Structure-Function Relationships in Fructose Dehydration to 5-Hydroxymethylfurfural under Mild Conditions by Porous Ionic Crystals Constructed with Analogous Building Blocks

Article abstract of DOI:10.1002/cctc.201900614

5-Hydroxymethylfurfural (HMF) is an important bio-based chemical, which has been recognized as a key platform chemical for furan-based polymers, biomass-derived fuels, fine chemicals, and pharmaceuticals. Catalytic conversion of fructose into HMF is an important reaction, and the development of an efficient heterogeneous catalysts with high selectivity towards HMF is highly desired. In this study, several porous ionic crystals with a Keggin-type polyoxometalate [PW₁₂O₄₀]^3? and a macrocation [Cr₃O(OOCCH₂CN)₆(H₂O)₃]⁺ as building blocks are synthesized, and the structures are solved by single crystal X-ray diffraction analysis. The catalytic activities are basically dependent on the porosity, and the compound with a mesopore (aperture: 3 nm×2 nm) showed the highest HMF yield (30 % at 353 K in methanol/toluene solvent).

Full text of DOI:10.1002/cctc.201900614

Accepted Article
Title: Structure-Function Relationships in Fructose Dehydration to
5-Hydroxymethylfurfural under Mild Conditions by Porous Ionic
Crystals Constructed with Analogous Building Blocks
Authors: Takumi Yamada, Keigo Kamata, Eri Hayashi, Michikazu Hara,
and Sayaka Uchida
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: ChemCatChem 10.1002/cctc.201900614
Link to VoR: http://dx.doi.org/10.1002/cctc.201900614
A Journal of
www.chemcatchem.org

10.1002/cctc.201900614
ChemCatChem
COMMUNICATION
Structure-Function Relationships in Fructose Dehydration to
5-Hydroxymethylfurfural under Mild Conditions by Porous Ionic
Crystals Constructed with Analogous Building Blocks
Takumi Yamada,[a] Keigo Kamata,[b] Eri Hayashi,[b] Michikazu Hara,[b] and Sayaka Uchida*[a]
Abstract: 5-Hydroxymethylfurfural (HMF) is an important bio-based
chemical, which has been recognized as a key platform chemical for
furan-based polymers, biomass-derived fuels, fine chemicals, and
pharmaceuticals. Catalytic conversion of fructose into HMF is an
important reaction, and the development of an efficient
heterogeneous catalysts with high selectivity towards HMF is highly
desired. In this study, several porous ionic crystals with a Keggin-type
accommodate fructose molecules, so that this reaction
presumably took place on the surface of the solid particles.
We have previously reported the synthesis and crystal
structure
Cr[Cr3O(OOCCH2CN)6(H2O)3]3[PW12O40]2∙69H2O [I] with
Keggin-type POM and macrocation
[PW12O40]3
[Cr3O(OOCCH2CN)6(H2O)3]+ with polar cyano groups (CN) and
water ligands.[13] Compound
possesses one-dimensional
of
a
mesoporous
ionic
crystal
a
a
polyoxometalate
[PW12O40]3
and
a
macrocation
I
[Cr3O(OOCCH2CN)6(H2O)3]+ as building blocks are synthesized, and
the structures are solved by single crystal X-ray diffraction analysis.
The catalytic activities are basically dependent on the porosity, and
the compound with a mesopore (aperture: 3 nm × 2 nm) showed the
highest HMF yield (30% at 353 K in methanol/toluene solvent).
mesopores with an aperture of 3 nm × 2 nm, and the mesoporous
structure is maintained by the water of crystallization as templates.
The void volume of I is 44.2% of the crystal lattice and the largest
among the porous ionic crystals reported to date. While I does not
contain any acidic functional groups, IR spectroscopy with
pyridine as a basic probe molecule suggests the existence of
Brønsted acid sites, and I shows moderate catalytic activity in
allylation of benzaldehyde with allyltin in water at 353 K.[13]
Based on the mesoporous structure of I and previous
reports on fructose dehydration to HMF, we reached an idea that
I may show promising performance as a catalyst for this reaction.
In this work, we show that I can catalyze fructose dehydration to
HMF without any apparent by-products and at mild conditions in
a methanol/toluene solvent (HMF yield 30% at 353 K and 30 min).
In order to prove the importance of the mesoporous structure, we
synthesized microporous ionic crystals with analogous building
Biofuels are promising alternative fuels for replacement of limited
fossil fuels and have fewer environmental problems. Among the
current biofuel resources, 5-hydroxymethylfurfural (HMF)
converted from C6 monosaccharides such as fructose is a
versatile and key intermediate that is attracting much attention.[1-
3] Solid acid catalysts such as zeolites,[4] sulfated zirconia,[5] metal
phosphates and phosphonates,[6] acidic resins,[7] and
heteropolyacids[8,9] have been reported to catalyze this reaction.
A drawback with these solid acid catalysts is that they cause
various side reactions, leading to by-products such as levulinic
acid and formic acid. A landmark work in this research area is a
biphasic system (water/methylisobutylketone modified with 2-
butanol) with hydrochloric acid or acidic resins: HMF is extracted
spontaneously into the organic phase and undesirable side
reactions are suppressed, leading to high fructose conversion
(90%) and HMF selectivity (80%) at 453 K and < 3 min.[10] On the
other hand, it has been recently reported by Yi et al.[11] that a
porous ionic crystal Cs2[Cr3O(OOCC2H5)6(H2O)3]2[SiW12O40]
synthesized by one of us[12] catalyzes the dehydration of fructose
to HMF (yield 3345%) at 403 K and 90 min, due to the mild
Brønsted acid sites generated from water molecules coordinating
to Cr3+ because of the polarizing effect due to the neighboring
Keggin-type polyoxometalate (POM) (i.e., [SiW12O40]4). However,
the pore size of this ionic crystal (5.2 Å) is too small to
blocks:
Cr[Cr3O(OOCCH2CN)6(CH3OH)2(H2O)]6[PW12O40]3
ꞏ18CH3OH [II] and Cr[Cr3O(OOCCH2CN)6(H2O)3]6[PW12O40]3
ꞏ50H2O [III]. Compounds II and III show lower catalytic activity
(HMF yield ca. 10%) than I in fructose dehydration to HMF.
Compound
[Cr3O(OOCCH2CN)6(H2O)3](NO3),
I
was
synthesized
Cr(NO3)3,
from
and
H3PW12O40∙nH2O in water with slight modification from our
previous work.[13] As shown in Figure 1a, POM [PW12O40]3 and
macrocation [Cr3O(OOCCH2CN)6(H2O)3]+ were arranged
alternately forming a ring, and a mesopore with an aperture of 3
nm × 2 nm was formed. Hydrogen-bonded water clusters served
as a template to construct and maintain the mesoporous structure,
and Cr3+ as a counter cation existed in the mesopore. Powder X-
ray diffraction pattern (PXRD) of I well agreed with the calculation
(Figures 2a and 2b) showing that the structure of the whole bulk
solid is well represented by the single crystal data.
Compound
II
was
synthesized
from
[Cr3O(OOCCH2CN)6(H2O)3](NO3) and H3PW12O40∙nH2O in
methanol with toluene as a poor solvent for crystallization.
Methanol was utilized as a solvent to construct micropores
instead of mesopores, because water and methanol have two and
one hydrogen atoms, respectively, to form hydrogen bonds, and
methanol molecules would form a smaller hydrogen-bonded
cluster as a template.[14] IR spectrum of II contained the
absorption bands of the two ionic components (see experimental
section and Figures S1a, S1b, and S1d). Elemental analysis of II
suggested that the macrocation and POM exist with a ratio of 6 :
3 (2 : 1) along with one Cr3+ as a counter cation to compensate
[a]
[b]
Dr. S. Uchida and T. Yamada
Department of Basic Science, School of Arts and Sciences,
The University of Tokyo
3-8-1 Komaba, Meguro-ku, Tokyo 153-8902 (Japan)
E-mail: csayaka@mail.ecc.u-tokyo.ac.jp
Dr. K. Kamata, E. Hayashi, Prof. Dr. M. Hara
Laboratory for Materials and Structures, Institute of Innovative
Research, Tokyo Institute of Technology
Nagatsuta-cho 4259, Midori-ku, Yokohama 226-8503 (Japan)
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/cctc.201900614
ChemCatChem
COMMUNICATION
the surplus anion charge. Thermogravimetry suggested that 18
CH3OH existed as solvent molecules (Figure S2a).
macrocation and POM exist with a ratio of 6 : 3 (2 : 1) along with
one Cr3+ as a counter cation to compensate the surplus anion
charge. Therefore, II and III have the same chemical formula and
can be considered as polymorphs. We have previously reported
that ionic crystals based on POMs often show polymorphism
because of the flexible crystal structure due to Coulomb
interaction, which works isotropically and in a long range.[15]
Elemental analysis of III showed that the amount of Mg is
negligible (0.25wt%). However, crystals did not appear without
the addition of anhydrous MgSO4, which probably acts as a buffer
to control the water content in the synthetic solution.
Thermogravimetry suggested that 50 H2O existed as solvent
molecules (Figure S2b).
Table 1. Crystallographic data
II
III
Crystal system
Space group
Triclinic
P-1
Monoclinic
C2
a = 17.116(17)
b = 18.776(19)
c = 21.76(2)
 = 112.649(6)
 = 89.983(16)
 = 117.265(4)
a = 27.941(4)
b = 19.1673(19)
c = 13.034(2)
 = 117.270(6)
Unit cell
Figure 1. Crystal structures of (a) I, (b)(c) II, and (d)(e) III. (c) and (e) show the
void analysis of II and III, respectively, with a probe diameter of 1.2 Å and grid
spacing 0.7 Å. Surface of the voids are shown in yellow. Solvent molecules are
removed from the crystal structure. Gray, dark blue, and green spheres in (a),
(b), and (d) show the carbon atoms, nitrogen atoms, and Cr3+ as counter cations,
respectively. Light green, orange, and purple polyhedra in (a), (b), and (d) show
the [WO6], [CrO6], and [PO4] units, respectively.
Volume
5603(9)
2
6204.6(16)
2
Z
Dcalc (g cm3
)
2.588
3909.84
2.431
F(000)
4086.08
19/19,
21/21,
23/24
33/26,
30/32,
27/27
h, k, l range
The crystal structure of II along with void analysis are shown
in Figures 1b and 1c, respectively. Crystallographic parameters
are summarized in Table 1. Compound II possessed winding one-
dimensional channels along the c-axis with an aperture of ca. 1
nm. The void volume was 16.1% of the crystal lattice. While the
positions of methanol solvent molecules could not be resolved by
single crystal XRD (SXRD), they probably existed in the one-
dimensional channels. Cr3+ as a counter cation could be located
in the one-dimensional channel and was disordered between two
positions. Two out of the three terminal water ligands of Cr3+ in
the macrocation were exchanged with methanol (i.e.,
[Cr3O(OOCCH2CN)6(CH3OH)2(H2O)]+). The macrocations were
arranged along the a-axis, and each water ligand was hydrogen-
bonded to the terminal oxygen of the neighboring POM (Figure
S3a). These hydrogen bonds as well as Coulomb interactions
between the oppositely charged ionic components probably
contribute to stabilize the crystal structure of II. PXRD of II well
agreed with the calculation (Figures 2e and 2f) showing that the
structure of the whole bulk solid is well represented by the single
crystal data.
 (Mo K) (mm1
R1(I > 2(I))
)
12.971
0.1457
0.4360
1.201
11.727
0.0452
0.1446
1.100
wR2 (all data)
GOF on F2
The crystal structure of III along with void analysis are
shown in Figures 1d and 1e, respectively. Crystallographic
parameters are summarized in Table 1. Compound III possessed
winding one-dimensional channels along the c-axis with minimum
and maximum apertures of ca. 0.7 nm and 1.3 nm, respectively.
The void volume was 30.6% of the crystal lattice. Thus the
porosity decreased in the order of I (aperture 3 nm × 2 nm, void
volume 44.2%) > III (aperture 1.3 nm (max) and 0.7 nm (min), void
volume 30.6%) > II (aperture 1 nm, void volume 16.1%). SXRD
could locate 42 out of the 50 solvent water molecules, which
existed in the one-dimensional channel. Cr3+ as a counter cation
was also located in the one-dimensional channel. The
macrocations were arranged along the c-axis, and cyano groups
of the neighboring macrocations were aligned parallel to each
other with C-C or C-N distances of 3.4873.617 Å, suggesting
weak dispersion force between the cyano groups. Such kind of
interaction with vinyl groups has been observed in our previous
work with a porous ionic crystal of [Cr3O(OOCCH=CH2)6(H2O)3]3
[PW12O40]∙15H2O.[16] PXRD of III well agreed with the calculation
Compound
III
was
synthesized
from
[Cr3O(OOCCH2CN)6(H2O)3](NO3), Cr(NO3)3, H3PW12O40∙nH2O,
and anhydrous MgSO4 in an aqueous methanol solution. IR
spectrum of III contained the absorption bands of the two ionic
components (see experimental section and Figures S1a, S1b,
and S1e). Elemental analysis of III suggested that the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201900614
ChemCatChem
COMMUNICATION
(Figures 2g and 2h) showing that the structure of the whole bulk
solid is well represented by the single crystal data.
Table 2 summarizes the results of fructose dehydration to
HMF with I as a solid catalyst. A mixed solution of methanol :
toluene = 1 : 2 (v/v) was chosen as a reaction solvent. Methanol
is necessary to dissolve fructose into the reaction solution and
toluene is added to prevent dissolution of the solid catalyst.
Temperature programmed desorption mass spectrometry (TPD-
MS) of the spent catalyst I showed that methanol and toluene
were absorbed in and excluded from the solid bulk, respectively.
Anhydrous MgSO4 is added to adsorb excess water, which
formed by the dehydration reaction of fructose, otherwise the solid
catalyst would become pasty and the reaction would stop. The
reaction did not proceed without a catalyst (entry 11), and ionic
components showed low activities (entries 9 and 10). Notably,
[Cr3O(OOCCH2CN)6(H2O)3](NO3) was slightly active (entry 10,
HMF yield 5%), suggesting that mild Brønsted acid sites
generated from water molecules coordinating to Cr3+ might exist
as in ref. 11. A maximum HMF yield of 30% was obtained after 30
min at 353 K (entries 1 and 2). Note that products other than HMF
(cf. levulinic acid and methyl levulinate) were not observed under
this reaction condition (see 1H-NMR of the reaction product,
Figure S4). We assume that the high HMF selectivity is due to the
mild acidity,[11] which prevents further hydrolysis to levulinic acid.
PXRD pattern of I after the reaction fairly agreed with that before
the reaction (Figures 2b and 2c), showing that the mesoporous
structure is maintained under the reaction condition. The
maximum HMF yield of this reaction at 353 K was 30% (30 min),
which slightly decreased upon prolongation of reaction time. Gas
chromatography mass spectrometry (GC-MS) suggested the
formation of polymerized products.
Furthermore, I can be easily recovered from the reaction
solution by filtration followed by washing with water to remove
MgSO4, and reused without significant loss of catalytic activity
(entries 4 and 5). Inductively coupled plasma optical emission
spectrometry (ICP-OES) analysis of the reaction solution showed
that leaching of I into the reaction solution is < 3%. In order to
prove the importance of the mesoporous structure, I was
evacuated at room temperature prior to the reaction, and PXRD
pattern showed that I became amorphous by this treatment
(Figure 2d). The solid catalyst became pasty immediately and
dissolved into the reaction solution, suggesting that the
mesoporous structure of I is the key to the structural stability as
well as the catalytic activity. When the reaction was carried out at
a higher temperature (373 K), the reaction rate increased while
the HMF yield remained almost the same (entry 3). At 383 K, by-
products due to the reaction between HMF and methanol (ether
and acetal) were observed (see GC-MS data, Figure S5), and
PXRD showed that the crystal structure of I changed under this
reaction condition.
(d)
(h)
(c)
(b)
(g)
(f)
(e)
(a)
3
5
10
2 Theta [deg]
15 3
5
10
2 Theta [deg]
15
Figure 2. Powder XRD patterns of (a) I (calc), (b) I (exp, before reaction), (c) I
(exp, after reaction), (d) I (exp, after evacuation at rt), (e) II (calc), (f) II (exp), (g)
III (calc), and (h) III (exp).
Table 2. Fructose dehydration to HMF with various catalysts[a]
Temp
[K]
Time
[min]
Conv.[d]
[%]
HMF yield[e]
[%]
Entry
Catalyst
1
2
3
I
353
353
15
30
30
35
10
30
I
I
373
353
353
353
353
353
353
15
30
30
30
75
30
30
35
35
30
15
15
15
5
30
25
22
5
4
5
6
7
8
9
I (2nd run)
I (3rd run)
II
II
10
12
< 2
III
[b]
K3PW12O40
10
[Cr3O(OOCCH2CN)6
(H2O)3](NO3)[b]
353
353
30
30
5
5
< 2
< 2
11
no catalyst[c]
[a]Reaction conditions: Fructose (0.4 g, 2.2 mmol), naphthalene (0.5 mmol,
internal standard), anhydrous MgSO4 (0.6 g, added to remove excess water
in the reaction solution), and catalysts (0.2 g) in methanol/toluene (5 mL/10
mL). [b]Partial dissolution was observed. [c]Anhydrous MgSO4 was added to
the reaction solution. [d]Amount of fructose was analyzed as follows: The
solution was filtered and the filtrate was evaporated to dryness. Water was
added to the residue, the solution was filtered again, and the filtrate was
collected and injected into the HPLC system with butanol as an internal
standard. [e]Only HMF was detected as a product. The discrepancy between
fructose conversion and HMF yield is probably due to fructose adsorbed in
the catalysts. We attempted to analyze the amount of fructose adsorbed in
the catalysts by dissolving the catalyst in methanol followed by evaporation
to dryness and dissolving the residue in water. However, unknown HPLC
peaks, which overlapped with fructose and the internal standard, were
observed.
The results of fructose dehydration to HMF with II and III as
solid catalysts are also summarized in Table 2. As for II, the HMF
yields after 30 and 75 min at 353 K were 5 and 10%, respectively
(entries 6 and 7). HMF yield did not increase further with the
reaction time. The reaction rate and activity were much lower than
that of I (HMF yield 30% at 353 K and 30 min). As for III, the HMF
yield after 30 min at 353 K was 12%, and did not increase further
with the reaction time. Therefore, the catalytic activities
decreased in the order of I (30%) >> III (12%) > II (10%), which is
in line with the porosity of the crystal structures I (aperture 3 nm
× 2 nm, void volume 44.2%) > III (aperture 1.3 nm (max) and 0.7
This article is protected by copyright. All rights reserved.

10.1002/cctc.201900614
ChemCatChem
COMMUNICATION
Cr(NO3)3∙9H2O (54 g, 0.14 mol) and cyano acetic acid (27 g, 0.32 mol)
were dissolved in 300 mL of acetone, and the solution was stirred at 323
K for 6 h. The solution was concentrated by evaporation, water was added
as a poor solvent, and [Cr3O(OOCCH2CN)6(H2O)3](NO3) was obtained as
nm (min), void volume 30.6%) > II (aperture 1 nm, void volume
16.1%). These results clearly show the significance of porosity of
the ionic crystals in this reaction system. In contrast, pinacol
rearrangement with I-III as catalysts was performed: Pinacol
rearrangement is a typical acid reaction, and the molecular size
of pinacol (C6H14O2, MW = 118) is smaller than that of fructose
(C6H12O6, MW = 180). The catalytic activities at 373 K were quite
similar among I-III, suggesting that the intrinsic acidities of I-III are
similar to each other.
a
green powder (21 g, yield 90% based on cyano acetic acid).
Compound I was synthesized as follows with slight modification from our
previous work:[13] [Cr3O(OOCCH2CN)6(H2O)3](NO3) (2.8 g, 3.6 mmol) was
dissolved in 200 mL of water followed by filtration. Cr(NO3)3∙9H2O (4.0 g,
10 mmol) and H3PW12O40∙nH2O (5.0 g, 1.7 mmol) were dissolved in a
minimum amount of water and added slowly to the filtrate. Green crystals
of I were obtained after 24 h (6 g, yield 55% based on H3PW12O40∙nH2O).
In conclusion, porous ionic crystals I-III were crystallized
Compound
[Cr3O(OOCCH2CN)6(H2O)3](NO3)
II
was
synthesized
(0.2 g,
as
follows:
mmol) and
with
[Cr3O(OOCCH2CN)6(H2O)3]+ as ionic components by the
appropriate choice of synthetic conditions. Compound
a a macrocation
Keggin-type POM [PW12O40]3 and
0.25
H3PW12O40∙nH2O (1.0 g, 0.34 mmol) were dissolved into 50 mL of
methanol with stirring at 333 K. The solution was cooled to rt (room
temperature) followed by the addition of 50 mL of toluene as a poor solvent.
Green crystals of II were obtained after 24 h (0.22 g, yield 38% based on
[Cr3O(OOCCH2CN)6(H2O)3](NO3)). Cr3+ as a counter cation probably
formed by the solvolysis of [Cr3O(OOCCH2CN)6(H2O)3]+ in hot methanol.
IR: 1656 asym(OCO), 1436 asym(CH), 1382 sym(OCO), 1079 asym(P-O),
981 asym(W=O), 897 asym(W-Oc-W), 814 asym(W-Oe-W), 597 asym(Cr3-
O). Elemental analysis (%) calcd for II: C 10.89, H 1.01, N 3.81, Cr 7.47,
P 0.70, W 50.01; found: C 10.54, H 0.79, N 3.83, Cr 7.47, P 0.70, W 49.05.
Prior to the elemental analysis, II was pretreated at 323 K under vacuum
to remove the methanol solvent molecules. Compound III was synthesized
as follows: [Cr3O(OOCCH2CN)6(H2O)3](NO3) (0.2 g, 0.25 mmol), MgSO4
(0.6 g, 5 mmol), and Cr(NO3)3∙9H2O (0.2 g, 0.5 mmol) were added to 100
mL of an aqueous methanol solution (25vol%) with stirring at rt. To the
solution, H3PW12O40∙nH2O (1.0 g, 0.34 mmol) was added with stirring.
Green crystals of III were obtained after 24 h and washed with water (0.53
g, yield 89% based on [Cr3O(OOCCH2CN)6(H2O)3](NO3)). IR: 1657
I
possessed mesopores while II and III possessed micropores, and
as expected, the catalytic activities on fructose dehydration to
HMF were I > II, III. The main features of the present reaction
system can be summarized as follows (Figure 3): No apparent by-
products due to mild acidity, monophase and easy work-up, and
reaction proceeding quickly under mild conditions (30 min at 353
K). Ongoing and future works are to increase the acidity slightly
by substituting the metal center of the macrocation (e.g., Cr3+

Fe3+)[17] and to increase the structural stability by strengthening
the Coulomb interaction among ionic components (e.g.,
[PW12O40]3  [SiW12O40]4). If we could increase the structural
stability of the porous ionic crystals and work at slightly higher
temperatures, it would be possible to make use of Lewis acid sites
generated from the metal center. These crystals may serve as a
bifunctional catalyst with both Brønsted and Lewis acidity, and
glucose as well as fructose may be used as a reactant for HMF
production.[18]


asym(OCO), 1439 asym(CH), 1381 sym(OCO), 1080 asym(P-O), 981
asym(W=O), 898 asym(W-Oc-W), 818 asym(W-Oe-W), 595 asym(Cr3-O).
Elemental analysis (%) calcd for III: C 9.92, H 0.83, N 3.86, Cr 7.56, Mg 0,
P 0.71, W 50.65; found: C 10.12, H 0.81, N 3.79, Cr 7.25, Mg 0.25, P 0.71,
W 49.62. Note that while the amount of Mg in III is negligible, crystals do
not appear without the addition of anhydrous MgSO4. Prior to the
elemental analysis, III was pretreated at 323 K under vacuum to remove
the
water
of
crystallization.
Single-Crystal X-ray Diffraction (SXRD) Analysis: X-ray diffraction data
of II and III were collected at 93 K with a CCD 2-D detector by using Rigaku
Saturn diffractometer with graphite monochromated Mo Kα radiation.
Structures were solved by direct methods (SHELX97), expanded using
Fourier techniques, and refined by full-matrix least squares against F2 with
the SHELXL-2014 package. Tungsten, chromium, and phosphorous
atoms were refined anisotropically. Chromium as a counter cation, oxygen,
carbon, and nitrogen atoms were refined isotropically. Hydrogen atoms
were not included in the model. Thermogravimetry suggested the
existence of 18 CH3OH and 50 H2O as solvent molecules in II and III,
respectively. Methanol solvent molecules in II were not located with SXRD.
42 out of the 50 water of crystallization in III could be located with SXRD.
Void analysis was carried out by a Mercury structure visualization software
(CCDC) with a probe radius of 1.2 Å and approximate grid spacing of 0.7
Å. Crystal data for II: triclinic, P-1, a = 17.116(17), b = 18.776(19), c =
21.76(2),  = 112.649(6),  = 89.93(16),  = 117.265(4), V = 5603(9), Z =
2, R1 = 0.1457, wR2 = 0.4360, GOF = 1.201. Crystal data for III: monoclinic,
C2/c, a = 27.941(4), b = 19.1673(19), c = 13.034(2),  = 117.270(6), V =
6204.6(16), Z = 1, R1 = 0.0452, wR2 = 0.1446, GOF = 1.100. CCDC-
1905293 and 1905294 contain the crystallographic data for II and III,
respectively.
Figure 3. Schematic illustration of the catalytic conversion of fructose to HMF
with I as a catalyst.
Experimental Section
Materials: Na2WO4∙9H2O, Na2HPO4∙9H2O, Cr(NO3)3∙9H2O, anhydrous
MgSO4, conc. HCl, cyano acetic acid, acetone, methanol, butanol, toluene,
diethylether and distilled water were purchased from Kanto Chemical Co.
Inc. and used as received. H3PW12O40∙nH2O was synthesized according
to the literature.[19] Fructose, naphthalene (internal standard), 5-
hydroxymethylfurfural, and levulinic acid (as reference standards) were
Characterization: Combustion analysis (Elementar, vario MICRO cube)
was used for the quantitative analysis of C, H, and N. Inductively coupled
plasma optical emission spectrometry (ICP-OES) (Agilent Technologies,
ICP-OES720) was used for the quantitative analysis of chromium,
magnesium, phosphorous, and tungsten. Prior to the ICP-OES
measurements, ammonium hydroxide solution (1 mL) was added to ca. 10
purchased
from TCI
Co.
Ltd.
and
used
as
received.
Synthesis: [Cr3O(OOCCH2CN)6(H2O)3](NO3) was synthesized as follows:
This article is protected by copyright. All rights reserved.

10.1002/cctc.201900614
ChemCatChem
COMMUNICATION
mg (accurately weighed) of II or III to dissolve the solid completely into
water (100 mL). FT-IR spectra were measured by the KBr pellet method
with a JASCO FT/IR 4100 spectrometer (JASCO) equipped with a TGS
MEXT of Japan and the Collaborative Research Project of
Laboratory for Materials and Structures, Institute of Innovative
Research, Tokyo Institute of Technology. Prof. M. Matsuo and Dr.
K. Shozugawa (Univ. of Tokyo) are acknowledged for providing
access to the ICP-OES instrument. Prof. Jun Terao (Univ. of
Tokyo) is acknowledged for providing access to the GC-MS
instrument.
detector. Thermogravimetry was conducted with
a Thermo Plus 2
thermogravimetric analyzer (Rigaku) with α-Al2O3 as a reference under a
dry N2 flow (100 mL min1) in the temperature range of 298773 K and an
increasing rate of 10 K min1. Powder XRD patterns were measured with
a New advance D8 X-ray diffractometer (Bruker) by using Cu K radiation
(
=
1.54056 Å, 40 kV-40 mA) at 1.8 deg min1
.
Catalytic Reaction: Reaction was carried out in a glass reactor equipped
with a magnetic stirrer. In a typical run of fructose dehydration, a mixture
of fructose (0.4 g, 2.2 mmol), naphthalene (0.5 mmol, internal standard),
anhydrous MgSO4 (0.6 g, added to remove excess water in the reaction
solution), and catalysts (IIII 0.2 g, 0.0150.020 mmol) in methanol/toluene
(5 mL/10 mL) was stirred under air at 353373 K. The amounts of products
(HMF and levulinic acid) were followed by gas chromatography using a
GC-2014 (Shimadzu) fitted with a TC-WAX (GL Sciences) or HP-5 (Agilent
J&W) capillary column and a flame ionization detector. The reaction
solution (with the solid catalyst) was sampled periodically and injected
directly into the GC system. Gas chromatography mass spectrometry (GC-
MS) was conducted with a GCMS-QP2010 SE (Shimadzu) to check the
possible formation of by-products (5-dimethoxymethyl-2-furylmethanol
and 5-methoxymethylfurfural). The amount of reactant (fructose) was
followed by high-performance liquid chromatography (HPLC) using a
Hitachi Chromaster 5450 (Hitachi) fitted with a Gelpack GL-C610-S
(Hitachi Chemical) or a Inertsil ODS-3 (GL Sciences) with pure water as
eluent and a RI detector. Prior to the HPLC analysis, the reaction solution
was filtered and the filtrate was evaporated to dryness. Water was added
to the residue, the solution was filtered again, and the filtrate was collected
and injected into the HPLC system with butanol as an internal standard.
1H-NMR spectra of the reaction solution (CDCl3) were recorded on a
Bruker Advance III 500 MHz (Bruker). Prior to the 1H-NMR measurement,
the reaction solution was filtered, the filtrate was evaporated to dryness,
and CDCl3 was added to dissolve the residue. Temperature programmed
desorption mass spectrometry (TPD-MS) of the spent catalyst I was
conducted with a BELCAT (Microtrac BEL) to check the absorption of the
reaction solvent molecules (methanol and toluene).
Keywords: Biomass • Chromium • Crystal engineering •
Polyoxometalates • Ionic crystals
[1]
A. A. Rosatella, S. P. Simeonov, R. F. M. Frade, C. A. M. Afonso, Green
Chem. 2011, 13, 754.
[2]
[3]
T. Wang, M. W. Nolte, B. H. Shanks, Green Chem. 2014, 16, 548.
M. Hara, K. Nakajima, K. Kamata, Sci. Technol. Adv. Mat. 2015, 16,
034903.
[4]
[5]
[6]
[7]
V. V. Ordomsky, J. van der Schaaf, J. C. Schouten, T. A. Nijhuis, J. Catal.
2012, 287, 68.
J. B. Joo, A. Vu, Q. Zhang, M. Dahl, M. Gu, F. Zaera, Y. Yin,
ChemSusChem 2013, 6, 2001.
D. –M. Gao, B. Zhao, H. Liu, K. Morisato, K. Kanamori, Z. He, M. Zeng,
H. Wu, J. Chen, K. Nakanishi, Catal. Sci. Technol. 2018, 8, 3675.
H. Tang, N. Li, G. Li, W. Wang, A. Wang, Y. Cong, X. Wang, ACS
Sustainable Chem. Eng. 2018, 6, 5645.
[8]
[9]
Q. Zhao, L. Wang, S. Zhao, X. Wang, S. Wang, Fuel 2011, 90, 2289
N. L. Mulik, P. S. Niphadkar, K. V. Pandhare, V. V. Bokade, Chem. Select
2018, 3, 832.
[10] Y. Román-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312,
1933.
[11] X. Yi, I. Delidovich, Z. Sun, S. Wang, X. Wang, R. Palkovits, Catal. Sci.
Technol. 2015, 5, 2496.
[12] A. Lesbani, R. Kawamoto, S. Uchida, N. Mizuno, Inorg. Chem. 2008, 47,
3349.
[13] R. Kawahara, K. Niinomi, J. N. Kondo, M. Hibino, N. Mizuno, S. Uchida,
Dalton Trans. 2016, 45, 2805.
[14] I. Bakó, T. Megyes, S. Bálint, T. Grósz, V. Chihaia, Phys. Chem. Chem.
Phys. 2008, 10, 5004.
[15] K. Mizuno, T. Mura, S. Uchida, Cryst. Growth Des. 2016, 16, 4968.
[16] R. Kawahara, S. Uchida, N. Mizuno, Inorg. Chem. 2014, 53, 3655.
[17] S. Uchida, A. Lesbani, Y. Ogasawara, N. Mizuno, Inorg. Chem. 2012, 51,
775.
Acknowledgements
This work was supported by grant-in-aids for scientific research
JP16K05742 and JP17H05356 (Coordination Asymmetry) from
[18] H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 2007, 316, 1597.
[19] J. C. Bailar Jr., H. S. Booth, M. Grennert, Inorg. Synth. 1939, 1, 132
This article is protected by copyright. All rights reserved.

10.1002/cctc.201900614
ChemCatChem
COMMUNICATION
Entry for the Table of Contents
COMMUNICATION
Takumi Yamada, Keigo Kamata, Eri
Hayashi, Michikazu Hara, and Sayaka
Uchida*
Page No. – Page No.
Structure-Function Relationships in
Fructose
Dehydration
to
5-
Three porous ionic crystals were synthesized with
a
Keggin-type
Hydroxymethylfurfural under Mild
Conditions by Porous Ionic Crystals
Constructed with Analogous Building
Blocks
polyoxometalate and a macrocation with Cr3+ as building blocks. The catalytic
activities of the crystals toward fructose dehydration to HMF were dependent on
the porosity, and the mesoporous crystal (aperture: 3 nm × 2 nm) showed the
highest HMF yield
.
This article is protected by copyright. All rights reserved.