Acid-Catalysed Conversion of Carbohydrates into Furan-Type Molecules in Zinc Chloride Hydrate

Article abstract of DOI:10.1002/cplu.201800650

Acid-catalysed conversion of biomass, specifically cellulose, holds promise to create value-added, renewable replacements for many petrochemical products. We investigated an unusual acid-catalysed transformation of cellulose and cellobiose in the biphasic solvent system zinc chloride hydrate (ionic liquid)/anisole. Here, furyl hydroxymethyl ketone and furfural are obtained as major products, which are valuable but less commonly formed in high yields in transformations of cellulosic substrates. We explored this chemistry in small-scale model reactions and applied the optimised methods to the conversion of cellulose in bench-scale processes. The optimum reaction system and preferred reaction conditions are defined to select for highly desirable furanoid products in the highest known yields (up to 46 %) directly from cellulose or cellobiose. The method avoids the use of added catalysts: the ionic solvent zinc chloride hydrate possesses the intrinsic acidity required for the hydrolysis and chemical transformation steps. The process involves inexpensive media for the catalytic conversion of cellulose into high-value products under mild processing conditions.

Full text of DOI:10.1002/cplu.201800650

Accepted Article
Title: Acid-catalysed conversion of carbohydrates into furan type
molecules in zinc chloride hydrate
Authors: Iurii Bodachivskyi, Unnikrishnan Kuzhiumparambil, and D.
Bradley Glen Williams
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemPlusChem 10.1002/cplu.201800650
Link to VoR: http://dx.doi.org/10.1002/cplu.201800650
A Journal of
www.chempluschem.org

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
Acid-Catalysed Conversion of Carbohydrates into Furan-Type
Molecules in Zinc Chloride Hydrate
[a]
[b]
[a]
Iurii Bodachivskyi, Unnikrishnan Kuzhiumparambil, and D. Bradley G. Williams *
Abstract: Acid-catalysed conversion of biomass, specifically
cellulose, holds promise to create value-added, renewable
replacements for many petrochemical products. We investigated an
unusual acid-catalysed transformation of cellulose and cellobiose in
the biphasic solvent system zinc chloride hydrate (ionic liquid)/anisole.
Here, furyl hydroxymethyl ketone and furfural are obtained as major
products, which are valuable but less commonly formed in high yields
in transformations of cellulosic substrates. We explored this chemistry
in small-scale model reactions and applied the optimised methods to
the conversion of cellulose in bench-scale processes. The optimum
reaction system and preferred reaction conditions are defined, to
select for highly desirable furanoid products in the highest known
yields (up to 46%) directly from cellulose or cellobiose. The method
avoids the use of added catalysts: the ionic solvent zinc chloride
hydrate possesses the intrinsic acidity required for the hydrolysis and
chemical transformation steps. The process involves inexpensive
media for the catalytic conversion of cellulose into high-value products
under mild processing conditions.
cellulose into low-molecular-weight reducing sugars in 99% yield
within 1 h at 100 °C (solvents 1-methyl-3-butylimidazolium
chloride and triethyl-(3-sulfopropyl)-ammonium hydrogen
sulfate),[ into 5-(hydroxymethyl)furfural (HMF) in 79% yield
within 2 h at 120 °C (solvent 1-butyl-3-methylimidazolium
chloride),[8] or into levulinic acid in 40% yield in 2 h at 120 °C
7]
(solvent
1-(4-sulfobutyl)-3-methylimidazolium
hydrogen
sulfate),[ all of which are arguably benchmark instances of the
chemical processing of cellulose into target platform molecules.
Drawbacks of the use of ILs include the complexity of recovery of
ionic solvents and products from the reaction media, along with
their high cost, and these pose potential hurdles to their
widespread acceptance in industrial scale processes.[6] There is
thus an imperative to develop and employ inexpensive ILs that
are able to facilitate the efficient processing of polysaccharides
and which are easy, or relatively so, to recover; this remains a
goal for workers in this field.
9]
While common ILs suffer some drawbacks as mentioned,
inorganic molten salt hydrates show promise to avoid some of
these issues. Such ionic systems possess inherent acidity, as well
as low viscosity compared to common quaternary ammonium
salts, and have been demonstrated to be suitable media for the
dissolution of cellulose.[10–12] With commodity level pricing of some
inorganic salts and the ease of preparation of their hydrates, these
ILs are budding candidates for the chemical conversion of
Introduction
In the past few decades there has been a significant and growing
interest in acid-catalysed transformations of carbohydrates into
value-added organic building block chemicals (platform
molecules). These molecules are diverse and represent a
platform with the potential to replace a large portion of crude oil-
based products such as fuels, plastics, coatings, detergents, and
other useful commodities.[1–6] The application of ionic liquids (ILs),
as alternative reaction media to common aqueous and organic
solvents, in the valorisation of polysaccharides has enabled
substantial progress towards mild reaction conditions and
improved yields and selectivities.[5,6] That many ILs fully dissolve
cellulosic material under mild conditions is a primary enabler
towards high yields and selectivity, although secondary reasons
exist for the success of ILs in this context.[6] Pertinent examples
of successful applications of ILs include the transformation of
cellulose. Zinc chloride hydrate with the formula ZnCl
–4; this system is considered to be an IL. Although the
conventional presentation of the molecular formula of zinc
chloride hydrate is ZnCl ∙nH O, its formula is more accurately
represented as [Zn(OH ][ZnCl
2 2
∙nH O (n =
3
2
2
2
)
6
4
] in the case of n = 3, based on
solid state and liquid state determinations.[13,14]) is useful in the
conversion of polysaccharides into furan-type molecules, typically
into HMF.[15–18] The processes detailed in these papers require
lengthy pretreatment of the substrates, the use of microwave
irradiation, or high loading of acid catalyst, all of which inhibit the
[6]
ability to scale up such processes. It remains an unsolved
problem to convert polysaccharides in such IL solvent systems
into one or two major products under mild reaction conditions
using techniques that are potentially scalable.
The present work discloses the functionality and potential of zinc
chloride hydrate solvent in the catalytic conversion of
carbohydrates into unusual furan-type molecules (i.e. products
that are valuable but less seldom obtained during the conversion
of cellulose) under mild reaction conditions.
[
a]
b]
I. Bodachivskyi, Prof. Dr. D.B.G. Williams
University of Technology Sydney
School of Mathematical and Physical Sciences
Broadway NSW 2007, PO Box 123 Broadway NSW 2007 (Australia)
E-mail: Bradley.Williams@uts.edu.au
[
Dr. U. Kuzhiumparambil
University of Technology Sydney
Climate Change Cluster (C3)
Broadway NSW 2007, PO Box 123 Broadway NSW 2007 (Australia)
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
Results and Discussion
FF of 16 and 13 mol%, respectively, but HMF was present only in
trace amounts (Table 1). It is worth noting that FF was found
predominantly in the xylene phase, while FHK was present in both
the organic solvent and the polar aqueous phase. The incomplete
separation of the polar product FHK from the acidic IL by xylene
would prolong its exposure to the acidic conditions, leading to its
partial decomposition.
To explore the course of the hydrolytic conversion of
polysaccharides towards platform molecules in ZnCl ∙3H O, the
2 2
processing of microcrystalline cellulose (MCC) was performed.
Firstly, MCC was dissolved in this IL (loading of cellulose 1 wt%
based on IL, 80 °C, 1 h) and then the solution was heated under
constant agitation at 120 °C for 1 h during which the mixture
became dark-brown coloured. Dilution of the reaction medium
after completion of the process, with aqueous sodium hydrogen
carbonate, led to precipitation of the unreacted cellulose and
conversion of zinc chloride into insoluble zinc salts. The aqueous
phase yielded three main products, being HMF, its isomer furyl
hydroxymethyl ketone (FHK), and furfural (FF), each in yields
below 3.0 mol% based on the number of anhydroglucose units
present in the cellulose substrate. It is worth noting that low-
molecular-weight sugars, and the product of rehydration of
furanoids, namely, levulinic acid, were identified in trace amounts
only. However, there was a mass of dark-brown colour solids in
the water phase: these are typically considered to be high-
molecular-weight by-product humins.[6] The portfolio of products
suggests that cellulose hydrolyses into monomer glucose under
these reaction conditions, and that the glucose transforms rapidly
into HMF and FHK in the acidic reaction media. We propose that
this process proceeds, under our conditions, via hydrolysis of
cellulose into glucose, followed by isomerisation of glucose into
ketohexoses and their subsequent dehydration into furanoid
products. To test this proposal, we performed liquid
chromatography-mass spectrometry (LC-MS) analysis of the
recovered reaction media. The results confirmed the presence of
glucose and its two isomers fructose and an unidentified
ketohexose sugar, as proposed in Scheme 1 and presented in
Figure S1. FF is most likely formed in a sequence that involves
conversion of the intermediate hexose sugar fructose into one or
more pentose products (e.g., arabinose) followed by dehydration
into furan derivatives, as has been disclosed in the literature.[
Table 1. Yields of furan derivative from the processing of cellulose in the
biphasic system ZnCl
products.
2 2
∙3H O/organic solvent and partition coefficients of
log P[b]
log P[c]
Solvent
FHK yield,
[mol%][
FF yield,
[mol%]
a]
[a]
m-Xylene
16
16
20
14
–0.93
–0.45
0.03
13
11
8
0.64
0.48
0.98
0.97
–0.63
1.79
tert-Butylbenzene
Anisole
4
1
2
-Methylanisole
–0.37
–0.57
0.79
1
,2-Dimethoxybenzene 12
-sec-Butylphenol
11
2
6
[a] mol% based on anhydroglucose unit. [b,c] Logarithm of partition
coefficient of product between the organic solvent and the water-diluted ionic
liquid (1:1 wt/v IL/water) for FHK and FF, respectively. Reaction conditions:
MCC (5 mg), ZnCl
h.
2 2
∙3H O (500 mg), solvent-extractant (2.00 mL), 120 °C, 1
It had become clear that one method to improve the yield of
desirable products, FHK in particular, was to extract them into a
phase separate from the reaction medium, such as into an organic
solvent. But a more efficient means of removal of products from
the IL phase was required, and so we investigated various organic
solvents for this purpose. It is worth noting that common polar high
boiling solvents, such as dimethylformamide, dimethylsulfoxide,
γ-valerolactone, and methyl isobutyl ketone, which are useful for
19]
This dehydration process would be catalysed by the ZnCl
is acidic in nature, and is driven by aromatic stabilisation. The
furan-type molecules are unstable in ZnCl ∙3H O and are
converted into high-molecular-weight by-products, arguably by
condensation with low-molecular-weight sugars as was recently
highlighted.[20]
2
, which
2
2
the conversion of carbohydrates in water-based biphasic
[4,6]
2 2
systems, are miscible with ZnCl ∙3H O, and cause precipitation
To improve the outcome towards platform molecules, we
investigated the influence of loading of cellulose in the medium,
and reaction temperature and time. In no case did the reaction
afford products in yields higher than 5 mol%, persistently
accompanied by humins. The only useful strategy was the
continuous extraction of furan derivatives by m-xylene during the
process, which separates reactive furan derivatives from the
acidic medium/catalyst. This was accomplished by the addition of
this solvent-extractant from the start of the reaction. Additionally,
of cellulose. In turn, aromatic solvents like tert-butylbenzene,
anisole, 4-methylanisole, 1,2-dimethoxybenzene, or 2-sec-
butylphenol form a second phase, and therefore maintain the
polar phase of the reaction system, in which cellulose remains
dissolved. All of these aromatic solvents were applied to the
conversion of MCC in a biphasic system; among them, anisole
afforded high overall yields of and selectivity to FHK and FF (20
and 8 mol%, respectively, Table 1). This result may be
rationalised by the higher partition coefficients of products in
anisole, which serves to efficiently remove them from the catalytic
system and therefore limits side reactions. A measure of
miscibility between the ionic solvent and the organic solvent is
deleterious to the course of the reaction. For instance, the
occurrence of humins was observed when 1,2-dimethoxybenzene,
or 2-sec-butylphenol, was used as the organic phase; the partial
2 2
we identified that dilution of ZnCl ∙3H O with a small amount of
water (1 equivalent volume based on the weight of IL) after
completion of the process serves to better extract products into
the organic solvent phase, likely due to disruption of the ionic
solvent and a salting out effect. Analysis of the products in the
polar and m-xylene phases showed improved yields of FHK and
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
OH
OH OH
O
O
OH
HO
-
3
H O
2
OH
O
FHK
OH
e o exose
K t h
.
H
O
OH OH
O
OH
O
ZnCl2 3H O
2
O
HO
O
O
O
OH
O
HO
HO
HO
O
HO
-
n
3
H O
OH
2
OH OH
OH OH
Fructose
D
-
-
HMF
-
Cellulose OH
lucose
G
D
OH
O
O
O
H
HO
O
3
H O
2
OH OH
A
ra inose
b
H
-
FF
D
2 2
Scheme 1. Catalytic conversion of cellulose into furan-type molecules in ZnCl ·3H O.
miscibility led to degradation of the furan-type products with
concomitant diminished yields. Conversely, the less polar
aromatic solvent anisole demonstrates lower miscibility with the
IL, which protects the extracted products from side reactions.
Aiming to improve the conversion of cellulose into FHK and FF in
2 2
the biphasic system ZnCl ∙3H O/anisole, we conducted some
parameter testing. Firstly, the influence of the loading of cellulose
to IL (1, 1.25, 1.5, 1.75 and 2 wt% based on the IL) with a low
amount of extractant (2 v/w based on IL) was examined. While
the yields of FHK and FF remained essentially unchanged in
these reactions (Figure S2), ranging from 18–20 mol% and 7–8
mol%, respectively, there was an increase in the amount of
humins formed when employing higher concentrations of
cellulose. This is most likely due to longer residence times of
furanoids in the IL phase, consistent with the literature,[21] which
would promote acid-catalysed by-product formation. This occurs
because of non-exhaustive extraction of the FHK and FF into the
organic phase, leaving open the possibility of reaction between
the furanoids and sugars present.
Figure 1. Production of furanoids from reaction of cellulose in the biphasic
system ZnCl ∙3H O/anisole at various reaction temperatures. Reaction
conditions: MCC (5 mg), ZnCl ∙3H O (500 mg), anisole (3.00 mL), 1 h. ●: yield
2 2
2
2
of FHK; ○: yield of FF.
In response, and to enhance the efficiency of the extraction, the
2 2
conversion of MCC at a loading of 1 and 2 wt% in ZnCl ∙3H O
was performed making use of larger volumes of anisole. The
yields of FHK and FF are slightly improved to 23 mol% and 10
mol%, respectively, at 6 v/wt (based on IL) of anisole and 1 wt%
cellulose in the IL (Figure S3). The overall yields of FHK are lower
at 2 wt% loading of MCC, and the formation of humins was
observed in all instances. Overall there was no significant
difference in yield between the two loadings of cellulose nor in the
formation of humins.
Reaction temperature and the time for which the reactions were
conducted substantially influence the outcomes (Figure 1, 2). For
instance, the highest yields of FHK and FF were produced at
120 °C (Figure 1) and yields of FHK reached a peak at 1.5 h (24
mol%, Figure 2). Longer reaction times led to diminished yields
through by-product formation. At 100 °C (low temperature) we
identified small amounts of HMF (2 mol%), and only at that
temperature. Presumably, elevated temperatures favour reaction
pathways leading to FHK and FF, in preference to those
producing HMF (Scheme 1).
Figure 2. Production of furanoids from reaction of cellulose in the biphasic
system ZnCl ∙3H O/anisole at various reaction times. Reaction conditions: MCC
5 mg), ZnCl ∙3H O (500 mg), anisole (3.00 mL), 120 °C. ●: yield of FHK; ○:
yield of FF.
2
2
(
2
2
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
To test this proposal, low molecular-weight sugars, such as
cellobiose, glucose, and fructose, respectively, which putatively
appear during the course of the transformation of cellulose,[ were
subjected to reaction in the IL. As is evident from Figure 3, HMF
is one of the products of the conversion of sugars at temperatures
below 100 °C, reaching yields up to 16 mol% (based on sugar
substrate) from cellobiose or glucose. At higher temperatures
FHK and FF become the dominant furanoid products, with yields
of HMF falling away dramatically above 100 °C.
The processes with cellobiose, glucose and fructose at elevated
temperature was accompanied by complete conversion of
substrates (99% conversion of all substrates at the temperature
above 100 °C, Figures 3a-c). Losses in yields of desirable
products is most likely associated with the formation of humins,
which were noted in some instances. Apart from furan derivatives
and humins, we identified glucose and fructose during the
transformation of cellobiose at lower temperatures (yields up to
19 and 4 mol%, respectively, Figure 3a); the occurrence of
fructose (yield up to 26 mol%) is also apparent during the
conversion of glucose (Figure 3b). The portfolio of compounds
identified maps perfectly onto the acid-catalysed cascade of
cellulose conversion proposed in Scheme 1, involving hydrolysis,
aldose-ketose isomerisation and dehydration steps, respectively.
It is likely that the dehydration of fructose into HMF is only
favourable at temperatures below 100 °C. At higher temperatures
the rates of fructose isomerisation into ketohexose or the pentose
intermediate are increased, leading to FHK and FF becoming
major products. From the outcomes it appears that the rates of
the fructose isomerisation and further dehydration into FHK are
higher, relative to the processes forming FF, accounting for the
fact that FHK often appears in somewhat higher yields than FF.
There is an interesting contrast in this set of reactions compared
6]
to those involving cellulose. With cellulose,
a
reaction
temperature of 120 °C affords highest yields of furan derivatives;
with low molecular weight substrates the optimum temperature is
110 °C. This finding presumably reflects the need for elevated
temperatures to hydrolyse the glycosidic bonds in cellulose at an
appreciable rate compared to subsequent reactions that lead to
furan derivatives. Reactions converting glucose into furans
appear to proceed at higher rate than the hydrolysis reaction of
cellulose into glucose, and the optimum temperature that we
identified is likely a trade-off between optimised hydrolysis,
optimised formation of furans and minimised conversion into
humins.
It is possible, in theory, to advantage the hydrolysis of
polysaccharides by the addition of water or an acid catalyst, and
thus improve the yields of target products. However, reaction of
cellulose in zinc chloride with higher ratios of water
(
ZnCl
2 2 2 2 2 2
∙3.25H O and ZnCl ∙3.5H O), or ZnCl ∙3H O with added
acid catalysts (p-toluenesulfonic acid, oxalic acid, aluminium(III)-,
lanthanum(III)-, gadolinium(III)-, ytterbium(III)- or hafnium(IV)
trifluoromethanesulfonates, or lanthanum(III)- or gadolinium(III)
chlorides) afforded similar or diminished yield of furanoids (Table
S1). Arguably, the excess of water, or added catalyst, increases
the overall acidity of the reaction media leading to decomposition
of the ketone before its extraction, or to rehydration reactions
potentially affording levulinic acid.[2,6] However, levulinic acid was
not detected during the analysis of the reaction media and so this
process appears not to prevail. Overall, the conversion in neat
2 2
ZnCl ∙3H O remains the best process to date.
All of the reactions performed to this stage had been conducted
at 5 mg scale. A subsequent set of reactions was performed at
Figure 3. Conversion of carbohydrates (a: cellobiose; b: glucose; c: fructose)
and the yields of the major products obtained in the biphasic system
1
0× this scale under identical reaction conditions (Table 2). The
scaling operations revealed that the yields of the desired products
were reduced to 15% for FHK and remained relatively stable at
ZnCl
5 mg), ZnCl
glucose; □: yield of fructose; ●: yield of FHK; ○: yield of FF; ▲: yield of HMF.
2
∙3H
2
O/anisole at various reaction temperature. Reaction conditions: sugar
(
2
∙3H O (500 mg), anisole (3.00 mL), 1 h. ×: conversion; ■: yield of
2
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
reproducible conversion of cellulose into HMF in 4 catalytic runs
in aqueous zinc chloride solvent.[18]
Table 2. Results of scaling and comparative literature data of the
2 2
conversion of carbohydrates in ZnCl ∙nH
O[
a]
Conclusions
Entry
Substrate
Time Temp
FHK
yield
FF
yield
[mol%]
S
[%]
X
[%]
[h]
[°C]
[mol%]
This work details the one-pot transformation of cellulose, into the
furan-type platform products FHK and FF in high yield, in biphasic
media containing ZnCl ∙3H O and anisole. These two products
2 2
1[b]
2[c]
3[c]
MCC
1.5
1.5
1.5
120
115
115
15
20
32
9
33
48
46
73
61
99
MCC
9
are valuable but are less commonly formed in high yield in the
transformation of cellulosic substrates. The optimum method
avoids the use of added catalysts because the ionic solvent
possesses the intrinsic acidity required to facilitate all requisite
transformations, namely hydrolysis, isomerization, and
dehydration. In turn, anisole promotes the extraction of products
from acidic media and preserves them from decomposition.
These outcomes improve on the state of the art and provide a
framework, involving inexpensive, recoverable media for the
efficient direct conversion of cellulose into platform chemicals
under mild conditions. The work highlights the ongoing challenge
of removal of the reactive (under the reaction conditions) furanoid
products from the catalyst, which is likely the current major
obstacle to larger scale reactions and improved yields and
selectivities.
Cellobiose
14
Cotton
fibre
4[d,15]
5 min
135
12
–
–
–
Cotton
cellulose
5[e,17]
6[f,22]
1.67
2.5
190
270
1
1
6
–
–
–
–
Cellulose
–
[
[
(
(
a] S = total selectivity of furan-type molecules. X = conversion of substrate.
b] Reaction conditions: MCC (50 mg), ZnCl ∙3H O (5,000 mg), anisole
30.0 mL). [c] Reaction conditions: MCC or cellobiose (50 mg), ZnCl ∙3H
5,000 mg), anisole (40.0 mL). [d] Reaction conditions: cotton fibre (1,000
∙3.36H O (140 mL), heating under microwave irradiation 600
2
2
2
2
O
mg), ZnCl
W.
(
2
2
[
15]
2 2
[e] Reaction conditions: cotton cellulose (650 mg), ZnCl ∙3.56H O
100 mL), n-butanol (600 mL).[ [f] Reaction conditions: cellulose (1,000
17]
2 4 2
mg), compressed steam and nitrogen (2.5 MPa), calcined Ca(H PO ) (0.1
[
22]
mmol).
Experimental Section
Preparation of ionic liquids and processing of carbohydrates were
performed in oven-dried glassware at atmosphere pressure. Reagents
were used as supplied from commercial sources. All details about
analytical methods and equipment are specified in the Supporting
Information (SI).
9
% for FF (Entry 1). Arguably, this relates to reduced mass
transfer of products into the anisole phase leading to
decomposition of unstable molecules in the polar acidic reaction
media. To improve the outcome, the volume of anisole was
increased to 8 vol/wt of IL and the reaction temperature was
slightly reduced (by 5 °C). This improved the yield of FHK to 20%
while the yield of FF remained constant at 9% (Entry 2). It is
noteworthy that total selectivity of furan-type molecules was 48%
Preparation of zinc chloride hydrate solvent and dissolution of
carbohydrates
Anhydrous zinc chloride and deionised water in a molar ratio 1 to 3.0, 3.25
(
Entry 2), based on conversion of substrate (recovered unreacted
substrate), indicating that the transformation of MCC in
ZnCl ∙3H O is a reasonably selective process. Even higher yields
2 2 2 2 2 2
or 3.5, to form ZnCl ∙3H O, ZnCl ∙3.25H O or ZnCl ∙3.5H O (Table S2),
respectively, were introduced to a round-bottom flask equipped with a
thermometer and magnetic follower. The mixture was heated and agitated
at 80 °C for 30 minutes, resulting in a transparent colourless ionic liquid.
Then the microcrystalline cellulose (100–200 mg, to form a 1–2 wt%
solution based on the solvent) was introduced to the flask with the
prepared zinc chloride hydrate solvent (10.00 g) and the mixture was
heated and agitated at 80 °C for 1 h, resulting in a yellowish transparent
solution. The same method was employed for the dissolution of low-
molecular-weight carbohydrates (cellobiose, glucose, or fructose) but the
processing temperature was 60 °C.
2
2
of FHK and FF can be achieved from cellobiose in bench-scale
processes. The total yield of furan products is excellent (46% yield
of furan products, being a combination of 32% FHK and 14% FF,
Entry 3). To the best of our knowledge, these outcomes set the
benchmark in terms of yields of FHK and FF during the conversion
of poly- or oligoglucans under mild reaction conditions. For
comparison purposes, the targeted synthesis of FHK from
polysaccharides was hitherto possible only for processes
conducted under microwave heating in comparably lower yield
than those achieved in this work (12% yield of FHK, Entry 4).[15]
Otherwise, FHK and FF are attainable only in small amounts as
by-products of the degradation of cellulose (Entries 5, 6);[
usually, the major product secured from cellulose is HMF. It is
worth noting that zinc chloride hydrate solvent is amenable to
recovery and can potentially be reused, as has been shown in the
Microscale model reactions
A solution of microcrystalline cellulose or low-molecular-weight sugars
(cellobiose, glucose or fructose, 5–10 mg) in ZnCl
2 2
∙nH O (n = 3.0–3.5, 500
17,22,23]
mg, 1–2 wt% solution of carbohydrate based on the ionic liquid) and
organic solvent (2.00–6.00 mL of m-xylene, tert-butylbenzene, anisole, 4-
methylanisole, 1,2-dimethoxybenzene or 2-sec-butylphenol, respectively)
were introduced to a two-neck round-bottom flask equipped with a
condenser, thermometer and magnetic follower. The mixture was heated
and stirred at a predetermined reaction temperature and for a fixed period
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
of time. In some instances, Brønsted acids (p-toluenesulfonic acid or oxalic
acid, 1–20 mol% based on the number of anhydroglucose units present in
cellulose) or Lewis acids (aluminium(III)-, lanthanum(III)-, gadolinium(III)-,
ytterbium(III)- or hafnium(IV) trifluoromethanesulfonates, or lanthanum(III)-
or gadolinium(III) chlorides, 1–20 mol% based on the number of
anhydroglucose units present in cellulose) were added to the reaction
media when investigating the role of the catalyst on the outcome of the
conversion of cellulose. Recovery of the product included dilution of the
ionic liquid with 1 equivalent volume of water based on the weight of ionic
liquid in the reactor, followed by agitation of the biphasic system at room
temperature for 30 minutes, after which the organic solvent phase was
removed from the reactor. Work-up of the aqueous phase was performed
by the addition of aqueous sodium hydrogen carbonate (0.50 mL, 1 M),
followed by centrifugation and decantation (20,000 × g for 10 minutes) to
remove precipitated zinc salts. The aqueous and organic solvent phases
were analysed using high performance liquid chromatography (HPLC), as
detailed in SI, to provide the results described in the main text. All
microscale reactions were repeated at least three times.
(s, 2H), 3.27 (br s, 1H, OH); 13C NMR: (125 MHz, CDCl
, 25 °C): δ = 187.6,
3
150.1, 147.0, 117.9, 112.6, 65.1; IR (neat): νmax = 3331, 3133, 3115, 1678,
1563, 1468, 1421, 1321, 1270, 1168, 1113, 1029, 976, 912, 880, 774, 636,
590, 504 cm–1; HRMS (ESI): m/z calcd for C
H
O
H [M+H] : 127.0390,
+
6
6
3
found: 127.0378.
Furfural.[24] TLC: R
1
f
= 0.532 (1.5:1 hexane/EtOAc; UV, KMnO ); H NMR
4
(
(
500 MHz, CDCl
3
, 25 °C): δ = 9.63 (s, 1H), 7.68 (t, J = 1.0 Hz, 1H), 7.23
13
dd, J = 3.5, 1.0 Hz, 1H), 6.58 (dd, J = 3.5, 1.0 Hz, 1H). C NMR: (125
MHz, CDCl , 25 °C): δ = 177.8, 152.9, 148.0, 121.0, 112.5. IR (neat):
3
ν
9
9
max = 3132, 2807, 1673, 1568, 1473, 1394, 1367, 1278, 1157, 1081, 1021,
30, 883, 759, 595 cm–1. HRMS (ESI): m/z calcd for C
H [M+H]+:
H O
5 4 2
7.0284, found: 97.0278.
[25]
Hydroxymethylfurfural.
TLC: R
f
= 0.213 (1.5:1 hexane/EtOAc; UV,
1
KMnO
4
); H NMR (500 MHz, CDCl
.5 Hz, 1H), 6.51 (d, J = 3.5 Hz, 1H), 4.70 (s, 2H), 2.97 (br s, 1H, OH). 13
3
, 25 °C): δ = 9.56 (s, 1H), 7.21 (d, J =
3
C
NMR: (125 MHz, CDCl , 25 °C): δ =177.7, 160.7, 152.3, 123.0, 110.0, 57.6.
3
IR (neat): νmax = 3339, 3120, 2841, 1657, 1582, 1519, 1396, 1368, 1336,
Bench scale processes
278, 1188, 1070, 1017, 986, 965, 806, 768, 511 cm–1. HRMS (ESI): m/z
1
+
6 6 3
calcd for C H O H [M+H] : 127.0390, found: 127.0379.
A solution of microcrystalline cellulose or cellobiose (50 mg) in ZnCl
2 2
∙3H O
(5.00 g, 1 wt% solution of carbohydrate based on the ionic liquid) and
anisole (30.0 or 40.0 mL) were introduced to a two-neck round-bottom
flask equipped with a condenser, thermometer and magnetic follower. The
mixture was heated and stirred at a predetermined reaction temperature
Acknowledgements
(115 or 120 °C) for 1.5 h. Recovery of the products included dilution of
We thank University of Technology Sydney for financial support.
ionic liquid with 1 equivalent volume of water based on the weight of ionic
liquid in reactor followed by agitation of the biphasic system at room
temperature for 30 minutes. The two phases were separated and a small
amount of each (0.5 mL) was processed, as detailed above, before
chromatographic analysis using an HPLC system to provide the outcome
shown in the main text. The remaining aqueous phase was diluted with
aqueous hydrochloric acid (25.0 mL, 0.2 M) to precipitate the unreacted
cellulose, which was recovered, washed with deionised water (3 × 25.0
mL), vacuum oven-dried (60 °C, 1 mbar, 12 h) and weighed to provide the
conversion of cellulose detailed in the manuscript.
Keywords: biomass • carbohydrates • catalysis • ionic liquids •
sustainable chemistry
The authors declare no conflict of interest.
[
1]
2]
R.J. Van Putten, J.C. Van Der Waal, E. De Jong, C.B. Rasrendra, H.J.
Heeres, J.G. De Vries, Chem. Rev. 2013, 113, 1499–1597.
M. Dusselier, M. Mascal, B.F. Sels, in Selective catalysis for renewable
feedstocks and chemicals. Top. Cur. Chem., Vol 353 (Ed.: K. Nicholas),
Springer, Cham, 2014, pp. 1–40.
[
Isolation of furan-type molecules and recovery of solvents
[3]
I. Delidovich, K. Leonhard, R. Palkovits, Energy Environ. Sci. 2014, 7,
2803–2830.
Isolation of furan-type molecules from anisole phase comprised adsorption
of products on the solid adsorbents. A solution of products in anisole (40.0
mL, combined anisole phases after microscale conversion of
carbohydrates) was diluted with 1 equivalent volume of hexane and was
passed through a glass funnel containing silica gel (10.00 g) and activated
carbon (1.00 g) under ambient pressure. The polar products were trapped
on the solid adsorbents, while the anisole and hexane were removed as a
filtrate. Then furan-type molecules were desorbed with ethyl acetate (3 ×
[4]
I. Delidovich, P.J.C. Hausoul, L. Deng, R. Pfützenreuter, M. Rose, R.
Palkovits, Chem. Rev. 2016, 116, 1540–1599.
[5]
[6]
Z. Zhang, J. Song, B. Han, Chem. Rev. 2017, 117, 6834−6880.
I. Bodachivskyi, U. Kuzhiumparambil, D.B.G. Williams, ChemSusChem
2018, 11, 642–660.
[7]
[8]
Y. Liu, W. Xiao, S. Xia, P. Ma, Carbohydr. Polym. 2013, 92, 218–222.
C. Van Nguyen, D. Lewis, W.H. Chen, H.W. Huang, Z.A. ALOthman, Y.
Yamauchi, K.C.W. Wu, Catal. Today 2016, 278, 344–349.
Y. Shen, J.-K. Sun, Y.-X. Yi, B. Wang, F. Xua, R.-C. Sun, Bioresour.
Technol. 2015, 192, 812–816.
20.0 mL) and collected as a mixture. The recovery of products from the
diluted with 1 equivalent volume of water ionic liquid (10.0 mL, combined
aqueous phases after microscale conversion of carbohydrates) comprised
extraction with ethyl acetate (3 × 20.0 mL) under vigorous agitation at room
temperature for 10 min. The achieved solutions of products in ethyl acetate
after recovery from anisole and aqueous phases were combined and dried
[9]
[10] S. Fischer, H. Leipner, E. Brendler, W. Voigt, K. Fischer, in
Polysaccharide Applications. Vol. 737 (Eds.: M.A. El-Nokaly, H.A. Soini),
American Chemical Society, Washington, 1999, pp. 143–150.
[11] S. Fischer, H. Leipner, K. Thümmler, E. Brendler, J. Peters, Cellulose
2003, 10, 227–236.
over MgSO
4
and the solvent was removed under reduced pressure. The
mixture of furan-type molecules was subjected to flash column
chromatography (hexane–ethyl acetate, 1.5:1) to isolate FHK, FF and
HMF. The HPLC analysis of the recovered anisole and diluted with water
ionic liquid showed only trace amounts of furanoids remained in both
solvents implying almost complete recovery.
[12] S. Sen, J.D. Martin, D.S. Argyropoulos, ACS Sustainable Chem. Eng.
2013, 1, 858–870.
[13] R.J. Wilcox, B.P. Losey, J.C.W. Folmer, J.D. Martin, M. Zeller, R.
Sommer, Inorg. Chem. 2015, 54, 1109–1119.
[14] S. Sen, B.P. Losey, E.E. Gordon, D.S. Argyropoulos, J.D. Martin, J.Phys.
Furyl hydroxymethyl ketone.[
1
15]
Chem. B 2016, 120, 1134–1141.
TLC: R
, 25 °C): δ = 7.63 (dd, J = 1.5, 0.5
Hz, 1H), 7.30 (dd, J = 3.5, 0.5 Hz, 1H), 6.60 (dd, J = 3.5, 1.5 Hz, 1H), 4.74
f
= 0.308 (1.5:1 hexane/EtOAc;
[15] L. Yang, G. Li, F. Yang, S.M. Zhang, H.X. Fan, X.N. Lv, Carbohydr. Res.
011, 346, 2304–2307.
4 3
UV, KMnO ); H NMR (500 MHz, CDCl
2
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
[
16] T. Deng, X. Cui, Y. Qi, Y. Wang, X. Hou, Y. Zhu, Chem. Commun. 2012,
8, 5494–5496.
[21] H. Xia, S. Xu, X. Yan, S. Zuo, Fuel Process. Technol. 2016, 152, 140–
146.
4
[17] X. Lv, G. Li, F. Yang, P. Gao, Z. Liu, L. Meng, X. Yu, Ind. Eng. Chem.
Res. 2013, 52, 297–302.
[22] N. Shi, Q. Liu, T. Wang, L. Ma, Q. Zhang, Q. Zhang, ACS Sustainable
Chem. Eng. 2014, 2, 637–642.
[
18] Y.-R. Zhang, N. Li, M.-F. Li, Y.-M. Fan, RSC Adv. 2016, 6, 21347–21351.
19] L. Zhang, G. Xi, Z. Chen, D. Jiang, H. Yu, X. Wang, Chem. Eng. J. 2017,
[23] F. Shafizadeh, Adv. Carbohydr. Chem. 1968, 23, 419–474.
[24] M. Mascal, E.B. Nikitin, ChemSusChem 2009, 2, 423–426.
[25] N. Murai, M. Yonaga, K. Tanaka, Org. Lett. 2012, 14, 1278–1281.
[
307, 868–876.
[20] G. Tsilomelekis, M.J. Orella, Z. Lin, Z. Cheng, W. Zheng, V. Nikolakis,
D.G. Vlachos, Green Chem. 2016, 18, 1983–1993.
This article is protected by copyright. All rights reserved.

ChemPlusChem
10.1002/cplu.201800650
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Iurii Bodachivskyi, Unnikrishnan
Kuzhiumparambil, D. Bradley G.
Williams*
Page No. – Page No.
Acid-Catalysed Conversion of
Carbohydrates into Furan-Type
Molecules in Zinc Chloride Hydrate
Cellulose is abundant and renewable, and is converted into value-added small
molecules to be employed for chemical synthesis, replacing some reliance upon
petrochemical sources. In the present work, furyl hydroxymethyl ketone and furfural
are obtained as major products, which are valuable but less commonly formed in
high yields in transformations of cellulosic substrates.
This article is protected by copyright. All rights reserved.