H2O (3:1) 1 wt% H2SO4 170 °C 87c 15.5/0.0 40.9/38.9 24.4/22.0 [110]
Poplar THF/H2O AlCl3-NaCl 160 °C 64c unknow unknow unknow [67]

a
Yield of furfural was based on the weight of C5 fraction in biomass.
b
Yield of furfural was based on the weight of the starting materials.
c
Yield of furfural was based on the moles of C5 fraction in biomass.
d
Hemicellulose percent before and after reaction.
e
Cellulose percent before and after reaction.
f
Lignin percent before and after reaction.
g
The percent of composition before reaction/ the percent of composition after reaction.

treatment using hot compressed water, with about 60% C5 monomers pubescens, the conversion of hemicellulose reached 72.6 wt% without
and oligomers obtained [97,98]. In these researches, an optimization significant degradation of cellulose and lignin (only 10.4 wt% and
study of the pre-treatment conditions led also to maximization of one- 13.3 wt%, respectively) after hydrothermal treatment at a low tem-
pot furfural yield, using water as sole solvent. Luo et al. proposed that perature of 120 °C for 4 h [88]. The above investigation showed that the
the liquid product with high furan content could be obtained at a addition of homogeneous catalysts such as mineral acid and metal salts
moderate temperature for a shorter time, while phenol compounds in water improved the selective dissolution of hemicellulose in biomass.
could be obtained at a higher temperature for a comparatively longer The utilization of heterogeneous catalysts has recently gained more
time [99]. Thus, the selective dissolution and conversion of hemi- attention because they can recycle compared with homogeneous cata-
cellulose to furfural from raw biomass was possible. Eucalyptus glo- lysts. A nanoscale heterogeneous catalyst, solid acid SO42−/Fe2O3, with
bulus wood samples were treated with hot, compressed water at 196 °C both lewis and Brønsted acidity, was found to effectively hydrolyze
to separate hemicellulose as soluble saccharides, leaving a solid phase 63.5% hemicellulose in wheat straw while keeping cellulose and lignin
mainly made up of cellulose and lignin [100]. Luo et al. found that inactive [106].
more than 80% hemicellulose was dissolved from pubescens, accom- The simultaneous dissolution of hemicellulose and lignin was gen-
panied with 30% lignin and 10% cellulose at 180 °C for 0.5 h [88]. It erally used first, then the dissolved lignin was separated from dissolved
could be seen that traditional heating method was not very efficient for hemicellulose by precipitation by the addition of water (Fig. 5(B)). For
the selective dissolution and conversion of hemicellulose to furfural in example, in the absence of acid, 79.4% of the lignin and 79.6% of the
raw biomass, because a large amount of lignin and cellulose were also hemicellulose in birch sawdust, could be removed using GVL/H2O (1:1)
dissolved. In order to improve the selective dissolution of hemi- from cellulose, and 33.3% lignin could be subsequently recovered by
cellulose, a green process for the microwave-assisted hydrothermal precipitation [107]. The addition of an acid catalyst (H2SO4) can im-
selective dissolution and utilisation of hemicellulose in pubescens was prove the efficiency of the organosolv fractionation process. He et al.
developed. The process facilitated the efficient dissolution of hemi- used an ethanol/H2O (1/1, v/v) co-solvent system with 0.050 M oxalic
cellulose at 200 °C (> 95%), while obtaining hemicellulose-free residue acid, to simultaneously fractionate 88.0 wt% of hemicellulose and
that could be further used as starting materials within many industrial 89.2 wt% of lignin in corn stover at 140 °C for 1 h, while cellulose was
processes. [101]. not obviously degraded [108]. After evaporation of ethanol from the
The mixed solvent played an important role in reducing the de- liquid, a yield of 83 wt% lignin (a rufous powder) could be recovered.
gradation of hemicellulose, cellulose and lignin, compared with water The investigations above have achieved the selective dissolution of
solvent, and the reduction of conversion rate of cellulose and lignin was hemicellulose from biomass by changing the solvent medium or using
more significant that benefited the selective conversion of hemi- catalysts to control the reaction. However, most researchers still used
cellulose [102]. The simultaneous separation and selective conversion mineral acid catalysts. The utilization of green solvent, alleviating the
of hemicellulose in pubescens samples was achieved in water-cyclo- need to purchase and transport petroleum-derived solvents to biomass
hexane solvent at 160 °C [102]. conversion, was a benefit for the development of sustainability in bio-
Some catalysts were also added in the solvent system to enhance the mass conversion. Therefore, the development of effective catalytic and
selective dissolution of hemicellulose in raw biomass. Dilute acid hy- solvent systems was still needed to achieve the selective dissolution and
drolysis was also used to remove hemicellulose selectively as sugars conversion of hemicellulose in biomass. Furthermore, the yield of fur-
[103]. Mateo et al. reported that acid hydrolysis of hemicellulose fural obtained from the selective dissolution and conversion of hemi-
fractions of olive tree pruning can also lead to a valuable solid residue cellulose in biomass was low.
mainly composed of cellulose and lignin [104]. Yi et al. achieved the To improve the yield of furfural, some organic solvents were added
selective dissolution of hemicellulose from corn stover in water using with water and mineral acids for the simultaneous dissolution and
AlCl3 catalyst [105]. Luo et al. investigated the selective dissolution of conversion of hemicellulose in biomass. Molina et al. studied the con-
hemicellulose from pubescens promoted by AlCl3 under solvent-thermal version of cynara cardunculus (cardoon) lignocellulose with 1 wt%
conditions. In the AlCl3 promoted dissociation of hemicellulose from H2SO4 catalyst in a cyclopentyl methyl ether(CPME)/H2O system,

20
Y. Luo et al. Catalysis Today 319 (2019) 14–24

obtaining a higher yield of furfural (near 100 mol%) at 170 °C, how- system was also an effective method to improve furfural yield in the
ever, the yield of furfural was only about 60 mol% in water under the second step. Mittal et al. demonstrated that a MIBK-H2O (2:1) biphasic
same reaction conditions [109]. Furthermore, they also found that the reaction system could be used to convert biomass-derived pentose hy-
addition of NaCl notably accelerated the reaction rate of furfural for- drolyzates to furfural with the yield of 80% [114]. Gurbuz et al.
mation from hemicellulosic pentoses and required a much shorter re- achieved a high yield of furfural by using biphasic reactors and alkyl-
action time to produce furfural at such high yields. Cai et al. obtained phenol solvents. Firstly, corn stover was treated for 5 h at 90 °C in an
87 mol% furfural from the conversion of maple wood in the THF/H2O aqueous solution containing 0.1 M HCl and saturated NaCl. This corn
(3:1) biphasic system at 170 °C with 1 wt% H2SO4, however, the yield of stover feed (1.1 wt% xylose) resulted in the maximum yield of 70% to
furfural was only 39 mol% in water under the same reaction conditions produce furfural with the addition of SBP [53]. Morais et al. developed
[110]. Due to the extraction of furfural to the organic phase in THF/ a novel approach to produce furfural from lignocellulosic biomass via
H2O (3:1) biphasic medium, the side reactions to produce undesirable two reaction stages. The first consisted of an extraction of hemicellulose
intermediates were suppressed, while the equilibrium of the reaction from wheat straw using high-pressure CO2 and H2O to produce a water-
system shifted to furfural production, so improved yield of furfural was soluble fraction containing pentoses in oligomeric and monomeric
achieved. Although high yield of furfural was achieved from the con- form. The second step was performed at 180 °C for 60 min in a water/
version of biomass, the problems associated with the need for mineral THF/MIBK biphasic system with 50 bar of initial CO2 pressure, which
acid made the process inefficent. Yang et al. obtained furfural yield of favored the production of furfural and allowed to obtain furfural at a
55%, 38%, 56% and 64%, respectively, from corn stover, pinewood, yield and selectivity of 43 mol% and 44 mol%, respectively [115].
switch grass and poplar at 160 °C in NaCl-THF/H2O systems using AlCl3 Therefore, it could be seen that the addition of catalysts or organic
catalyst. The aqueous phase containing AlCl3·6H2O and NaCl could be solvents both benefitted the production of furfural for the further de-
recycled multiple times (> 5) without any loss of activity or selectivity polymerization and conversion of dissolved hemicellulose.
for furfural. [67]. To obtain furfural with high yields, catalysts and organic solvents
Catalysts and organic solvent can be added to improve the selective were simultaneously added for the further depolymerization and con-
dissolution of hemicellulose, and the simultaneous conversion of version of the dissolved hemicellulose in the second step. For example,
hemicellulose to furfural by using a one-step method. However, the Xing et al. presented a new process to produce furfural (89.8%) with the
yield and selectivity of furfural needed to be improved. Therefore, the addition of THF and NaCl by the further depolymerization and con-
simultaneous selective dissolution and conversion of hemicellulose to version of waste aqueous hemicellulose solutions (mainly xylose oli-
furfural needs to be investigated further. gomers) using a continuous two zone biphasic reactor [116]. With the
addition of NaCl and THF, a liquid system containing hemicellulose
4.3.2. Two-step methods to improve furfural yield from hemicellulose in raw derived monomers and oligomers was further heated to 200 °C,
biomass achieving the maximum yield of 76.9 mol % with 82.2% selectivity to
Hemicellulose in biomass can be selectively dissolved in solvent. furfural based on the moles of converted hemicellulose using a 5 wt %
However, the dissolved hemicellulose as water-soluble carbohydrates pubescens to solvent ratio. Some organic solvents and heterogenous
were mainly in the form of oligomers [94,103,111]. Bai and co-workers catalysts instead of homogenous catalysts were added for the further
obtained a high xylo-oligosaccharides (XOS) yield of 54.7 wt% based on conversion of dissolved hemicellulose. Li et al. studied the further hy-
xylan content from the catalytic conversion of bamboo hemicellulose at drolysis and dehydration of hydrolysates (C5 monosaccharide and oli-
150 °C for 45 min using novel solid acid catalysts of sulfonated bamboo- gosaccharides) obtained from corncob using Sn-MMT as a solid catalyst
based carbon materials [112]. Therefore, the production of furfural in a biphasic system (SBP/NaCl-DMSO) [117]. Furthermore, Sn-MMT
from lignocellulosic biomass with high yields imposed high demands on catalyst also demonstrated excellent catalytic performance in the con-
the development of depolymerization processes for the effective utili- version of water-insoluble hemicelluloses (WIH, average molecular
zation of hemicellulose in raw biomass. Two-step methods for the weight(Mw) = 64,700) and water-soluble fraction (WSF, average mo-
conversion of hemicellulose in raw biomass has been proposed as one of lecular weight(Mw) = 10,770) in the SBP/NaCl-DMSO system [62].
the efficient approaches to obtain furfural with high yield and se- Gao et al. produced a high yield of furfural (82.4%) from waste aqueous
lectivity. This included the selective dissolution of hemicellulose from hemicellulose solution of a hardwood kraft-based dissolving pulp pro-
other two components and then further depolymerization and conver- duction processing (mainly oligomers), when carried out in Toluene/
sion of the dissolved hemicellulose. Therefore, many researchers focus H2O (30:15) solvent with the addition of ZSM-15 and NaCl catalyst at
on the production of furfural from raw biomass via a two-step method. 190 °C for 3 h [118]. Luo et al. obtained 72.6 wt% dissolved hemi-
Kim et al. performed microwave conversion of biomass (corn stover, cellulose without significant degradation of cellulose and lignin after
switch grass, and poplar) to furfural using maleic acid in a two-step hydrothermal treatment at a low temperature of 120 °C for 4 h for the
procedure [113]. In the first step at the reasonably mild temperature of first step [88]. The extracted hemicellulose could be divided mainly
160 °C, the efficient conversion of hemicellulose in biomass to xylose into two parts, that is, small molecular products such as xylose, furfural
with high yields was achieved. In the second step at higher tempera- and acetic acid (about 36.1%) and oligomers (about 63.9%). Therefore,
tures of 200 °C, higher selectivity of furfural (67%) was obtained by the the further conversion of the oligomers and monomers derived from
further reaction of xylose derived from corn stover. Heterogeneous hemicellulose was carried out at 160 °C for 2 h with the addition of THF
catalysts were often added to improve the furfural yield instead of using and SiO2 by forming a SiO2–AlCl3–H2O/THF system, and 33.1 wt%
homogeneous catalysts in the second step. A microwave-assisted, two- furfural based on the weight of hemicellulose in pubescens was obtained.
stage cascade process, using only corn stover and water in the presence It was demonstrated that the co-contribution of NaCl and co-solvent
of niobium phosphate (NbP) as a catalyst, lead to good furfural yields depolymerized the oligomers to small molecular products (Mw < 150
[51]. In the first stage of the process, corn stover was hydrothermally Da) and improved the selectivity to furfural. [78]. The synergetic ef-
fractionated to obtain soluble hemicellulosic sugars (mostly oligomers) fects of specific catalysts with certain solvents may be one reason to
and a solid residue containing cellulose and lignin. In the second step of produce furfural with high yield and high selectivity (Scheme 1).
the process, the obtained hemicellulosic sugars (mostly oligomers) were Lê et al. reported that water exhibited better performance on the
hydrolyzed and dehydrated over NbP catalysts to yield furfural at hydrolysis of hemicellulose while GVL was benefit for the dissolution of
moderate temperature (< 200 °C), with both steps being performed in lignin in GVL/H2O co-solvent system, leaving cellulose intact [119].
water. Furfural yield was only up to 23 mol% with respect to the Based on the above literature, it was found that H2O played important
starting raw biomass. roles in the selective dissolution of hemicellulose. Besides, in our pre-
The addition of only organic solvents in water to form biphasic vious work, it was observed that little hemicellulose dissolved in pure

21
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Scheme 1. (A) The conversion of pubescens to furfural with high yield and lower molecular weight lignin derivatives in H2O/GVL/THF system; (B) Proposed reaction
network for the NaCl promoted depolymerization and conversion of hemicellulose in pubescens to furfural with high yield in H2O/GVL/ THF system. Dimer is as a
representative for oligomers [70].

GVL [78]. Water was an obvious choice as green solvent and hydrolysis 5. Conclusions and future perspectives
was the key reaction to degrade biomass especially hemicellulose
[120]. Furthermore, because H2O acted as nucleophile agent with Today, there is a need to rapidly develop new technologies that
higher hydrogen bonds accepting ability, so it could be speculated that allow us to convert biomass to value-added chemicals. In the domain of
H2O might form hydrogen bonds with intermolecular linkages con- value-added chemicals, furfural possesses significant advantages using
necting hemicellulose, cellulose and lignin, which contributed to the biomass as a starting material, which could alleviate the utilization of
cleavage of these linkages. The addition of catalysts, such as mineral petroleum-derived chemicals. However, the production of furfural with
acid or solid acid catalysts, might also form hydrogen bonds with in- high yield and selectivity from raw biomass is challenging. A large
termolecular linkages, which enhanced the cleavage of intermolecular volume of work has investigated the conversion of hemicellulose to
bonds and then promoted hemicellulose dissolution. In the further re- furfural. However, the use of pure hemicellulose model components as
action of dissolved hemicellulose-derivatives, H2O promoted the clea- starting materials is not an economically viable option as separation
vage of intramolecular linkages (such as β-1,4 glucosidic bond) in and purification of hemicellulose from biomass is not easy.
hemicellulose with the help of catalysts and organic solvents, which Furthermore, the performance of pure hemicellulose and the actual
was benefit to produce furfural. lignocellulosic biomass is quite different. Here, two strategies to pro-
Therefore, an environmentally benign two-step method was pro- duce furfural directly from raw biomass are discussed in detail. One is
posed to improve the yield and selectivity of furfural as follows: (1) the conversion of hemicellulose in biomass to furfural by using the one-
Adding acid catalysts, especially replacing the homogeneous catalyst by pot method. Although high yields of furfural are obtained, the se-
heterogenous catalysts, may be helpful for the dissolution and conver- lectivity to furfural is still low because of the products obtained are a
sion of hemicellulose to oligo- and mono-saccharides under acid cata- mixture including those from the conversion of both cellulose and
lytic systems. The addition of organic solvent, such as biomass derived lignin. Another one is the selective dissolution and conversion of
green solvent (GVL), helped H2O break down the intermolecular lin- hemicellulose in actual biomass to furfural. One-step and two- step
kages and dissolved hemicellulose selectively. (2) Using a biphasic methods for the selective dissolution and conversion of hemicellulose in
system by the addition of organic solvent (Fig. 5(C)). In biphasic actual biomass to furfural are usually used. The selective dissolution of
system, because furfural produced was continuous extracted into the hemicellulose in biomass is achieved using a one-step method, which
organic phase, so the side reactions producing undesirable inter- could simply the reaction processes, but the yield of furfural needs to be
mediates were suppressed and the equilibrium of reaction system was improved. In contrast, a two-step method for the conversion of hemi-
shifted to furfural production, which improved the yield of furfural cellulose in raw biomass is identified as one of the efficient approaches
[78]. (3) NaCl showed not only salt effects but also as promoter effects to obtain furfural with high yield and selectivity. The effective utili-
to enhance the reacction of xylose to furfural in biphasic system. In zation of hemicellulose in lignocellulosic biomass to furfural includes
contrast with homogenous catalytic processes, it was desirable for the the selective dissolution of hemicellulose from raw biomass and the
development of heterogeneous catalysts to promote the depolymeriza- selective formation of furfural from dissolved hemicellulose-deriva-
tion of hemicellulose derivatives to furfural. tives. For the sake of the production of furfural with high yield and

22
Y. Luo et al. Catalysis Today 319 (2019) 14–24

Scheme 2. The role of solvents and catalysts in the hemicellulose selective dissolution, depolymerization and conversion to furfural needs more further attention.

selectivity, catalysts and organic solvents are the most commonly re- solvent and catalyst is necessary. It also needs to develop high effective
quired for the further depolymerization and conversion of dissolved catalytic system for the further conversion of hemicellulose derivatives
hemicellulose. Especially, the biphasic system formed by the addition of to produce furfural by the design of novel efficient catalysts and new
organic solvents and NaCl is benefit to produce furfural. Because fur- efficient solvent system. Hence, we expect that replacement of fossil
fural produced in biomass conversion could be continuously extracted resource-based production of furfural by more sustainable technologies
into the organic phase in biphasic system, so the side reactions pro- based on renewable biomass will continue to grow.
ducing undesirable intermediates are suppressed and the equilibrium of
reaction system are shifted to furfural production, which improve the Acknowledgements
yield of furfural.
The production of furfural with high yields and selectivity directly This work is financially supported by the National Natural Science
from hemicellulose in raw biomass is challenging but crucial. In order Foundation of China (No. 21536007), 111 Project (No. B17030) and
to obtain furfural with high yield and selectivity from hemicellulose, Sichuan Science and Technology Program (No.2018JY0207).
two strategies are highly considered (Scheme 2): The selective dis-
solution of hemicellulose in biomass, and simultaneous selective con- References
version of hemicellulose to produce furfural with high yield and se-
lectivity directly via one-step method are recommended. The one-step [1] J.S. Luterbacher, D.M. Alonso, J.A. Dumesic, Green Chem. 16 (2014) 4816–4838.
method to achieve furfural with high yield and selectivity, appeals the [2] P. Peng, D. She, Carbohydr. Polym. 112 (2014) 701–720.
[3] E.L. Kunkes, D.A. Simonetti, R.M. West, J.C. Serrano-Ruiz, C.A. Gärtner,
development of multifunctional catalytic systems. By adjusting and J.A. Dumesic, Science 322 (2008) 417–421.
controlling the function of the catalysts and solvents, the selectivity of [4] J.Q. Bond, D.M. Alonso, D. Wang, R.M. West, J.A. Dumesic, Science 327 (2010)
hemicellulose dissolution and of the production of furfural is controlled. 1110–1114.
[5] T.P. Vispute, H.Y. Zhang, A. Sanna, R. Xiao, G.W. Huber, Science 330 (2010)
The use of biomass-derived green solvent such as GVL, 2-MTHF and so 1222–1227.
on, is highly expected in the reaction system. Alternatively, considering [6] J.R. Rostrup-Nielsen, Science 308 (2005) 1421–1422.
the selective dissolution of hemicellulose, as well as the production of [7] D.R. Dodds, R.A. Gross, Science 318 (2007) 1250–1251.
[8] N. Sella Kapu, H.L. Trajano, Biofuels, Bioprod. Biorefin. 8 (2014) 857–870.
furfural with high efficiency, two-step conversion of biomass is highly [9] B. Kamm, P. R. Gruber, M. Kamm, Wiley-VCH Verlag GmbH & Co. KGaA, 2006.
recommended. In the two-step reaction, it is expected to achieve the [10] L. Negahdar, I. Delidovich, R. Palkovits, Appl. Catal., B 184 (2016) 285–298.
selective dissolution of hemicellulose while cellulose and lignin re- [11] K. Sanderson, Nature 474 (2011) 12–14.
[12] S.V. Vassilev, D. Baxter, L.K. Andersen, C.G. Vassileva, T.J. Morgan, Fuel 94
tained firstly, and then the further depolymerization and conversion of
(2012) 1–33.
hemicellulose derivatives to furfural. For the selective dissolution of [13] H.V. Scheller, P. Ulvskov, Annu. Rev. Plant Biol. 61 (2010) 263–289.
hemicellulose, the main challenges are to inhibit the dissolution of [14] L. Negahdar, I. Delidovich, R. Palkovits, Appl. Catal. B 184 (2016) 285–298.
cellulose and lignin, which improve the selectivity of hemicellulose [15] K. Tekin, S. Karagöz, S. Bektas, Renew. Sustain. Energy Rev. 40 (2014) 6733-687.
[16] D.O. Otieno, B.K. Ahring, Carbohydr. Res. 360 (2012) 84–92.
dissolution that simplifying the complexity of the starting materials. [17] R. Sahu, P.L. Dhepe, ChemSusChem 5 (2012) 751–761.
Simultaneously, the inhibition of the formation of side products, which [18] G.S. Yi, Y.G. Zhang, ChemSusChem 5 (2012) 1383–1387.
are not benefit to produce furfural, is needed. Furthermore, the further [19] J. Bian, F. Peng, X.P. Peng, P. Peng, F. Xu, R.C. Sun, Bioresour. Technol. 127
(2013) 236–241.
valorization of hemicellulose derivatives to furfural with high yield and [20] B.S. Weeks, P.P. Perez, Med. Sci. Monit. 15 (2009) 43–46.
selectivity would be improved. Therefore, designing strategy for the [21] N.M. Hansen, D. Plackett, Biomacromolecules 9 (2008) 1493–1505.
selective dissolution of hemicellulose to obtain xylan or xylose type [22] J.W. Agger, T. Isaksen, A. Várnai, S. Vidal-Melgosa, W.G. Willats, R. Ludwig,
S.J. Horn, V.G.H. Eijsink, B. Westereng, Proc. Acad. Sci. 111 (2014) 6287–6292.
products that benefit for the follow-up reaction by the assistance of [23] P. Langan, L. Petridis, H.M. O’Neill, S.V. Pingali, M. Foston, Y. Nishiyama,

23
Y. Luo et al. Catalysis Today 319 (2019) 14–24

R. Schulz, B. Lindner, B.L. Hanson, S. Harton, W.T. Heller, V. Urban, B.R. Evans, [74] D.Q. Ding, J.J. Wang, J.X. Xi, X.H. Liu, G.Z. Lu, Y.Q. Wang, Green Chem. 16
S. Gnanakaran, A.J. Ragauskas, J.C. Smith, Green Chem. 16 (2014) 63–68. (2014) 3846–3853.
[24] F. Carvalheiro, L.C. Duarte, F.M. Gírio, J. Sci. Ind. Res. (2008) 849–864. [75] J.J. Fan, M. De Bruyn, V.L. Budarin, M.J. Gronnow, P.S. Shuttleworth, S. Breeden,
[25] H.M. Hernández-Hernández, J.J. Chanona-Pérez, A. Vega, P. Ligero, D.J. Macquarrie, J.H. Clark, J. Am. Chem. Soc. 135 (2013) 11728–11731.
J.A. Mendoza-Pérez, G. Calderón-Domínguez, E. Terres, R.R. Farrera-Rebollo, Ind. [76] Y. Sun, J.Y. Cheng, Bioresour. Technol. 83 (2002) 1–11.
Crops Prod. 86 (2016) 163–172. [77] M. Sasaki, T. Adschiri, K. Arai, Bioresour. Technol. 86 (2003) 301–304.
[26] A.R.C. Morais, M.D.D. Matuchaki, J. Andreaus, R. Bogel-Lukasik, Green Chem. 18 [78] Y.P. Luo, Z. Li, Y.N. Zuo, Z.S. Su, C.W. Hu, ACS Sustain. Chem. Eng. 5 (2017)
(2016) 2985–2994. 8137–8147.
[27] T. A. Werpy, G. Petersen, O. Ridge, US Department of Energy, Oak Ridge, USA, [79] D.M. Alonso, S.G. Wettstein, M.A. Mellmer, E.I. Gurbuz, J.A. Dumesic, Energy
2004. Environ. Sci. 6 (2013) 76–80.
[28] D. Montané, J. Salvadó, C. Torras, X. Farriol, Biomass Bioenergy 22 (2012) [80] J. Li, D.J. Ding, L.J. Xu, Q.X. Guo, Y. Fu, RSC Adv. 4 (2014) 14985–14992.
295–304. [81] P. Brazdausks, N. Vedernikovs, M. Puke, I. Kruma, Key Eng. Mater. 64 (2014)
[29] R. Mariscal, P. Maireles-Torres, M. Ojeda, I. Sádaba, M.L.ópez Granados, Energy 245–248.
Environ. Sci. 9 (2016) 1144–1189. [82] J.S. Luterbacher, J.M. Rand, D.M. Alonso, J. Han, J.T. Youngquist,
[30] M.C. Mccann, N.C. Carpita, J. Exp. Bot. 66 (2015) 4109–4118. C.T. Maravelias, B.F. Pfleger, J.A. Dumesic, Science 343 (2014) 277–280.
[31] M.E. Himmel, S.Y. Ding, D.K. Johnson, W.S. Adney, M.R. Nimlos, J.W. Brady, [83] T.J. Wang, K. Li, Q.Y. Liu, Q. Zhang, S.B. Qiu, J.X. Long, L.G. Chen, L.L. Ma,
T.D. Foust, Science 315 (2007) 804–807. Q. Zhang, Appl. Energy 136 (2014) 775–780.
[32] M.S. Mettler, D.G. Vlachos, P.J. Dauenhauer, Energy Environ. Sci. 5 (2012) [84] X.Y. Li, Q.L. Liu, C.H. Luo, X.C. Gu, L.F. Lu, X.B. Lu, ACS Sustain. Chem. Eng. 5
7797–7809. (2017) 8587–8593.
[33] X.B. Zhao, L.H. Zhang, D.H. Liu, Biofuels, Bioprod. Biorefin. 6 (2012) 561–579. [85] P. LaxmikantáDhepe, RSC Adv. 4 (2014) 26215–26221.
[34] Y. Zheng, J. Zhao, F.Q. Xu, Y.B. Li, Prog. Energy Combust. Sci. 42 (2014) 35–53. [86] W.Z. Li, Y.S. Zhu, Y.J. Lu, Q.Y. Liu, S.N. Guan, H.M. Chang, H. Jameel, L.L. Ma,
[35] L. Kratky, T. Jirout, Chem. Eng. Technol. 34 (2011) 391–399. Bioresour. Technol. (245) (2017) 258–265.
[36] J. Lee, J. Biotechnol. 56 (1997) 1–24. [87] L.X. Zhang, G.Y. Xi, J.X. Zhang, H.B. Yu, X.C. Wang, Bioresour. Technol. (224)
[37] Y. Sun, J.Y. Cheng, Bioresour. Technol. 83 (2002) 1–11. (2017) 656–661.
[38] V.B. Agbor, N. Cicek, R. Sparling, A. Berlin, D.B. Levin, Biotechnol. Adv. 29 (2011) [88] Y.P. Luo, L.B. Hu, D.M. Tong, C.W. Hu, RSC Adv. 4 (2014) 24194–24206.
675–685. [89] M. Stöcker, Angew. Chem. Int. Ed. 47 (2008) 9200–9211.
[39] R.P. Swatloski, S.K. Spear, J.D. Holbrey, R.D. Rogers, J. Am. Chem. Soc. 124 [90] Q. Qing, B. Yang, C.E. Wyman, Bioresour. Technol. 101 (2010) 9624–9630.
(2002) 4974–4975. [91] Y. Kim, E. Ximenes, N.S. Mosier, M.R. Ladisch, Enzyme Microb. Technol. 48
[40] Z. Liu, F.S. Zhang, Energy Convers. Manage. 49 (2008) 3498–3504. (2011) 408–415.
[41] D.M. Alonso, S.G. Wettstein, J.A. Dumesic, Green Chem. 15 (2013) 584–595. [92] R.R. Singhania, R.K. Sukumaran, A.K. Patel, C. Larroche, A. Pandey, Enzyme
[42] S. Brethauer, C.E. Wyman, Bioresour. Technol. 101 (2010) 4862–4874. Microb. Technol. 46 (2010) 541–549.
[43] L. Shuai, J. Luterbacher, ChemSusChem 9 (2016) 133–155. [93] J.D. DeMartini, S. Pattathil, J.S. Miller, H. Li, M.G. Hahn, C.E. Wyman, Energy
[44] F. Carvalheiro, L.C. Duarte, F.M. Girio, J. Sci. Ind. Res. 67 (2008) 849–864. Environ. Sci. 6 (2013) 898–909.
[45] M.J. Taherzadeh, K. Karimi, BioResources 2 (2007) 707–738. [94] K. Dussan, B. Girisuta, M. Lopes, J.J. Leahy, M.H. Hayes, ChemSusChem 9 (2016)
[46] I. Finore, A. Poli, P. Di Donato, L. Lama, A. Trincone, M. Fagnano, M. Mori, 492–504.
B. Nicolaus, A. Tramice, Green Chem. 18 (2016) 2460–2472. [95] A.V. Carvalho, A.M. da Costa Lopes, R. Bogel-Łukasik, RSC Adv. 5 (2015)
[47] P. Kaur, N.K. Bhardwaj, J. Sharma, Carbohydr. Polym. 145 (2016) 95–102. 47153–47164.
[48] P.D. Carà, M. Pagliaro, A. Elmekawy, D.R. Brown, P. Verschuren, N.R. Shiju, [96] P. Zhang, S.J. Dong, H.H. Ma, B.X. Zhang, Y.F. Wang, X.M. Hu, Ind. Crops Prod. 76
G. Rothenberg, Catal. Sci. Technol. 3 (2013) 2057–2061. (2015) 688–696.
[49] M.S. Holm, S. Saravanamurugan, E. Taarning, Science 328 (2010) 602–605. [97] C.K. Nitsos, K.A. Matis, K.S. Triantafyllidis, ChemSusChem 6 (2013) 110–122.
[50] R. Karinen, K. Vilonen, M. Niemelä, ChemSusChem 4 (2011) 1002–1016. [98] C.K. Nitsos, T.C. Papadopoulou, K.A. Matis, Kostas S. Triantafyllidis, ACS Sustain.
[51] H.G. Bernal, L. Bernazzani, A.M.R. Galletti, Green Chem. 16 (2014) 3734–3740. Chem. Eng. 4 (2016) 4529–4544.
[52] S. Dutta, S. De, B. Saha, M.I. Alam, Catal. Sci. Technol. 2 (2012) 2025–2036. [99] J. Luo, Y. Xu, L.J. Zhao, L.L. Dong, D.M. Tong, L.F. Zhu, C.W. Hu, Bioresour.
[53] E.I. Gürbüz, S.G. Wettstein, J.A. Dumesic, ChemSusChem 5 (2012) 383–387. Technol. 101 (2010) 8873–8880.
[54] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411–2502. [100] S. Peleteiro, V. Santos, G. Garrote, J.C. Parajo, Carbohydr. Polym. 146 (2016)
[55] M.A.T. Hansen, H. Jørgensen, K.H. Laursen, J.K. Schjørring, C. Felby, Biomass 20–25.
Bioenergy 56 (2013) 572–581. [101] Y.P. Luo, J.J. Fan, V.L. Budarin, C.W. Hu, J.H. Clark, Green Chem. 19 (2017)
[56] J.P. Lange, E. Van der Heide, J. Van Buijtenen, R. Price, ChemSusChem 5 (2012) 4889–4899.
150–166. [102] Y. Xu, L.B. Hu, H.T. Huang, D.M. Tong, C.W. Hu, Carbohydr. Polym. 88 (2012)
[57] M.A. Mellmer, C. Sener, J.M.R. Gallo, J.S. Luterbacher, D.M. Alonso, J.A. Dumesic, 1342–1347.
Angew. Chem. Int. Ed. 53 (2014) 11872–11875. [103] L.Q. Jiang, A.Q. Zheng, Z.L. Zhao, F. He, H.B. Li, W.G. Liu, Bioresour. Technol. 182
[58] J.M.R. Gallo, D.M. Alonso, M.A. Mellmer, J.H. Yeap, H.C. Wong, J.A. Dumesic, (2015) 364–367.
Solvent. Top. Catal. 56 (2013) 1775–1781. [104] S. Mateo, J.G. Puentes, S. Sánchez, A.J. Moya, Carbohydr. Polym. 93 (2013)
[59] E.I. Gürbüz, J.M.R. Gallo, D.M. Alonso, S.G. Wettstein, W.Y. Lim, J.A. Dumesic, 416–423.
Angew. Chem. Int. Ed. 52 (2013) 1270–1274. [105] J. Yi, T. He, Z.C. Jiang, J.M. Li, C.W. Hu, Chin. J. Catal. 34 (2013) 2146–2152.
[60] G. Marcotullio, W. De Jong, Green Chem. (12) (2010) 1739–1746. [106] C. Zhong, C.M. Wang, F. Huang, F.X. Wang, H.H. Jia, H. Zhou, P. Wei, Carbohydr.
[61] J.M.R. Gallo, D.M. Alonso, M.A. Mellmer, J.H. Yeap, H.C. Wong, J.A. Dumesic, Polym. 131 (2015) 384–391.
Top. Catal. (56) (2013) 1775–1781. [107] W. Fang, H. Sixta, ChemSusChem 8 (2015) 73–76.
[62] H.L. Li, J.L. Ren, L.J. Zhong, R.C. Sun, L. Liang, Bioresour. Technol. 176 (2015) [108] T. He, Z.C. Jiang, P. Wu, J. Yi, J.M. Li, C.W. Hu, Sci. Rep. 6 (2016) 38623–38633.
242–248. [109] M.C. Molina, R. Mariscal, M. Ojeda, M.L. Granados, Bioresour. Technol. 126
[63] O. Yemis, G. Mazza, Bioresour. Technol. 102 (2011) 7371–7378. (2012) 321–327.
[64] N. Sweygers, J. Harrer, R. Dewil, L. Appels, J. Clean. Prod. 187 (2018) 1014–1024. [110] C.M. Cai, T. Zhang, R. Kumar, C.E. Wyman, Green Chem. 15 (2013) 3140–3145.
[65] L.X. Zhang, H.B. Yu, P. Wang, Y. Li, Bioresour. Technol. 151 (2014) 355–360. [111] B.S. Santucci, P. Maziero, S.C. Rabelo, A.A. Curvelo, M.T.B. Pimenta, Bioenergy
[66] W.J. Wang, J.L. Ren, H.L. Li, A.J. Deng, R.C. Sun, Bioresour. Technol. 183 (2015) Res. 8 (2015) 1778–1787.
188–194. [112] Y.Y. Bai, L.P. Xiao, R.C. Sun, Biomass Bioenergy 75 (2015) 245–253.
[67] Y. Yang, C.W. Hu, M.M. Abu-Omar, ChemSusChem 5 (2012) 405–410. [113] E.S. Kim, S. Liu, M.M. Abu-Omar, N.S. Mosier, Energy Fuels 26 (2012) 1298–1304.
[68] T. Yang, Y.H. Zhou, S.Z. Zhu, H. Pan, Y.B. Huang, ChemSusChem 10 (2017) [114] A. Mittal, S.K. Black, T.B. Vinzant, M.O. Brien, M.P. Tucker, D.K. Johnson, ACS
4066–4079. Sustain. Chem. Eng. (5) (2017) 5694–5701.
[69] Q.X. Lin, H.L. Li, X.H. Wang, L.F. Jian, J.L. Ren, C.F. Liu, R.C. Sun, Catalysts 7 [115] A.R.C. Morais, M.D.D. Matuchaki, J. Andreaus, R. Bogel-Lukasik, Green Chem. 18
(2017) 118–131. (2016) 2985–2994.
[70] A.S. Mamman, J.M. Lee, Y.C. Kim, I.T. Hwang, N.J. Park, Y.K. Hwang, J.S. Chang, [116] R. Xing, W. Qi, G.W. Huber, Energy Environ. Sci. 4 (2011) 2193–2205.
J.S. Hwang, Biofuels, Bioprod. Biorefin. 2 (2008) 438–454. [117] H.L. Li, X.F. Chen, J.L. Ren, H. Deng, F. Peng, R.C. Sun, Biotechnol. Biofuels 8
[71] A.M.C. Lopes, A.R.C. Morais, R.M. Lukasik, Production of Platform Chemicals from (2015) 127–138.
Sustainable Resources, Chapter 2 Springer, Singapore, 2017, pp. 45–80. [118] H.L. Gao, H.T. Liu, B. Pang, G. Yu, J. Du, Y.D. Zhang, H.S. Wang, X.D. Mu,
[72] S. Van den Bosch, W. Schutyser, R. Vanholme, T. Driessen, S.F. Koelewijn, Bioresour. Technol. (172) (2014) 453–456.
T. Renders, B.D. Meester, W.J.J. Huijgen, W. Dehaen, C.M. Courtin, B. Lagrain, [119] H.Q. Lê, A. Zaitseva, J. Pokki, M. Ståhl, V. Alopaeus, H. Sixta, ChemSusChem (9)
W. Boerjan, B.F. Sels, Energy Environ. Sci. 8 (2015) 1748–1763. (2016) 2939–2947.
[73] M. Hakola, A. Kallioinen, M. Kemell, P. Lahtinen, E. Lankinen, M. Leskelä, T. Repo, [120] D. Steinbach, A. Kruse, J. Sauer, Biomass Conv. Bioref., Biomass Conv. Bioref. (2)
T. Riekkola, M. Siika-aho, J. Uusitalo, S. Vuorela, N.V. Weymarn, ChemSusChem 3 (2017) 247–274.
(2010) 1142–1145.

24