The first molecular level monitoring of carbohydrate conversion to 5-hydroxymethylfurfural in ionic liquids. B2O3-an efficient dual-function metal-free promoter for environmentally benign applications

Article abstract of DOI:10.1002/cssc.201100670

The mechanistic nature of the conversion of carbohydrates to the sustainable platform chemical 5-hydroxymethylfurfural (5-HMF) was revealed at the molecular level. A detailed study of the key sugar units involved in the biomass conversion process has shown that the simple dissolution of fructose in the ionic liquid 1-butyl-3-methylimidazolium chloride significantly changes the anomeric composition and favors the formation of the open fructoketose form. A special NMR approach was developed for the determination of molecular structures and monitoring of chemical reactions directly in ionic liquids. The transformation of glucose to 5-HMF has been followed in situ through the detection of intermediate species. A new environmentally benign, easily available, metal-free promoter with a dual functionality (B₂O ₃) was developed for carbohydrate conversion to 5-HMF.

Full text of DOI:10.1002/cssc.201100670

DOI: 10.1002/cssc.201100670
The First Molecular Level Monitoring of Carbohydrate
Conversion to 5-Hydroxymethylfurfural in Ionic Liquids.
B O —An Efficient Dual-Function Metal-Free Promoter for
2
3
Environmentally Benign Applications
Elena A. Khokhlova, Vadim V. Kachala, and Valentine P. Ananikov*
[
a]
The mechanistic nature of the conversion of carbohydrates to
the sustainable platform chemical 5-hydroxymethylfurfural (5-
HMF) was revealed at the molecular level. A detailed study of
the key sugar units involved in the biomass conversion process
has shown that the simple dissolution of fructose in the ionic
liquid 1-butyl-3-methylimidazolium chloride significantly
changes the anomeric composition and favors the formation
of the open fructoketose form. A special NMR approach was
developed for the determination of molecular structures and
monitoring of chemical reactions directly in ionic liquids. The
transformation of glucose to 5-HMF has been followed in situ
through the detection of intermediate species. A new environ-
mentally benign, easily available, metal-free promoter with a
dual functionality (B O ) was developed for carbohydrate con-
2
3
version to 5-HMF.
Introduction
[
9]
Atmospheric carbon dioxide can be considered as the ultimate
carbon source for the sustainable future of the chemical indus-
try and for the elimination of barriers in the production of
transportation fuels. The photosynthetic fixation of carbon pro-
duces carbohydrates accounting for 75% of the world’s renew-
and others. An excellent promoter has been discovered re-
cently based on boric acid (B(OH) )—a weak, noncorrosive, and
3
[18,19]
nontoxic acid.
Zhu and coworkers have shown that B(OH)3
in combination with an organic ligand and Ru nanoparticles fa-
[18]
cilitates the conversion of cellulose. Riisager and coworkers
11
[1]
able biomass (ca. 1.7ꢀ10 tonnes). A central research chal-
lenge in the coming century is to develop practical and effi-
cient procedures for converting cellulose into platform chemi-
cals and biofuels. It has been estimated that up to one third of
the world’s demand for transportation fuels can be obtained
have demonstrated that B(OH) alone is an efficient metal-free
3
[19]
reagent for 5-HMF formation. The combination of a techno-
logical advantage (noncorrosive) with biocompatibility (non-
toxic and metal-free) and easy availability (low price reagent)
makes boric compounds very promising promoters for bio-
mass conversion.
[
2]
from cellulose biomass.
Continuous efforts have focused on the conversion of vari-
ous carbohydrates (fructose, glucose, sucrose, cellulose, etc.)
into 5-hydroxymethylfurfural (5-HMF)—a single renewable bio-
Two pathways—cyclic and acyclic—may be proposed based
on the sugar chemistry known for water and common organic
solvents (Scheme 1). The cyclic pathway is often discussed for
transformations involving ILs, whereas a considerably more
[
3]
mass-derived building block. A variety of relevant target
chemicals and petroleum-based industrial applications are ac-
[9,20]
complicated picture appears for the acyclic route.
A knowl-
[
4]
cessible through the use of 5-HMF. A diverse range of chemi-
cal and industrial applications take advantage of the nontoxic
edge of the mechanistic pathways of sugar conversion is of
principal importance for further development of the field to
create efficient procedures, control the selectivity of the reac-
[
5]
properties of 5-HMF. The material safety data sheet for 5-HMF
lists it as a low health hazard (Category 1) compound; further-
more, it is present in practically every food that contains carbo-
[9–20]
tion, and to avoid side reactions leading to humins.
In
spite of the tremendous interest in this topic, mechanistic
studies in IL systems are rather complicated due to the lack of
a powerful nondestructive analytic tool. The optimization of
[
6]
[7]
hydrates, it is used as a flavor component, and has been
[
8]
studied as an antitumor agent.
The one-pot formation of 5-HMF using hexoses derived di-
rectly from crude biomass has been achieved in several chemi-
cal systems with a high efficiency demonstrated in ionic liquid
[
a] E. A. Khokhlova, Dr. V. V. Kachala, Prof. Dr. V. P. Ananikov
Zelinsky Institute of Organic Chemistry
Russian Academy of Sciences
Leninsky Prospekt, 47, Moscow, 119991 (Russia)
Fax: (+7)499-1355328
[
9]
(
IL) media. Several systems have been developed in recent
À
years involving ILs with Cl anions to increase the solubility of
carbohydrates and various metal-containing promoters, for ex-
E-mail: val@ioc.ac.ru
[
10]
[11]
[12]
[13]
GeCl4,
[16]
ample, CrCl and CrCl ,
CoSO4,
FeCl2,
[15]
lanthanide
2
3
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201100670.
[
14]
[17]
salts, Cr/N-heterocyclic carbenes, WCl /HCl, Ta O /H O,
6
2
5
2
ChemSusChem 2012, 5, 783 – 789
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
783

V. P. Ananikov et al.
1
Figure 1. (a) H NMR spectrum of glucose in [BMIM]Cl with conventional
sample preparation; (b) the same sample after treatment in our NMR reactor,
which is shown in Figure 2. Conditions: 808C, 600 MHz. The inserts show
the magnified region containing the glucose signals.
Scheme 1. Cyclic and acyclic pathways for fructose conversion to 5-HMF.
NMR spectroscopy in ILs did not improve the quality of the
spectra of the studied system.
the reaction conditions for 5-HMF production in ILs mainly
relies on indirect analytical measurements based on extraction.
The extraction of 5-HMF from IL systems is a standalone ques-
tion, and a routinely used procedure involves treatment with
Surprisingly, a more detailed study has shown that it is the
microheterogeneous nature of the sample that causes the
challenging problems in the observation of NMR signals in
[
9]
[25]
water and an organic solvent. Such treatment destroys the IL
system and leads to the production of toxic waste containing
IL components and the possible contamination of the environ-
ILs. High-quality NMR spectra were recorded by using a spe-
cial reactor that eliminates microheterogeneity below the spec-
troscopic influence level (Figure 1b). Stirring the IL sample di-
rectly in the NMR tube was the key issue to improve the quali-
ty of the NMR spectra. Importantly, stirring in the test tube fol-
lowed by transfer of the sample into the NMR tube often led
to NMR spectra of an unacceptable quality. For better accuracy
and reproducibility of the measurements, it was of principal
importance to perform stirring directly in the NMR tube prior
to recording the spectra.
[
21]
ment. In addition, mechanistic studies and gaining a better
insight into the nature of the chemical processes in IL systems
are impossible without direct monitoring of the reactions.
The goal of our research was to understand the mechanistic
nature of carbohydrate conversion to 5-HMF at the molecular
level and to find efficient promoters for environmentally
benign applications. A reliable NMR approach was developed
for the direct spectral investigation of IL systems and charac-
terization of the intermediate species.
The NMR setup developed in the present study (Figure 2)
can be easily assembled and utilized with standard NMR hard-
ware (verified on NMR spectrometers operating at 300, 500,
1
13
and 600 MHz for H and 75, 125, and 150 MHz for C). Not
only 1D experiments, but also 2D heteronuclear multiple-bond
correlation (HMBC), heteronuclear single-quantum correlation
Results and Discussion
Development of an NMR methodology to study IL systems
(
HSQC), and COSY spectra were successfully recorded and uti-
[26]
Our continuous efforts to record NMR spectra of acceptable
quality in IL/carbohydrate systems failed in the initial stages
due to very broad spectral lines
lized to assign the NMR signals. When the temperature of
the sample was kept constant, the NMR study could be per-
and the inability to reproduce
the chemical shifts of spectral
patterns (Figure 1a). Indeed, the
difficulties of obtaining high-res-
olution NMR spectra in ILs asso-
ciated with high viscosity, the
conductive ionic character of the
medium, and absorption of radi-
ofrequencies are well document-
Figure 2. (a) Assembled reactor for native-state NMR spectroscopy in ILs. (b) Magnified images of glass stirrer and
NMR tube. (c) Example of a sample used to monitor the formation of 5-HMF in an IL by performing NMR spectros-
[
22–24]
ed.
Applying all the known
solutions to the issues related to copy.
7
84
www.chemsuschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 783 – 789

Monitoring Conversion of Carbohydrates in Ionic Liquids
formed for a few hours after stir-
ring (sufficient time to record 2D
spectra). When the quality of the
NMR spectrum decreased over
time or after
a temperature
change, the homogeneity of the
sample could be restored by stir-
ring again.
The elimination of microheter-
ogeneity made it possible to
record high-resolution NMR
spectra for a broad range of IL
systems: we have verified the
scope for individual solutions of
fructose, glucose, sucrose, and 5-
HMF as well as reaction mixtures
in ILs. Thus, this approach
should be very useful for con-
ducting mechanistic studies in
native-state IL systems.
Monitoring of carbohydrate
conversion to 5-HMF
Scheme 2. Interconversion of carbohydrates and formation of 5-HMF.
Using this NMR approach, we
were able to record high-resolution NMR spectra directly in
[
27]
ILs. Thus, for the first time, the signals of individual anomers
were completely resolved (Scheme 2). Surprisingly, dissolution
of fructose in 1-butyl-3-methylimidazolium chloride ([BMIM]Cl)
leads to a significant change in the anomeric composition of
1
/2/3/4/5=14:24:1:55:6 compared to the known data for a
[
26]
water solution of 8:40:1:50:<1. In contrast to fructose, the
dissolution of glucose in ILs did not result in a significant
change in the ratio between the anomers (cf. Tables S1 and S2
in the Supporting Information).
An increase in the amount of the open fructoketose 5 is no-
ticeable, as well as the difference in the amounts of 1 and 2.
All of the species 1–8 have been characterized, and the data
are provided in Tables S1–S9 (Supporting Information). The de-
hydration of fructose takes place directly in the IL and leads to
the formation of 5-HMF (Scheme 2). The conversion of glucose
is more difficult and requires a promoter to furnish the trans-
formation into fructose, which is followed by dehydration.
Other carbohydrates, such as sucrose, undergo splitting into
glucose and fructose in ILs. The changes in the anomeric com-
positions in ILs give rise to both acyclic and cyclic pathways in
the formation of 5-HMF and humins.
1
3
1
Figure 3. C{ H} NMR spectrum (150 MHz, 808C) of the native reaction mix-
ture: [BMIM]Cl (&
in Scheme 3.
), glucose (*), 5-HMF (~), and borate complex 9 (#) shown
showed the expected order of relative reactivity: fructose>su-
crose>glucose (Table 1, entries 1–3). Interestingly, phenylbor-
onic acid also exhibited moderate activity in this reaction (en-
tries 4 and 5). This suggests that the coordination of B(OH)3
with a single sugar unit is sufficient for the reaction and rules
[28]
out the assumption that two sugar units are involved.
Unexpectedly, boron trioxide (B O ) showed excellent activity
This NMR approach was efficient not only for the characteri-
zation of molecular systems in ILs, but also for direct monitor-
ing of the chemical transformations. Stirring inside the NMR
tube played a dual role—to eliminate microheterogeneity and
to ensure continuous mixing of the reaction solution. We re-
corded high-resolution NMR spectra of the reaction mixtures,
which allowed us to monitor the initial reagents, the products,
and even to identify a borate/sugar complex (9, Figure 3).
NMR monitoring was successfully performed with the 5-HMF
2
3
in 5-HMF formation leading to 98–99% conversion for all of
the carbohydrates studied (Table 1, entries 6–8). The reaction
À
was highly anion dependent with Cl necessary for 5-HMF for-
À
mation (entries 8–10). The PF6 anion was found to inhibit the
À
À
reaction (entry 11), whereas BF4 was compatible with Cl and
resulted in excellent conversion (entry 12). The mixture of
[BMIM]Cl and [BMIM][BF ] has advantages for some practical
4
applications because it is considerably less viscous than pure
[BMIM]Cl.
[19]
formation reaction using the known promoter B(OH)3, which
ChemSusChem 2012, 5, 783 – 789
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
785

V. P. Ananikov et al.
Thus, to develop an efficient IL system and to understand
the reaction mechanism it is important to perform an NMR
study of the IL systems during the reaction and before the iso-
lation. If native-state NMR spectroscopy is not applied to the
characterization of IL systems, the reactions are optimized
based on the “indirect” yields determined after workup (this
may not always correspond to the real yields after reaction).
This again justifies the high importance of our NMR procedure.
Table 1. Direct monitoring of carbohydrate conversion to 5-HMF under
native-state IL conditions in the NMR reactor.
[
a]
Entry IL Sugar Conditions
Conver-
sion [%]
1
2
3
4
5
6
7
8
9
0
[BMIM]Cl
fructose B(OH)
sucrose B(OH)
glucose B(OH)
sucrose PhB(OH)
glucose PhB(OH)
3
3
3
, 1008C, 2 h
, 1208C, 3 h
, 1208C, 5 h
90
85
61
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
[BMIM]Cl
[BMIM][PF
[BMIM][BF
2
, 1208C, 1 h 69
, 1208C, 1 h 42
2
fructose
sucrose
glucose
glucose
glucose
glucose
glucose
B
B
B
B
B
B
B
2
O
2
O
2
O
2
O
2
O
2
O
2
O
3
3
3
3
3
3
3
, 1008C, 1 h
, 1208C, 2 h
, 1208C, 5 h
, 1208C, 5 h
, 1208C, 5 h
, 1208C, 5 h
, 1208C, 5 h
98
99
99
0
0
0
99
98
93
In situ study of carbohydrate conversion to 5-HMF
From a mechanistic point of view, the B O -promoted conver-
6
]
]
2
3
1
4
sion of glucose to fructose involves the formation of 1,2-
borate derivatives. One such 1,2-borate complex with glucose,
[
[
b]
b]
1
1
[BMIM]Cl/[BMIM][PF
[BMIM]Cl/[BMIM][BF
[BMIM]Cl
6
]
]
12
14
15
4
13
1
9
, has been detected in situ by performing a C and H NMR
sucrose BF
glucose BF
3
3
, 908C, 0.5 h
, 908C, 0.5 h
spectroscopy study on an IL (Scheme 3). The difference be-
tween B O and B(OH) is governed by the ability of the former
[BMIM]Cl
2
3
3
[
a] The formation of 5-HMF was studied by using NMR spectroscopy, and
1
3
1
to facilitate dehydration by the trapping of water.
the conversion of sugars was calculated based on C{ H} spectra. Condi-
tions: IL (0.6 g), sugar (0.56 mmol), and B (0.28 mmol) or B(OH)
PhB(OH) (0.56 mmol). For a detailed description of all experimental pro-
cedures see the Experimental Section. [b] 1:1 mixture of the ILs.
2
O
3
3
/
Complex 9 was rather stable and most likely represented a
resting state in the studied reaction (Scheme 3). After the com-
plete conversion of glucose and fructose to 5-HMF, there was
still a detectable amount of 9, and the conversion of 9 to 5-
HMF was rather slow. Even if the reaction was started with
fructose (Table 1, entries 1 and 6), 9 was still detected when
performing NMR spectroscopy (due to the equilibrium be-
tween fructose and glucose in solution and trapping of the
latter as a borate complex). The decomposition of 9 resulted in
2
The complete conversion of carbohydrates as determined by
our NMR procedure was independently confirmed by perform-
ing chromatography. None of the reagents (fructose, glucose,
sucrose) were identified in the separation experiments after
full conversion was measured by using NMR spectroscopy. The
separation and quantitative recycling of boric reagents was
achieved by treatment with water. Good yields of the isolated
products were found by using B O as the promoter (60%).
the elimination of B(OH) as shown by NMR spectroscopy.
3
In addition to glucose-to-fructose isomerization, B O facili-
2
3
tated the dehydration of fructose into 5-HMF by trapping
water and shifting the equilibrium. Both the cyclic and acyclic
pathway for 5-HMF formation require the elimination of three
water molecules for each sugar unit (Scheme 1). As detected
2
3
Although the optimization of the isolation procedure is out
of the scope of this study, we performed a preliminary investi-
gation to understand the issues related to the side reactions
[
29]
11
and the decrease of the isolated yield of 5-HMF. After the re-
actions, we isolated a dark-brown solid consisting of humins,
which were insoluble in most organic solvents. Molecular
weights of around 2700 Da were measured by using ESI–MS.
The formation of humins as byproducts during the conversion
of carbohydrates to 5-HMF has been reported and has been
shown to decrease the selectivity and yield of 5-HMF
by performing B NMR spectroscopy, the reaction of B O with
2
3
water furnished the formation of B(OH) , which may also act as
3
a promoter. Thus, B O was a highly efficient reagent with two
2
3
functions—a promoter for glucose conversion into fructose
and a dehydration agent for 5-HMF formation (Scheme 3). To
further test the hypothesis, the process was investigated in the
presence of BF , which is a more efficient Lewis acid and
3
[
30]
production.
water-trapping compound and should possess higher activity.
In total agreement, the reaction was performed under mild
conditions at 908C and resulted in 93–98% conversion in 0.5 h
Interesting results were obtained from an ESI–MS study of
the isolation procedure, indicating a possible increase of the
yield of humins during the isolation step: intense polymer sig-
nals were observed after workup, whereas the intensities of
the polymer signals were considerably lower in the crude mix-
tures. These findings show that the isolation procedure itself
may affect the composition of the final products.
[32]
(Table 1, entries 14–15).
Thus, using our NMR approach, direct monitoring of the re-
action with glucose was successfully achieved and resulted in
the comprehensive detection of reagent conversion, product
formation, interconversion of B O and B(OH) , a borate com-
2
3
3
1
Another important effect—the slow release of 5-HMF—was
observed upon heating/ultrasonic treatment of the isolated
brown solid and repeated extraction with organic solvents.
This observation may indicate the presence of weak interac-
tions or reversible adsorption involving 5-HMF and humins
and/or the slow decomposition of humins resulting in the for-
mation of 5-HMF. This preliminary data deserves further atten-
tion as 5-HMF recovery from humins would be a topic of con-
siderable practical importance.
plex, and the anomeric forms of carbohydrates by using H,
11
13
B, and C NMR spectroscopy.
Conclusions
A new environmentally benign, easily available, metal-free pro-
moter with dual functionality (B O ) has been found for the
2
3
conversion of carbohydrates to 5-HMF. We have shown that
the complete conversion of the carbohydrate can be achieved
7
86
www.chemsuschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 783 – 789

Monitoring Conversion of Carbohydrates in Ionic Liquids
[
31]
Scheme 3. In situ NMR study of glucose conversion to 5-HMF using our NMR reactor.
NMR experiments
in the IL system studied. The evaluation of the 5-HMF extrac-
tion procedure calls for a better understanding of the process
of humin formation and more careful consideration of the iso-
lation/purification protocols.
1
13
H and C NMR chemical shifts are given relative to external
1
13
[D ]DMSO (capillary insert), d=2.50 and 39.5 ppm for H and C,
6
11
respectively. B NMR chemical shifts are given relative to external
.1m H BO in a capillary at d=0.0 ppm. 908 pulses were recali-
0
The anomeric compositions of fructose and glucose were
determined for the first time in an IL. Quite surprisingly, simple
dissolution of fructose in the IL led to the appearance of the
open fructoketose form. This is of particular significance for
sugar chemistry in ILs beyond the reaction studied here.
We have developed an efficient NMR approach with the abil-
ity to characterize molecular transformations in IL systems, to
resolve individual species in complex equilibrium, and to
detect reaction intermediates. This NMR method can reveal
previously inaccessible mechanistic details of native-state IL
systems at the molecular level.
3
3
brated for the samples in ionic liquids (ILs).
1
Typical parameters for H NMR spectra were a spectral width of
12000 Hz and a pulse width of 20 ms at a power level of 17 W. Typ-
13
ical parameters for C NMR spectra were a spectral width of
3 kHz and a pulse width of 13.9 ms at a power level of 63 W and a
3
[33]
relaxation delay of 2 s. Inverse gate WALTZ-16 broadband proton
13
decoupling was used for C NMR spectra, the free induction
decays (FIDs) were multiplied by performing an exponential
weighting (LB=2 Hz) before Fourier transformation. Typical param-
11
eters for B NMR spectra were a spectral width of 7600 Hz and a
11
pulse width of 13.9 ms at a power level of 58 W. The B NMR spec-
tra were recorded without proton decoupling. 1D selective gradi-
ent-selected (gs)-NOESY spectra were recorded using 80 ms
sinc.1000 1808-shaped pulse with gradient selection.
Experimental Section
1
1
2
D ( H– H) gs-COSY and inverse proton detected heteronuclear
General experimental procedures and methods
1
13
spectra ( H– C) gs-heteronuclear single-quantum correlation
(HSQC), and gs-heteronuclear multiple-bond correlation (HMBC)
were recorded and processed by using standard Bruker NMR soft-
ware. Parameters for COSY were a spectral width of 3000–4800 Hz,
an acquisition data size of 1024 points and eight transients accu-
mulated per increment with 1 s relaxation for a total of 128–256
experiments, data processing by using zero filling in the F1
domain, and shifted sine-bell apodization of factor 0 in both di-
The NMR reactor was assembled directly in a standard NMR tube
as shown in Figure 2 and Figure S1 (Supporting Information). All
NMR measurements were performed on native IL systems without
alteration for the purpose of spectral study. All H, C, and B NMR
spectra were recorded by using a Bruker Avance II 600 instrument
at 303 and 353 K. The NMR spectra and tables of chemical shifts
are provided in the Supporting Information. Spectral data for
water solutions measured under the same conditions are also pro-
vided in the Supporting Information for comparison.
1
13
11
1
13
mensions. Parameters for ( H– C) gs-HSQC and gs-HMBC were a
1
13
spectral width of 4800–6600 Hz for H and 22.0–35 kHz for C,
ChemSusChem 2012, 5, 783 – 789
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
787

V. P. Ananikov et al.
1
024ꢀ256 data sets, four or eight scans (gs-HSQC) and eight or 16
The structure of 9 was established by using 1D selective gs-NOESY
1
13
[26]
scans (gs-HMBC), and a relaxation delay of 1 s. The H– C coupling
value in the HMBC spectrum was set to 8 Hz. The FIDs were pro-
cessed using zero filling in the F1 domain, and a sine-bell (gs-
HMBC) and sine-squared (gs-HSQC) window function in both di-
mensions was applied prior to Fourier transformation. C GARP
decoupling was used in gs-HSQC experiments.
and 2D gs-HSQC and COSY spectra. The signals of protons H-1
1
and H-2 in the H NMR spectrum were shifted to a lower field in
comparison to free glucose in the IL. Due to the strong overlap of
IL protons with the H-3, H-4, and H-5 protons of glucose, they
were assigned using 1D NOESY spectra, where the following corre-
lations were discovered: H-2 and H-4, H3 and H-5, and H-1 and H-
13
[33]
2
. The last correlation, as well as a J value of 3.8 Hz between H-1
and H-2, indicated the a-anomeric configuration of glucose. The
same complex was detected in the reaction involving B O .
2
3
Electrospray ionization mass-spectrometry
High resolution mass spectra (HR–MS) were measured by using
Bruker MicroTOF II and Maxis instruments using electrospray ioni- Acknowledgements
zation. The measurements were performed in the positive ion
mode (interface capillary voltage=4500 V) and in the negative ion
mode (3200 V) for the studied mass range; internal calibration was
performed using an electrospray calibrant solution (Fluka). A sy-
ringe injection was used for solutions in acetonitrile, methanol, or
The work was supported by the Russian Foundation for Basic Re-
search (10-03-00370), grant MD-4969.2012.3, and the Programs
of Division of Chemistry and Material Sciences of RAS.
À1
water (flow rate 3 mLmin ). Nitrogen was applied as a dry gas,
and the interface temperature was set at 1808C. MS was applied
to confirm the polymeric nature of the material, but not for the
measurements of precise masses as fragmentation/destruction of
the polymer may occur.
Keywords: biomass
spectroscopy · sustainable chemistry
·
furfural
·
ionic liquids
·
nmr
[
1] a) L. D. Schmidt, P. J. Dauenhauer, Nature 2007, 447, 914; b) A. Corma, S.
Iborra, A. Velty, Chem. Rev. 2007, 107, 2411; c) H. Rçper, Starch/Staerke
2
002, 54, 89.
[
[
2] M. Chidambaram, A. T. Bell, Green Chem. 2010, 12, 1253.
3] Top Value Added Chemicals from Biomass. Vol. I. Results of Screening for
Potential Candidates from Sugars and Synthesis Gas (Eds.: T. Werpy, G.
Petersen), U.S. Department of Energy, Golden, 2004, http://www.nrel.-
gov/docs/fy04osti/35523.pdf.
Carbohydrate conversion to 5-hydroxymethylfurfural
In a typical experiment, a reaction mixture containing glucose
(
0.56 mmol, 0.1 g), boron promoter (B O : 0.28 mmol, 19.5 mg;
2 3
B(OH)3 0.56 mmol, 34.6 mg; BF ·Et O: 0.56 mmol, 79.5 mg) and
3
2
[
4] Selected representative studies: a) J. O. Metzger, Angew. Chem. 2006,
[
BMIM]Cl (3.43 mmol, 0.6 g) was transferred under Ar into an NMR
118, 710; Angew. Chem. Int. Ed. 2006, 45, 696; b) T. Thananatthanachon,
tube. The resulting sample was placed in an oil bath at 90–1208C.
The NMR reactor with stirrer was assembled and operated at 90–
T. B. Rauchfuss, Angew. Chem. 2010, 122, 6766; Angew. Chem. Int. Ed.
2010, 49, 6616; c) A. A. Rosatella, S. P. Simeonov, R. F. M. Frade, C. A. M.
Afonso, Green Chem. 2011, 13, 754; d) M. J. Climent, A. Corma, S. Iborra,
Green Chem. 2011, 13, 520; e) J. J. Verendel, T. L. Church, P. G. Ander-
sson, Synthesis 2011, 11, 1649; f) B. Kamm, M. Kamm, M. Schmidt, T.
Hirth, M. Schulze in Biorefineries: Industrial Processes and Products, Vol. 2
1
208C with continuous stirring (800 rpm) for the indicated time in-
tervals (Table 1). The NMR tube was taken out of the oil bath,
placed into a 200 mL glass beaker containing 100 mL of water at
8
08C and within <3 min transferred into the NMR spectrometer
(
Eds.: B. Kamm, P. R. Gruber, M. Kamm), Wiley-VCH, Weinheim, 2006,
with the probe temperature stabilized at 808C. Standing the
sample in the NMR tube for a long time after stirring caused dete-
rioration of the spectra quality; thus, microheterogeneity should
be eliminated by stirring directly before measurement.
pp. 97–149; g) F. K. Kazi, A. D. Patel, J. C. Serrano-Ruiz, J. A. Dumesic,
R. P. Anex, Chem. Eng. J. 2011, 169, 329; h) A. I. Torres, P. Daoutidis, M.
Tsapatsis, Energy Environ. Sci. 2010, 3, 1560; i) X. Tong, Y. Ma, Y. Li, Appl.
Catal. A: Gen. 2010, 385, 1; j) T. Buntara, S. Noel, P. H. Phua, I. Meliꢂn-
Cabrera, J. G. de Vries, H. J. Heeres, Angew. Chem. 2011, 123, 7221;
Angew. Chem. Int. Ed. 2011, 50, 7083; k) F. M. A. Geilen, T. vom Stein, B.
Engendahl, S. Winterle, M. A. Liauw, J. Klankermayer, W. Leitner, Angew.
Chem. 2011, 123, 6963; Angew. Chem. Int. Ed. 2011, 50, 6831.
The 5-HMF yield was calculated on the basis of the integrals of the
anomeric carbon atoms (C-1) of a-d-glucopyranose, b-d-glucopyra-
nose, a-d-glucopyranose-1,2-borate and C-4 of 5-HMF in
1
3
1
1
13
1
C{ H}NMR spectra. The H and C{ H} NMR spectra were recorded
by applying standard pulse length and shim adjustments. The
À1
[
5] The acute toxicity of 5-HMF is rather low, LD50 =3.1 gkg body weight
in rats: a) R. J. Ulbricht, S. J. Northup, J. A. Thomas, Fundam. Appl. Toxi-
col. 1984, 4, 843. For biological and metabolic effects see: b) H. R. Glatt,
Y. Sommer in Acrylamide and Other Health Hazardous Compounds in
Heat-Treated Foods (Eds.: K. Skog, J. Alexander), Woodhead Publishing,
Cambridge, 2006, pp. 328–353.
1
3
1
C{ H} NMR spectra for quantitative analysis and determination of
yields were recorded by using an inverse gated pulse sequence.
Each experiment (Table 1) was performed three times, and the re-
sults were totally reproducible. After complete conversion, the
products were separated and isolated in 60% yield by using pub-
À1
À1
[
6] For example, up to 0.08 mgg in cookies, 1.9 mgg in roasted coffee,
9
À1
.5 mgg in caramel products, etc.: a) M. Murkovic, M.-A. Bornik, Mol.
[19,20]
lished procedures (selectivity not determined).
Nutr. Food Res. 2007, 51, 390; b) L. A. Ameur, G. Trystram, I. Birlouez-
Aragon, Food Chem. 2006, 98, 790; c) M. Murkovic, N. Pichler, Mol. Nutr.
Food Res. 2006, 50, 842; d) S. Bachmann, M. Meier, A. Kꢃnzig, Lebensmit-
telchemie 1997, 51, 49.
Detection of the borate complex 9
[
[
[
7] J. M. de Man, Principles of Food Chemistry, 3rd ed., Springer, Heidelberg,
1
999.
Complex 9 was prepared as follows: 1-butyl-3-methylimidazolium
chloride ([BMIM]Cl, 4.01 mmol, 0.70 g), glucose (0.28 mmol,
8] K. Michail, V. Matzi, A. Maier, R. Herwig, J. Greilberger, H. Juan, O.
Kunert, R. Wintersteiger, Anal. Bioanal. Chem. 2007, 387, 2801.
5
0.4 mg), and B(OH) (0.84 mmol, 51.9 mg) were transferred under
3
9] Recent reviews: a) M. E. Zakrzewska, E. Bogel-Łukasik, R. Bogel-Łukasik,
Chem. Rev. 2011, 111, 397; b) T. Stꢄhlberg, W. Fu, J. M. Woodley, A. Riis-
ager, ChemSusChem 2011, 4, 451; c) S. Van de Vyver, J. Geboers, P. A.
Jacobs, B. F. Sels, ChemCatChem 2011, 3, 82; d) H. Olivier-Bourbigou, L.
Magna, D. Morvan, Appl. Catal. A: Gen. 2010, 373, 1; e) O. A. El Seoud, A.
Ar into an NMR tube and heated in an oil bath at 1008C with con-
tinuous stirring for 1 h. The sample was transferred into the NMR
spectrometer as described above, and the measurements were
conducted at 808C.
7
88
www.chemsuschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2012, 5, 783 – 789

Monitoring Conversion of Carbohydrates in Ionic Liquids
Koschella, L. C. Fidale, S. Dorn, T. Heinze, Biomacromolecules 2007, 8,
[25] Possible origins/initiators of heterogeneity in IL (selected examples):
a) D. Jeong, D. Kim, M. Y. Choi, H. J. Kim, Y. J. Jung, Dynamic Heterogenei-
ty in Room-Temperature Ionic Liquids in Ionic Liquids: Theory, Properties,
New Approaches (Ed.: A. Kokorin), InTech, 2011, pp. 167–182; b) Y.-F.
Hu, Z.-C. Liu, C.-M. Xu, X.-M. Zhang, Chem. Soc. Rev. 2011, 40, 3802;
c) M. F. C. Gomes, J. N. C. Lopes, A. A. H. Padua, Top. Curr. Chem. 2009,
290, 161; d) Z. Hu, C. J. Margulis, Proc. Natl. Acad. Sci. USA 2006, 103,
831; e) Y. T. Wang, G. A. Voth, J. Am. Chem. Soc. 2005, 127, 12192.
[26] NMR spectra, signal assignment, anomeric composition, and tables of
chemical shifts are provided in the Supporting Information.
[27] It was shown previously that even small amounts of additives used for
spectral studies (such as deuterated solvents, internal standards, etc.)
can significantly change the properties of ILs (see Ref. [22]). Therefore,
the IL systems should be studied in their native state by NMR.
[28] See Scheme S1 in the Supporting Information for further details.
[29] In spite of its apparent simplicity, the separation of 5-HMF from ILs is a
challenging task with many controversial reports (see Ref. [9] and refer-
ences therein). Complete and nondestructive separation of 5-HMF leav-
ing the IL intact (rather than destruction by water/organic solvent ex-
traction) is of particular importance.
2629; f) J. Lewkowski, ARKIVOC 2001, 1, 17; g) H. Wang, G. Gurau, R. D.
Rogers, Chem. Soc. Rev. 2012, 41, 1519.
[
10] a) H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 2007, 316,
1597; b) E. A. Pidko, V. Degirmenci, R. A. van Santen, E. J. M. Hensen,
Angew. Chem. 2010, 122, 2584; Angew. Chem. Int. Ed. 2010, 49, 2530;
c) J.-A. Chun, J.-W. Lee, Y.-B. Yi, S.-S. Hong, C.-H. Chung, Starch/Staerke
2
2
010, 62, 326; d) Y. Zhang, E. A. Pidko, E. J. M. Hensen, Chem. Eur. J.
011, 17, 5281; e) Z. Zhang, Z. Zhao, Bioresour. Technol. 2011, 102, 3970;
f) J.-W. Lee, M.-G. Ha, Y.-B. Yi, C.-H. Chung, Carbohydr. Res. 2011, 346,
77; g) M. Tan, L. Zhao, Y. Zhang, Biomass Bioenergy 2011, 35, 1367;
1
h) Q. Cao, X. Guo, S. Yao, J. Guan, X. Wang, X. Mu, D. Zhang, Carbohydr.
Res. 2011, 346, 956.
[
[
[
11] F. Tao, H. Song, L. Chou, Carbohydr. Res. 2011, 346, 58.
12] F. Tao, H. Song, L. Chou, ChemSusChem 2010, 3, 1298.
13] Z. Zhang, Q. Wang, H. Xie, W. Liu, Z. K. Zhao, ChemSusChem 2011, 4,
131.
[
[
14] T. Stꢄhlberg, M. G. Sørensen, A. Riisager, Green Chem. 2010, 12, 321.
15] G. Yong, Y. Zhang, J. Ying, Angew. Chem. 2008, 120, 9485; Angew. Chem.
Int. Ed. 2008, 47, 9345.
[
[
[
16] L. Lai, Y. Zhang, ChemSusChem 2010, 3, 1257.
[30] a) B. R. Caes, R. T. Raines, ChemSusChem 2011, 4, 353; b) F. S. Asghari, H.
Yoshida, Ind. Eng. Chem. Res. 2006, 45, 2163; c) C. Sievers, M. B. Valen-
zuela-Olarte, T. Marzialetti, D. Musin, P. K. Agrawal, C. W. Jones, Ind. Eng.
Chem. Res. 2009, 48, 1277; d) S. K. R. Patil, C. R. F. Lund, Energy Fuels
2011, 25, 4745; e) See also Refs. [2,4k,9,14,20].
17] F. Yang, Q. Liu, M. Yue, X. Bai, Y. Du, Chem. Commun. 2011, 47, 4469.
18] Y. Zhu, Z. N. Kong, L. P. Stubbs, H. Lin, S. Shen, E. V. Anslyn, J. A. Maguire,
ChemSusChem 2010, 3, 67.
[
[
19] a) T. Stꢄhlberg, S. Rodriguez-Rodriguez, P. Fristrup, A. Riisager, Chem.
Eur. J. 2011, 17, 1456; b) T. S. Hansen, J. Mielby, A. Riisager, Green Chem.
011, 13, 109.
20] a) J. B. Binder, A. V. Cefali, J. J. Blank, R. T. Raines, Energy Environ. Sci.
010, 3, 765; b) S. J. Dee, A. T. Bell, ChemSusChem 2011, 4, 1166; c) R.
[31] Only two anomers of fructose are shown in the scheme (see Scheme 2),
+
+
2
and H /H
3
O
as counterions are omitted for clarity.
[32] Although BF
3
showed higher activity, further studies are needed to
2
reveal the potential applicability for 5-HMF production as reaction with
water may produce toxic and corrosive HF.
[33] A. J. Shaka, Decoupling Methods in Encyclopedia of Nuclear Magnetic Res-
onance, Vol. 3 (Eds.: D. M. Grant, R. K. Harris), Wiley, Chichester, 1996,
p. 1558.
Karinen, K. Vilonen, M. Niemela, ChemSusChem 2011, 4, 1002 and refer-
ences therein.
[
[
[
21] T. Welton, Green Chem. 2011, 13, 225.
22] V. P. Ananikov, Chem. Rev. 2011, 111, 418.
23] W. R. Carper, Malten Salts in Encyclopedia of Magnetic Resonance, Wiley
Online Library, 2011, DOI:10.1002/9780470034590.emrstm0316.pub2.
24] R. Giernoth in Ionic Liquids in Chemical Analysis (Ed.: M. Koel), CRC
Press, Boca Raton, 2009, p. 355.
[
Received: October 27, 2011
Published online on February 22, 2012
ChemSusChem 2012, 5, 783 – 789
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
789