OSDA-Free Zeolite Beta with High Aluminum Content Efficiently Catalyzes a Tandem Reaction for Conversion of Glucose to 5-Hydroxymethylfurfural

Article abstract of DOI:10.1002/cctc.201500837

Organic structure-directing agent (OSDA)-free zeolite Beta with high Al content exhibit remarkably high catalytic performance in the conversion of glucose to 5-hydroxymethylfurfural (HMF) by virtue of their appropriate acid properties; specifically, a sufficient number of Lewis acid sites were generated by calcination of the NH₄-form zeolite, while the original Bronsted acid sites were substantially maintained, with both acid sites being in close proximity. The OSDA-free Beta catalyst, having a large number of Bronsted acid sites and sufficient Lewis acid sites, showed superior catalytic performance to a physical mixture of each type of acid catalyst with a similar number of each acid site. This behavior could be ascribed to the high reactivity and slow intrazeolitic diffusion of fructose. The structure of the Al species active in the isomerization of glucose to fructose is discussed based on relationships between catalytic activities and changes in Al species by the calcination, changes that were observed by ²⁷Al magic-angle spinning (MAS) NMR and IR spectroscopies. Lewis and Bronsted, working together: Organic structure-directing agent (OSDA)-free zeolite Beta with high Al content effectively promotes conversion of glucose to 5-hydroxymethylfurfural (HMF). The high aluminum content of the zeolite provides suitable acid properties; specifically, a large number of Bronsted acid sites with a sufficient number of Lewis acid sites in close proximity. The zeolite calcined at 773 K showed >70 % yield of HMF.

Full text of DOI:10.1002/cctc.201500837

DOI: 10.1002/cctc.201500837
Full Papers
OSDA-Free Zeolite Beta with High Aluminum Content
Efficiently Catalyzes a Tandem Reaction for Conversion of
Glucose to 5-Hydroxymethylfurfural**
[
a]
Ryoichi Otomo, Toshiyuki Yokoi,* and Takashi Tatsumi
Organic structure-directing agent (OSDA)-free zeolite Beta with
high Al content exhibit remarkably high catalytic performance
in the conversion of glucose to 5-hydroxymethylfurfural (HMF)
by virtue of their appropriate acid properties; specifically, a suf-
ficient number of Lewis acid sites were generated by calcina-
showed superior catalytic performance to a physical mixture of
each type of acid catalyst with a similar number of each acid
site. This behavior could be ascribed to the high reactivity and
slow intrazeolitic diffusion of fructose. The structure of the Al
species active in the isomerization of glucose to fructose is dis-
cussed based on relationships between catalytic activities and
changes in Al species by the calcination, changes that were
tion of the NH -form zeolite, while the original Brønsted acid
4
sites were substantially maintained, with both acid sites being
in close proximity. The OSDA-free Beta catalyst, having a large
number of Brønsted acid sites and sufficient Lewis acid sites,
27
observed by Al magic-angle spinning (MAS) NMR and IR
spectroscopies.
Introduction
It has long been recognized that 5-hydroxymethylfurfural
tuning of acid properties is indispensable for achieving a high
yield of HMF from glucose.
(
HMF) is a bio-based “platform” material for producing useful
[
1]
chemicals. Increasing research interest has been paid to the
synthesis of HMF from sugars, particularly from glucose be-
cause it is the most abundant component of woody biomass.
However, the one-pot conversion of glucose to HMF is still
a challenging subject because it is a multistep reaction by way
of fructose as an intermediate and each step requires different
In previous work, we studied the conversion of glucose to
HMF by using zeolite Beta synthesized with tetraethylammoni-
[
3]
um (TEA) cations. We found that calcination or steam treat-
ment of zeolite Beta increased the number of Lewis acid sites
at the expense of Brønsted acid sites and that the generated
Lewis acid sites promoted the isomerization of glucose, the
rate-determining step, followed by the dehydration of fructose
to HMF over the remaining Brønsted acid sites. However, it
was also found that the Lewis acid sites inevitably promoted
[2]
types of active species; the isomerization step requires, for
example, Lewis acids and the dehydration step requires Brønst-
ed acids (Scheme 1). Thus, a qualitative and quantitative fine-
[
3]
undesired reactions of fructose. It occurred to us that the se-
lectivity to HMF in the dehydration step would be determined
by the ratio of Brønsted acid sites to Lewis acid sites (B/L ratio)
and that a large number of Al atoms could be used for gener-
ating a sufficient number of Lewis acid sites and a large
number of Brønsted acid sites.
Zeolite Beta can be synthesized by a variety of methods,
typically by using TEA cations as organic structure-directing
[
4]
agents (OSDAs). In 2008, Xiao et al. reported the OSDA-free
[
5]
synthesis of zeolite Beta. Since this pioneering work, other
groups have also reported the OSDA-free synthesis of zeolite
[
6]
Beta. According to these reports, the Si/Al ratios of the prod-
ucts ranged from 4 to 6, which is lower than that of conven-
tional zeolite Beta synthesized with TEA cations. Today, OSDA-
free synthesis is one of the hottest topics in zeolite science,
and it has attracted considerable attention because OSDA-free
synthesis has several advantages over the conventional meth-
Scheme 1. Tandem reaction for conversion of glucose to HMF.
[
a] Dr. R. Otomo, Prof. Dr. T. Yokoi, Prof. Dr. T. Tatsumi
Chemical Resources Laboratory
Tokyo Institute of Technology
ods using OSDAs. For example, it reduces costs, CO and NOx
2
4
259 Nagatsuta-cho, Midori-ku, Yokohama 226-8503 (Japan)
emissions, and the energy consumption arising from the use
of OSDAs. Therefore, OSDA-free synthesis of zeolites will cer-
tainly contribute to the development of potential green chemi-
cal processes if the thus-prepared zeolites can exhibit high cat-
E-mail: yokoi@cat.res.titech.ac.jp
[
**] OSDA=Organic structure-directing agent.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201500837.
ChemCatChem 2015, 7, 4180 – 4187
4180
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
alytic performance. There are several research papers on the
was not changed by subsequent calcination (Table 1). XRD
analysis confirmed that the framework structure was retained
even after calcination at 7008C (Figure 1).
[7]
catalytic applications of OSDA-free zeolite Beta. Taking ad-
vantage of the high cation-exchange capacity, Fe-exchanged
[
7d,e]
OSDA-free Beta was applied in NO decomposition.
The
x
high concentration of Brønsted acid sites resulted in high cata-
lytic activity in the cracking of hydrocarbons, hydroconversion
Table 1. Structural and acid properties of catalyst samples.
[6d,7b,e]
of decane, and hydroamination of styrene.
In contrast, it
[
a]
[b]
[c]
Sample
Si/Al BAS
LAS
B/L
S
BET
2
V
micro
was reported that high Al content resulted in fast deactivation
and that the combination of steam treatment and acid treat-
ment provided high-silica zeolite Beta to delay the deactiva-
À1
À1
À1
À1
[
mmolg ] [mmolg
]
[m g ] [mlg
]
Na-form Beta(OF)
Beta(OF)-Cal450
Beta(OF)-Cal500
Beta(OF)-Cal550
Beta(OF)-Cal600
Beta(OF)-Cal700
Beta(TEA)-Cal500
Beta-HBL
5.4
–
–
–
643
0.25
0.24
0.24
0.23
0.23
0.22
0.24
0.24
0.22
0.21
0.33
5.5 0.61
5.5 0.71
5.5 0.62
5.5 0.31
5.5 0.20
0.21
0.20
0.22
0.24
0.19
0.13
0.013
0.18
0.19
0.11
2.9 607
3.6 603
2.9 606
1.3 591
1.1 575
1.9 632
8.5 663
0.78 581
0.95 547
3.7 890
[
7b,c]
tion.
In this way, applications of OSDA-free zeolite Beta in
acid catalysis have already been reported. However, they only
focused on catalytic reactions over the Brønsted acid sites. Al-
though Lewis acid sites should be present in the reported cat-
15
75
0.25
0.11
0.14
0.18
[
7a–c,e]
alysts,
there has been no report on the Lewis acid cataly-
sis of OSDA-free Beta, still less on bifunctional catalysis em-
ploying both the Lewis and Brønsted acid sites.
Beta(OF)-Cal700-AT 52
Beta-CV
USY
13
5.4 0.42
Herein, we demonstrate the first effective conversion of glu-
cose to HMF over OSDA-free zeolite Beta, the acid properties
of which are tuned by a simple method making the best use
of its high Al content. The high Al composition of OSDA-free
zeolite Beta provides a sufficient number of Lewis acid sites
with a large number of Brønsted acid sites in close proximity
to each other. Such a combination of acid sites should be
a key factor for the efficient production of HMF from glucose.
[
a] Number of Brønsted acid sites. [b] Number of Lewis acid sites. [c] Ratio
of the number of Brønsted acid sites to Lewis acid sites.
The state of the Al atoms before and after the calcination
27
was investigated by Al magic-angle spinning (MAS) NMR
analysis (Figure 2). The NH -form sample showed a main peak
4
Results and Discussion
Characterization of Beta(OF) zeolites
Na-form Beta(OF) zeolite showed a typical XRD pattern as-
signed to *BEA-type structure without any other crystal phase
contaminates (Figure 1). The Si/Al ratio of the Na-form
Beta(OF) zeolite was 5.4, which was much lower than that of
the conventional zeolite Beta (typically, >10). After ion-ex-
change, the Si/Al ratio of the sample was 5.5, and the ratio
2
7
Figure 2. Al MAS NMR spectra of Beta(OF) zeolites calcined at different
temperatures.
at 57 ppm and a shoulder peak at 54 ppm, which were as-
signed to tetrahedral Al atoms at the T1, T2, and T3–T9 sites in
[
8]
the *BEA-type framework. For Beta(OF)-Cal450, the intensity
of the peak at 57 ppm decreased as a result of the dealumina-
tion and a new peak appeared at 0 ppm, which is assigned to
an octahedral Al species. For Beta(OF)-Cal500, the intensity of
the peak at 57 ppm was further decreased and the peak at
0
ppm weakened, broadening to the high field region. Calcina-
tion above 6008C led to a further decrease of the peak at
7 ppm with the peak at 54 ppm mostly retained, indicating
5
that dealumination mainly occurred with Al atoms at the T3–
T9 sites, as reported for zeolite Beta synthesized with TEA cat-
Figure 1. XRD patterns of a) Na-form Beta(OF), b) NH
c) Beta(OF)-Cal500, d) Beta(OF)-Cal600, and e) Beta(OF)-Cal700.
4
-form Beta(OF),
[
9]
ions. A very broad peak was observed in the range of À20 to
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4181 ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
1
0 ppm for Beta(OF)-Cal600 and -Cal700, indicating the forma-
tion of species with strong quadrupolar interactions. In addi-
tion, new weak peaks appeared at approximately 30 and
+
6
0 ppm, which are assigned to Al(OH)2 species in a hydrated
form counterbalancing the negative charge on the framework
and extra-framework Al species in tetrahedral coordination, re-
[
9,10]
spectively.
Figure 3 shows the IR spectra of the samples measured at
2
58C after the pre-treatment. Beta(OF)-Cal450 exhibited ab-
À1
sorption bands attributed to bridging OH groups at 3612 cm
À1
and silanol groups at approximately 3740 cm , superposed on
Scheme 2. Dealumination process through sequential hydrolysis of SiÀOÀAl
bonds.
The acid properties of the catalyst samples were evaluated
by IR observation using pyridine as a probe molecule (Table 1).
The numbers of Brønsted and Lewis acid sites of Beta(OF)-
À1
Cal450 were 0.61 and 0.21 mmolg , respectively. Beta(OF)-
Cal500 showed an increased number of Brønsted acid sites be-
+
cause below 5008C the NH4 cation is not completely decom-
posed to form Brønsted acid sites. Thus, the ratio of the
number of Brønsted acid sites to that of Lewis acid sites (B/L
ratio) was increased from 2.9 to 3.6. As the calcination temper-
ature was further increased, Brønsted acid sites were de-
creased; in particular, there was a remarkable difference be-
tween Beta(OF)-Cal550 and Beta(OF)-600. We assume that in
addition to the dealumination occurring below 5508C, the for-
+
mation of Al(OH)2 partly accounts for this marked decrease in
Figure 3. IR spectra after the pre-treatment of a) Beta(OF)-Cal450,
b) Beta(OF)-Cal500, c) Beta(OF)-Cal550, d) Beta(OF)-Cal600, and e) Beta(OF)-
Cal700.
Brønsted acid sites. Although the dealumination was enhanced
and Brønsted acid sites were decreased with calcination tem-
perature, the number of Lewis acid sites was not much
changed, suggesting that the Al species formed by the deep
dealumination do not act as Lewis acid sites. Thus, the total
number of acid sites and the B/L ratio were decreased.
À1
a broad band (3000–3700 cm ) derived from the hydrogen
bonding of OH groups. Moreover, a small peak appeared at
À1
3
782 cm , which is assigned to partially hydrolyzed Al species
[11]
bonded to the framework.
In the spectrum of Beta(OF)-
À1
Cal500, a new band appeared at 3665 cm , which is assigned
Dehydration of fructose to HMF by using Beta(OF) zeolites
[12]
to extra-framework clusters.
For Beta(OF)-Cal600 and
À1
-
Cal700, the band at 3665 cm intensified, with a concomitant
First, we compared the catalytic performance of the Beta(OF)
zeolites in the dehydration of fructose at a short reaction time
(0.5 h) because the dehydration reaction is much faster than
the isomerization reaction (Table 2). Beta(OF)-Cal450, -Cal500
and -Cal550, which have relatively large numbers of Brønsted
acid sites, showed higher conversion of fructose and higher
yields of HMF than Beta-Cal600 and -Cal700 with relatively
small numbers of Brønsted acid sites. The yield of HMF plotted
against the number of Brønsted acid sites shows that the de-
hydration rate clearly increased along with the number of
Brønsted acid sites (Figure 4). Even when the reaction time
was short (about 0.5 h), HMF was formed in high yield, leading
to a slope less than 1, which was probably due to the de-
creased concentration of fructose and subsequent reactions of
HMF. As the undesired reactions of fructose seemed to pro-
ceed over the Lewis acid sites in parallel with the dehydration
and as the Beta(OF) catalysts had similar numbers of Lewis
acid sites, the selectivity to HMF was determined by the B/L
ratio, as seen in Figure 5. A similar relationship has been re-
ported by Huber et al. in the conversion of glucose to levulinic
À1
decrease in the band at 3612 cm .
27
Al MAS NMR and IR spectroscopies revealed that a type of
dealumination—a cleavage of SiÀOÀAl bonds of the frame-
work—was caused by autosteaming during the calcination
(
Scheme 2). Tetrahedral Al atoms at the T3–T9 sites in the
framework were first converted to octahedral species in weak
quadrupolar interaction below 5008C [Eqs. (1), (2), and (3)], as
indicated by the sharp peak at 0 ppm in the NMR spectra. The
Al species shown in Equations (2) and (3) were experimentally
À1
verified by the IR peak at 3782 cm . Above 5008C, a large
amount of octahedral Al species formed and interacted with
each other, resulting in broadening of the peak at 0 ppm. A
part of these species condensed to form small clusters, as
À1
shown by the peak at approximately 3665 cm in the IR spec-
[12,13]
tra.
In the deep dealumination processes occurring above
6
008C, SiÀOÀAl bonds were further hydrolyzed and the species
bearing OH groups condensed to form other species, including
cationic species [Eqs. (4) and (5)], as shown by the NMR peaks
at approximately 30 ppm.
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4182
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Lewis acid sites but significantly different numbers of Brønsted
acid sites (Figure 6). A variety of products were produced,
among which fructose, anhydroglucose, HMF, and furfural
were the main products and are thus plotted in the figures. In
[
a]
Table 2. Dehydration of fructose to HMF by using Beta(OF) zeolites.
[
b]
[c]
[d]
Entry
Catalyst
Conv.
%]
Select.
[%]
Yield
[%]
[
1
2
3
4
5
6
Beta(OF)-Cal450
Beta(OF)-Cal500
Beta(OF)-Cal550
Beta(OF)-Cal600
Beta(OF)-Cal700
Beta(TEA)-Cal500
Beta-HBL
81
91
80
79
71
68
97
76
83
74
61
60
70
85
62
76
60
48
43
48
82
[e]
7
[
4
[
a] Reaction conditions: catalyst, 0.1 g; fructose, 0.67 mmol; water,
.5 mL; DMSO, 0.5 mL; THF, 15 mL; temperature, 1808C; time, 0.5 h.
b] Conversion of fructose. [c] Selectivity for HMF. [d] Yield of HMF. [e] Cat-
alyst, 0.5 g.
Figure 4. Dependence of the HMF yield on the number of Brønsted acid
sites.
Figure 6. Evolution of glucose conversion (circles), fructose yield (squares),
HMF yield (triangles), anhydroglucose (diamonds), and furfural (crosses) in
the presence of a) Beta(OF)-Cal500 and b) Beta(OF)-Cal700. Reaction condi-
tions: catalyst, 0.1 g; glucose, 0.67 mmol; water, 4.5 mL; DMSO, 0.5 mL; THF,
15 mL; temperature, 1808C.
the early stages of the reaction, fructose was produced over
the Lewis acid sites and anhydroglucose was produced over
the Brønsted acid sites. The yield of fructose increased along
with the glucose conversion up to the maximum and then
continuously decreased, accompanied by an increase in the
yield of HMF. The yield of anhydroglucose changed in a similar
manner to fructose; anhydroglucose is hydrolyzed to glucose
and then converted to fructose and other products. The reac-
tion proceeded by the same routes for both catalysts, but the
time course of the yields of fructose and HMF was different;
for Beta(OF)-Cal500, the yield of fructose increased up to 10%
at a conversion of 44% and then decreased, whereas Beta(OF)-
Cal700 showed the maximum fructose yield of 15% at a con-
version of 53%. Beta(OF)-Cal500 exhibited a higher yield of
HMF up to a maximum of 72% at 6 h. With prolonged the re-
action time, the HMF yield gradually decreased, which was
probably due to its subsequent reactions. For Beta(OF)-Cal700,
the HMF yield slowly increased along with reaction time and fi-
nally reached 57%. These differences could be attributed to
Figure 5. Correlation between selectivity for HMF and B/L ratio.
[14]
acid over metal–phosphate catalysts. Comparison with Beta(-
TEA)-Cal500 with the Si/Al ratio of 15 confirms that for
Beta(OF) zeolites the large number of Brønsted acid sites re-
sulting from the high Al content led to the high catalytic per-
formance in the dehydration of fructose (Table 2, entries 2 and
6
).
Conversion of glucose to HMF by using Beta(OF) zeolites
Next, time course profiles of the reaction were compared for
Beta(OF)-Cal500 and -Cal700, which had similar numbers of
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4183
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
different reaction rates in the dehydration of fructose to HMF.
For Beta(OF)-Cal500, fructose was promptly consumed by its
dehydration reaction to HMF. In contrast, the dehydration rate
for Beta(OF)-Cal700 was slower, resulting in the relatively large
amount of fructose remaining. For Beta(OF)-Cal700, the lower
selectivity to HMF in the dehydration step, as shown in Table 2,
led to the decreased maximum yield of HMF in the conversion
of glucose.
As mentioned above, a large number of Brønsted acid sites
is a key factor for the selective production of HMF from glu-
cose. This prompted us to ask: can we obtain HMF in a high
yield by using a mixture of zeolite Beta with a low B/L ratio
and that with a high B/L ratio? Thus, we performed reaction
using a mixture of Beta(OF)-Cal700 (B/L ratio=1.1) and zeolite
Beta with a high B/L ratio (about 8.5), designated as “Beta-
HBL”. The combination of 0.1 g of Beta(OF)-Cal700 and 0.5 g of
Beta-HBL led to an increase in the selectivity of HMF, accompa-
nied by a decrease in the selectivity of fructose compared with
the single use of Beta(OF)-Cal700 (Table 3, entries 5 and 6).
However, the conversion of glucose was not much changed;
Beta-HBL mainly promoted the dehydration of fructose
(Table 3, entry 7). Clearly, the whole product selectivity is large-
ly governed by Beta(OF)-Cal700. Additionally, the combination
of 0.1 g of Beta(OF)-Cal700 and 0.1 g of Beta(OF)-Cal500
showed 56% selectivity for HMF, indicating that the apparent
selectivity is determined by the individual selectivity
The solvent system is also a key factor for the high yield of
HMF; dimethyl sulfoxide (DMSO) and tetrahydrofuran sup-
pressed the subsequent reactions of HMF into levulinic acid
[
15]
[3a]
and humins. Our previous results and the data in Table S1
in the Supporting Information) clearly demonstrated the ef-
(
fects of the mixed solvents.
The selectivities of fructose and HMF were compared at high
conversions of glucose (80–88%) and the different selectivities
for the Beta(OF) zeolites can be seen in Table 3. Beta(OF)-
of each Beta(OF) component (Table 3, entries 2, 5,
Table 3. Conversion of glucose to HMF by using Beta(OF) and other types of
catalysts.
and 8). This behavior could be ascribed to slow intra-
zeolitic diffusion as a result of a relatively large mo-
lecular dimension of fructose compared with the
pore diameter of zeolite Beta. In fact, Beta(OF) zeo-
lite, which has a large number of Brønsted acid sites
in close proximity to the Lewis acid sites, has an ad-
vantage over mixtures of each type of catalysts.
From an industrial viewpoint, a high concentration
of substrate is feasible. Notably, Beta(OF)-Cal500 still
exhibited high catalytic performance in a concentrat-
ed reaction system; 68% selectivity for HMF was ach-
ieved at 97% conversion of glucose in 10 wt% solu-
tion (Table 3, entry 9).
[a]
Entry
Catalyst
Conv.
Fructose
Select. [%]
HMF
Select. [%]
[
b]
[c]
[d]
[
%]
1
2
3
4
5
Beta(OF)-Cal450
Beta(OF)-Cal500
Beta(OF)-Cal550
Beta(OF)-Cal600
Beta(OF)-Cal700
Beta(OF)-Cal700+ Beta-HBL
Beta-HBL
Beta(OF)-Cal500+ Beta(OF)-Cal700
Beta(OF)-Cal500
Beta(OF)-Cal700-AT
Beta(TEA)-Cal500
Beta-CV
81
86
88
86
80
85
19
96
97
40
81
65
83
4
3
4
12
15
4
3
3
1
2
65
66
67
55
45
55
39
56
68
24
50
43
49
[
e]
6
7
8
9
1
[
[
f]
g]
0
[h]
1
1
1
9
14
1
1
2
3
The Al content of aluminosilicate zeolites synthe-
sized by OSDA-free routes, except for RTH-type zeo-
[
i]
USY
[
16]
lites, is generally uncontrollable and too high for
direct catalytic applications; for example, the Si/Al
ratio for OSDA-free Beta ranges from 4 to 6. Such
high-Al-content zeolites are normally used as solid
[
0
a] Reaction conditions: catalyst, 0.1 g; glucose, 0.67 mmol; water, 4.5 mL; DMSO,
.5 mL; THF, 15 mL; temperature, 1808C; time, 3 h. [b] Conversion of glucose. [c] Se-
lectivity for fructose. [d] Selectivity for HMF. [e] The combination of 0.1 g of Beta(OF)-
Cal700 and 0.5 g of Beta-HBL. The total numbers of Brønsted and Lewis acid sites
were 0.075 and 0.026 mmol, respectively, B/L=2.9. [f] The combination of 0.1 g of
Beta(OF)-Cal500 and 0.1 g of Beta(OF)-Cal700. The total numbers of Brønsted and
acid catalysts after being converted to low-Al-content
À1
[7b,c]
Lewis acid sites were 0.091 and 0.039 mmolg , respectively, B/L=2.3. [g] Reaction
ones by post-synthesis treatments.
Therefore, we
conditions: catalyst, 0.75 g; glucose, 10 mmol; water, 4.5 mL; DMSO, 0.5 mL; THF,
prepared a high-silica Beta(OF) zeolite by treating
1
5 mL; temperature, 1808C; time, 3 h. [h] Reaction time, 5 h. [i] Reaction time, 1 h.
Beta(OF)-Cal700 with 1m HNO at 808C (designated
3
as “Beta(OF)-Cal700-AT”). However, Beta(OF)-Cal700-
AT, with a Si/Al ratio of 52, showed poor catalytic per-
Cal450, -Cal500, and -Cal550, which had a large number of
Brønsted acid sites, showed high selectivities to HMF (65, 66,
and 67%, respectively). On the other hand, Beta(OF)-Cal600
and -Cal700, with a small number of Brønsted acid sites,
showed low selectivities to HMF (55 and 45%) but high selec-
tivities to fructose (12 and 15%). These results verified that
a large number of Brønsted acid sites is an essential factor for
achieving the selective production of HMF. At the same time,
a sufficient number of Lewis acid sites are essential for the
rate-determining isomerization step, as revealed by our previ-
formance (Table 3, entry 10). Beta(OF)-Cal500 showed higher
selectivity for HMF than Beta(TEA)-Cal500 with a Si/Al ratio of
15. Furthermore, it took 5 h for Beta(TEA)-Cal500 to achieve
a conversion over 80% (Table 3, entries 2 and 11). In addition,
a commercially available zeolite with a Si/Al ratio of 13, desig-
nated as Beta-CV, showed poor performance (Table 3,
entry 12). These findings clearly indicate that the removal of Al
atoms from Beta(OF) leads to decreased catalytic activity and
that Beta(OF) can be used as a highly effective acid catalyst
when the high Al composition is kept. Beta(OF)-Cal500 was
also compared with USY zeolite, which can be synthesized
without OSDAs. USY, with a Si/Al ratio of 5.4, showed a faster
reaction rate than all of the Beta(OF) zeolites; however,
[
3a]
ous work. Thus, Beta(OF)-Cal500, which had a large number
of Brønsted acid sites with a sufficient number of Lewis acid
sites, showed the highest yield of HMF.
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4184
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Beta(OF)-Cal500 showed a higher selectivity and a higher yield
of HMF than USY, although the B/L ratios were similar (Table 3,
entries 2 and 13). Beta(OF)-Cal500 and USY were characterized
by IR spectroscopy with CO as the probe molecule to estimate
the acid strength (Figure S1 in the Supporting Information).
Based on the shift of the absorption band for adsorbed CO
molecules, USY has stronger Lewis acid sites compared with
Beta(OF)-Cal500. Such strong Lewis acid sites would escalate
sub-reactions of fructose as well as the rapid isomerization of
glucose, resulting in decreased selectivity of HMF. Thus,
Beta(OF)-Cal500 was found to be more suitable than the USY
zeolite for the conversion of glucose to HMF.
a broad NMR peak at dꢀ30 ppm, it showed a slightly lower
conversion of glucose. Thus, neither the species at n˜ =
À1
3665 cm nor the extra-framework cationic species exhibited
a clear correlation with the catalytic activity in the conversion
of glucose. This behavior suggests that active Lewis acid sites
were already generated by calcination at a relatively low tem-
perature and that the number of active sites was not much
changed by calcination above 6008C, corresponding to the
À1
finding that the IR band at n˜ =3782 cm was commonly ob-
served for Beta(OF) zeolites in similar intensities. Therefore, we
propose that the Al species bearing OH groups, showing
À1
a peak at n˜ =3782 cm , are the active sites for the isomeriza-
The USY zeolite was calcined at different temperatures to
prepare a series of USY zeolites with different acid properties
tion of glucose. Such Al species bearing mono- and di-hydroxyl
groups are generated in the hydrolysis of SiÀOÀAl bonds
[Scheme 2, Eqs. (2) and (3)], but these species are subsequently
converted to other species under severe conditions. Therefore,
(
Table S2 in the Supporting Information). The prepared USY
zeolites with different B/L ratios showed a clear correlation be-
tween the B/L ratio and the HMF selectivity, similar to the
Beta(OF) zeolites, indicating that the correlation could be gen-
erally applied to other zeolites (Figure S2 in the Supporting In-
formation). Note that the correlation for the USY zeolites was
seen in a lower selectivity range compared with the Beta(OF)
zeolites. These results suggested that the selectivity for HMF
was determined not only by the B/L ratio but also by the activ-
ity of each acid site.
À1
the band at n˜ =3782 cm remained approximately constant
for all the samples.
Note that these species are hydrated in reaction systems
containing water. One can assume that such hydrated species
do not show any activity because water molecules are irrever-
sibly adsorbed on the Lewis acid sites and such water mole-
cules cannot be replaced with glucose molecules. In fact,
2
+
Al(OH)(aq) species in aqueous solutions can exchange ligand
4
water molecules in a dissociative mechanism at a rate 10
3
+
times higher than Al(aq) species, which are formed in acidic
Active Al species for isomerization of glucose
[
19]
solutions of AlCl3. This finding suggests the possibility that
Al species bearing OH groups on zeolites can act as Lewis acid
sites to polarize carbonyl groups even in the presence of
water. Moreover, there are several recent research papers re-
porting that Ti, Cr, or Sn species bearing OH groups are active
1
3
C NMR isotope experiments revealed that on zeolite Beta,
glucose is isomerized to fructose through an intramolecular hy-
[3a,17]
dride transfer over the Lewis acid sites.
The Meerwine–
Ponndorf–Verley and Oppenauer (MPVO) reaction is an inter-
molecular hydride transfer between carbonyl and alcohol
groups over Lewis acid sites, which is analogous to the isomer-
ization of glucose. Dealuminated zeolite Beta have been re-
ported to be excellent catalysts for the MPVO reaction, where
three types of Al species have been proposed as the active
[
20]
for the isomerization of monosaccharides.
We have checked for the leaching of Al species during the
reaction by using Beta(OF)-Cal500. We found a very small con-
centration of Al species in the reaction solution by inductively
coupled plasma analysis (Figure S3 in the Supporting Informa-
tion). Furthermore, the reaction did not significantly proceed
after the removal of the solid catalyst, confirming that the re-
action was heterogeneously promoted by the Beta(OF) cata-
lyst. This result is consistent with previous literature reports in
which the use of zeolite catalyst in the absence of NaCl and
acids showed only a very small amount of Al and Si atoms
leached into the reaction solution, which did not affect the
[
18]
species.
Al species bearing OH groups observed at
in IR spectra were proposed by van Bekkum
Wichterlovµ et al. attributed the catalytic activity to
À1
3
782 cm
[
18b]
et al.
extra-framework AlO counter cations inside the pores of zeo-
x
[
18c]
lite Beta.
Sacco et al. suggested that the species associated
À1
with the 3665 cm IR band were responsible for the selective
[18d]
production of the cis-alcohol in the MPV reduction.
In our
[
21]
work, the number of Lewis acid sites estimated by IR measure-
ments with pyridine correlated with the activity for the conver-
sion of glucose. However, this method does not estimate the
activity of particular active sites, but only estimates the global
acid sites in the catalyst because pyridine is a highly basic mol-
ecule. Therefore, the structure of the Al species active in the
isomerization of glucose was considered based on the relation-
ship between the catalytic activity and the change of Al spe-
cies by calcination.
conversion of glucose to fructose and HMF.
Reuse of Beta(OF) zeolite
Finally, the reusability of Beta(OF)-Cal500 was tested. Figure 7
shows the conversion of glucose and the product distributions
in four consecutive runs, indicating that the conversion of glu-
cose was fairly constant (about 86–91%) and that the product
distributions did not significantly change either; for example,
the HMF selectivity was nearly constant at 65–67%. Further-
more, no significant change in the structural properties of the
catalyst was observed even after four runs (Table S3, Figure S4
in the Supporting Information).
Looking at the conversions summarized in Table 3, we could
infer that Beta(OF)-Cal450 has active sites for the isomerization
of glucose. However, it showed neither an IR band at n˜ =
À1
3
665 cm nor broad peaks in the NMR spectra. Although
À1
Beta(OF)-Cal700 showed an intense band at n˜ =3665 cm and
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4185
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
(
2.5m) at 808C to allow ion-exchange to take place and to produce
the NH -form free of Na ions. The NH -form zeolite was heated to
4
4
the desired temperature (450–7008C) in a muffle furnace at a rate
À1
of 38Cmin and calcined at the final temperature for 5 h, followed
by cooling to ambient temperature. The obtained catalysts are des-
ignated as “Beta(OF)-Calx”, where x is the calcination temperature.
Zeolite Beta synthesized in previous work by using TEA cations
[3]
was used as Beta(TEA). A commercial zeolite Beta with a Si/Al
ratio of 75 was supplied by Clariant, Japan, and designated as
“
Beta-HBL”. A commercial zeolite Beta with a Si/Al ratio of 13, was
purchased from Clariant, and designated as “Beta-CV”. A commer-
cial USY zeolite “HSZ-350HUA” was purchased from Tosoh, Japan.
Powder X-ray diffraction (XRD) patterns of the prepared samples
Figure 7. Reusability test of Beta(OF)-Cal500 in the conversion of glucose to
HMF. Reaction conditions: catalyst, 0.1 g; glucose, 0.67 mmol; water, 4.5 mL;
DMSO, 0.5 mL; THF, 15 mL; temperature, 1808C; time, 3 h. The solid catalyst
was recovered by filtration, washed with water and ethanol, and dried at
ambient temperature after each run.
were collected on a Rigaku Ultima III diffractometer using CuK ra-
a
diation (40 kV, 40 mA). Elemental analyses of the samples were per-
formed on an inductively coupled plasma-atomic emission spec-
trometer (ICP-AES, Shimadzu ICPE-9000). Nitrogen adsorption
measurements were performed on a BEL-mini analyzer, BEL Japan.
Prior to the measurements, all samples were degassed at 4008C
for 1 h.
Conclusions
2
7
Solid-state Al MAS NMR spectra of hydrated samples were record-
ed on a JEOL ECA-600 spectrometer at a resonance frequency of
1
1
We demonstrated that the high aluminum composition of
OSDA-free zeolite Beta can be successfully utilized by taking
advantage of the acid properties appropriate for bifunctional
catalysis in the conversion of glucose to HMF. Calcination of
56.4 MHz using a 4 mm sample rotor with a spinning rate of
5.0 kHz. Chemical shifts were referenced relative to an aqueous
Al(NO ) solution. The p/12 pulse was used for excitation with 1 s
3
3
of relaxation delays, which was long enough to attain saturation
[
22]
the NH -form Beta(OF) zeolite at 5008C generated a sufficient
recovery.
4
number of Lewis acid sites that are active for the isomerization
of glucose with the original Brønsted acid sites substantially
maintained. Calcination at higher temperature did not increase
the number of the active Lewis acid sites. The large number of
Brønsted acid sites gave high selectivity to HMF in the dehy-
dration step, leading to a high HMF yield of over 70%. The
proximity of both types of acid sites, derived from a high Al
content, was also found to be a key factor for the selective
The acid properties of the zeolites were evaluated by FTIR tech-
niques using pyridine as a probe molecule. FTIR spectra were ob-
tained by using a Jasco FTIR 4100 spectrometer equipped with
a TGS detector. The powdered samples were pelletized into a self-
supporting disk 1 cm in diameter, which was held in a glass cell.
After pre-treatment at 4508C, the sample was cooled to 258C and
a spectrum was recorded. Then, the sample was heated to 1508C
and an excess of pyridine (~1 kPa) was introduced into the cell.
The gas phase was evacuated at 2508C for 10 min and a spectrum
was recorded at 1508C. For quantitative calculation of the ad-
sorbed pyridine, the reported molar extinction coefficients of the
2
7
production of HMF. Al MAS NMR and IR spectroscopies
showed a change of Al species caused by calcination. It was
proposed that Al species bearing OH groups are the active
Lewis acid sites for the isomerization of glucose. The Beta(OF)
zeolite calcined at 5008C showed stable activity during four
consecutive runs.
À1
absorption bands at n˜ =1455 and 1545 cm for the Lewis and
[23]
Brønsted acid sites were used, respectively. The IR measurements
had differences of Æ5%.
The catalytic reactions were performed in mixed solvents (water/di-
methyl sulfoxide/tetrahydrofuran) by using a Teflon-lined stainless
steel autoclave (50 mL). At a set point, the reaction was quenched
by cooling the autoclave in an ice bath. Time course profiles were
obtained by collecting the reaction results at different time points.
Reaction runs were repeated several times, achieving similar results
with differences of Æ3%. The reaction mixture was filtered prior to
the quantitative analysis to remove the solid catalyst. The homoge-
neous solution after the filtration was separated into an aqueous
phase and an organic phase by adding excess NaCl (~3 g) as a salt-
ing-out agent. Sugar substrates and water-soluble products in the
aqueous phase were analyzed by HPLC (Shimadzu, LC-20 A series
with RI and UV (280 nm) detectors in the ion-exclusion mode,
REZEX RHM-Monosaccharide column, Phenomenex). HMF and
other products in the organic phase were also analyzed by HPLC
(Shimadzu LC-10A with a UV detector (260 nm) in the reverse
phase mode, Gemini C18 column, Phenomenex). For the reusability
tests, the solid catalyst recovered by filtration after each run was
thoroughly washed with water and ethanol. After being dried at
ambient temperature, the catalyst was used in the next run.
Experimental Section
All the materials were used as received from the respective chemi-
cal companies. Fumed silica Cab-O-Sil M5 (Cabot, USA) and sodium
aluminate (EP, Wako, Japan) were used as Si and Al sources, respec-
tively, in the OSDA-free synthesis. Glucose and fructose of reagent
grade were purchased from Kanto Chemical, Japan. Other materials
were of reagent grade and purchased from Wako.
OSDA-free zeolite Beta was prepared by following the report by
[5]
Xiao et al.: Sodium aluminate was added to a NaOH aqueous so-
lution and completely dissolved. After the addition of fumed silica,
the resulting slurry was stirred for 1 h. Finally, Beta seed crystals
(
3 wt%) were added and the thus-obtained aluminosilicate gel
with the molar composition of 1.0SiO :0.02Al O :0.3Na O:25H O
2
2
3
2
2
was transferred to a PTFE-lined stainless steel autoclave and hydro-
thermally treated at 1408C for 24 h. The solid product was recov-
ered by filtration, washed with distilled water, and dried overnight
at 1008C. The Na-form zeolite was stirred in NH NO3 solution
4
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4186
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
9] J. A. van Bokhoven, D. C. Koningsberger, P. Kunkeler, H. van Bekkum,
A. P. M. Kentgens, J. Am. Chem. Soc. 2000, 122, 12842–12847.
Acknowledgements
[
10] a) S. Li, A. Zheng, Y. Su, H. Fang, W. Shen, Z. Yu, L. Chen, F. Deng, Phys.
Chem. Chem. Phys. 2010, 12, 3895–3903; b) Z. Yu, A. Zheng, Q. Wang, L.
Chen, J. Xu, J.-P. Amoureux, F. Deng, Angew. Chem. Int. Ed. 2010, 49,
We thank Dr. Ulrich Müller (BASF SE) and Feng-Shou Xiao (Zhe-
jiang University) for helpful discussions. This work was partly per-
formed under the framework of the INCOE (International Network
of Centers of Excellence) project coordinated by BASF SE.
8
657–8661; Angew. Chem. 2010, 122, 8839–8843.
11] A. Vimont, F. Thibault-Starzyk, J. C. Lavalley, J. Phys. Chem. B 2000, 104,
86–291.
12] a) C. Jia, C. P. Massiani, D. Barthomeuf, J. Chem. Soc. Faraday Trans.
993, 89, 3659–3665; b) I. Kiricsi, C. Flego, G. Pazzuconi, W. O. Parker, R.
[
[
2
1
Keywords: biomass · glucose · heterogeneous catalysis ·
tandem reaction · zeolites
Millini, C. Perego, G. Bellussi, J. Phys. Chem. 1994, 98, 4627–4634.
13] A. Omegna, M. Vasic, J. A. van Bokhoven, G. Pirngruber, R. Prins, Phys.
Chem. Chem. Phys. 2004, 6, 447–452.
[
[
1] a) Y. Romµn-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312,
933–1937; b) J. N. Chheda, G. W. Huber, J. A. Dumesic, Angew. Chem.
Int. Ed. 2007, 46, 7164–7183; Angew. Chem. 2007, 119, 7298–7318; c) X.
Tong, Y. Ma, Y. Li, Appl. Catal. A 2010, 385, 1–13; d) M. J. Climent, A.
Corma, S. Iborra, Green Chem. 2011, 13, 520–540; e) A. A. Rosatella, S. P.
Simeonov, R. F. M. Frade, C. A. M. Afonso, Green Chem. 2011, 13, 754–
[14] R. Weingarten, Y. T. Kim, G. A. Tompsett, A. Fernµndez, K. S. Han, E. W.
Hagaman, W. C. Conner Jr., J. A. Dumesic, G. W. Huber, J. Catal. 2013,
304, 123–134.
[15] a) S. H. Mushrif, S. Caratzoulas, D. G. Vlachos, Phys. Chem. Chem. Phys.
2012, 14, 2637–2644; b) Y. Romµn-Leshkov, J. A. Dumesic, Top. Catal.
2009, 52, 297–303; c) J. Wang, J. Ren, X. Liu, G. Lu, Y. Wang, AIChE J.
2013, 59, 2558–2566.
1
7
93; f) R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra,
H. J. Heeres, J. G. de Vries, Chem. Rev. 2013, 113, 1499–1597.
2] a) A. Takagaki, M. Ohara, S. Nishimura, K. Ebitani, Chem. Commun. 2009,
[16] T. Yokoi, M. Yoshioka, H. Imai, T. Tatsumi, Angew. Chem. Int. Ed. 2009, 48,
9884–9887; Angew. Chem. 2009, 121, 10068–10071.
[
6
276–6278; b) R. Huang, W. Qi, R. Su, Z. He, Chem. Commun. 2010, 46,
[17] J. S. Kruger, V. Choudhary, V. Nikolakis, D. G. Vlachos, ACS Catal. 2013, 3,
1279–1291.
1115–1117; c) E. Nikolla, Y. Romµn-Leshkov, M. Moliner, M. E. Davis, ACS
Catal. 2011, 1, 408–410.
3] a) R. Otomo, T. Yokoi, J. N. Kondo, T. Tatsumi, Appl. Catal. A 2014, 470,
[18] a) E. J. Creyghton, S. D. Ganeshie, R. S. Downing, H. van Bekkum, J. Mol.
Catal. A 1997, 115, 457–472; b) P. J. Kunkeler, B. J. Zuurdeeg, J. C. van
der Waal, J. A. van Bokhoven, D. C. Koningsberger, H. van Bekkum, J.
Catal. 1998, 180, 234–244; c) O. Bortnovsky, Z. Sobalík, B. Wichterlovµ,
Z. Bastl, J. Catal. 2002, 210, 171–182; d) B. Akata, J. Warzywoda, A. Sac-
co, Jr., J. Catal. 2004, 222, 397–403.
[
[
3
4
18–326; b) R. Otomo, T. Tatsumi, T. Yokoi, Catal. Sci. Technol. 2015, 5,
001–4007.
4] a) R. L. Wadlinger, G. T. Kerr, E. J. Rosinski (Mobil Oil Corporation),
US3308069, 1967; b) R. B. Borade, A. Clearfield, Micropor. Mater. 1996, 5,
2
1
89–297; c) P. R. Hari Prasad Rao, M. Matsukata, Chem. Commun. 1996,
441–1442; d) M. A. Camblor, A. Corma, S. Valencia, J. Mater. Chem.
[19] T. W. Swaddle, J. Rosenqvist, P. Yu, E. Bylaska, B. L. Phillips, W. H. Casey,
Science 2005, 308, 1450–1453.
1
998, 8, 2137–2145.
5] B. Xie, J. Song, L. Ren, Y. Ji, J. Li, F.-S. Xiao, Chem. Mater. 2008, 20, 4533–
535.
6] a) G. Majano, L. Delmotte, V. Valtchev, S. Mintova, Chem. Mater. 2009,
1, 4184–4191; b) Y. Kamimura, W. Chaikittisilp, K. Itabashi, A. Shimoji-
[20] a) M. Moliner, Y. Romµn-Leshkov, M. E. Davis, Proc. Natl. Acad. Sci. USA
2010, 107, 6164–6168; b) R. Bermejo-Deval, R. S. Assary, E. Nikolla, M.
Moliner, Y. Romµn-Leshkov, S.-J. Hwang, A. Palsdottir, D. Silverman, R. F.
Lobo, L. A. Curtiss, M. E. Davis, Proc. Natl. Acad. Sci. USA 2012, 109,
9727–9732; c) R. Gounder, M. E. Davis, ACS Catal. 2013, 3, 1469–1476;
d) V. Choudhary, S. H. Mushrif, C. Ho, A. Anderko, V. Nikolakis, N. S. Mar-
inkovic, A. I. Frenkel, S. I. Sandler, D. G. Vlachos, J. Am. Chem. Soc. 2013,
135, 3997–4006; e) P. Wolf, M. Valla, A. J. Rossini, A. Comas-Vives, F.
NfflÇez-Zarur, B. Malaman, A. Lesage, L. Emsley, C. CopØret, I. Hermans,
Angew. Chem. Int. Ed. 2014, 53, 10179–10183; Angew. Chem. 2014, 126,
10343–10347.
[
[
4
2
ma, T. Okubo, Chem. Asian J. 2010, 5, 2182–2191; c) Y. Kamimura, S. Ta-
nahashi, K. Itabashi, A. Sugawara, T. Wakihara, A. Shimojima, T. Okubo, J.
Phys. Chem. C 2011, 115, 744–750; d) B. Xie, H. Zhang, C. Yang, S. Liu, L.
Ren, L. Zhang, X. Meng, B. Yilmaz, U. Müller, F.-S. Xiao, Chem. Commun.
2011, 47, 3945–3947; e) K. Itabashi, Y. Kamimura, K. Iyoki, A. Shimojima,
T. Okubo, J. Am. Chem. Soc. 2012, 134, 11542–11549.
[
7] a) B. Yilmaz, U. Müller, M. Feyen, S. Maurer, H. Zhang, X. Meng, F.-S.
Xiao, X. Bao, W. Zhang, H. Imai, T. Yokoi, T. Tatsumi, H. Gies, T. De Baer-
demaeker, D. De Vos, Catal. Sci. Technol. 2013, 3, 2580–2586; b) T. De
Baerdemaeker, B. Yilmaz, U. Müller, M. Feyen, F.-S. Xiao, W. Zhang, T. Tat-
sumi, H. Gies, X. Bao, D. De Vos, J. Catal. 2013, 308, 73–81; c) Y. Kubota,
K. Itabashi, S. Inagaki, Y. Nishita, R. Komatsu, Y. Tsuboi, S. Shinoda, T.
Okubo, Chem. Mater. 2014, 26, 1250–1259; d) M. Ogura, K. Itabashi, J.
Dedecek, T. Onkawa, Y. Shimada, K. Kawakami, K. Onodera, S. Nakamura,
[21] a) D. W. Gardner, J. Huo, T. C. Hoff, R. L. Johnson, B. H. Shanks, J.-P. Tes-
sonnier, ACS Catal. 2015, 5, 4418–4422; b) T. Wang, J. A. Glasper, B. H.
Shanks, Appl. Catal. A 2015, 498, 214–221.
[22] A. Omegna, J. A. van Bokhoven, R. Prins, J. Phys. Chem. B 2003, 107,
8854–8860.
[23] C. A. Emeis, J. Catal. 1993, 141, 347–354.
ˇ
T. Okubo, J. Catal. 2014, 315, 1–5; e) P. Sazama, B. Wichterlovµ, S.
Sklenµk, V. I. Parvulescu, N. Candu, G. Sµdovskµ, J. D eˇ de cˇ ek, P. Klein, V.
ˇˇ
´
Received: July 28, 2015
Pashkova, P. Stastny, J. Catal. 2014, 318, 22–33.
[
8] G. Valerio, A. Goursot, R. Vetrivel, O. Malkina, V. Malkin, D. R. Salahub, J.
Am. Chem. Soc. 1998, 120, 11426–11431.
Revised: August 16, 2015
Published online on October 16, 2015
ChemCatChem 2015, 7, 4180 – 4187
www.chemcatchem.org
4187
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim