Discovery and Exploration of the Efficient Acyclic Dehydration of Hexoses in Dimethyl Sulfoxide/Water

Article abstract of DOI:10.1002/cssc.201902322

Current gaps in the development of sustainable processes include a lack of strategies to systematically identify and optimize the formation of new products. The dehydration of hexoses to 5-hydroxymethylfurfural (HMF) is a particularly widely studied process. In an attempt to identify a new high-selectivity conversion of glucose, quantitative NMR spectroscopy is used to screen conditions that have been reported to yield high conversions of glucose but low formation of HMF. In this manner, an olefinic six-carbon byproduct is identified. By adding water, selectivity for the compound was nearly tripled relative to previous reports. The detection of high-yielding side reactions in the formation of HMF is remarkable, considering how extensively HMF formation has been studied. High selectivity for the acyclic pathway allows hitherto unobserved intermediates in this pathway to be identified by using in situ NMR spectroscopy. An additional, presumably cyclic, pathway contributes to HMF formation.

Full text of DOI:10.1002/cssc.201902322

Accepted Article
Title: Discovery and Exploration of the Efficient Acyclic Dehydration of
Hexoses in DMSO/Water
Authors: Esben Taarning, Irantzu Sádaba, Pernille Rose Jensen, and
Sebastian Meier
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: ChemSusChem 10.1002/cssc.201902322
Link to VoR: http://dx.doi.org/10.1002/cssc.201902322
A Journal of
www.chemsuschem.org

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
Discovery and Exploration of the Efficient Acyclic Dehydration of
Hexoses in DMSO/Water
Esben Taarning,[b] Irantzu Sádaba,[b] Pernille Rose Jensen,[c] and Sebastian Meier*[a]
Abstract: Current gaps in the development of sustainable processes
include lacking strategies to systematically identify and optimize the
formation of new products. The dehydration of hexoses to 5-
hydroxymethylfurfural (HMF) is a particularly widely studied process.
In an attempt to identify a new high-selectivity conversion of glucose,
quantitative NMR spectroscopy was used to screen conditions that
were reported to yield high conversion of glucose, but low formation
used to produce post-functionalized polyesters and upgraded
monomers.
The formation of HMF has been studied extensively for
several decades, but the mechanism by which it forms has
remained a matter of debate.[1b, 5] HMF may either form from five-
membered fructofuranose through a route encompassing cyclic
intermediates[6] or through
a route involving open-chain
of HMF. In this manner, an olefinic six-carbon byproduct was identified. intermediates.[7] Acyclic routes from glucose or fructose have
By water addition, selectivity for the compound was nearly tripled
relative to previous reports. The detection of high-yielding side
reactions in the formation of HMF is remarkable, considering how
extensively HMF formation has been studied. High selectivity for the
acyclic pathway allows hitherto unobserved intermediates in this
been experimentally supported by the detection of new sets of
byproducts[4b] that are best rationalized through hypothetical
acyclic intermediates including 3,4-di-deoxyglucosone-3-ene
(3,4-DGE).[4b, 6] Low populations of open-chain carbohydrates and
their poorly characterized properties, in particular their reactivity
pathway to be identified using in situ NMR spectroscopy. An additional, and catalyst binding, have complicated the mechanistic
presumably cyclic, pathway contributes to HMF formation.
understanding of carbohydrate chemistry under reaction
conditions.[8] The best suited experimental approach to such
understanding arguably involves NMR spectroscopy, as isomeric
forms of carbohydrates can be distinguished in solution by NMR
spectroscopy. NMR approaches have also proven useful to
determine chemical structures in unpurified reaction mixtures,[9]
which is particularly beneficial for detecting transient
intermediates that are difficult or impossible to purify.[10]
Here, we apply quantitative NMR spectroscopy on reaction
mixtures that use cheap and abundant chemicals under
conditions where reported yields of HMF were low at high glucose
conversion. In this manner, reactions producing unexpected
major byproducts are identified, which are systematically
optimized to result in a non-commercial bio-based chemical at
record high selectivity (Scheme 1). Using these rationales, a
reaction yielding almost 50 % of the non-commercial olefinic C6-
acid THA directly from glucose was developed (Figure 1).
Biomass production from carbon dioxide exceeds 100·1012 kg per
year[1] with an average content of carbohydrates near 75% (in
addition to 20% lignin and 5% other compound classes).[2]
Contrary to other renewable biomass, carbohydrates are
obtainable as cheap bulk chemicals. As the primary product of
glycolysis and as the monomeric constituent of starch and
cellulose, glucose is the most abundant carbohydrate building
block. Hence, benign, efficient and tunable means to obtain fuels
and plastic building blocks particularly from glucose are desirable.
Processes for the conversion of carbohydrates to chemicals
often include stoichiometric dehydration, in order to reduce the
oxygen content in the product molecule. Acid catalyzed
dehydration of hexoses has attracted attention in the production
of the cyclic product HMF[1b, 3] and of linear acids.[4] Both cyclic
and linear compounds are investigated as possible bio-platform
molecules, from which new groups of useful chemicals can be
derived. Thus, HMF has been converted to various furan
derivatives, while unsaturated alpha-hydroxy acids have been
[a]
Dr. S. Meier
Department of Chemistry
Technical University of Denmark
Kemitorvet Building 207, 2800-Kgs. Lyngby, Denmark
E-mail: semei@kemi.dtu.dk
Scheme 1. Flowchart for the identification of a process giving high selectivity
for a new glucose derived product from serendipitous and systematic steps.
[b]
[c]
Dr. E. Taarning, Dr. I. Sádaba
Haldor Topsøe A/S,
Haldor Topsøes Allé 1, 2800-Kgs. Lyngby, Denmark
Dr. P.R. Jensen
Various metal salts reported to have low selectivity for HMF
formation from carbohydrates were found to catalyze instead the
formation of unforeseen byproducts. The main byproduct in the
reaction mixture was identified as a trans-olefinic compound
(Figure S1). This trans-olefinic compound was preferentially
observed using SnCl4 in DMSO, with a selectivity for this
Department of Health Technology
Technical University of Denmark
Elektrovej 349 2800-Kgs. Lyngby, Denmark
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.

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
Figure 2. (A) Quantitative 13C NMR spectra showing the substrate signal and
the product mixture run under identical conditions. Integration is relative to the
DMSO signal as the internal standard (red numbers). (B) Schematic overview
of the product distribution (%carbon) in the presence of water (10% vol/vol).
Figure 1. (A) Chemical structures of HMF and THA. (B) Product composition in
a reaction of glucose (1 M) and SnCl4 (0.1 M) in DMSO/water binary mixtures
with the indicated water content (%vol/vol at 100 C, 20 hours). The 1H NMR
spectral region showing narrow HMF signals and the multiplet of the THA olefins
is shown. (C) Selectivity for THA and HMF and conversion of glucose
determined by integrals of spectra shown in (B) and validated by quantitative
13C NMR.
1
Quantitative H and 13C NMR spectra were used to quantify the
residual reaction products, which are mostly comprised of
polymeric humins, 3-deoxy compound byproducts (Figure 2,
Figure S3) and lactate. Approximately 23% of carbon balance are
not accounted for by the main soluble products and are ascribed
to humins and minor byproducts. Water concentrations above
10% (w/v) were found to increase humin formation. The formation
of furanics, THA, 3- deoxy compounds and retro-aldol products
mirrors the pathways catalyzed by Sn-containing zeolites near
160°C.[4b] The tuning of product distribution by water resembles
recent discoveries of water[12] and alkali ion[13] effects at the active
site of heterogeneous Sn(IV) catalysts under such industrially
relevant conditions.
byproduct near 20%. In situ NMR identified this olefinic compound
as trans-THA (Figure S2). No cis-THA was observed, suggesting
that a cis-intermediate preferably reacts to HMF, while a
corresponding trans-intermediate reacts to THA (Figure 1A).
Around 38% selectivity for HMF was observed at 100 °C,
consistent with previous reports on formation of HMF in
SnCl4/DMSO.[3d, 3f]
Compared to water as the solvent, polar aprotic solvents like
DMSO are known to strongly affect activation energies in aldose
conversion to furanics.[11] Hence, we hypothesized that the
addition of water to the SnCl4/DMSO system may stabilize acidic
protons relative to the transition state and decelerate possible
Brønsted acid catalyzed pathways to HMF. In addition, a water
effect on product selectivity was anticipated, as the formation of
both HMF and THA are stoichiometric dehydration reactions.
Water addition was therefore used to tune the product selectivity
in the stoichiometric dehydration reaction, stipulated by the fact
that previous studies on various reaction systems had shown
significant water effects on product distribution. HMF is principally
considered stable in water/DMSO mixtures with very limited
rehydration and conversion to levulinic acid and formic acid.
Accordingly, no increase in formic acid or levulinic acid formation
was observed when adding water (5-20% v/v) to the reaction
mixture (Figure 1B). In contrast, the ratio of THA to HMF changed
up to fourfold, giving NMR yields of more than 44% THA at a water
content of 10% (w/v) (Figure 1 B, C). This yield corresponds to a
threefold increase in THA yield relative to the previous reports
using Sn-Beta catalyzed conversions of glucose, sucrose or
fructose in water or methanol.[4a, 4b]
High selectivity for acyclic products of glucose conversion
forms a promising basis for new mechanistic insight into the
relevant pathways. Low reaction temperatures and the use of
homogeneous catalyst made SnCl4/DMSO/water systems
amenable to mechanistic studies by in situ NMR spectroscopy.
The reactivity of glucose and fructose was compared under
identical reaction conditions by using both substrates in an
equimolar mixture (Figure 3) The reaction was followed by 1H-13C
HSQC NMR spectroscopy to yield high-quality baseline-resolved
signals. Resultant reaction time courses are displayed in Figure
3, showing that fructose reacted 30-fold faster than glucose. The
increased reactivity of fructose is a possible consequence of the
higher open-chain fraction of fructose as compared to glucose.
Sn(IV) has previously been suggested to exert its catalytic activity
by coordinating open chain glucose, fructose or enol species as
five-membered chelate structures.[3d] Consistent with this
functionality, free open-chain fructose was not observed in the
reaction mixture, possibly due to its chelation and reactivity
(Figure S4). The rapid conversion of fructose to THA by
SnCl4/DMSO/water permits full conversion within 24 hours at
temperatures as low as 40 C. The resultant THA product is a
diastereomeric compound with chiral centers at C2 and C5
(Figure S5).
This article is protected by copyright. All rights reserved.

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
Figure 4. Kinetic profile for the conversion of all aldohexoses at 100 C (A) and
Figure 3. (A,B) Projections of the 2D 1H-13C HSQC spectra onto the 13C
dimension for a reaction comparing fructose and glucose conversion under
identical reaction conditions. 1H-13C HSQC was used to to obtain 13C spectra
with increased sensitivity while tracking the reaction in situ (0.5 M glucose, 0.5
M fructose, 0.1 M SnCl4, 90 C). Spectra of 10 minutes duration were recorded
and the first ten spectra are shown. (B) Reaction progress showing higher
reactivity of fructose as compared to glucose.
correlation of reaction rates to previously determined acyclic fractions (drawn
on a double logarithmic scale in B). Conversion was tracked in situ by 1H-13
C
HSQC NMR spectroscopy on a reaction mixture of all aldohexoses (10 mg each,
0.1 M SnCl4 in 600 l d6-DMSO) to warrant identical reaction conditions for all
substrates.
Higher reactivity of fructose than of glucose is inherent to
both open-chain and cyclic reaction pathways, as the open and
the furanose form are higher populated in fructose than in glucose.
In situ 1H-13C HSQC NMR spectroscopy at optimized reaction
conditions was thus employed for a competitive in situ assay
between isomeric species to probe the conversion of all linear
aldohexoses in a reaction mixture (Figure S7). Figure 4 shows the
resultant reaction profiles with a significantly different kinetics of
cyclic hexose conversion, where glucose reacted slowest and
idose fastest. This trend parallels the highest expected stability of
pyranose forms for glucose and lowest stability of pyranose forms
for idose as a consequence of the spatial distribution of non-
hydrogen substituents at the 6-membered ring (all-equatorial for
glucose and all-axial for idose).[14]
A comparison of the initial rate of carbohydrate conversion
and of open chain populations previously determined in water
indicated a correlation between open chain population and
reaction rate (Figure 4B). Metal chlorides have previously been
suggested to catalyze the anomerization of glucose and fructose
in acyclic form.[3b] The observed correlation between the rate of
conversion and fraction of acyclic aldohexose is consistent with a
rapid pre-equilibrium between cyclic and open aldohexose forms,
prior to a rate-limiting conversion of the open substrate to the
reactive complex between acyclic form and metal. If the rate-
limiting formation of reactive complex from open substrate is
designated with rate constant 푘푐표푚푝푙푒푥, then the rate of complex
Figure 5. Hyperpolarized 13C NMR experiment of initial steps of [2-13C]fructose
conversion at 70 C in DMSO by SnCl4 (0.1 M ; top). A time series of 1D 13C
spectra is displayed, where the overall signal fades due to the loss of transient
nuclear spin hyperpolarization. Time resolved 13C spectra show the formation
of an enol species as the initial detectable intermediate.
influencing the rate of conversion (here through 푘푐표푚푝푙푒푥)
[15] may
thus account for kinetic differences displayed in Figure 4.
Combined, these studies indicate the value of employing
competitive in situ experiments for determining plausible routes of
hexose conversion.
In order to obtain further mechanistic insight into the routes
of hexose conversion, in situ NMR was conducted to track
reactions and address the following: (1) what are the earliest
intermediates detectable upon carbohydrate-catalyst interaction,
(2) is it possible to rationalize different selectivity to cis and trans
olefinic bonds upon dehydration and (3) is HMF formed through
the linear pathway? The detection of early intermediates requires
[
]
[
]
formation 푑 푐표푚푝푙푒푥 /푑푡 = 푘푐표푚푝푙푒푥 ∙ 표푝푒푛 depends on the
equilibrium constant 퐾 between open and dominant cyclic
[
]
[
]
species as 푑 푐표푚푝푙푒푥 /푑푡 = 푘푐표푚푝푙푒푥 ∙ 퐾 ∙ 푐푦푐푙푖푐 .
Faster
conversion of substrates with higher acyclic population (higher 퐾)
and possible structural details in hydroxyl group orientation
This article is protected by copyright. All rights reserved.

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
compared to 110.5 kJ/mol, Figure S9), predicting rates of
formation that differ by eEa/RT≈3 as observed.
Kinetic data for the conversion of fructose to trans-3,4 DGE
and THA fitted well with a mechanism of sequential irreversible
elementary reactions on one pathway (Figure S10). In contrast,
reaction profiles did not fit well with a sole conversion of cis-3,4
DGE to HMF, indicating that HMF in part forms through a cyclic
mechanism from fructose, not unexpected in the presence of
accumulating Brønsted acidic THA (Figure S10). A kinetic model
describing the formation of THA via trans-3,4 DGE and of HMF
from cis-3,4 DGE or from a cyclic pathway is shown in Figure 7.
In situ experiments were conducted in the presence and in the
absence of added water (Figure S11) to probe the effect of water
on pathway steps. Resultant fits of the model to kinetic data in the
presence and in the absence of added water are shown in Figure
7. The apparent rate constant k1 of the kinetic model integrates
contributions from open chain substrate concentration, catalyst
complexation and dehydration, while rate the apparent rate
constant k2 includes a contribution from water concentrations. As
expected, k2 is thus increased in the presence of water. In contrast,
cyclic, Brønsted acid catalyzed dehydrations to HMF are
strongest reduced in rate (k4 and k5) in the presence of water. This
finding is consistent with the increased selectivity for THA and
with the hypothesized stabilization of acidic protons and indirect
deceleration of possible Brønsted acid catalyzed pathways to
HMF in the presence of water.[11]
Scheme 2. Plausible reaction scheme consistent with in situ kinetic data
described herein for the conversion of glucose to THA (trans-2,5,6-trihydroxy-
3-hexenoic acid) and HMF (5-hydroxymethylfurfural) in DMSO/SnCl4. Directly
observed intermediates and products are highlighted in green. Partial hydrolysis
of SnCl4 to active species has previously been reported[16] and catalytic [Sn]
Lewis acid sites are shown in a generalized manner.
a short experimental dead time and sufficiently rapid, sensitive
and information-rich experiments. Hyperpolarized NMR is a
means to temporarily creating non-equilibrium nuclear spin
magnetization that provides ~10,000-fold enhanced signal on the
sub-minute timescale. Hyperpolarized 13C NMR was employed
due to the rich structural information of the 13C NMR signal, due
to the high reactivity of fructose and due to its commercial
availability as [2-13C]-fructose. The experiment shows the initial
formation of an enol at C2 in 2,4,5,6-tetrahydroxyhex-2-enal
(Figure 5), with the 13C2 signal split in a characteristic manner by
the adjacent aldehyde proton at C1. This species is converted to
trans-3,4 DGE (Scheme 2, Figure 6), which accumulates to levels
that make it conveniently detectable in conventional NMR
reaction progress kinetic analysis. Assignment spectra of the
intermediates identified by conventional NMR are shown in
Figures S2 and S7.
Spectroscopic observation of the acyclic pathway makes it
evident that the formations of trans-3,4 DGE is largely favoured
over the formation of cis-3,4 DGE (Figure 6). Due to the high
reactivity of the fructose in SnCl4/DMSO/water, kinetic
experiments were conducted in the range between 40 C and 100
C to establish the activation energy for trans-3,4 DGE and cis-
3,4 DGE formation from fructose (Figure S8). Trans-3,4 DGE
forms approximately threefold faster than cis-3,4 DGE (Figure 6),
with the activation energy of trans-3,4 DGE formation being ~3
kJ/mol lower than that of cis-3,4 DGE formation (107.5 kJ/mol as
Figure 7. Kinetic analysis (model displayed on top) of in situ NMR spectra for
the conversion of 1 M fructose (0.1 M SnCl4, 323 K) in 600 l d6-DMSO (middle)
and in 600 l d6-DMSO/D2O (85/15 v/v, bottom).
Figure 6. Fructose conversion at 40 C as monitored by in situ 1H NMR (fructose
(1 M), SnCl4 (0.1 M), DMSO/H2O (9:1 (v/v))).
In conclusion, we find that an approach combining
serendipitous screening by quantitative NMR and systematic
improvement by modulation of solvent properties affords high
This article is protected by copyright. All rights reserved.

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
Hyperpolarized 13C NMR experiments were conducted by forcefully
injecting 600 l of the aqueous fructose solution into a solution of SnCl4 in
DMSO (100 mg in 4 ml) equilibrated to 70 ºC in an Oxford 400 MHz magnet.
Acquisition of 13C spectra was started prior to the injection. Detailed
experimental procedures are described in the Supporting Information.
selectivities for the non-commercial C6-compound THA and
sheds light on pathways that compete with HMF production.
Mechanistic studies are consistent with a chelation of acyclic
species by the Sn(IV) catalyst, followed by dehydration at C3 to
form a 2-enol-1-aldehyde species as the first intermediate. Kinetic
data were consistent with a sequential reaction of trans-3,4 DGE
to THA and of cis-3,4 DGE to HMF, with another (presumably
cyclic) pathway contributing to HMF formation from fructose.
Acyclic pathways to HMF through 3-deoxyglucosone and cis-3,4
DGE have previously been suggested based on computational
studies[3e, 6b, 17] and on the experimental observation of cis-3,4
DGE in carbohydrate- and HMF-containing solutions.[18] DMSO is
an attractive solvent for the reaction due to its ability to solubilize
carbohydrates and to reduce humin formation compared to
aqueous solvents. Further improvements using biphasic
systems[3a] in order to minimize humin formation and to facilitate
purification, or use of buffer in order to avoid mechanistic changes
due to the formation of free Brønsted acid could be envisioned.
The suitability of heterogeneous Sn(IV) catalysts in aprotic
solvents for the catalysis of acyclic carbohydrate conversion
pathways is under investigation. We finally note that the pathways
described herein have biochemical counterparts in the formation
of highly reactive glucose degradation products linked to
cytotoxicity and diabetic complications.[18b, 19] Mechanistic studies
of these biochemical processes in water may be facilitated by the
current pathway study.
Acknowledgements
This work was supported by Innovation Fund Denmark (case
number 5150-00023B). All 800 MHz NMR spectra were recorded
on the spectrometer of the NMR center DTU supported by the
Villum Foundation.
Keywords: carbohydrate • Lewis acid • qNMR• reaction kinetics
• reaction mechanism
[1]
a) C. B. Field, M. J. Behrenfeld, J. T. Randerson, P. Falkowski, Science
1998, 281, 237-240; b) R.-J. van Putten, J. C. van der Waal, E. de Jong,
C. B. Rasrendra, H. J. Heeres, J. G. de Vries, Chemical Reviews 2013,
113, 1499-1597.
[2]
[3]
a) B. Kamm, M. Kamm, CABEQ 2004, 18, 1-6; b) K. J. Ptasinski, in
Efficiency of Biomass Energy, John Wiley & Sons, Inc., 2015, pp. 673-
706.
a) Y. Roman-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312,
1933-1937; b) H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science
2007, 316, 1597-1600; c) A. S. Amarasekara, L. D. Williams, C. C. Ebede,
Carbohydr. Res. 2008, 343, 3021-3024; d) S. Hu, Z. Zhang, J. Song, Y.
Zhou, B. Han, Green Chem. 2009, 11, 1746-1749; e) H. Jadhav, C. M.
Pedersen, T. Solling, M. Bols, ChemSusChem 2011, 4, 1049-1051; f) C.
B. Rasrendra, B. A. Fachri, I. G. Makertihartha, S. Adisasmito, H. J.
Heeres, ChemSusChem 2011, 4, 768-777; g) G. R. Akien, L. Qi, I. T.
Horvath, Chem. Comm. 2012, 48, 5850-5852; h) G. Tsilomelekis, T. R.
Josephson, V. Nikolakis, S. Caratzoulas, ChemSusChem 2014, 7, 117-
126.
Experimental Section
Reaction mixtures containing carbohydrate in d6-DMSO with catalyst (10-
mol%) and defined water fraction were incubated under shaking (600 rpm
at 100 ºC for 20 hours in an Eppendorf Thermomixer). Samples were
transferred to 5 mm NMR sample tubes after the reaction and immediately
analysed at 30 ºC by 1H and 13C NMR spectroscopy. Samples yielded
some discolouring due to humin formation, but remained transparent
(albeit slightly coloured) for the experiments with best THA selectvities.
Selectivities were estimated by comparing 13C NMR signal (inverse gated
13C 1D experiment using a recycling delay of 10 seconds, comparing
primary alcohol signals of glucose, fructose and THM C6) integrals of a
substrate solution with the product mixture (both normalized to the d6-
DMSO signal).
[4]
a) H.-S. Chen, A. Wang, H. Sorek, J. D. Lewis, Y. Román-Leshkov, A. T.
Bell, ChemistrySelect 2016, 1, 4167-4172; b) S. Tolborg, S. Meier, I.
Sádaba, S. G. Elliot, S. K. Kristensen, S. Saravanamurugan, A. Riisager,
P. Fristrup, T. Skrydstrup, E. Taarning, Green Chem. 2016, 18, 3360-
3369; c) S. G. Elliot, C. Andersen, S. Tolborg, S. Meier, I. Sádaba, A. E.
Daugaard, E. Taarning, RSC Adv. 2017, 7, 985-996.
[5]
[6]
a) M. J. Antal Jr, T. Leesomboon, W. S. Mok, G. N. Richards, Carbohydr.
Res. 1991, 217, 71-85. b) X. Qian, M. R. Nimlos, M. Davis, D. K. Johnson
and M. E. Himmel, Carbohydr. Res., 2005, 340, 2319–2327.
a) S. H. Mushrif, S. Caratzoulas, D. G. Vlachos, Phys. Chem. Chem.
Phys. 2012, 14, 2637-2644; b) S. Kunnikuruvan, N. N. Nair, ACS Catal.
2019, 9, 7250-7263.
In situ experiments were performed by transferring reaction mixtures from
1.5 ml Eppendorf safelock tubes directly to NMR tubes and heating
samples to the desired temperature in the spectrometer. Reaction
progress was then followed by pseudo-2D spectra recording a series of 1H
or 13C NMR spectra in real time. Comparison of glucose and fructose and
comparison of reactivity of the linear aldohexoses was conducted using
[7]
[8]
T. Wang, M. W. Nolte, B. H. Shanks, Green Chem. 2014, 16, 548-572.
a) S. Meier, M. Karlsson, P. R. Jensen, ACS Sustainable Chem. Eng.
2017, 5, 5571-5577; b) W. Plazinski, A. Plazinska, M. Drach, Phys. Chem.
Chem. Phys. 2016, 18, 9626-9635.
1H-13
C HSQC spectra and initial rates were approximated by an
[9]
a) M. Bojstrup, B. O. Petersen, S. R. Beeren, O. Hindsgaul, S. Meier,
Anal. Chem. 2013, 85, 8802-8808; b) S. Meier, Anal. Bioanal. Chem.
2014, 406, 7763-7772; c) B. O. Petersen, O. Hindsgaul, S. Meier, Analyst
2014, 139, 401-406.
exponential fit. NMR spectra were recorded, processed and analysed with
Bruker Topspin software. Kinetic data of reaction progress of fructose
conversion at 40, 70 or 100 ºC were integrated in Bruker Topspin 3.0 and
fitted to a model of sequential elementary reactions in proFit 7 (Quantum
Soft, Switzerland).
[10] R. R. Forseth, F. C. Schroeder, Curr. Opin. Chem. Biol. 2011, 15, 38-47.
[11] M. A. Mellmer, C. Sener, J. M. R. Gallo, J. S. Luterbacher, D. M. Alonso,
J. A. Dumesic, Angew. Chem. Int. Ed. 2014, 53, 11872-11875.
[12] a) D. Padovan, L. Botti, C. Hammond, ACS Catalysis 2018, 8, 7131-
7140; b) D. Padovan, S. Tolborg, L. Botti, E. Taarning, I. Sádaba, C.
Hammond, React. Chem. Eng. 2018, 3, 155-163; c) I. Tosi, A. Riisager,
E. Taarning, P. R. Jensen, S. Meier, Catal. Sci. Technol. 2018, 8, 2137-
2145.
Isotope-enriched [2-13C]-fructose (Cambridge Isotope Laboratories,
Tewksbury, MA) was hyperpolarized using
a HyperSense (Oxford
Instruments, England) polarizer with a magnetic field of 3.35 T. Solid-state
polarizations of ~30% were obtained in the self-glassing carbohydrate
syrup. After 1 hour of polarization, the samples were dissolved with heated
Milli-Q water to yield liquid hyperpolarized [2-13C]-fructose samples.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
[13] S. Tolborg, I. Sádaba, C. M. Osmundsen, P. Fristrup, M. S. Holm, E.
Taarning, ChemSusChem 2015, 8, 613-617.
[18] a) M. Hellwig, A. Nobis, S. Witte, T. Henle, J. Agric. Food Chem. 2016,
64, 2746-2753; b) T. Linden, A. Cohen, R. Deppisch, P. Kjellstrand, A.
Wieslander, Kidney Int. 2002, 62, 697-703.
[14] Y. Zhu, J. Zajicek, A. S. Serianni, J. Org. Chem. 2001, 66, 6244-6251.
[15] R.-J. van Putten, J. C. van der Waal, M. Harmse, H. H. van de
Bovenkamp, E. de Jong, H. J. Heeres, ChemSusChem 2016, 9, 1827-
1834.
[19] P. J. Thornalley, A. Langborg, H. S. Minhas, Biochemical J. 1999, 344,
109-116.
[16] K. R. Enslow, A. T. Bell, Catalysis Science & Technology 2015, 5, 2839-
2847.
[17] H. Rasmussen, H. R. Sørensen, A. S. Meyer, Carbohydrate Research
2014, 385, 45-57.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201902322
ChemSusChem
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
Serendipitous and systematic
approaches are combined to
triple the yield of a non-
Esben Taarning, Irantzu Sádaba,
Pernille Rose Jensen, and Sebastian
Meier*
commercial compound directly
from glucose and to make the
acyclic pathway of carbohydrate
Page No. – Page No.
O
conversion accessible to
O
HO
Discovery and Exploration of the
Efficient Acyclic Dehydration of
Hexoses in DMSO/Water
observation.
OH
O
HO
OH
OH
Layout 2:
COMMUNICATION
Author(s), Corresponding Author(s)*
((Insert TOC Graphic here))
Page No. – Page No.
Title
Text for Table of Contents
This article is protected by copyright. All rights reserved.