Selective Dehydration of Mannitol to Isomannide over Hβ Zeolite

Article abstract of DOI:10.1021/acscatal.7b01295

Isomannide is a potential feedstock for the production of super engineering plastics. A prospective route to obtain isomannide is dehydration of mannitol derived from lignocellulosic biomass, but homogeneous acid catalysts reported in the literature produce a large amount of 2,5-sorbitan as a byproduct in the dehydration reaction. In this work, we initially studied the mechanism of proton-induced dehydration of mannitol by density functional theory calculations, which suggested that local steric hindrance around acid sites designed at the angstrom level can tune the selectivity toward isomannide formation. Based on this prediction, we found that the precisely defined microporous confinement offered by Hβ provides improved selectivity and high catalytic activity for the production of isomannide, where 1,4-dehydration is favored by 20 kJ mol^-1 of activation energy. The optimization of the Si/Al ratio of Hβ to balance the acid amount and hydrophobicity improved the catalytic activity and achieved 63% yield of isomannide, far exceeding the best result reported previously (35% yield).

Full text of DOI:10.1021/acscatal.7b01295

Subscriber access provided by Binghamton University | Libraries
Article
Selective Dehydration of Mannitol to Isomannide over H-Beta Zeolite
Haruka Yokoyama, Hirokazu Kobayashi, Jun-ya Hasegawa, and Atsushi Fukuoka
ACS Catal., Just Accepted Manuscript • Publication Date (Web): 13 Jun 2017
Downloaded from http://pubs.acs.org on June 13, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
Selective Dehydration of Mannitol to Isomannide
over H-Beta Zeolite
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Haruka Yokoyama,†,‡ Hirokazu Kobayashi,† Jun-ya Hasegawa,*,† Atsushi Fukuoka*,†
†Institute for Catalysis, Hokkaido University, Sapporo, Hokkaido 001-0021, Japan
‡Graduate School of Chemical Sciences and Engineering, Hokkaido University, Sapporo,
Hokkaido 060-8628, Japan
KEYWORDS
Sugar alcohol; Mannitol; Dehydration; Isomannide; Solid acid catalyst; Zeolite
ABSTRACT
Isomannide is a potential feedstock for the production of super engineering plastics. A
prospective route to obtain isomannide is dehydration of mannitol derived from lignocellulosic
biomass, but homogeneous acid catalysts reported in the literature produce a large amount of 2,5-
sorbitan as a by-product in the dehydration reaction. In this work, we initially studied the
mechanism of proton-induced dehydration of mannitol by density functional theory (DFT)
calculations, which suggested that local steric hindrance around acid sites designed at the
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 33
1
2
3
4
5
6
7
8
9
Angstrom level can tune the selectivity toward isomannide formation. Based on this prediction,
we found that the precisely-defined microporous confinement offered by Hß provides improved
selectivity and high catalytic activity for the production of isomannide, where 1,4-dehydration is
favored by 20 kJ mol-1 of activation energy. The optimization of the Si/Al ratio of Hß to balance
the acid amount and hydrophobicity improved the catalytic activity and achieved 63% yield of
isomannide, far exceeding the best result reported previously (35% yield).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
Utilization of renewable biomass has recently attracted significant attention as a strategy to
decrease CO2 emission and lessen the use of finite fossil resources.1,2 Woody biomass is an
especially attractive feedstock for chemicals that does not compete with the use of biomass as
food.3 Cellulose and hemicellulose, which are both sugar polymers (Scheme 1), account for ca.
70% of the dry weight of wood stems, thus indicating that synthesis and utilization of sugar
should be the primary subject in biorefineries.4 Our group reported the hydrolytic hydrogenation
of cellulose to sugar alcohols by supported platinum catalysts.5 The reaction affords sorbitol and
mannitol in 25% and 6% yields, respectively. Mannitol is formed via fructose produced by the
isomerization of a glucose intermediate.6 After substantial studies on the optimization of the
reaction conditions, the yield of sorbitol exceeded 90% in the conversion of pure cellulose, and
sorbitol can also be obtained from real biomass.7-9 Furthermore, production of mannitol in good
yields is possible through the conversion of woody biomass. A niobium-molybdenum mixed
oxide catalyst selectively epimerizes glucose to mannose,10 and the subsequent hydrogenation
gives mannitol as the sole product.11 Softwood hemicellulose is also a promising feedstock for
ACS Paragon Plus Environment
2

Page 3 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
the synthesis of mannitol as it predominantly consists of glucomannan, a polymer of glucose and
mannose. Shirai et al. succeeded in producing mannitol in 14% yield based on the total sugar
content in cedar chips.12 Currently, 30,000 tons per year of mannitol is manufactured mostly
from starch;13 however, the source may be replaced by lignocellulosic biomass.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Production of diols from woody biomass.
Diols such as isosorbide and isomannide synthesized by the dehydration of sugar alcohols
(Scheme 1) are precursors to plastics with outstanding properties. Elimination of two water
molecules from sorbitol gives isosorbide, and isosorbide-derived polycarbonate has already been
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 33
1
2
3
4
5
6
7
8
9
commercialized by Mitsubishi (DURABIO). This plastic shows excellent transparency, coloring
properties, ultraviolet-resistance, impact-resistance, and surface hardness, which are useful for
car coatings and surface protection of liquid crystal displays. Likewise, dehydration of mannitol
gives isomannide, and isomannide-derived plastics exhibit similar remarkable properties found
in isosorbide-based ones.14 Moreover, isomannide-based polyamides and polyethers possess
higher thermal stability than isosorbide-based plastics.15,16 Therefore, isomannide is a potential
precursor to engineering plastics.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Studies on the dehydration of mannitol to isomannide have, however, been very limited in
contrast to the intensive research on the conversion of sorbitol to isosorbide.17-24 In previous
reports, dehydration of mannitol showed low selectivity toward the formation of isomannide.
Isomannide is formed by 1,4-dehydration via 1,4-mannitan, but 2,5-dehydration of mannitol
producing 2,5-sorbitan occurs in parallel (Scheme 2). For example, use of sulfuric acid resulted
in 35% yield of isomannide and 40% yield of 2,5-sorbitan.25 Subcritical water converted
mannitol to isomannide and 1,4-mannitan in 31% yield and 2,5-sorbitan in 41% yield.26
Therefore, selective dehydration of mannitol to isomannide is a challenge. In this work, we first
studied the reaction mechanism of proton-induced dehydration of mannitol by density functional
theory (DFT) calculations and then explored heterogeneous catalysts for selective synthesis of
isomannide based on the clarified mechanism.
ACS Paragon Plus Environment
4

Page 5 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Reaction pathways of mannitol dehydration.
EXPERIMENTAL
DFT Calculations. The potential energy profile of mannitol dehydration in the presence of a
proton was calculated using the Gaussian 09 program. The structure was optimized by DFT
calculations at the B3LYP/6-31G(d) level.27-29 The solvation effect was taken into account using
a self-consistent reaction field (SCRF) method with a polarized continuum model,30 for which
the dielectric constant of bulk methanol was used.
Materials and Characterization. The zeolites used in this study were as follows: Hß (Si/Al =
12.5), Hß(25), Hß(75), and HZSM-5(45) from Clariant; Hß(50), Hß(250), and HUSY(250) from
Tosoh; Hß(19), Hß(150), HY(2.55), and HUSY(40) from Zeolyst; and HMOR(45) from the
Catalysis Society of Japan (JRC-HM-90). All the zeolite samples were calcined at 823 K for 8 h
prior to use. N2 adsorption measurements were performed at 77 K with a BEL-SORP Mini (BEL
Japan) to determine the BrunauerEmmetTeller (BET) surface areas and micropore volumes of
the zeolites. The amounts of acid sites were quantified by NH3-temperature programmed
desorption (NH3-TPD; BEL Japan, BELCAT-A, mass spectrometer, m/z = 16 and 15).
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 33
1
2
3
4
5
6
7
8
9
Mannitol Dehydration. Mannitol dehydration was conducted in a three-necked flask (100
mL) with no solvent at 700 hPa (Figure S1). Reduced pressure was used to improve the
reproducibility, although atmospheric pressure gave similar results. Mannitol (182 mg) and
catalyst (50 mg) mixed for 5 min on a mortar were placed into the flask. When sulfuric acid was
used as a catalyst, mannitol was impregnated with 0.42 wt% sulfuric acid and dried at room
temperature under vacuum instead of mixing on a mortar. The flask was kept at 423 K,
monitored by a Teflon-coated thermocouple sustained in the reaction mixture, in an oil bath for 1
h under stirring with a magnetic stir bar. After the reaction, water-soluble products were
extracted with 20 mL of water. The solution was filtered to separate the solid catalyst and
analyzed using a high-performance liquid chromatograph (HPLC; Shimadzu LC10-ATVP,
refractive index detector) equipped with a Shodex Sugar SH-1011 column (⌀8 × 300 mm, mobile
phase: water 0.5 mL min-1, 323 K). Products were identified with nuclear magnetic resonance
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
spectroscopy (NMR; JEOL ECX-400, H 400 MHz). The coke content in the used zeolite was
measured by a total organic carbon analyzer (Shimadzu TOC-V CSN).
RESULTS AND DISCUSSION
Computation of the Mechanism for Mannitol Dehydration. Understanding the mechanistic
details of mannitol dehydration is useful to design a catalytic system that selectively promotes
the formation of isomannide. For this purpose, mannitol dehydration was simulated with DFT
calculations in the presence of a proton as a model acid catalyst, where two pathways were
hypothesized: 1,4-dehydration to 1,4-mannitan and 2,5-dehydration to 2,5-sorbitan (Scheme 2).
The dehydration occurs via an SN2-type reaction, evidenced by steric inversion that formed only
2,5-sorbitan with no generation of 2,5-mannitan (see below). An SN2-type reaction also takes
ACS Paragon Plus Environment
6

Page 7 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
place in the dehydration of a similar sugar alcohol (sorbitol) by sulfuric acid.31 In Figure 1, the
relative potential energy (E) is defined to be 0 kJ mol-1 for protonated mannitol (A). In the 1,4-
dehydration, an OH group at C4 attacks C1 bearing a protonated OH group via a transition state
having a trigonal bipyramidal structure at C1 with E of 92 kJ mol-1 (D). Subsequently,
elimination of a water molecule produces a stable protonated cyclic compound (E = 22 kJ mol-
1) (E).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Energy diagram of mannitol dehydration in the presence of a proton. DFT calculations
at the B3LYP/6-31G(d) level with SCRF: red, mannitol to 1,4-mannitan through 1,4-
dehydration; black, mannitol to 2,5-sorbitan through 2,5-dehydration.
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 33
1
2
3
4
5
6
7
8
9
Mannitol conversion by 2,5-dehydration was compared with the 1,4-dehydration. Protonation
of the C2 OH group gives E of 30 kJ mol-1 (F), which is more stable than the corresponding
intermediate in 1,4-dehydration mainly due to an additional hydrogen bond between the proton
and C1 OH group. Attack of the C5 OH group on C2 leads to a transition state with E of 52 kJ
mol-1 (G). Thus, the transition state in the 2,5-dehydration is 40 kJ mol-1 more stable than that in
the 1,4-dehydration. The higher stability of the transition state in 2,5-dehydration is due to the
formation of hydrogen bonds and a trans configuration of the two hydroxyl groups at C3 and C4
in the five-membered ring, whereas 1,4-dehydration results in a more sterically hindered cis
configuration at C2 and C3 in the ring (enlarged image, Figure S2). Next, the five-membered
ring product with inversion of stereochemistry at the C2 position (protonated 2,5-sorbitan) has E
of 4 kJ mol-1 (H). The final product of 2,5-dehydration has a higher potential energy than that of
1,4-dehydration despite the lower steric hindrance of the OH groups. This is attributed to the
final configuration of 2,5-dehydration that enforces an eclipsed conformation (dihedral angle of
CCCO: 2.5, Figure S3). The conformation causes steric repulsion and also undermines
hyperconjugative resonance stabilization,32 thus resulting in higher potential energy. Note that
the product (H) is less stable than intermediate (F), but 2,5-dehydration can occur due to
irreversible removal of water and a negative G value in the total reaction (Table S1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The most important factor that determines the selectivity of the two pathways is the difference
in the relative energies of the intermediates and transition states. Even taking into account errors
in the DFT calculations,33 the estimated activation energies of the elementary steps (C to D for
1,4-dehydration and F to G for 2,5-dehydration) are very close to each other (both 82 kJ mol-1).
However, the energy of the 2,5-dehydration pathway is relatively lower than that of the 1,4-
dehydration route, which gives a smaller apparent activation energy for the 2,5-dehydration (See
ACS Paragon Plus Environment
8

Page 9 of 33
ACS Catalysis
1
2
3
4
5
6
Equations section in the Supporting Information). Hence, 2,5-dehydration mainly occurs in
mannitol conversion.
7
8
9
Accordingly, an additional controlling factor is necessary to change the selectivity toward 1,4-
dehydration, which could be a steric effect that selectively raises the energies of the intermediate
F and transition state G. At the transition state, the peripheral structures of the reaction center are
significantly different between the two reaction pathways (Figure 2). In the 2,5-dehydration, the
reaction center C2 has a CH2OH group (dimensions 2.1  3.4  4.2 Å determined by the van der
Waals radius) directed through the side of the leaving water molecule. In contrast, C1 in 1,4-
dehydration is terminated by only two hydrogen atoms (1.0  1.5  2.3 Å). Therefore, 2,5-
dehydration needs more space around the side of the reaction center. Based on this consideration,
we propose that steric hindrance around acid sites may elevate the potential energies of the
intermediate F and transition state G and suppress the 2,5-dehydration. As a result, the relative
selectivity of less bulky 1,4-dehydration should increase. We assume that such steric
confinement would be realized by a microporous acid catalyst with a precisely-defined structure
at the Angstrom level.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Conformations of the transition states of mannitol dehydration.
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 33
1
2
3
4
5
6
7
8
9
Screening of Zeolite Catalysts in Dehydration of Mannitol. The DFT results motivated us
to use microporous zeolites as catalysts. Table 1 shows the results of mannitol dehydration over
various zeolites at 423 K for 1 h. No reaction proceeded in the absence of catalyst under the
conditions (entry 1). Hß(50) gave 96% conversion of mannitol and produced 60% yield of
isomannide (entry 2). Other identified products were 1,4-mannitan (3.0% yield; an intermediate
to isomannide) and 2,5-sorbitan (17% yield). Remaining part (16%) was not identified, but it
might contain coke and inter-molecular dehydration products. It is notable that Hß gave 63%
yield of 1,4-dehydration products in total, which was 3.7-fold higher than that of the 2,5-
dehydration product. In the 2,5-dehydration, the stereochemistry was inverted to form 2,5-
sorbitan and no 2,5-mannitan was observed, indicating an SN2 reaction for the dehydration. Note
that the SN2 reaction is preferred in 1,4-dehydration due to the lower steric hindrance and higher
energy barrier for the formation of a carbocation at the primary position. HMOR(45) converted
only 14% of mannitol and the isomannide yield was 1.7% (entry 3). HZSM-5(45) showed as
good selectivity for 1,4-dehydration as Hß, but the conversion was lower (entry 4). HUSY(40)
provided a high conversion, 97%, but the main product was 2,5-sorbitan (52%) (entry 5), and the
selectivity was similar to that by sulfuric acid (entry 6).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The differences in the selectivities and activities among the zeolites are ascribed to structural
characteristics of the zeolite pores. We have previously shown that dehydration of sorbitol
predominantly takes place in the pores of zeolites.21 The *BEA framework of Hß has a pore
dimension of 6.6 × 6.7 Å that is slightly larger than the molecular diameter of mannitol (6.3 Å).
The steric confinement disfavors the bulky reaction center of 2,5-dehydration (Figure S4A), but
accepts the compact center of 1,4-dehydration (Figure S4B) in the transition states. Indeed, 2,5-
dehydration was the primary reaction occurring in the pores of HUSY(40), which has super
ACS Paragon Plus Environment
10

Page 11 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
cages (12 Å) providing enough space for the dehydration reaction. In addition, mannitol and
dehydration products can diffuse into the three-dimensional pores of *BEA. In contrast, the one-
dimensional pores of MOR and small pores of ZSM-5 (5.1 × 5.5 Å) lead to low catalytic activity
probably due to slow diffusion of sugar molecules. Therefore, we conclude that the high activity
and selectivity of Hß is attributed to the three-dimensional pore structure with a suitable size to
preferably accommodate mannitol molecules proceeding to 1,4-dehydration. Hß shows high
catalytic activity also for the dehydration of sorbitol,21,22 but the 1,4-dehydration selectivity of
Hß in the sorbitol conversion was similar to that of H2SO4. Sorbitol easily undergoes 1,4-
dehydration without steric control by catalysts, which is in stark contrast to the mannitol
dehydration. Therefore, the results represented here is one of the rare examples in which a zeolite
changes regio-selectivity in biomass conversion reactions.34,35 Meanwhile, the acid strength of
the zeolite has a minor effect; NH3 desorption enthalpy (HMOR: 140 kJ mol-1 > HZSM-5: 135
kJ mol-1; HUSY: 135 kJ mol-1 > Hß: 130 kJ mol-1)36 has no correlation with the catalytic activity
or selectivity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 33
1
2
3
4
Table 1. Catalytic dehydration of mannitol at 423 K for 1 ha
5
6
7
8
9
b
Yield (%)
SBET Acid
amountc
Conv. Sel.d
(m2 g-
1,4-
2,5-
(mmol g-1) (%) (%) Isomannide Mannitan Sorbitan Otherse
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Entry Catalyst
)
1
2
3
4
5
6
7
8
9
None
-
-
<1
96
14
26
97
98
99
6
-
0.0
60
1.7
13
0.0
3.0
3.6
3.0
6.5
0.0
17
<1
16
2
Hß(50)
560
540
0.28
0.48
0.32
0.42
-
66
39
59
29
40
65
38
41
HMOR(45)
6.3
2.6
HZSM-5(45) 670
8
HUSY(40)
830
-
22
28
63
52
17
8
f
H2SO4
11
52
12
Hß(75)g
Naß(75)h
Liß(75)i
600
580
600
0.22
n.d.j
n.d.j
2.1
1.7
2.9
22
4
3.3
0.53
2.3
1.6
8
3
aMannitol 182 mg (1.00 mmol), catalyst 50 mg. bBET specific surface area. cCalculated by NH3-
d
e
f
TPD using h-peak. Selectivity to 1,4-dehydration. (conversion) – (total yields of products).
H2SO4 0.77 mg, 413 K. g413 K, 3 h. hIon-exchanged with NaNO3.37 iIon-exchanged with LiCl.38
jNo h-peak.
We studied the effect of the Si/Al ratio of Hß zeolite on the dehydration of mannitol at 413 K
for 1 h (Figure 3). Conversion of mannitol was only 10% at a Si/Al ratio of 12.5, but the
conversion increased with increasing Si/Al ratio. Hß(75) showed the highest catalytic activity;
the conversion was 96% and isomannide yield was 54%. The best Si/Al ratio was the same as
that in the dehydration of sorbitol.21,22 A further increase in the Si/Al ratio decreased the catalytic
activity due to low acidity. A Brønsted acid was necessary for this reaction, as evidenced by the
deactivation of Hß by ion-exchange with Li+ and Na+ (Table 1, entries 8 and 9). As a result, a
volcano-type curve is depicted for the conversion against the Si/Al ratio (Figure 3, left axis). It
has been hypothesized that a hydrophobic acid site shows high catalytic activity in water-
ACS Paragon Plus Environment
12

Page 13 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
involved reactions, as proposed in the literatures.39-43 Therefore, activity is increased with
increasing Si/Al ratio over a specific range.21,22,43 Moreover, the turnover frequency of acid sites
for the dehydration was constantly elevated from 2.6 h-1 to 130 h-1 with increasing Si/Al ratio
(Figure 3, right axis). Thus, in this work, the hydrophobicity of Hß with different Si/Al ratios
was measured by water vapor adsorption. In Figure 4, the right axis represents the adsorption
amount of N2 at 77 K per that of H2O at 298 K at a relative pressure p/p0 of 0.1, which correlates
with hydrophobicity.44 The analysis clearly shows that an increase in the Si/Al ratio improves the
hydrophobicity. In contrast, zeolites with lower Si/Al ratios have a larger amount of acid sites
(Figure 4, left axis). Hence, we conclude that Hß(75) shows high catalytic activity due to a good
balance between acid amount and hydrophobicity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
By considering the above factors, we found that Hß(75) gave 63% yield of isomannide under
the optimized conditions of 413 K for 3 h (Table 1, entry 7), which is 1.8-fold higher than the
best result ever reported (35%).25 In the reuse experiments, Hß(75) was reusable three times, but
the activity was slightly decreased in the fourth run probably due to formation of extra-
framework aluminum oxide (Figures S5-S7). Since the selectivity was constant in the reuse
experiments, a slight increase in the reaction temperature could compensate for the small
deactivation.
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Effect of Si/Al ratio of Hß on the dehydration of mannitol at 413 K for 1 h. Red
diamonds, mannitol conversion; black triangles, isomannide yield; blue circles, turnover
frequency [= (amount of isosorbide  2 + amount of 1,4-mannitan)/(amount of acid site)/1 h].
Figure 4. Dependence of hydrophobicity and acid amount on Si/Al ratio of Hß zeolites. Red
diamonds, acid amount determined by NH3-TPD; blue triangles, amount of water vapor
adsorption at 298 K compared to that of N2 adsorption at 77 K, p/p0 = 0.1.
ACS Paragon Plus Environment
14

Page 15 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
Mechanistic Study on Mannitol Dehydration over Hß(75). The time course was measured
in the mannitol dehydration over Hß(75) at 413 K to clarify the reaction pathway (Figure 5). The
curve of the remaining amount of mannitol had an inflection point at ca. 15 min prior to a
smooth decrease in a first-order manner. The presence of an induction period was due to the
phase change of mannitol. Namely, the reaction temperature was lower than the melting point of
mannitol (440443 K), and therefore mannitol was initially in the solid state. Slow dehydration
of solid mannitol formed dehydration products and water, which gradually transformed the solid-
phase mannitol into the liquid phase by dissolution and/or cryoscopy. Hence, dehydration of
mannitol was accelerated in the first 15 min. Among the products, 1,4-mannitan and 2,5-sorbitan
were first produced in parallel. The 1,4-mannitan yield reached its maximum value of 10% at 40
min and then decreased by its dehydration to isomannide. The amount of 2,5-sorbitan also had a
local maximum at 1 h, showing that 2,5-sorbitan underwent further reaction. Therefore, we
tested the reaction of 2,5-sorbitan in the presence of a small amount of water, which did not give
mannitol, 1,4-mannitan, or isomannide. Thus, dehydration of mannitol to 2,5-sorbitan is
irreversible under the conditions used in this study, and successive reaction of 2,5-sorbitan
results in degradation. Notably, the amount of isomannide increased linearly in the beginning
and was further enhanced with time. A plot of selectivity versus conversion clearly showed the
formation of isomannide from the initial stage (Figure S8). This is a typical example of a rake
mechanism:45 dehydration of mannitol produces 1,4-mannitan on the catalyst, and part of 1,4-
mannitan is successively converted to isomannide before leaving the catalyst (Scheme S1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Time course of mannitol dehydration over Hß(75) at 423 K.
We performed mannitol dehydration at 413–428 K to evaluate the effect of ß-zeolite on the
kinetic parameters. Since direct analysis of the rake mechanism is complicated, we approximated
the rake mechanism as a pseudo first-order reaction to determine the rate constants for 1,4-
dehydration (k1) and 2,5-dehydration (k2) shown in Scheme 2. The k1 value was determined from
a total formation rate of 1,4-mannitan and isomannide because part of 1,4-mannnitan is
converted to isomannide before leaving the catalyst. The pseudo first-order rate constants were
reasonably determined as summarized in Table 2 by curve fitting of the time courses after
inflection points at different temperatures (Figure S9). The apparent activation energies
calculated from the Arrhenius plot (Figure 6) were 107+13 kJ mol-1 for k1 and 126+12 kJ mol-1
for k2. Therefore, k1 is smaller than k2 even after taking into account possible errors. The error is
mainly derived from a small deviation in reaction rate affected by the melting state of mannitol.
For better clarification of the difference in activation energy between 1,4- and 2,5-dehydration,
we plotted the logarithm of the product ratio (1,4-mannitan + isomannide)/(2,5-sorbitan) against
ACS Paragon Plus Environment
16

Page 17 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
T-1 at similar conversion values of 5580% (eqs. 13, Figure 7). We verified that the product
ratio is independent of the conversion values in the range studied (Figure S8). In this case, the
error can be removed because the ratio does not depend on the absolute reaction rate but is equal
to the relative value of k1/k2. This analysis more clearly shows that the apparent activation energy
for 1,4-dehydration is 11+6 kJ mol-1 lower than that for 2,5-dehydration. In contrast, sulfuric acid
gave 11+2 kJ mol-1 higher apparent activation energy for 1,4-dehydration than that for 2,5-
dehydration (Figure 7). This result was qualitatively consistent with the DFT calculations, which
indicated that proton-catalyzed dehydration of mannitol in the open space prefers 2,5-
dehydration. Thus, we propose that the steric confinement of the ß-zeolite pore disfavors 2,5-
dehydration of mannitol by ca. 20 kJ mol-1 compared to open space, which leads to the selective
synthesis of isomannide over the Hß catalyst.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Rate constants of mannitol dehydration over Hß(75)
a
b
T /K
413
418
423
428
k1
k2
0.027
0.041
0.066
0.078
0.011
0.019
0.030
0.039
aMannitol to 1,4-mannitan. bMannitol to 2,5-sorbitan.
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Arrhenius plots of 1,4-dehydration (black) and 2,5-dehydration (red) for Hß(75).
1,4 - mannitan  isommanide k1

(1)
2,5 - sorbitan
k2
E
RT
a1
(E E
)
a1
a2
k1 A e
A
1
1
RT


e
(2)
E
a2
RT
k2
A2
A2e
Accordingly,
 (Ea1  Ea2 )
A
1
1,4 - mannitan  isomannide
ln


(3)
2,5 -sorbitan
RT
A2
Figure 7. Temperature dependence of the ratio of 1,4-dehydration/2,5-dehydration for Hß(75)
(black) and sulfuric acid (red). ΔEa = Ea1-Ea2
ACS Paragon Plus Environment
18

Page 19 of 33
ACS Catalysis
1
2
3
4
5
6
CONCLUSIONS
7
8
9
DFT calculations indicate that 2,5-dehydration is energetically more favored than 1,4-
dehydration in open space, which is consistent with the experimental result with sulfuric acid.
Focusing on the transition state, the computations show that 2,5-dehydration needs more space
around the reaction center than 1,4-dehydration. Based on this finding, we succeeded in selective
conversion of mannitol to isomannide using the precisely-defined microporous confinement of
Hß zeolite. Since the pores of ß-zeolites are slightly larger than mannitol, 2,5-dehydration is
selectively inhibited by an increase in apparent activation energy of 20 kJ mol-1, and favors 1,4-
dehydration toward the formation of isomannide as the final product. Optimization of the Si/Al
ratio increased the catalytic activity due to the well-balanced acid amount and hydrophobicity,
maximizing the yield of isomannide to 63% at Si/Al = 75.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information. Schematic diagram of reaction setup, enlarged images of
intermediates in DFT calculations, additional experimental results, and scheme of the rake
mechanism.
AUTHOR INFORMATION
Corresponding Authors
*E-mail: hasegawa@cat.hokudai.ac.jp (Jun-ya Hasegawa)
*E-mail: fukuoka@cat.hokudai.ac.jp (Atsushi Fukuoka)
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 33
1
2
3
4
Notes
5
6
7
The authors declare no competing financial interests.
8
9
ACKNOWLEDGMENT
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
This work was supported by Japan Science and Technology Agency (JST) ALCA.
REFERENCES
(1) Sheldon, R. A. Green Chem. 2014, 16, 950-963.
(2) Gallezot, P. Chem. Soc. Rev. 2012, 41,1538-1558.
(3) Luterbacher, J. S.; Alonso, D. M.; Dumesic, J. A. Green Chem. 2014, 16, 4816-4838.
(4) Kobayashi, H.; Fukuoka, A. Green. Chem. 2013, 15, 1740-1763.
(5) Fukuoka, A.; Dhepe, P. L. Angew. Chem. Int. Ed. 2006, 45, 5161-5163.
(6) Kobayashi, H.; Ito, Y; Komanoya, T.; Hosaka, Y.; Dhepe, P. L.; Kasai, K.; Hara, K.;
Fukuoka, A. Green Chem. 2011, 13, 326-333.
(7) Palkovits, R.; Tajvidi, K.; Procelewska, J.; Rinaldi, R.; Ruppert, A. Green Chem. 2010, 12,
972-978.
(8) Hilgert, J.; Meine, N.; Rinaldi, R.; Schüth, F. Energy Environ. Sci. 2013, 6, 92-96.
(9) Kobayashi, H.; Yamakoshi, Y.; Hosaka, Y.; Yabushita, M.; Fukuoka, A. Catal. Today
2014, 226, 204-209.
(10) Takagaki, A.; Furusato, S.; Kikuchi, R.; Oyama, S. T. ChemSusChem 2015, 8, 3769-3772.
(11) Makkee, M.; Kieboom, A. P. G.; van Bekkum, H. Starch-Starke 1985, 37, 136-141.
(12) Yamaguchi, A.; Sato, O.; Mimura, N.; Hirosaki, Y.; Kobayashi, H.; Fukuoka, A.; Shirai,
M. Catal. Commun. 2014, 54, 22-26.
(13) Lichtenthaler, F. W. Acc. Chem. Res. 2002, 35, 728-737.
ACS Paragon Plus Environment
20

Page 21 of 33
ACS Catalysis
1
2
3
4
5
6
(14) Fenouillot, F; Rousseau, A.; Colomines, G.; Saint-Loup, R.; Pascault, J.-P. Prog. Polym.
Sci. 2010, 35, 578-622.
7
8
9
(15) Thiem, J.; Bachmann, F. Macromol. Chem. 1991, 192, 2163-2182.
(16) Kricheldorf, H. R.; Masri, M. A. J. Polym. Sci. Part A Polym. Chem. 1995, 33, 2667-2671.
(17) Montassier, C.; Ménézo, J. C.; Moukolo, J.; Naja, J.; Hoang, L. C.; Barbier, J.; Boitiaux, J.
P. J. Mol. Catal., 1991, 70, 65-84.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(18) Andrews, A.; Bhatia, K. K.; Fagan, P. J. US Pat. 2002, 6 689 892.
(19) Xi, J.; Zhang, Y.; Ding, D.; Xia, Q,; Wang, J.; Liu, X.; Lu, G.; Wang, Y. Appl. Catal. A:
2014, 469, 108-115.
(20) Morita, Y.; Furusato, S.; Takagaki, A.; Hayashi, S.; Kikuchi, R.; Oyama, S. T.
ChemSusChem 2014, 7, 748-752.
(21) Kobayashi, H.; Yokoyama, H.; Feng, B.; Fukuoka, A. Green Chem., 2015, 17, 2732-2735.
(22) Otomo, R.; Yokoi, T.; Tatsumi, T. Appl. Catal. A: Gen. 2015, 505, 28-35.
(23) Shi, J.; Shan, Y.; Tian, Y.; Wan, Y.; Zheng, Y.; Feng, Y. RSC Adv. 2016, 6, 13514-13521.
(24) Cubo, A.; Iglesias, J.; Morales, G.; Melero, J. A.; Moreno, J.; Sánchez-Vázquez, R. Appl.
Catal. A Gen. 2017, 531, 151-160.
(25) Bock, K.; Pedersen, C.; Thøgersen, H. Acta Chem. Scand. 1981, 35b, 441-449.
(26) Yamaguchi, A.; Sato, O.; Mimura, N.; Shirai, M. RSC Adv. 2014, 4, 45575-45578.
(27) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B, 1988, 37, 785-789.
(28) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257-2261.
(29) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(30) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105, 2999-3093.
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 33
1
2
3
4
5
6
(31) Yabushita, M.; Kobayashi, H.; Shrotri, A.; Hara, K.; Ito, S.; Fukuoka, A. Bull. Chem. Soc.
Jpn., 2015, 88, 996-1002.
7
8
9
(32) Weinhold, F. Nature 2001, 411, 539-541.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(33) Lynch, B. J.; Truhlar, D. G. J. Phys. Chem. A 2001, 105, 2936-2941.
(34) Cheng, Y.-T.; Wang, Z.; Gilbert, C. J.; Fan, W.; Huber, G. W. Angew. Chem. Int. Ed.,
2012, 51, 11097-11100.
(35) Dusselier, M.; van Wouwe, P.; Dewaele, A.; Jacobs, P. A., Sels, B. F. Science, 2015, 349,
78-80.
(36) Niwa, M.; Katada, N. Chem. Rec., 2013, 13, 432-455.
(37) Wang, Y.; Kamiya, Y.; Okuhara, T. Water Res. 2007, 41, 269-276.
(38) Accardi, R. J.; Lobo, R. F. Microporous Mesoporous Mater. 2000, 40, 25-34.
(39) Namba, S.; Hosonuma, N.; Yashima, T. J. Catal. 1981, 72, 16-20.
(40) Okuhara, T. Chem. Rev. 2002, 102, 3641-3666.
(41) Corma, A.; Domine, M. E.; Valencia, S. J. Catal. 2003, 215, 294-304.
(42) Zapata, P. A.; Faria, J.; Ruiz, M. P.; Jentoft, R. E.; Resasco, D. E. J. Am. Chem. Soc. 2012,
134, 8570-8578.
(43) Siddiki, S. M. A. H.; Toyao, T.; Kon, K.; Touchy, A. S.; Shimizu, K. J. Catal. 2016, 344,
741-748.
(44) J. Pires, M. L. Pinto, A. Carvalho, M. B. de Carvalho, Adsorption, 2003, 9, 303-309.
(45) Li, D.; Bui, P.; Zhao, H. Y.; Oyama, S. T.; Dou, T.; Shen, Z. H. J. Catal., 2012, 290, 1-12.
SYNOPSIS
TOC graphic
ACS Paragon Plus Environment
22

Page 23 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1
94x65mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 25 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2
38x18mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3
45x27mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 27 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4
60x47mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 33
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5
70x51mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 29 of 33
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6
51x43mm (300 x 300 DPI)
ACS Paragon Plus Environment