Mechanism and stereoselectivity of zeolite-catalysed sugar isomerisation in alcohols

Article abstract of DOI:10.1039/c6cc05592c

Glucose isomerisation to fructose can occur by different pathways and the mechanism of zeolite-catalysed glucose isomerisation in methanol has remained incompletely understood. Herein, the mechanism is studied using an ¹H-¹³C HSQC NMR assay resolving different fructose isotopomers. We find that zeolite-catalysed glucose isomerisation proceeds predominantly via a hydride shift into the pro-R position of fructose, thus resembling the stereoselectivity of the enzymatic isomerisation process.

Full text of DOI:10.1039/c6cc05592c

ChemComm
View Article Online
View Journal
COMMUNICATION
Mechanism and stereoselectivity of zeolite-
catalysed sugar isomerisation in alcohols†
Cite this:DOI: 10.1039/c6cc05592c
ab
a
c
Shunmugavel Saravanamurugan, Anders Riisager,* Esben Taarning and
Sebastian Meier*
a
Received 6th July 2016,
Accepted 7th October 2016
DOI: 10.1039/c6cc05592c
www.rsc.org/chemcomm
Glucose isomerisation to fructose can occur by different pathways The interconversion of glucose and fructose especially by
and the mechanism of zeolite-catalysed glucose isomerisation Sn-Beta in water has been studied both experimentally and
5
a,f,8
in methanol has remained incompletely understood. Herein, the computationally.
In contrast, weakly Lewis acidic
1
13
mechanism is studied using an H– C HSQC NMR assay resolving Al-zeolites, such as H-USY (6), isomerise glucose into fructose
different fructose isotopomers. We find that zeolite-catalysed in alcohol but not in water. This differential reactivity of tin and
glucose isomerisation proceeds predominantly via a hydride shift aluminium active sites lead us to experimentally study their
into the pro-R position of fructose, thus resembling the stereo- reaction mechanism in alcohols to extend earlier studies on the
selectivity of the enzymatic isomerisation process.
reaction of tin-containing zeolites in water. Specifically, we
explore the role of Lewis acidic sites in Al-zeolites H-USY (6)
Glucose isomerisation to fructose constitutes the largest indus- and H-USY (30) and compare their reactivity to Sn-Beta and to
1
trial biocatalytic process on earth with a current production the homogeneous Lewis acidic catalyst AlCl3.
7
2
volume on the order of 10 tons per year. Glucose is the most
The glucose isotopomer deuterated at the C2 position has been
3
abundant and cheapest hexose, while fructose is a widely used employed to demonstrate intramolecular hydride shift reactions
sweetener and a prospective substrate for the processing of by tracking the fate of the deuteron by NMR spectroscopy.
biomass to chemicals and fuels. Solid acid catalysts containing Experimental evidence for the 1,2-hydride shift mechanism in
Lewis acid sites emerge as alternatives to immobilised enzyme water has relied upon the changing spectral appearance of H and
formulations of xylose isomerase (E.C. 5.3.1.5) for aldose-to-
ketose isomerisation, including glucose to fructose, at relatively when using [2- H] glucose as the substrate. Here, a H– C HSQC
mild conditions. Hence, microporous crystalline materials NMR assay is employed to distinguish different carbohydrate
5a,9
4
1
13
C NMR signals in fructose due to deuteration of the C1 position
2
1
13
5
such as zeolites have proven to catalyse the glucose-to-fructose isotopmers in situ, specifically deuterated C1-positions in fructose
2
isomerisation in water under equilibrium-limited conditions and fructosides formed from [2- H] glucose with the Lewis acidic
(42% fructose, 8% other sugars and 50% glucose) or in alcohols zeolites in methanol. The two-dimensional NMR assay provides
using processes that can yield higher conversion to fructose due sufficient sensitivity to accurately assess the stereoselectivity of
to the rapid sequestering of fructose as methyl fructosides the detected 1,2-hydride-shift operating with aluminium-based
5c,6
2
(Scheme 1).
However, the structural integrity of zeolites in zeolites in alcohol by detecting a major shift of the glucose 2- H
water is not promising and zeolites are prone to disintegrate. In into the pro-R position of fructose. Similar stereoselectivity is also
contrast, the stability of the zeolites is much better retained found when using tin-based zeolites or homogeneous AlCl -based
3
when using organic solvents.
Stannosilicates, such as Sn-Beta, with strong Lewis acid sites the enzyme-catalysed reaction mechanism.
can catalyse an intramolecular hydride shift resulting in aldose-
catalysis. The preferential transfer of the pro-R proton resembles
9
1
13
Using H– C HSQC NMR spectroscopy, we characterise the
C1 positions in fructose and fructosides formed upon isomer-
4
c,5b,c,7
to-ketose isomerisation in both water and alcohol.
2
isation of [2- H] glucose by H-USY (6) and Sn-Beta in protonated
methanol. Experiments are performed in the kinetic regime
where the major glucoside species formed are glucofuranosides
and where fructose species accumulate to above 60% due to
their rapid sequestration as fructosides. All hexoses and their
corresponding glycosides can be spectroscopically resolved by
a
Technical University of Denmark, Department of Chemistry, Kemitorvet,
2
800-Kgs. Lyngby, Denmark. E-mail: ar@kemi.dtu.dk, semei@kemi.dtu.dk
b
Center of Innovative and Applied Bioprocessing (CIAB),
Mohali 160 071, Punjab, India
c
Haldor Topsøe A/S, Haldor Topsøes All ´e 1, 2800-Kgs. Lyngby, Denmark
1
13
10
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/c6cc05592c
H– C HSQC NMR without the need for sample fractionation.
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
Scheme 1 Overview of possible glucose and fructose species formed in the zeolite catalysed conversion of glucose to fructose in methanol solution.
The pro-R position of fructose is highlighted in red.
1
3
In addition, a primary isotope effect on the C chemical shift of
1
1
À0.35 ppm and a secondary isotope effect on the H
1
chemical
shift on the order of À0.02 ppm are observed relative to protonated
1
13
fructose using H– C HSQC NMR. Such a shielding effect and
concurrent signal shift to smaller frequencies is characteristic
for the introduction of heavier isotopes, such as deuterons
instead of protons. These isotope effects are sufficient to resolve
C -deuterated fructose isotope signals spectroscopically from
1
1
2
1
protonated signals (C H H and C H groups, Fig. 1).
2
Beyond the characteristic change in chemical shift of
1
13
C
1
-deuterated fructose isotopomers in H– C HSQC signals,
deuteration is evident when using edited HSQC spectra to distin-
1
2
1
guish C H H and C H
2
groups by the sign of the NMR signal (see
group thus exhibits the
Fig. 1). As expected, the deuterated C
1
inverse sign of the protonated version in edited HSQC spectra. In
addition, the deuterated C H signal exhibits a characteristic
1
1
1
13
splitting due to J couplings during C chemical shift evolution
CD
2
1
and due to JHD couplings during H detection (see Fig. 1a). These
couplings then result in a characteristic E. COSY type triplet signal
with inverse sign and characteristic chemical shift change relative
to a protonated species as shown in Fig. 1.
Fig. 1 Comparison of fructofuranoside obtained by gently heating com-
2
mercial [1-pro-S- H] fructose in methanol in the presence of 0.1% sulphuric
Only one out of the two possible stereoisomers emerge upon
,2-hydride shift in the isomerisation of [2- H] glucose by H-
2
acid (a) and fructofuranoside obtained by isomerisation of [2- H] glucose in
2
1
2
methanol using H-USY (6). The latter yields [1-pro-R- H] fructofuranoside
2
USY (6). Using a commercial fructose isotopomer deuterated at due to the transfer of the glucose deuteron H-2 into the 1-pro-R position
of fructose (b). Different NMR signal colors represent positive and negative
the pro-S position as a reference (Fig. 1a), aldose-to-ketose
1
2
1
2
signals, respectively, distinguishing C H H and C H groups.
isomerisation using aluminium containing H-USY (6) zeolite
in methanol was found to selectively catalyse a hydride shift into
2
the pro-R position of fructose, yielding [1-pro-R- H] fructoside
In order to substantiate the activity of partially hydrolysed
in
was employed for comparison, yielding the same
zeolite get partially hydrolysed to form sites with an adjacent stereoselectivity as H-USY (6). The same stereoselectivity was also
2
isotopomer from [2- H] glucose (Fig. 1b and Scheme 2). It was framework sites, the homogeneous Lewis acid catalyst AlCl
3
previously suggested that framework aluminium sites in the MeOH-d
4
1
1
silanol group as indicated in Scheme 2. These aluminium sites found when using an aluminium-based zeolite with an altered
can accommodate hexa-coordinated cyclic intermediates with Si/Al ratio (H-USY (30)) and Sn-Beta in methanol (Fig. 2). Notably,
the C-2-hydroxy group and the aldehyde prior to the 1,2-hydride the stereoselectivity for the zeolite-catalysed hydride transfer into
shift. Such a mechanism would be the intramolecular equivalent the pro-R position coincides with the stereoselectivity of the xylose
to the stereoselective Meerwein–Ponndorf–Verley and Oppenauer isomerase catalysed reaction, which is thought to proceed as a
11
12
(MPVO) reaction.
metal-catalysed intramolecular hydride transfer.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
ChemComm
Communication
hydride shifts from protonated and from deuterated C-2 posi-
tions of glucose in the same sample, using equal amounts of
2
protonated and [2- H] glucose as the substrates under identical
process conditions. Hence, the kinetic isotope effect k /kD for
H
the shift of hydride and deuteride could be obtained from a
single reaction mixture (see Scheme S1, ESI†). In addition, the
glucose to fructose isomerisation reaction in alcohol can be
approximated as irreversible, witnessed by the lack of mannose
and mannoside formation from glucose due to the rapid
formation and sequestration of fructosides upon isomerisation
in alcohol. Hence, the reverse hydride shift from fructose to
glucose can be neglected in the kinetic analysis of forming the
two different fructose isotopomers from glucose.
2
Scheme 2 Overview of the hydride shift of H-2 into the 1-pro-R position
by aluminium-containing zeolites.
H D
The kinetic isotope effect k /k for the isomerisation of
glucose in methanol is found to be 2.2 Æ 0.2 using H-USY (6)
and similarly 2.0 Æ 0.2 using Sn-Beta (see Fig. S4, ESI†).
Previous reports for the Lewis acid site catalysis of glucose
isomerisation to fructose have yielded kinetic isotope effects
5
f
13
k /k of 1.7 Æ 0.1 to 2.7 Æ 0.3, depending on the catalytic
H
D
system. The obvious and significant kinetic isotope effect (Fig. 3)
shows that the stereoselective hydride shift constitutes the rate-
limiting step for the glucose isomerisation using aluminium-
containing zeolites in methanol. The H– C HSQC-based assay
employed to follow the Lewis acid catalysed hydride shift in glucose
provides approximately 30-fold higher sensitivity as compared to
1
13
13
conventional C NMR. In addition, the use of two-dimensional
NMR provides significant gains in resolution as compared to one-
dimensional spectra. Due to these analytical benefits, it is possible to
estimate the fractions of hydrogen transfer to the pro-R and to the
pro-S positions using glucose molecule labelled only by deuteration
at position 2. Thus, transfer of the deuteron was detected to minor
(0.5%) extent into the pro-S position when using Lewis acid catalysts
in deuterated methanol (Fig. 4).
Using the fraction of pro-R and to the pro-S forms and
assuming kinetic control of this selectivity, the transfer into
À1
the pro-S position encounters a 16.3 Æ 0.4 kJ mol higher free
energy of activation for the transfer in methanol catalysed by
H-USY (6). The glucose substrate was deuterated at position 2 to
2
Fig. 2 Hydride shift of the [2- H] glucose deuteron into the 1-pro-R
position by aluminium-containing zeolites and by Sn-Beta as well as
dissolved AlCl
sides all contain the H
the deuteron into the 1-pro-R position of fructose. Fructose and fructoside
3
(a and b). As for H-USY (6), resultant fructose and fructo-
1
1
pro-S hydrogen, due to stereoselective transfer of
1
13
signals containing deuterated 1-pro-R positions yield 1-pro-S H–
C
1
13
signals with chemical shift change and splitting in the H and C dimen-
sions when compared to the protonated reference signals (which are split
2
into doublets through
JHH couplings). Fructose rather than fructoside
signals are shown for Sn-Beta (b) due to the lower Brønsted acidity and
slower glycoside formation with Sn-Beta.
Fig. 3 Kinetic profiles for the determination of the kinetic isotope effect in
the isomerisation of natural abundance glucose and [2- H] glucose in
2
The concurrent resolution of C-1 protonated and singly deuterated
fructose positions affords the simultaneous measurement of methanol at 120 1C using H-USY (6).
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
glucose-to-fructose isomerisation encompasses solvent exchange
rather than an intramolecular hydride-shift for H-USY (6).
The work was funded by the Innovation Fund Denmark
(case number 5150-00023B). SM gratefully acknowledges
funding by Grant 2013_01_0709 of the Carlsberg Foundation.
SS thanks the Department of Biotechnology (Government of
India), New Delhi, India, for support. 800 MHz NMR spectra
were recorded on the spectrometer of the National Instrument
Center for NMR Spectroscopy of Biological Macromolecules at
the Technical University of Denmark. Prof. Mads Clausen is
acknowledged for helpful discussions.
Notes and references
1
2
3
4
Y. Rom ´a n-Leshkov, M. Moliner, J. A. Labinger and M. E. Davis,
Angew. Chem., Int. Ed., 2010, 49, 8954.
R. DiCosimo, J. McAuliffe, A. J. Poulose and G. Bohlmann, Chem.
Soc. Rev., 2013, 42, 6437.
E. Nikolla, Y. Roman-Leshkov, M. Moliner and M. E. Davis, ACS
Catal., 2011, 1, 408.
(a) J. B. Binder and R. T. Raines, J. Am. Chem. Soc., 2009, 131, 1979;
(b) S. G. Wettstein, D. M. Alonso, E. I. G u¨ rb u¨ z and J. A. Dumesic,
Curr. Opin. Chem. Eng., 2012, 1, 218; (c) M. Moliner, Y. Rom ´a n-
Leshkov and M. E. Davis, Proc. Natl. Acad. Sci. U. S. A., 2010,
Fig. 4 Stereoselectivity of hydride shift into the pro-R position using
H-USY (6). C1H1 signals of a-fructofuranosides are labelled by their
abundance (a). To a small extent (B4.6%), isomerisation encompasses
protonation of C1 from the solvent rather than through an intramolecular
1
07, 6164.
5
(a) Y. Roman-Leshkov, M. Moliner, J. A. Labinger and M. E. Davis,
Angew. Chem., Int. Ed., 2010, 49, 8954; (b) R. Bermejo-Deval,
R. S. Assary, E. Nikolla, M. Moliner, Y. Rom ´a n-Leshkov, S.-J.
Hwang, A. Palsdottir, D. Silverman, R. F. Lobo, L. A. Curtiss and
M. E. Davis, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 9727;
2
hydride shift. [1-pro-S- H] a-fructofuranoside reference signals are dis-
played with a single bold contour. (b) 3D plot of the spectral region shown
in (a) as contour plot.
(
c) S. Saravanamurugan, M. Paniagua, J. A. Melero and A. Riisager,
1
13
J. Am. Chem. Soc., 2013, 135, 5246; (d) S. Saravanamurugan and
A. Riisager, Catal. Sci. Technol., 2014, 4, 3186; (e) M. Paniagua,
S. Saravanamurugan, M. Melian-Rodriguez, J. A. Melero and
A. Riisager, ChemSusChem, 2015, 8, 1088; ( f ) V. Choudhary,
A. B. Pinar, R. F. Lobo, D. G. Vlachos and S. I. Sandler,
ChemSusChem, 2013, 6, 2369.
(a) L. M. Ren, Q. Guo, P. Kumar, M. Orazov, D. D. Xu, S. M. Alhassan,
K. A. Mkhoyan, M. E. Davis and M. Tsapatsis, Angew. Chem., Int. Ed.,
2015, 54, 10848; (b) S. Saravanamurugan, A. Riisager, E. Taarning
and S. Meier, ChemCatChem, 2016, 8, 3107.
at least 99% as estimated from H– C HSQC NMR spectra of
the substrate (see Fig. S5, ESI†). Nevertheless, fructose that only
is protonated at the C-1 position forms to almost 5% using
H-USY (6) in protonated methanol due to the incorporation of
hydrogen from the solvent, presumably through the formation
of an enol intermediate.
6
1
13
In conclusion, HÀ C HSQC NMR assays were applied to
detect the stereoselective hydride shift of the glucose hydrogen
H-2 into the pro-R position of fructose, both for solid Al-zeolites
H-USY (6) and (30)) and Sn-Beta zeolite as well as for dissolved
AlCl . The hydride shift into the pro-R position resembles
the stereochemistry of the xylose isomerase catalysed reaction,
and the stereochemical outcome results from the concerted
transition state involving hydride transfer from C2 to C1. Such
7
8
R. Bermejo-Deval, M. Orazov, R. Gounder, S.-J. Hwang and
M. E. Davis, ACS Catal., 2014, 4, 2288.
G. N. Li, E. A. Pidko and E. J. M. Hensen, Catal. Sci. Technol., 2014,
(
4
, 2241.
3
9 K. Bock, M. Meldal, B. Meyer and L. Wiebe, Acta Chem. Scand., Ser. B,
1983, 37, 101.
0 (a) M. Bøjstrup, B. O. Petersen, S. R. Beeren, O. Hindsgaul and
S. Meier, Anal. Chem., 2013, 85, 8802; (b) B. O. Petersen,
O. Hindsgaul and S. Meier, Analyst, 2014, 139, 401.
1
stereoselective catalysis by zeolites indicates a possible role 11 (a) E. J. Creyghton, S. D. Ganeshie, R. S. Downing and H. van
Bekkum, J. Mol. Catal. A: Chem., 1997, 115, 457; (b) J. Brus,
L. Kobera, W. Schoefberger, M. Urbanova, P. Klein, P. Sazama,
in enantioselective synthesis, not least of biomolecules. The
1
13
HÀ C HSQC NMR assays hinge on the distinction of carbo-
E. Tabor, S. Sklenak, A. V. Fishchuk and J. Dedecek, Angew. Chem.,
2015, 54, 541; (c) J. A. van Bokhoven, A. M. van der Eerden and
D. C. Koningsberger, J. Am. Chem. Soc., 2003, 125, 7435.
2 (a) K. N. Allen, A. Lavie, G. K. Farber, A. Glasfeld, G. A. Petsko and
D. Ringe, Biochemistry, 1994, 33, 1481; (b) A. Y. Kovalevsky,
L. Hanson, S. Z. Fisher, M. Mustyakimov, S. A. Mason, V. T.
Forsyth, M. P. Blakeley, D. A. Keen, T. Wagner, H. L. Carrell,
A. K. Katz, J. P. Glusker and P. Langan, Structure, 2010, 18, 688.
3 J. Q. Tang, X. W. Guo, L. F. Zhu and C. W. Hu, ACS Catal., 2015,
5, 5097.
hydrate isotopomers, thus permitting the direct quantification
of isotopomers and of their respective rates of formation.
1
1
1
13
HÀ C HSQC NMR both affords accurate determinations of
H D
kinetic isotope effects (k /k = 2.2 for H-USY (6)) and sufficient
resolution and sensitivity to determine the difference in activa-
tion energy between hydrogen transfer to the pro-R and pro-S
À1
positions (16.3 Æ 0.4 kJ mol ). To a minor degree (B4.6%),
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016