See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/264613303

Insights into the Cr(III) catalysed isomerization mechanism of glucose to

fructose in the presence of water using ab initio molecular dynamics

Article in Physical Chemistry Chemical Physics · July 2014

DOI: 10.1039/C4CP02095B

CITATIONS READS

57 1,736

3 authors:

Samir H. Mushrif Jithin John Varghese

University of Alberta Indian Institute of Technology Madras
96 PUBLICATIONS 3,338 CITATIONS 16 PUBLICATIONS 569 CITATIONS

SEE PROFILE SEE PROFILE

Dionisios G Vlachos
University of Delaware
761 PUBLICATIONS 32,960 CITATIONS

SEE PROFILE

All content following this page was uploaded by Samir H. Mushrif on 30 July 2015.

The user has requested enhancement of the downloaded file.

PCCP RSCPublishing

ARTICLE

Insights into the Cr(III) catalysed isomerization

Cite this: DOI: 10.1039/x0xx00000x
mechanism of glucose to fructose in the
presence of water using ab initio molecular
dynamics
Received 00th January 2012,
Accepted 00th January 2012 Samir H. Mushrif,*a Jithin J. Varghesea and Dionisios G. Vlachos b
DOI: 10.1039/x0xx00000x
The mechanism of glucose ring opening and isomerization to fructose, catalyzed by the Lewis
www.rsc.org/ acid catalyst CrCl 3 in the presence of water, is investigated using Car–Parrinello molecular
dynamics with metadynamics. Minimum energy pathways for the reactions are revealed and
the corresponding free energy barriers are computed. Addition of glucose replaces two water
molecules in the active [Cr(H 2 O)5OH] +2 complex, with two hydroxyl groups of glucose taking
their place. Ring opening and isomerization reactions can only proceed if the first step
involving the deprotonation of glucose is accompanied by the protonation of the OH  group in
the partially hydrolyzed metal center ([Cr(C 6H12O6)(H2 O)3OH]+2  [Cr(C 6H11O6)(H2O) 4]+2).
This provides further evidence that the partially hydrolyzed [Cr(H 2O)5OH] +2 is the active
species catalyzing ring opening and isomerization reactions and that unhydrolysed Cr +3 may
not be able to catalyze the reactions. After the ring opening, the isomerization reaction
proceeds via deprotonation, followed by hydride shift and the back donation of the proton from
the metal complex to the sugar. Water molecules outside the first coordination sphere of the
metal complex participate in the reaction for mediating the proton transfer. Hydride shift in the
isomerization is the overall rate limiting step with a free energy barrier of 104 kJ/mol. The
simulation computed barrier is in agreement with experiments.

Introduction Glucose is the monomer unit of cellulose and a readily

available and less expensive sugar.
As a renewable source for hydrocarbons, biomass is a Isomerization of glucose to fructose is the largest
promising candidate.1,2 Carbohydrates constitute more than immobilized biocatalytic process worldwide for the production
75% of the vast amount of biomass produced by nature and of high fructose corn syrup. However, enzymatic processes face
barely 5% of them are consumed in food and nonfood several limitations:7 (1) they require extremely high purity of
sectors.3 For the utilization of biomass, it is imperative to glucose; (2) they are slow and expensive; (3) the enzyme
develop processes that can convert these carbohydrates into catalyst goes through an irreversible decay; and (4) they can
useful chemicals and fuels. However, carbohydrates are highly only operate within a very narrow temperature and pH (basic)
functional molecules owing to the presence of oxygen. Hence, range. This makes a single pot conversion of glucose to HMF
it is vital to selectively remove the oxygenated groups so that (and subsequently to levulinic acid, if –valerolactone is the
they can be transformed into a variety of chemical desired product) difficult, as after the isomerization, fructose
intermediates in fine chemicals, polymers, plastics and fuels needs to be dehydrated to HMF in an acidic environment at
sectors. 5-hydroxymethyl-furfural (HMF) and its 2,5- temperatures higher than 90 °C. Alternate approaches to this
disubstituted furan derivatives are such key building-block process were proposed by Davis and co-workers using the
molecules, currently derived from petrochemicals. Dehydration zeolite Sn-beta, a heterogeneous Lewis acid catalyst, 8 and
of fructose to HMF in aqueous media has been successfully recently by Vlachos and co-workers using CrCl3, a
performed by Dumesic and coworkers. 4,5 However, economic homogeneous Lewis acid catalyst. 9-12 They also demonstrated
analysis of the process has shown that the cost of the product promising results for a “single pot” conversion of glucose to
HMF is most sensitive to the initial cost of fructose. 6 Hence, it fructose to HMF by carrying in tandem isomerization and
becomes crucial to efficiently convert glucose to fructose. dehydration reactions by combining Lewis and Brønsted acid

This journal is © The Royal Society of Chemistry 2013 J. Name., 2013, 00, 1-3 | 1
ARTICLE Journal Name

catalysts in aqueous media. In general, a heterogeneous catalyst corresponding free energy barriers in ring opening and
would be a preferred choice since it is easy to separate and isomerization are reported. The role of OH  group, coordinated
reuse. However, the dehydration reaction of fructose to HMF with Cr in [Cr(H2O)5OH]+2, in reducing the free energy barrier
and rehydration of HMF to levulinic acid are associated with of the ring opening and the isomerization reaction is discussed.
side reactions that lead to the formation of soluble polymers Participation of water in the reaction chemistry is demonstrated.
and insoluble humins. These by-products can deactivate the To the best of our knowledge, this is the first comprehensive
catalyst by clogging its pores. In addition to Cr (III), various paper reporting the molecular level mechanism and energetics
metals (as salts) like Al (III), Zn (II), Sn (IV) carry out this of metal salt catalyzed glucose isomerization reaction while
chemistry13-18 with the highest yield of HMF from glucose considering the dynamics of explicit water.
(~60%) obtained from Al (III) and Cr (III). Al (III) is less active
than Cr (III) for glucose isomerization to fructose. 11
Computational Methods
In the homogeneously catalyzed glucose to HMF All the simulations were performed using the CPMD
conversion, it was thought that the metal ion was the active package; version 3.15.1.22 The CPMD package implements the
Lewis acid center for isomerization and that the addition of an Car–Parrinello scheme23 for ab initio molecular dynamics
inorganic acid like HCl, provided the Brønsted acidity for the calculations. The first–principles calculations were performed
dehydration of fructose to HMF. Using a speciation model using the planewave–pseudopotential implementation of Kohn–
together with kinetic experiments for CrCl3, it was recently Sham density functional theory. 24 The Troullier–Martins
suggested11 that the active species might not be hexahydrated pseudopotential25 with the Perdew–Burke–Ernzerhof
Cr+3, but the partially hydrolyzed [Cr(H 2O)5OH]+2 complex. generalized gradient approximation 26 was used for all the atoms
Additionally, even without the addition of any inorganic acid, in the simulation system. Only the Γ-point was used for
the hydrolysis of Cr+3 by water molecules releases a proton and integration over the Brillouin zone in the reciprocal space. An
reduces the pH of the medium,11, 19, 20 providing Brønsted energy cut–off of 100 Ryd. was used. The simulation system
acidity to drive fructose dehydration. contained one Cr+3, an OH ion, 2 Cl ions, a pyranosic glucose
molecule and 11 water molecules. 3 of the water molecules
[Cr(H2O)6]+3 + H2O ↔ [Cr(H2O)5OH]+2 + H3O+
were directly coordinated with the metal and 8 water molecules
Car–Parrinello molecular dynamics (CPMD) simulations and were out of the inner coordination sphere. Extra water
EXAFS analysis suggest that Cr+3 remains coordinated with 6 molecules were taken to allow the possibility of water mediated
oxygens of water in the absence of glucose. Upon the addition proton transfer during reaction steps. The size of the simulation
of glucose, two oxygen atoms of the hydroxyl groups in cell was 18 18 18 Å3 and more than 3 Å space between the
glucose replace 2 water molecules within the inner coordination outermost atom and the border of the simulation cell was kept
sphere. The metal center is then coordinated with 3 water to avoid any artefacts, as suggested in the CPMD manual.
molecules, one OH , and two oxygen atoms of the glucose Unlike the condensed phase environment, a smaller simulation
molecule. It is hypothesized that this complexation of glucose system was implemented to render these expensive calculations
with the active [Cr(H 2O)5OH]+2 species is the first step in the tractable. However, the rate determining step in the reaction
isomerization of glucose to fructose. Similar behaviour is was studied in the presence of 124 explicit water molecules,
expected for other metals like Al, Zn, and Sn. Importantly, under experimental density. Each and every atom in the
isotopic labelling and kinetic isotope effects indicate that the simulation system was treated quantum mechanically.
mechanism of glucose isomerization in Crcontaining solution Calculations were performed with spin polarization on and the
could be analogous to that of heterogeneous Sn-beta zeolite and ground state spin multiplicity of the system was taken as 4. The
is an intramolecular hydride transfer from the C 1 to the C2 metal cation is coordinated with OH , 3 water molecules and
carbon.21 This provides promise for understanding two oxygen atoms of the hydroxyl groups of glucose, and Cl 
homogeneous Lewis acid isomerization chemistry and to ions are not present within the first coordination sphere of the
extend the findings to heterogeneous catalysts. metal (cf. Scheme 1).
To gain a molecular level understanding of CrCl 3 catalyzed OH
isomerization of glucose to fructose and to explain the catalytic
5 O H
role of the partially hydrolyzed metal center in the reaction, we H
H 1
investigate the mechanism of glucose isomerization to fructose, 4 OH H
catalyzed by the active [Cr(H 2O)5OH]+2 species, in the presence HO O H -
3 2 O
of water, using ab initio molecular dynamics. Simulations are H
H O 3+
performed with explicit, quantum mechanically treated water
H Cr
molecules, glucose, and the active catalyst species. Since O H
isomerization is preceded by glucose ring opening and H O H
O H
[Cr(H2O)3OH]+2 is also believed to catalyze the ring opening H
H
process, the catalytic transformation of closed ring Scheme 1 Octahedral coordination of Cr +3 with glucose, OH
glucose/glucopyranose to its open chain form is also studied. and 3 water molecules. Carbon atoms in glucose are numbered
Reaction mechanisms, minimum free energy pathways and the from 1 to 6.

2 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
Journal Name ARTICLE

This is in agreement with EXAFS data that suggests that the paper, but general guidelines provided in the literature 31, 32 were
metal center is coordinated with 6 oxygen atoms in an followed. The height of the metadynamics potential was kept
octahedral fashion.11 Our previous work11 on CPMD fixed at 1.25 kJ. Analogous to the original Car–Parrinello
simulations of the metal coordination with glucose, water and scheme, the dynamics of the collective variables were separated
OH also suggested that the coordination of glucose replaces from the ionic and fictitious electronic motion by choosing an
two water molecules in [Cr(H2O)5OH]+2. As this particular appropriate value for the fictitious mass of the collective
arrangement of Cr with glucose and water is also believed to be variables. The temperature of the collective variables is set to
the first step in glucose ring opening and thus isomerization, we 375 K (same as the physical temperature of the system) and is
chose this as our starting configuration (cf. Scheme 1). In the controlled in a window of ± 200 K using velocity rescaling. A
cascade of reactions including glucose to fructose isomerization sample input file of the CPMD–metadynamics simulation in
followed by fructose dehydration to HMF, 11 the concentration this work is provided as supporting information.
of this species can be low; thus, our simulations here focus on
the potentially active species in the local environment of water Results and Discussion
and glucose only, rather than simulating all Cr-containing
species. It has to be noted that there exists a dynamic Glucose Ring Opening
equilibrium between OH and water molecules attached to Cr +3. The first step in glucose isomerization is the opening of the
During the CPMD run, we observed that the OH  group pyranosic ring (cf. Scheme 2). More than 99.5% of glucose
abstracts a proton from a neighbouring water molecule exists in water in the pyranosic form, and the six membered
coordinated with Cr +3, thus becoming a water molecule itself pyranosic ring is coordinated with [Cr(H 2O)3OH]+2, as shown
and making the original water molecule an OH. Nosé-Hoover in Scheme 1, and as suggested earlier using EXAFS data and
chain thermostat was used for controlling ionic and electronic CPMD calculations.11 Hence, this was taken as the starting
temperatures. The frequency for the ionic thermostat was set to structure for simulating the ring opening step. One picosecond
1800 cm-1 (characteristic of a C–C bond vibration frequency) of CPMD simulation without metadynamics was performed on
and for the electron thermostat to 10000 cm -1. The fictitious the starting configuration to allow the system to equilibrate at
electron mass parameter in CPMD was set to 400 a.u. Short finite temperature. During equilibration, it was observed that
molecular dynamics runs without the thermostat were the hydroxyl group of the C 1 carbon of the glucose ring gets
performed to obtain an approximate value around which the deprotonated and loses the proton to a neighbouring water
fictitious electronic kinetic energy oscillates. Based on the molecule that is not in the first coordination sphere of the metal.
observations from these runs, a value of 0.01 a.u. was chosen It has to be noted that this reactive event of deprotonation was
for the electronic kinetic energy. The timestep of a single captured within a small timescale during the CPMD simulation
molecular dynamics step was set to 0.0964 fs. Geometry without any metadynamics acceleration. All other steps in
optimization was performed on the system before molecular glucose ring opening and isomerization require metadynamics
dynamics and the system was equilibrated for 1 ps. Energies, acceleration.
including the fictitious electronic kinetic energy were
monitored throughout the simulation to ascertain that the O
1
system does not deviate from the Born–Oppenheimer surface
6
OH 2
during the molecular dynamics simulation. Molecular dynamics H OH
trajectories were visualized using the VMD software. 27 5 OH
H HO
3
H
CPMD simulations are computationally expensive and H
4 1
hence it is not feasible to run a simulation for experimental OH H H
4
OH
reaction timescales. The metadynamics technique 28-30 was thus HO OH
3 2 5
implemented in conjunction with CPMD to accelerate the H OH
dynamics and avoid being trapped in deep minima of the H OH
energy surface. It allows the construction of multidimensional 6
free energy surface along selected reaction coordinates. This OH
technique, as described by Laio and Gervasio,29 is based on
Scheme 2 Ring opening of glucose molecule. Carbon atoms are
“filling up” the free energy surface by dropping potentials at
labelled from 1 to 6.
small time intervals in the coordinate space of interest. The
flattening of the free energy surface thus helps the system to
Glucose ring opening involves the cleavage of C 1–O (ring
overcome the activation energy barriers. The details of how
oxygen) bond and the transfer of „H‟ from the hydroxyl group
metadynamics is implemented by extending the Car–Parrinello
of C1 to the ring oxygen (cf. Scheme 2). However, two different
Lagrangian in CPMD can be found in the literature. 28 The
pathways are possible for the „H‟ transfer. One does not involve
collective variables (or reaction coordinates) implemented to
the OH group on the metal center and the proton which is
simulate the glucose ring opening and isomerization reactions
already removed from the C 1 hydroxyl group is transferred to
are described in the respective sections. The parameters in
the ring oxygen, via water mediated proton transfer (cf. Fig.
metadynamics were not optimized for the system studied in this

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 3
ARTICLE Journal Name

1a). The other entails the protonation of the OH  group in that the metadynamics simulation was terminated shortly after
[Cr(H2O)3OH]+2, thus forming glucose coordinated the ring opening was observed.
[Cr(H2O)4]+3 as an intermediate that later transfers the proton to
the ring oxygen (direct transfer or water mediated) as the C 1–
ring oxygen bond cleaves (cf. Fig. 2a).
To find the minimum energy pathway for glucose ring
opening and to understand the role of OH  group of Cr, two
different metadynamics simulations were performed. In both
simulations, one of the metadynamics collective variables
(CV1) was the coordination number of C 1 with the ring oxygen.
In the case where the OH of the Cr+3–OH group is not
involved in ring opening (cf. Fig. 1), the other collective
variable (CV2) was the coordination number of ring oxygen
with respect to hydrogens in water molecules. This would allow
the water mediated proton transfer to the ring oxygen. In the
case where the OH of the Cr+3–OH group facilitates glucose
ring opening (cf. Fig. 2), the second collective variable (CV2) is
the difference between coordination number of Cr +3–OH
oxygen atom and ring oxygen with respect to all hydrogens,
including those of water molecules.
The free energy surface of the ring opening of glucose,
without the involvement of the OH of the Cr+3–OH is shown
in Fig. 3. Even though Cr +3 remains coordinated to glucose, the
pathway does not involve the protonation of Cr+3–OH to form
an additional water molecule on the metal center.

Fig. 2 a) Ring opening mechanism of glucose facilitated by the

OH in the Cr+3–OH group. b) Metadynamics collective
variables implemented to simulate the ring opening.

Fig. 1 a) Ring opening mechanism of glucose without the

involvement of OH in Cr+3–OH. b) Metadynamics collective Fig. 3 Metadynamics calculated free energy landscape for the
ring opening reaction of glucose without the involvement of the
variables implemented to simulate the ring opening.
OH of the [Cr(H2O)3OH]+2 group. Structures and free energy
values at local minima are shown. The free energy barrier in the
The free energy barrier for this pathway (272 kJ/mol) is
reaction step is shown over the arrow. Refer to Fig. 1(b) for the
prohibitively high, thus strongly suggesting that Cr +3 alone or definition of collective variables CV1 and CV2.
unhydrolysed [Cr(H2O)4] +3 coordinated with glucose cannot
catalyze the ring opening of glucose in water. It has to be noted

4 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
Journal Name ARTICLE

The product free energy well was not allowed to fill catalyzing the glucose ring opening reaction. The presence of
completely, and hence, based on the difference in the free the base OH on the Lewis acid center appears to play an
energies shown in Fig. 3, no conclusion can be drawn about the important role in ring opening, in qualitative agreement with ab
equilibrium of open vs. closed chain forms of glucose. A large initio calculations.33 The back donation of the proton to the ring
difference in the free energy is however expected since more oxygen of glucose to complete ring opening is mediated by
than 99.5% of glucose exists in the closed ring pyranosic form water molecules in the immediate vicinity (but not in the inner
in water. coordination sphere) of the hydrolyzed metal complex.
To further investigate the role of Cr +3–OH in catalyzing the
ring opening of glucose, a metadynamics simulation with a Mechanism of Glucose Isomerization
collective variable (CV2) taking into account the oxygen of the The mechanism of the isomerization reaction is shown in
OH group on Cr+3, i.e., allowing the formation of Fig. 5. The isomerization reaction is initiated by the
[Cr(H2O)4]+3, was performed (cf. Fig. 2b). Fig. 4 shows the deprotonation of the open chain glucose molecule at C 2 carbon,
corresponding free energy surface. The free energy barrier for followed by hydride transfer from C 2 to C1 and back
the ring opening, with [Cr(H 2O)4]+3 as an intermediate, is 125 protonation at C1 oxygen. The metadynamics collective
kJ/mol. Comparison of the mechanisms shown in Fig. 1 and variables implemented to simulate individual isomerization
Fig. 2 and the corresponding free energy barriers in Fig. 3 and steps are defined in Fig. 6. Details are discussed below.
Fig. 4, respectively, clearly indicate that glucose ring opening
may not proceed unless an intermediate state involving the
protonation of the OH  group on Cr ([Cr(C6H12O6)(H2O)3OH]+2
 [Cr(C6H11O6)(H2O)4]+2), is formed.

Fig. 4 Metadynamics calculated free energy landscape for the

ring opening reaction of glucose, facilitated by the OH of the
[Cr(H2O)3OH]+2 group. Structures and free energy values at
local minima are shown. The free energy barrier in the reaction
step is shown over the arrow. Refer to Fig. 2(b) for the
definition of collective variables CV1 and CV2.

Direct (or water mediated) transfer of proton from C1

hydroxyl group to ring oxygen for ring opening is an extremely
high barrier step and hence unhydrolysed Cr +3 would not be Fig. 5 Mechanism of isomerization of glucose, catalyzed by
able to catalyze the ring opening and thus, the subsequent [Cr(H2O)3OH]+2 in the presence of water molecules.
isomerization reaction. It has to be noted that since the initial
deprotonation of glucose was observed in the CPMD Deprotonation at C2: The first step in the isomerization
equilibration period, the barrier for the same is not reported in reaction is the deprotonation of C 2 hydroxyl group of the open
the free energy surfaces shown in Fig. 3 and Fig. 4. However, chain glucose molecule. Based on our observation for the ring
an additional metadynamics simulation was performed to opening, it was believed that the presence of an intermediate
estimate the barrier and was found to be 79 kJ/mol. [Cr(H2O)4]+3 would be needed for the subsequent hydride
To summarize, analogous to heterogeneously catalyzed transfer step from C2 to C1. One of the collective variables (cf.
glucose ring opening by the Sn–beta zeolite,33 the partially Fig. 6a) for the deprotonation step was the difference in the
hydrolyzed metal complex appears to be an active site in coordination of glucose oxygen and oxygen in Cr +3–OH, with

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 5
ARTICLE Journal Name

respect to the hydrogen. However, since the orientation of the The deprotonation of glucose and protonation of Cr +3–OH
hydroxyl hydrogen of glucose is in the opposite direction of the involves a free energy barrier of 44 kJ. It has to be noted that
metal complex, the C2–O–H angle had to be incorporated as an since the proton transfer from glucose oxygen to Cr+3–OH is
additional collective variable (CV2). The free energy surface, water mediated (similar to that for the closed ring glucose), the
as a function of these 2 collective variables is shown in Fig. 7. value of CV1 (cf. Fig. 7) goes from ~0.9 to ~0.1. If Cr+3–OH
would have picked up the same „H‟ that was removed from the
C2 hydroxyl group, CV1 would have gone from ~0.9 to ~ 0.9.
The water mediated proton transfer was also confirmed from
the molecular dynamics trajectory. Since the hydrogen which
was part of CV2 (angle bend) is picked up by a water molecule
and the water molecule moves around in the simulation cell
during the metadynamics simulation, several peaks at CV1  0
(for different values of CV2) of approximately the same free
energy level can be seen in Fig. 7. The free energy barrier
between these peaks is 25–30 kJ/mol. With the proton transfer
from C2–OH to Cr+3–OH, the average Cr–O distance for the C2
oxygen decreases from ~2.1 Å to ~1.88 Å, thus making it more
strongly coordinated with the metal center. The oxygen in Cr–
OH, which now becomes Cr+3 ----- OH2, on the contrary, moves
slightly away from the metal center. The enhanced interaction
of C2 oxygen with the metal center is crucial for the next
elementary step of isomerization, which is the hydride transfer
from C2 to C1.

Fig. 6 Definition of collective variables (CV) for a)

deprotonation, b) hydride shift, and c) back donation of proton.
Atoms participating in CV definitions are shown in bold.

Fig. 8 Metadynamics calculated free energy landscape for the

hydride transfer from C2 to C1. Structures corresponding to two
minima in the free energy surface are shown. The free energy
barrier in the reaction step is shown over the arrow. Refer to
Fig. 6b) for the definition of collective variables CV1 and CV2.

Hydride Transfer from C2 to C1: Hydride transfer is usually the

rate limiting step in sugars isomerization. 9,21,33-35 In the recent
isotope labelling studies of glucose to fructose isomerization 21
using CrCl3 and AlCl3 as catalysts, a prominent kinetic isotope
effect was observed when the deuterium was at the C2 position,
Fig. 7 Metadynamics calculated free energy landscape for the thus suggesting that hydride shift from the C2 to C1 position is
deprotonation of C2 hydroxyl group of glucose. Structures the rate-limiting step with both catalysts. The implemented
corresponding to different minima in the free energy surface are metadynamics collective variables to simulate the hydride
shown. The free energy barrier in the reaction step is shown transfer are defined in Fig. 6b. One hydrogen is bonded to each
over the arrow. Refer to Fig. 6a for the definition of collective C1 and C2 before the hydride transfer and hence, CV1  CV2 
variables CV1 and CV2.
0.9. However, after the hydride transfer, CV1 is  1.8 and CV2

6 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
Journal Name ARTICLE

is  0.1. The Cr–O distance for the C2 oxygen is ~1.88 Å and from the glucose molecule upon relaxation. This indicates that,
for the C1 oxygen is ~2.1 Å before the hydride transfer, and the unlike ionic liquids, the hydride shift is probably catalyzed by a
trend reverses after the hydride transfer. We believe that the single Cr moiety in water and the cooperativity is created by the
presence of an intermediate water molecule on the metal Lewis acid center, Cr, and base, OH.
complex after the deprotonation of open chain glucose
([Cr(C6H12O6)(H2O)3OH]+2  [Cr(C6H11O6)(H2O)4]+2) plays a
key role in hydride transfer as well. This allows the Lewis acid
metal center (Cr+3) to have the strongest interaction with the C 2
oxygen (before the hydride transfer) and with the C 1 oxygen
(after the hydride transfer) since C1/C2 oxygen is the only
negatively charged oxygen in its inner coordination sphere.
Since hydride shift in sugars isomerization is suggested to be a
concerted transfer of neutral „H‟ atom from C 2 to C1 and of an
electron from C2 oxygen to C1 oxygen,36 a stronger interaction
of the oxygen atom(s) with the metal center would enhance the
catalytic activity of the metal center. The free energy surface
for hydride transfer as a function of these collective variables is
shown in Fig. 8. Hydride shift was observed to be an
energetically slightly unfavourable step with a free energy
barrier of 148 kJ/mol. Since hydride shift is shown to be the
Fig. 9 Metadynamics calculated free energy landscape for the
rate limiting step in the glucose isomerization reaction (with a hydride transfer from C2 to C1 in a fully condensed phase.
free energy barrier higher than any other step)33,34 in water,37, 38 Structures corresponding to two minima in the free energy
we compare the calculated free energy barrier with the surface are shown. The free energy barrier in the reaction step
experimental data. 11 The enthalpy of activation, as reported is shown over the arrow. Refer to Fig. 6b for the definition of
from the experimental data for glucose to fructose collective variables CV1 and CV2.
isomerization by [Cr(H 2O)3OH]+2, is 64 kJ/mol.11 Using the
Eyring equation, if we plot the experimental kinetics data as
ln(k/T) vs. 1/T, the entropy of activation can be calculated from
the intercept [ln(kB/h + ∆S/R)]. The entropy of activation from
the experimental data is 129 J/mol. Thus, at T=375 K the free
energy of activation for [Cr(H2O)3OH]+2 catalyzed glucose
isomerization is 112 kJ/mol. The metadynamics calculated free
energy of activation (148 kJ/mol) is significantly higher than
the experimental value. We believe that the reason for this
difference is that metadynamics simulations were performed
with a few explicit water molecules, and do not fully account
for the condensed phase environment effects. Hence, in order to
fully account for the condensed phase reaction environment and
to get an accurate estimate of the free energy barrier for the rate
limiting step, we simulated the hydride transfer step in the
presence of 124 explicit water molecules at the experimental
density. The hydride transfer step, with the same metadynamics
collective variables defined in Fig. 6b was simulated and the
free energy surface is shown in Fig. 9. In the condensed phase
environment, the hydride transfer step has a free energy barrier
of 104 kJ/mol. This is in excellent agreement with the
Fig. 10 Metadynamics free energy landscape for the
experimental value (112 kJ/mol) and indicates the important reprotonation of the glucose molecule to complete the
role of solvent in accurately estimating free energy barriers of isomerization reaction cycle. Structures corresponding to
these reactions. In comparison, the free energy barrier reported minima in the free energy surface are shown. The free energy
for hydride transfer catalyzed by the heterogeneous Sn–beta barrier in the reaction step is shown over the arrow. Please refer
zeolite catalyst is 80 kJ/mol at 298 K.33 The barrier for hydride to Fig. 6c for the collective variables CV1 and CV2.
transfer catalyzed by Cr (III) in ionic liquids ranges from 50–70 Reprotonation at C1: The last step in glucose to fructose
kJ/mol39,40 and further reduces in the presence of two Cr (III) isomerization is the back donation of proton from the metal
centres. However, in the present work, when we attempted to complex to the C1 oxygen of the sugar. The collective variables
put two Cr (III) moieties to form a binuclear complex, similar implemented for this step are given in Fig. 6c. At the beginning
to that in ionic liquids, one of the Cr (III) ions moved away

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 7
ARTICLE Journal Name

of the simulation, CV1 is equal to ~ 1.8 and CV2 is almost subsequent isomerization effectively. In addition, the
equal to zero. If the oxygen of the water molecule, which is mechanism for isomerization was detailed, and we found that
coordinated with Cr +3 and is part of CV2, loses the proton for the hydride transfer is the rate limiting step in the mechanism.
back donation (through water mediated proton transfer) to the Furthermore we have revealed the exact role of the OH  group
sugar molecule, the final product state would have been CV1  in the active species, which is to accept the proton from the
0.9 and CV2  0.9. However, if any other water molecule hydroxyl group of glucose to form an intermediate water
that is not part of CV2 (but coordinated with Cr) gets molecule and thereby to reduce the free energy barrier for ring
deprotonated, then the final product would have CV1  0.9 opening and hydride transfer. Finally, we have shown the
and CV2  1.8. During the metadynamics run, both cases are explicit participation of water molecules to mediate the proton
observed and transfer of proton is always water mediated. The transfers in the reaction cycle.
free energy barrier in the former case is 58 kJ and in the latter
case is 68 kJ/mol (cf Fig. 10). Additionally, this step is Acknowledgements
energetically slightly unfavourable. JJV and SHM would like to acknowledge financial support
provided by Nanyang Technological University, Singapore.
Conclusions The authors would like to thank Dr. Vinit Choudhary, Dr.
Stavros Caratzoulas, and Prof. Douglas Doren for useful
Car–Parrinello molecular dynamics simulations were
discussions. DGV acknowledges support from the Catalysis
performed for the first time to investigate glucose isomerization
Centre for Energy Innovation, an Energy Frontier Research
to fructose using CrCl3 as a Lewis acid catalyst in water.
Centre funded by the U. S. Department of Energy, Office of
Reactive events were captured using metadynamics and
Science and Office of Basic Energy Sciences under award
minimum energy pathways and corresponding free energy
number DE-SC0001004.
barriers were calculated. Coordination of glucose with partially
hydrolyzed octahedral Cr(III) complex (C6H12O6 + Notes and references
[Cr(H2O)5OH]+2  [Cr(C6H12O6)(H2O)3OH]+2 + 2H2O) is a
School of Chemical and Biomedical Engineering,
required for catalyzing ring opening and isomerization
Nanyang Technological University,
reactions. Hydroxyl groups attached to C 1 and C2 carbon atoms
62 Nanyang Drive,
of glucose replace two water molecules in the inner
Singapore, 637459, Singapore
coordination sphere of the metal. The OH  group in the metal
Corresponding Author Email: [email protected]
complex is in dynamic equilibrium with the rest of the three b
Center for Catalytic Science and Technology and Catalysis Center for
water molecules and has a stronger interaction with the metal Energy Innovation,
center than that of the water molecules and glucose hydroxyl Department of Chemical & Biomolecular Engineering,
groups. The first step in isomerization is the deprotonation of University of Delaware,
glucose at one of the hydroxyl groups coordinated with the Newark, DE 19716, USA.
metal center, thus forming a glucose alkoxide type species with
enhanced interaction between the metal center and glucose
1 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem. Int.
oxygen. This step is associated with a small free energy barrier.
Ed., 2007, 46, 7164-7183.
However, if the OH group attached to the metal complex does
2 M. J. Climent, A. Corma and S. Iborra, Green Chem., 2011, 13, 520-
not pick up this proton to form an additional intermediate water
540.
molecule on the metal complex, the overall free energy barrier
3 H. Roper, Starch-Starke, 2002, 54, 89-99.
for ring opening becomes prohibitively high. With the
4 J. N. Chheda, Y. Roman-Leshkov and J. A. Dumesic, Green Chem.,
formation of the intermediate water molecule, the free energy
2007, 9, 342-350.
barrier for ring opening is significantly reduced. Similarly, for
5 Y. Roman-Leshkov, J. N. Chheda and J. A. Dumesic, Science, 2006,
the isomerization reaction, the formation of this intermediate
312, 1933-1937.
water molecule activates the hydride transfer. The intermediate
6 A. I. Torres, P. Daoutidis and M. Tsapatsis, Energy Environ. Sci.,
water molecule later gives the proton back to the glucose
2010, 3, 1560-1572.
molecule, via water mediated proton transfer and completes the
7 S. H. Bhosale, M. B. Rao and V. V. Deshpande, Microbiol Rev, 1996,
isomerization reaction. This further solidifies our understanding
60, 280-300.
that, in addition to the Lewis acid center, the attached OH 
8 A. Corma, L. T. Nemeth, M. Renz, S. Valencia, Nature, 2001, 412,
Brønsted base is also essential for cooperative catalysis.
423-425.
Hydride transfer is the overall rate limiting step in glucose
9 M. Moliner, Y. Roman-Leshkov and M. E. Davis, Proc. Natl. Acad.
isomerization with a free energy barrier of 104 kJ/mol in good
Sci. USA, 2010, 107, 6164-6168.
agreement with the estimated experimental free energy of
10 Y. Roman-Leshkov, M. Moliner, J. A. Labinger and M. E. Davis,
activation of 112 kJ/mol.
Angew. Chem. Int. Ed., 2010, 49, 8954-8957.
Our simulations further support that [Cr(H2O)5OH]+2 is the
11 V. Choudhary, S. H. Mushrif, C. Ho, A. Anderko, V. Nikolakis, N. S.
active species for glucose isomerization and that unhydrolysed,
Marinkovic, A. I. Frenkel, S. I. Sandler and D. G. Vlachos, J. Am.
hexahydrated Cr cannot catalyze glucose ring opening and
Chem. Soc., 2013, 135, 3997-4006.

8 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 2012
Journal Name ARTICLE

12 . ikolla . om n-Leshkov, M. Moliner and M. E. Davis, ACS

Catal., 2011, 1, 408-410.
13 S. K. Tyrlik, D. Szerszen, B. Kurzak and K. Bal, Starch - Stärke,
1995, 47, 171-174.
14 A. A. Efremov, G. G. Pervyshina and B. N. Kuznetsov, Chem.. Nat.
Comp., 1998, 34, 182-185.
15 C. B. Rasrendra, I. G. B. N. Makertihartha, S. Adisasmito and H. J.
Heeres, Top. Cat., 2010, 53, 1241-1247.
16 Y. J. Pagán-Torres, T. Wang, J. M. R. Gallo, B. H. Shanks and J. A.
Dumesic, ACS Catal., 2012, 2, 930-934.
17 S. De, S. Dutta and B. Saha, Green Chem., 2011, 13, 2859-2868.
18 L. Peng, L. Lin, J. Zhang, J. Zhuang, B. Zhang and Y. Gong,
Molecules, 2010, 15, 5258-5272.
19 D. Hiroishi, C. Matsuura and K. Ishigure, Mineral Mag., 1998, 62A,
626-627.
20 J. C. Bailar, H. J. Eméleus, R. Nyholm, A. F. Trotman-Dickenson, A.
Eds. Comprehensive Inorganic Chemistry; Pergamon Press, Oxford,
1973; Vol. 3.
21 V. Choudhary, A. B. Pinar, R. F. Lobo, D. G. Vlachos and S. I.
Sandler, ChemSusChem, 2013, 6, 2369-2376.
22 CPMD, http://www.cpmd.org/, Copyright IBM Corp 1990-2008,
Copyright MPI für Festkörperforschung Stuttgart 1997-2001.
23 R. Car and M. Parrinello, Phys. Rev. Lett., 1985, 55, 2471-2474.
24 W. Kohn and L. J. Sham, Phys. Rev., 1965, 140, 1133-1138.
25 N. Troullier and J. L. Martins, Phys. Rev. B, 1991, 43, 1993-2006.
26 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77,
3865-3868.
27 W. Humphrey, A. Dalke and K. Schulten, J. Mol. Graphics Modell.,
1996, 14, 33-38.
28 M. Iannuzzi, A. Laio and M. Parrinello, Phys. Rev. Lett., 2003, 90,
238302.
29 A. Laio and F. L. Gervasio, Rep. Prog. Phys., 2008, 71, 126601.
30 A. Laio and M. Parrinello, Proc. Natl. Acad. Sci. USA, 2002, 99,
12562-12566.
31 B. Ensing, A. Laio, M. Parrinello and M. L. Klein, J. Phys. Chem. B,
2005, 109, 6676-6687.
32 A. Laio, A. Rodriguez-Fortea, F. L. Gervasio, M. Ceccarelli and M.
Parrinello, J. Phys. Chem. B, 2005, 109, 6714-6721.
33 R. S. Bermejo-Deval, R. S. Assary, E. Nikolla, M. Moliner, Y.
Roman-Leshkov, S. J. Hwang, A. Palsdottir, D. Silverman, R. F.
Lobo, L. A. Curtiss and M. E. Davis, Proc. Natl. Acad. Sci. USA,
2012, 109, 9727-9732.
34 R. S. Assary and L. A. Curtiss, J. Phys. Chem. A, 2011, 115, 8754-
8760.
35 R. S. Bermejo-Deval, R. Gounder, M. E. Davis, ACS Catal., 2012, 2,
2705-2713.
36 V. Choudhary, S. Caratzoulas and D. G. Vlachos, Carbohydr. Res.,
2013, 368, 89-95.
37 R. Xiong, M. Leon, V. Nikolakis, S. I. Sandler and D. G. Vlachos,
ChemSusChem, 2014, 7, 236-244.
38 M. Leon, T. D. Swift, V. Nikolakis and D. G. Vlachos, Langmuir,
2013, 29, 6597-6605.
39 Y. Zhang, E. A. Pidko and E. J. M. Hensen, Chem. Eur. J., 2011, 17,
5281–5288.
40 E. A. Pidko, V. Degirmenci and E. J. M. Hensen, ChemCatChem,
2012, 4, 1263–1271.

This journal is © The Royal Society of Chemistry 2012 J. Name., 2012, 00, 1-3 | 9

View publication stats